Jump to content

Wikipedia talk:Manual of Style/Medicine-related articles/Archive 11

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 5←Archive 9Archive 10Archive 11Archive 12Archive 13→Archive 15
Note: Continued at Wikipedia talk:Manual of Style/Medicine-related articles/Archive 12

Product pricing

This has become unwieldy. Let's see if a clean start below helps. Best, Barkeep49 (talk)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Should there be a separate section or just the current brief mention? QuackGuru (talk) 17:31, 1 October 2019 (UTC)

  • I think it merits discussion. One issue is that WP articles are usually about the active molecule and not about any product(s) containing that molecule. Another one is that list prices in cases of innovative drugs can be very misleading, as actual, negotiated prices are confidential and frequently carry rebates up to 70% (and more) from the list price. Then, there are also risk-sharing agreements that affect the effective price; pay-per-result arrangements; bundle pricing, etc. etc.
Until we agree on how we present the prices keeping in mind the requirement to be WP:GLOBAL, I suggest to leave out a separate section. Deinitely, adding it after just a few hours of "discussion" between just two participants, even if highly respected as they are, seems premature. — kashmīrī TALK 18:38, 2 October 2019 (UTC)
I was trying to find guidance on product pricing and I missed it. I moved one sentence and just added "The pharmaceutical industry has tried to conceal medication prices because of their continuing legal cases in the United States. There is a lack of transparency regarding medication and vaccine prices among non-governmental organizations." Numerous articles have content about pricing. It needs a separate section. If you disagree with the content I added you can just comment it out or delete it for now. QuackGuru (talk) 19:13, 2 October 2019 (UTC)
The prices in the BNF and from medicaid are fairly sound. Sure a bit of work may be required to find appropriate sources and craft the appropriate wording. Doc James (talk · contribs · email) 21:45, 2 October 2019 (UTC)
Again, the world is not just the US - in Europe, prices are not disclosed for an entirely different reason - and also the way you used "pharmaceutical industry" is incorrect: the majority of pharmaceutical companies are manufacturers of generic drugs and dietary supplements (sic!). — kashmīrī TALK 22:15, 2 October 2019 (UTC)
Please edit the section to improve the wording or you can just deleted the content I wrote. I'm not sure what the best wording should be. I think the separate section should remain because so many articles discuss pricing. QuackGuru (talk) 01:30, 3 October 2019 (UTC)
I have re-written it. WhatamIdoing (talk) 00:08, 12 October 2019 (UTC)
Thanks and looks good. Doc James (talk · contribs · email) 02:53, 24 October 2019 (UTC)
  • I'm not convinced we need a pricing section or that the text there is understandable. Do we generally want prices on our drug articles or treatment sections? I think this section might encourage trivia. It isn't clear from the text that the do-this/don't-do-this follows. It asks lots of questions but the consequence would appear to be original-research. I certainly don't support QuackGuru's text, which was replaced by WhatamIdoing. The unreasonable increase in the cost of insulin is a notable issue in the US only and perhaps properly belongs in some article on drug company pricing. Is this sort of thing relevant in general (which a guideline should cover) or just a special case of including information per WP:WEIGHT. There may be merit in noting that some treatments are popular/uncommon due to pricing issues if that issue is covered by reliable sources. But to be honest, that seems to me to be pretty standard WP editing practice. -- Colin°Talk 10:03, 12 November 2019 (UTC)
    • I think that the consequence of answering those questions, in most cases, is going to be accurate reporting of what your source says. One of the biggest problems in the foreseeable-and-solvable category is someone getting a good source on third-world wholesale prices and writing "The price is two cents per pill", which some reader is going to interpret as "My prescription will cost 60 (local) cents at the pharmacy". The meaning of those questions is "don't write 'the price'. Instead, write down which price your source is talking about". I do not think there is any scope for OR in that. I think that it militates against OR.
    • As to whether we want prices, we probably do, in some cases. Those cases are IMO mostly WHO essential medicines (for which "The estimated average wholesale price worldwide, according to the WHO, is around two cents per pill" is a not unreasonable answer) and fancy new medicines (for which the answer probably sounds like "The company announced a list price of $120,000K per year in the US, but the revenue per patient is expected to be lower due to negotiated discounts and the patient assistance program", assuming that my [probably business magazine] source mentioned that kind of reality). This information isn't being included for medical reasons. It's being included because some of our readers are interested in money. WhatamIdoing (talk) 15:29, 12 November 2019 (UTC)
      • In what way are they interested in money? Is it "Can I afford to buy this" (paying privately for a drug); "Will my insurance/health-service provide this?"; or just generally interest in which drugs are very very cheap, fairly cheap, expensive, eye-wateringly expensive... At the moment I can see the text provoking some kind of table with all countries and with wholesale, retail and discount prices if such were available. Ibuprofen lead (only) says "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose. In the United States, it costs about US$0.05 per dose". It isn't covered by the body nor do we list details about generic vs brand pricing. Aspirin also only covers cost in the lead, and only wholesale cost for "developing world". It lists the monthly cost in the US, but per your notes does not say whether that is wholesale, but also doesn't say what that is treating. Valproate is the same. Hmm, information about developing world prices added only to the lead. Hmm, could that possibly be done by a certain editor with a focus on a medical translation project? The prices are of course now 4-5 years old and not maintained. The fact that this is being only added by one editor, only to the lead, not maintained, very specific prices that get out-of-date, all rings alarm bells for whether Wikipedia really is the place for this. Isn't there a better way to cover this, and one that doesn't waste precious space in our lead sections for what is a trivia detail. The "very cheap ... borderline acceptable to NHS" scale may be significant, but whether the wholesale price in the US in 2014 was 4c or 5c really isn't. -- Colin°Talk 21:44, 12 November 2019 (UTC)
        • The "money" audiences that I happen to care about are public policy (the "very cheap ... borderline acceptable to NHS" scale) and business stuff (e.g., someone trying to figure out why this company's in the news for over charging). If it's easier to write US$0.04 than "very cheap" (which is additionally a value judgment and therefore not something we could say without a source saying that), then that doesn't worry me.
          I'm not overly worried about keeping them up to date (although I mostly wouldn't put prices in the lead). As most of the prices for generics are stable over the years, being strictly up to date isn't terribly important. People aren't going to have radically different understandings of a drug if it turns out that the price has changed from one to four cents to now being two to five cents. For fancy new drugs, what gets the most attention in sources (and therefore what's most WP:DUE) is their announced initial price, which is always going to be its initial price, even though it will likely drop later. WhatamIdoing (talk) 02:31, 13 November 2019 (UTC)
          • Limited guidance on drug pricing was already in WP:PHARMMOS before the section in question was added here. Does it need covering at both MOSs? (My guess is "probably".) If not, which should be the one to talk about drug pricing?
            Having guidance on pricing might prove useful when discussing medical tourism, and the reasons why grey and black markets exist in medical care. Little pob (talk) 09:42, 13 November 2019 (UTC)
            • Although WP:PHARMMOS says "Economics: Global sales, distribution, cost in major English speaking countries, etc" it doesn't really offer guidance. I see now valproate does mention the cost later in a Cost section. I think "Econonomics" is a better title, which would encourage other economic data such as sales (particularly for blockbuster drugs that influence company shares) and whether the drug has been rejected on economic grounds (such as NICE rejection for drugs too expensive). I don't think there is any WP:WEIGHT/MOS:LEADREL argument to include specific cost details in the lead at all, though the other economic factors may be relevant for a small set of drugs. The cost-per-dose is not widely covered by reliable sources on drugs or medical treatments, largely because it is so variable per country and also over time. I think it is very difficult to justify including "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose.[13] In the United States, it costs about US$0.05 per dose." in the lead. This is noisy trivia which in a reliable source would merely get "this very cheap drug" or some such one/two word adjective. If we lack reliable sources to turn price details into an adjective, then probably better to simply remove from the lead on the grounds that nobody is commenting on the price as being something worth noting. I suspect it has been added to the lead because the medical translation project only translates the lead and the information was felt relevant to developing countries. I reject that argument, if that is the case, on the grounds that en:wp is here for its readers and not to be some source-code for other wp -- the translators can always be advised to include cost details from the body if that is felt relevant on other WPs. -- Colin°Talk 10:38, 13 November 2019 (UTC)

Proposal

Add the following to the "product pricing" section: "Detailed cost information does not belong in the lead, though some significant economic points may be relevant for some drugs (blockbusters, drugs rejected as too expensive, etc)." The "product pricing" section be renamed "Economics" to fit with the earlier heading examples.

The current wording could be shortened to "The cost of medicine or procedures should include specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the year." I don't think it worth including "if covered by reliable sources" because that is true of everything. I don't think the "indirect costs such as lost wages for the patient" is relevant to product pricing. This seems more relevant to disease articles, covering time off due to to sickness or to attend hospital for treatment or tests. -- Colin°Talk 10:45, 13 November 2019 (UTC)

  • Strong support. I am dead against having drug prices in the lead section and only mention them in the article body where it is specifically discussed in reliable sources as a matter of public interest (like Zolgensma being the most expensive drug in the world). — kashmīrī TALK 15:27, 13 November 2019 (UTC)

I've rewritten the section mostly per the proposal above. I agree with Kashmiri that it is hard to justify the routine addition of specific pricing details even in the article body. We have perhaps a difficulty translating raw figures into a more appropriate adjective that our sources would do. Few of our sources mention specific costs, so WP:WEIGHT makes it hard to argue for. There are of course drugs whose price is especially notable.

I've written the lead advice as: "Detailed cost information is rarely appropriate for inclusion in the lead." To be honest, I feel somewhat that we are having to include this guidance merely because one editor chose to ignore our basic rules for WP:LEAD: that they summarise article bodies. Any such prices in the lead could be moved/removed simply per WP:LEAD. -- Colin°Talk 13:58, 14 November 2019 (UTC)

It is irrelevant you rewritten the content. That's not the consensus across med articles. QuackGuru (talk) 14:06, 14 November 2019 (UTC)
QuackGuru articles do not have "consensus"; editors do. That is why we are discussing this here. You tried to add the text "Cost information may be included in the lead, but this is done on a case-by-case basis.", this was reverted, and you restored it again without attempting to achieve consensus. That's edit warring, which will get you blocked. I see from your earlier edit to include "The pharmaceutical industry has tried to conceal medication prices because of their continuing legal cases in the United States. There is a lack of transparency regarding medication and vaccine prices among non-governmental organizations." that you have an agenda. That's even more reason for you to avoid editing or warring over guidelines. The statement you added about the lead is meaningless because all information may be included on a case-by-case basis. Further the specific issue is "detailed cost information" whereas in our reliable sources, a lead section/page would merely note if a drug is cheap or expensive: our reliable sources know how to summarise.
MEDMOS cannot supersede WP:LEAD. Let's examine this consensus of editors. I looked at the first bunch of anti-infective medicines on the WHO list:
So this is a practice followed by exactly one editor, who also has an agenda about information in the lead.
The information doesn't even seem to be correct. Albendazole says "The wholesale cost in the developing world is between 0.01 and 0.06 USD per dose". The 2014 source gives the dose at 400mg and the table strength at 200mg and from $0.0115 to $0.0341 per table (i.e., from $0.023 to 0.0682, which would be rounded to "0.02 and 0.07 USD per dose"). By the following year the highest price is only 0.04 USD, so already out-of-date. Who knows what 2018 or 2019 is. The article claims "in the developing world" though I can't see where this is is indicated in the source. Their own sources of buyers and suppliers does seem to be focused on such nations, but I don't know if that is representative of the developing world. The website does not seem to include prices any newer than 2015, which is a strong concern. For the US price, the source given is a database which is not an allowed source. Readers need to search through the database for the drug then average the individual records themselves.
I see this issue was discussed Wikipedia talk:WikiProject Medicine/Archive 84#Price of medications and concluded "Except in the cases where the sources note the significance of the pricing (which did have consensus), there is no consensus to add the pricing to the articles". Therefore it seems that Doc James and QuackGuru are editing against consensus. It's going to take me a while to read through all the discussions, but once again I see a few folk at WP:MED having ideas the rest of Wikipedia do not support. We don't include the prices of potatoes, televisions or package holidays to Greece. It seems fairly obvious that specific price information is not the role of an encyclopaedia for the general reader. We have a "source" for wholesale prices that appears to be no longer maintained as of 2015. This "source" material is raw data, not meaningful data or commentary that could be used in a lead summary. If we have no consensus to add pricing to articles, except for significant notable cases, then there is very much no consensus to add this to the lead. I request QuackGuru reverts themselves. -- Colin°Talk 18:44, 14 November 2019 (UTC)

I see from Doc James talk page that this issue has become a recent hot topic. James edit warred four times with two editors. He also claimed that the RFC requirement "where the sources note the significance of the pricing" was met merely by a US Database including the drug price among its records. That is a strange way of assessing WP:WEIGHT. And a very deep rejection of what constitutes a "source" -- a pointer to the first page of database results showing 25,095 records. James writes "We know that the pharmaceutical industry is trying really hard to hide medication prices with ongoing legal cases in the United States. Many NGOs including Doctors Without Borders and UNICEF struggle with the lack of transparency around medication and vaccine prices. WP:NOTCENSORED applies here." I am reminded what a wise editor once wrote about people citing NOTCENSORED as a justification for including material: "Anyone who defends their edits by citing WP:NOTCENSORED doesn't have the first clue." The key requirement made by WhatAmIDoing in 2010 was "cost is discussed by significant sources". A database is not a discussion. A number in the BNF (a drug database) is not a discussion.

I think we should summarise and clarify (for those who think a database is a source) the RFC conclusion as "Except in the cases where reliable sources discuss the significance of drug pricing, drug price information is not appropriate in articles. In such cases where price information is relevant, the article should indicate why. The lead, as a summary of article content, should not include specific detailed price information." -- Colin°Talk 19:15, 14 November 2019 (UTC)

It appears that QuackGuru does not wish to justify their edits or explain why the consensus at the RFC Wikipedia talk:WikiProject Medicine/Archive 84#Price of medications is being ignored. I'm also guessing that User:DocJames is ignoring the discussion because the current text supports his position. It seems very clear from the RFC that James does not understand that WP:WEIGHT is not met by the inclusion of a figure in a database. I can get multiple sources for the bus timetable and journey times, yet this raw data is not encyclopaedic. In their discussion with User:Rhododendrites, James also claims that a wide interest in drug prices (particularly those on the WHO essential list) is sufficient to include detailed wholesale pricing for all such medicines. But this is the difference between raw data and encyclopaedic information. The encyclopaedic issue that people are discussing is whether and where the price is low or high, why the price is enormously higher in some countries (e.g. US) and whether price has affected availability of the treatment to patients (e.g., NICE does not consider it's cost/benefit justified). This is the sort of information our articles might include, case by case. Most of the price sources suggested are either now no longer maintained, or incomplete or are merely a link to a database result-set of 25,000 records, none of which meet the standards required for sourcing. User:WhatamIdoing reminded us in 2010 that "cost is discussed by significant sources" is the requirement.

I think that unless those who lost the argument in the RFC desist from edit warring on this page, we may need some administrative action here or another community RFC. I'll reword my text from just above, since it seems some think general discussion of drug pricing justifies absolute inclusion of drug pricing everywhere. Here's what I think the RFC demands MEDMOS says:

"Except in the cases where reliable sources discuss the significance of the price of a particular drug, drug price information is not appropriate in an article. In such cases where price information is relevant, the article should explain why. The lead, as a summary of article content, should not include specific detailed price information."

-- Colin°Talk 12:34, 16 November 2019 (UTC)

You pinged the incorrect user name. The sources being used qualify per WP:MEDRS. Lots of sources that discuss the price in greater detail are avaliable. There is an in depth discussion by WHO for example for each medication listed as essential. Sure I could write an entire section on this for every medication. Wikipedia is also a work in progress. Doc James (talk · contribs · email) 18:47, 21 November 2019 (UTC)
The 2016 RFC concluded, “Except in the cases where the sources note the significance of the pricing (which did have consensus), there is no consensus to add the pricing to the articles.”
Colin, could you have a look at Epipen as an example? Both of your versions are truer to the RFC and more helpful than QuackGuru’s version, which I oppose, but what a useless mess of information at Epipen, presented in an unhelpful fashion, while the pricing controversy is not summarized to the lead at all. Your phrase “... should not include specific detailed price information ...” is still warranted, as the Epipen pricing controversy could be summarized to the lead as a due weight portion of the article without getting into detailed and cumbersome pricing exammples. For an example like Epipen, where there was a pricing controversy, I prefer your 12:34 16 Nov proposal above to either of these.. QuackGuru’s detailed examples are just a repeat of basic editing practice, but are apparently misleading some editors to go against the RFC. Could you bold your new proposal please? SandyGeorgia (Talk) 14:58, 16 November 2019 (UTC)
I also prefer parts of what Colin wrote to what I wrote, especially his list of factors ["include specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average)"], rather than the questions I wrote. I think that including an example (although the specific example likely could be improved) will be useful to people whose writing skills are less developed than Colin's and SandyGeorgia's (and that's probably about 99% of us). WhatamIdoing (talk) 04:47, 19 November 2019 (UTC)
  • Add prices I have commented at many of these drug price discussions over the years. Prices are necessary context for understanding drugs and treatment. Consumers, physicians, journalists, and policy makers all want this information. Challenges for Wikipedia are that (1) prices vary by time, place, and by medical insurance (2) there are international price catalogs, but these are challenging for individual readers to interpret and (3) Wikipedia still does not have good policy of when to include or exclude primary information, and pricing has to come from primary sources. The information that I really want is context of the order of magnitude. Wikipedia articles should clearly communicate whether the price of a dose of a drug is closer to US$0.001, 0.01, 0.10, 1.00, 10, 100, or 1000. If the reader is left thinking that a $100 dose drug is generally affordable, or that a very inexpensive drug is inaccessible, then that reader is lacking a fundamental understanding of the nature of the drug. The most reasonable proposal we have for communicating order of magnitude pricing is to import and publish some price catalog for everything. Blue Rasberry (talk) 17:28, 16 November 2019 (UTC)

What I wrote originally was:

"The cost of medicine or procedures should include specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the year. Some medicines contribute significantly to a pharmaceutical company's turnover (blockbuster drugs), or their high price is a factor in their rejection or difficult acceptance by state health services or insurance companies. Detailed cost information is rarely appropriate for inclusion in the lead."

Upon reading the RFC and thinking more about what a lead should be, I proposed:

"Except in the cases where reliable sources discuss the significance of the price of a particular drug, drug price information is not appropriate in an article. In such cases where price information is relevant, the article should explain why. The lead, as a summary of article content, should not include specific detailed price information."

There is perhaps merit in merging the two. I agree Epipen is a mess and a sort of OR. Same with the WHO essential drug leads that James has added prices to. Here we have the leads of drug articles juxtaposing "developing world" and "US" prices, making some point but lacking a source that makes any point. Bluerasberry, I'm afraid I disagree with some of your points and don't think this is the place to rerun the RFC discussion. Wikipedia does have policies against raw data being dumped in articles and does have guidelines about leads. While price of drugs clearly is a concern, if our sources aren't making commentary on international drug price of drug XXX, then our articles also should not include that information. This is just basic WP:WEIGHT stuff. We are only writing this price information so badly because of using bad sources (online databases containing multiple prices from multiple sources) and the lack of skill of the writer. It smells of the kind of Wikipedia editing where folk just add random factoids and data without thinking about how to present information to the reader and where to put it in the article. Currently we have QuackGuru edit warring and just chucking out all of what I wrote and inserting his own agenda. There doesn't appear to be anyone who likes his edits, but nobody fixing it. I don't edit war, and right now I have a few real-world issues on my plate. So I hope you can together try to agree on a form of words that respects that RFC and our WP:WEIGHT and WP:LEAD guidelines. If you don't like the outcome of the RFC then it needs to be raised again for the whole community to comment, not just the handful with this page watchlisted. -- Colin°Talk 13:11, 19 November 2019 (UTC)

When faced with edit warring, I am disinclined to revert the damage, but support anyone who attempts to restore to RFC-based consensus wording. Also, I will be traveling over the US holiday and unlikely to follow this for a while, so hope someone addresses the issue. I, too, disagree with Bluerasberry's points. SandyGeorgia (Talk) 14:54, 19 November 2019 (UTC)
@Colin: I agree with you on all points about the difficulty of adding price: we lack quality data, the data requires intense regular updating at massive scale, our diverse audience will have challenges interpreting data, and we in Wikipedia lack the social norms, technical infrastructure, and labor base to manage this to our normal standards of quality. In so many ways you have said all these things and your case is well made.
Somehow, this is the development path I want:
  1. We have discussion where somehow, we place a value on including price information. I value general price information highly, and roughly, would like to place all drugs into categories of "nearly free, accessible by most, moderate price, expensive, very expensive", or some such scale. I care less about particular prices and more about informing readers about cost barriers. Other people may have another goal in mind, such as exact price information, but I care more about interpreting price for general readers than having exact financial data at points in time and place. The point here is to develop consensus about what is appropriate to include in infoboxes.
  2. We scope the ideal price system in Wikipedia, which we would have if we had all data, technology, etc.
  3. We identify whomever has the best drug price communication system in the world, and compare ourselves to them, and not necessarily to that ideal pricing system
  4. Based on the existing global norms of whom communicates pricing and how, we compromise from our ideal, and start experiments in communicating pricing in some circumstances with intent to develop our price communication policy. All wiki practices develop over years, and if we will communicate price in 10 years, the time to experiment is now.
Colin, this conversation could go lots of ways. Which way seems most productive to you? What do you see as the long term ending consensus in this? Blue Rasberry (talk) 14:58, 19 November 2019 (UTC)
Recognizing and understanding your wants/wishes/desires as good faith, nowhere in your list of wishes do you mention the policies and guidelines Wikipedia has in place that cover content. Colin has named several of those policies and guidelines ... I can name more (WP:NOT, WP:RS, WP:WEIGHT, WP:V, WP:LEAD, etc.). The best way forward (in this, and in any discussion) is to focus not on wants and wishes, but actual policies and guidelines governing content. Colin's position is rooted in policy and guideline. I gave Epipen as an example hopefully to help focus on the policy-driven addition of pricing information. When MEDRS sources discuss pricing, we can include it, to the extent we also respect WP:WEIGHT, WP:NOT and WP:LEAD. Getting ahead of Wikipedia-wide policy and guideline via "the time to experiment is now" would be folly for WPMED. SandyGeorgia (Talk) 15:30, 19 November 2019 (UTC)
Bluerasberry I appreciate your list of desirable information. It represents an aim that a publisher of original material could target if they hire the right people and pay for access or research. But on Wikipedia we are led by sources. We have seen the rather odd interpretation of WP:WEIGHT and 'reliable sources' to mean that just because someone has a database of arbitrarily-sourced prices from a handful of developing countries in 2014, or a database of US wholesale prices in any given week, that means this information is warranted in not only articles but also the lead. We can only give price information the weight that reliable sources do, when discussing price for that article. A database cannot 'discuss' and has no weight. I'm not sure that infoboxes will work for this, nor does that meet the RFC consensus. The price of a drug is too nuanced and varied to just add 'Low' to an infobox. This is a wiki, however, so it is possible that in future we do identify a good source or measure of weight for this. We currently don't appear to have one. --Colin°Talk 10:24, 21 November 2019 (UTC)
@SandyGeorgia and Colin: I agree with both of what you are saying about the present. Colin expresses that he understands me by saying that we could have reliable drug information from databases in the future, and in the future we might have our choice of expressing that information in any way we like. My objective in this discussion is to seek comments on what realistic expectations anyone has for including prices in Wikipedia if we have the right content quality and technology. I take for granted that consumers need price information, such as what Consumer Reports and similar publish, and that Wikipedia / Wikidata are going to have this information. If an argument needs to be made that reliable sources publish prices and expect consumers to have access to them, then it is possible to make a strong case for that. We are talking about drug prices and the pharma industry here, so it is plausible to imagine that there could be a multi-million dollar investment in Wikipedia just to sort prices any way that we imagined was appropriate if we had the best of everything available to us. I am not suggesting a price roll out everywhere now, but I want experiments and policy development now, and I want to talk through what we would do and how we would do it if we could be the best at this. 10 years ago Wikipedia had to generally dismiss database content, but with the advent of AI publishing / web 3.0 / Semantic Web, publishing is getting weird fast, and algorithmic interpretation is becoming routine in reliable publishing. I see no harm in small controlled live experiments and for anyone who wishes to speculate to develop policy which does not fit now, but which could in the future. Blue Rasberry (talk) 13:44, 23 November 2019 (UTC)
Here's my attempt at merging what we have currently and Colin's two paragraphs above: Unless reliable sources discuss the significance of the price of a particular drug or procedure, it is not appropriate to include that price information in an article. In articles where price information is relevant, the article must explain why it is being mentioned (e.g. claims of price gouging). Some medicines contribute significantly to a pharmaceutical company's turnover (blockbuster drugs), or their high price is a factor in their rejection or difficult acceptance by health services or insurance companies. Prices of medications and procedures can vary significantly between countries, for different brand names, and other factors. So include the specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the year. The lead, as a summary of article content, should not include specific or detailed price information. Little pob (talk) 13:46, 20 November 2019 (UTC)
WhatamIdoing, Bluerasberry, SandyGeorgia, what do you think of Little pob's proposal? I think it is a good mix and also adds 'price gouging' to 'blockbuster' and the health-service/insurance rejection issues. So we have some pointers as to what kind of commentary editors should look for when thinking about price. If we have a consensus then we can move forward to updating the guideline. Currently QuackGuru is refusing to engage in discussion and citing 'competence is required' in his edits, which is a clear insult towards us. Let's demonstrate how collaborative editors work. --Colin°Talk 10:24, 21 November 2019 (UTC)
On iphone at airport, briefly agree with Little pob but first sentence merely repeats policy ... must we restate policy to get the point across? I guess so ... but it seems strange to me that we even need to say any of this ... it is policy. But I defer so we can move forward. SandyGeorgia (Talk) 12:09, 21 November 2019 (UTC)
PS, prices in infobox, never, adamant oppose. SandyGeorgia (Talk) 12:11, 21 November 2019 (UTC)
I like much of what Little pob wrote.
User:SandyGeorgia, it is sometimes useful to repeat policy.
The main point that concerns me this morning is that I'm not entirely certain that "Unless reliable sources discuss the significance of the price of a particular drug or procedure" is something that holds as strong a consensus as it did during the last major discussion, which was nearly a decade ago. Since that time, there has been so much discussion about the cost of drugs in general, or the cost of drugs in terms of categories (e.g., the new anti-hypertensives are so x% more expensive than the old anti-hypertensives) that I think editors might think that drug costs are more worth mentioning.
They might also hold a different opinion about the costs of specific procedures. These can be quite interesting and discussed in detail by reliable sources (e.g., ultrasounds are much cheaper in Japan than in the US; MRI prices can vary by an order of magnitude within the same metropolitan area in the US), but there is much less emphasis on these. (Also, prices aren't Wikipedia:Biomedical information, so a business magazine is a reliable source for most price information.)
The sentence is also incomplete, because it doesn't mention disease costs. The economic costs of diseases is a routine calculation in the medical literature (and reflected in headlines about "Depression costs the national economy billions of dollars per year"), but that's an easy fix.
I think we might need to go back to editors and ask them where they fall among a range of options. I don't think we're going to get a response of "never, even if extraordinarily well-sourced" or "always, assuming you have any half-decent source at all", but I'm uncertain whether their answer will be closer to "usually yes" or "usually no" these days. To put it another way, encyclopedia articles need to cover the topic comprehensively (the best we can, which is constrained by the existence of sources), and I don't know whether editors currently believe that having some information about the price would be necessary for a comprehensive understanding of the topic. WhatamIdoing (talk) 18:24, 21 November 2019 (UTC)
Yes, I agree with the little pob text here. There are more issues to sort - this little pob text is for prose in Wikipedia, and the biggest part of this price discussion is about the circumstances under which we put prices in the infobox. That infobox discussion is another issue, and while we use default policy for infoboxes usually now, I am expecting another set of policy for infoboxes in the future and in development now. Blue Rasberry (talk) 13:27, 23 November 2019 (UTC)
Thanks for comments. I've taken on board what has been said, and hopefully addressed the majority of the points:
second pass
When adding price information to medical articles, it is important to establish the reason why the price is being stated. Some examples of when it might be appropriate to mention cost information in an article include:

Prices of medications and procedures can vary significantly; so include the specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the year:

checkY Do this: In the US, between 1987 and 2017, the wholesale price of long-acting insulin increased from US$170 to $1,400 per vial.[1]

☒N Don't do this: Insulin costs US$1,400.[1]

The lead, as a summary of article content, should not include specific or detailed price information.
I switched the examples to a list layout for easier reading and stylistic consistency with the rest of the MOS. Copied in the do and don't from current, as I think they're useful. I'm not 100% sold on that last sentence, but I'm not sure why. Might just be where I've placed it, or it could be that it's written as an absolute yet there may well be exceptions needed. Little pob (talk) 12:59, 23 November 2019 (UTC)
See "When adding price information to medical articles, it is important to establish the reason why the price is being stated." I don't get this. This suggests if I can't establish the reason why the price is being stated then it should not be included.
See "The lead, as a summary of article content, should not include specific or detailed price information." Prices are already in the lede in numerous med articles. The Current wording does not have these obstacles in place. Please stop trying to add content to this page that would result in deleting pricing from the lede.
Including the examples is fine. QuackGuru (talk) 13:45, 23 November 2019 (UTC)
@QuackGuru: This suggests if I can't establish the reason why the price is being stated then it should not be included. Precisely! If any editor can't justify the reason for adding content – i.e. give the reliable sources without interpretation – they shouldn't be adding that content. It should go without saying, but for clarity; I'm in no way trying to say that same content can't be added by an editor who does have the sources.
Prices are already in the lede in numerous med articles. That doesn't mean they should be in there. Several policies and guidelines have been listed, by much more competent editors than I, as to why there should not be detailed price information in the lead. It's also important to understand that the inclusion of "specific or detailed" is significant, used deliberately, and shouldn't be ignored. <aside>I think it's the fact that the sentence changes meaning if you do, accidentally, skim over the "specific or detailed" bit that has me unhappy with it. Thank you, QuackGuru, for helping clarify that thought process.</aside>
Please stop trying to add content to this page that would result in deleting pricing from the lede. As this is the talk page, no. Discussion around having prices in the lead is ongoing. Some editors are not seeing an issue. Some are strongly against prices in the lead (and apparently infoboxes). Some, myself included, think there is a middle ground to be found. I don't know what that middle ground is, wouldn't dare to guess what numbers each viewpoint has, and nor would I ask those whose opinion differs from mine to hold their tongue. That's not how we find consensus. Little pob (talk) 15:13, 23 November 2019 (UTC)
Proposing content that puts stiff barriers in place is not a middle ground. QuackGuru (talk) 13:31, 27 November 2019 (UTC)
Couple of points, as I know we've butted-heads previously elsewhere:
Re "stiff barriers" specifically – This might come back to bite me on the arse; but WP:IAR would still apply regardless of the wording we all settle on.
Re "the whole": Middle ground is found through discourse. We have to allow all concerned editors to set out their stall, and be willing to engage with them, to find that fabled land.
Little pob (talk) 15:13, 27 November 2019 (UTC)
  • Add prices per Blue. Prices in the infobox would be excellent, especially if we can automatically upload them based on high quality sources. User:Seppi333 was working on something like this. Doc James (talk · contribs · email) 18:27, 21 November 2019 (UTC)
  • add prices per Blue Rasberry--Ozzie10aaaa (talk) 19:08, 21 November 2019 (UTC)
  • Add prices is the way forward. Adding specific wording to this guideline to delete prices from all med articles is not going to work. QuackGuru (talk) 20:12, 21 November 2019 (UTC)
    • Could we please stop voting? Really, this situation is lot more complicated than just "add prices". You can "just add prices" to articles about some subjects, but can any of you premature voters tell me what the price is for a pregnancy?
      The conversation that we need to have here is (a) a conversation, not a vote, and (b) about all the complicated and diverse details, not just some one-size-fits-none "add prices and quit bothering me about the details" opinion. I don't necessarily mind if you want to add prices, but you need to tell me which prices you want to add, and which articles you want to add them to. WhatamIdoing (talk) 21:43, 21 November 2019 (UTC)
      • You are asking what is the typical "cost" of an inhospital delivery in various countries? Average in the US in 2013 was $32,000 but it varies fairly widely.[1] Doc James (talk · contribs · email) 22:34, 21 November 2019 (UTC)
        • No, that's the price of hospital-based childbirth and postpartum care in the US. If it's so easy to "just add prices" that we don't need to have an actual conversation about which prices to add, which articles benefit from prices, and what kinds of situations we should be including and excluding, then you should be able to tell me what the price of the pregnancy itself is. There's millions of pregnancies each year. You want to "add prices". Okay, what price would you add to Pregnancy to explain the costs of getting pregnant? WhatamIdoing (talk) 22:50, 21 November 2019 (UTC)
          • One is directly by the avaliable sources. There is the field of health economics that revolves around this. Would need to look at the sources. We have an entire article called Price of oil. Doc James (talk · contribs · email) 00:16, 22 November 2019 (UTC)
            • Okay, so that's something we should be talking about: "Add prices", but only if and when we've got good sources. But it's more complicated than that, isn't it? The price of pregnancy can be free. It can be hundreds or thousands or tens or even hundreds of thousands of dollars, if the couple is infertile. The price can also be measured in terms of morbidity and mortality for the pregnant woman: the price is weeks of vomiting, or months of depression, or years of worsened body dysmorphia, not to mention the number of women who die by suicide or murder as a result of an unintended pregnancy. And then there are straight economic costs, of lost wages and lost jobs and increased expenses for everything from bigger clothes to more food to prenatal healthcare costs (if any is obtained). So you want to "add prices". Pregnancy is well-studied; we can get sources for any of these. Which of these prices do you actually want to add? WhatamIdoing (talk) 00:36, 22 November 2019 (UTC)
              • Most stuff is much more well defined like the average price of a C-section versus in hospital vaginal birth or the price of a hip replacement or the price of a single medication. Doc James (talk · contribs · email) 01:59, 22 November 2019 (UTC)
                • Does that mean that you want to "add prices" only to these more discrete items, and not to "just add prices" to every article? Do you want to add only certain kinds of prices, like the price of the surgical procedure or the hip prosthesis, or do you want to include the price of the pain, medical equipment, caregiving, physical therapy, lost wages, etc. that go along with that hip replacement surgery? WhatamIdoing (talk) 02:28, 22 November 2019 (UTC)
@WhatamIdoing: Some years ago I looked for and could not find "price of pain" or total cost of medical care info.
I still think we need a general article on non-financial costs. I am not aware of any source which presents this concept. Blue Rasberry (talk) 13:19, 23 November 2019 (UTC)
Perhaps you had the wrong key words for your search? If you are trying to put a dollar value on the experience, I undestand that you measure the pain in DALYs and multiply that by whatever number your economist tells you is the relevant price for a year of (healthy) potential life lost. You can also state the cost directly: "Most people experience moderate to severe levels of pain for n weeks after the surgery". This may be preferable, just like "most people can't work for n weeks" is usually more informative to individuals (especially individuals outside the US) than "People lose US $X in wages". WhatamIdoing (talk) 02:59, 24 November 2019 (UTC)
We include a lot of DALY data. Agree for surgery the typical length of pain is more useful when looking at that measure. The average cost or range of costs for a procedure is also useful. Doc James (talk · contribs · email) 14:47, 26 November 2019 (UTC)
Informative to some (e.g., Americans), but less to others (e.g., anyone in the developed world). "Two-hour surgery that can usually be handled at an outpatient surgical center" translates across time and place better than "Cost US$12,000 in the US in 2015". The same isn't really true for mass-produced products (drugs, supplies, devices, etc.), although if we wanted to avoid numbers altogether, it'd be possible to place most products in some sort of price range ("higher price than most similar devices"). WhatamIdoing (talk) 15:54, 26 November 2019 (UTC)

What's wrong

No editor has explained what is the problem with the current wording and if an editor wants to add content that appears to violate WP:CREEP then that editor should explain why they want to violate CREEP. Telling editors what should or should not be included in every med article is not productive. For example, Onasemnogene abeparvovec says "It carries a list price of US$2.125 million per treatment, making it the most expensive medication in the world as of 2019.[6] In its first four months of sales US$160 million of medical was sold.[7]" We should not add content that would delete this from the article for no good reason. If anyone wants to add content to this guideline that would delete the US$2.125 million price tag from the article then that would be a violation of CREEP. QuackGuru (talk) 14:02, 21 November 2019 (UTC)

Quack, you are not actually entitled to an explanation that you can understand or that will WP:SATISFY you. At some level, I want you to remember that Wikipedia:Policy writing is hard and that it's a skillset that most editors, including you, don't have to any significant degree. But since you have asked, an incomplete list of the problems includes:
  • Redundancy: Some versions say the same thing twice. For example (NB: I'm only giving one example of this problem. This is not an exhaustive list of all instances of redundancy), the current version tells editors to mention the "scope (e.g., US-specific, worldwide average)" and then asks them to consider "Is it in a particular country, or a worldwide average?" Saying this twice is pointless.
  • Incompleteness: Some versions do not include all of the information that editors think would be helpful. For example (NB: again, just one example), some versions do not mention whether (or under which circumstances) a price should be mentioned in the lead.
  • Divergence from consensus: Some versions do not align with the current consensus in all points. In some cases, I am even uncertain what editors would recommend. For example (NB: just one example), perhaps editors have changed their minds, and now they think that the price of drugs is worth mentioning in a greater proportion of articles than they thought several years ago.
The bottom line is that even though these problems are (apparently) not obvious to you, they do exist, and they are worth fixing. WhatamIdoing (talk) 18:08, 21 November 2019 (UTC)

Current wording

Economics may include prices of medications or the cost of procedures, if covered by reliable sources. Cost information may be included in the lead, but this is done on a case-by-case basis. The cost of medicine or procedures can include specific details regarding the kind of price (e.g., wholesale, retail, discounted), scope (e.g., US-specific, worldwide average) and the date or year. Certain medicines contribute significantly to a pharmaceutical company's turnover (blockbuster drugs), or their high price is a factor in their rejection or difficult acceptance by state health services or insurance companies. Prices may vary significantly between countries, for different brand names, and other factors, so include specific information. For example, is your source reporting the list price or the actual selling price? Is it in a particular country, or a worldwide average? Is it the total cost, including indirect costs such as lost wages for the patient, or the initial price?

checkY Do this: In the US, between 1987 and 2017, the wholesale price of long-acting insulin increased from US$170 to $1,400 per vial.[1]

☒N Don't do this: Insulin costs US$1,400.[1]

  • Comment: please guys let's do a proper RfC, a local consensus cannot overturn a previous RfC, so this vote is bound to be pointless anyway. Please let's revert back to the RfC state and follow a proper BRD procedure, the MED community can do better than that. --Signimu (talk) 15:32, 21 November 2019 (UTC)
    • WP:RFCs do not result in binding decisions. An RFC is an advertising process for an otherwise normal talk-page discussion. Wikipedia:Consensus can change (that's the actual policy; neither WP:RFC nor WP:BRD are policies), and if the consensus disagrees with the outcome of a previous RFC-advertised discussion, then any convincing discussion is adequate to "overturn" it – just like any other discussion. WhatamIdoing (talk) 17:18, 21 November 2019 (UTC)
      Ah just saw your reply, we discussed this on my talk page, but just to write it down here too, yes consensus can change, but it should be at least on a similar level of "strongness" to avoid relying on a WP:SHAM consensus. It's just that I'm sure we can reach a better consensus by discussing more, and more broadly Signimu (talk) 19:34, 27 November 2019 (UTC)
  • This isn't the best we can do. Let's keep talking above about how to improve it, rather than just voting. WhatamIdoing (talk) 17:52, 21 November 2019 (UTC)

It is complicated and you are forgetting WP:WEIGHT

I'm really disappointed at the level of argument from QuackGuru and DocJames. QuackGuru insists the "current wording" be left because "If it ain't broken then don't fix it". We didn't have anything on Price a month ago, but it is worth reminding ourselves of what QuackGuru added:

"Economics may include prices of medications or cost of procedures if covered by reliable sources. The pharmaceutical industry has tried to conceal medication prices because of their continuing legal cases in the United States. There is a lack of transparency regarding medication and vaccine prices among non-governmental organizations."

The first sentence is in direct contradiction to and misunderstanding of policy. Our policy doesn't just say (as James further above claims) that if a "reliable source" mentions something, then we can also mention it, in whatever detail we choose and however prominently we choose. Whether or not to include things, and whether they are significant enough to appear in the lead, and how much detail to afford to them, is determined by WP:WEIGHT: "in proportion to the prominence of each viewpoint in the published, reliable sources". Our policy goes on to say "Undue weight can be given in several ways, including but not limited to depth of detail, quantity of text, prominence of placement, juxtaposition of statements and imagery". Weight isn't just about viewpoints, but about any aspect of a subject and how it is presented. This is explained in policy: "An article should not give undue weight to minor aspects of its subject, but should strive to treat each aspect with a weight proportional to its treatment in the body of reliable, published material on the subject". The second two sentences added by QuackGuru are his own agenda and not appropriate text for any guideline on Wikipedia. The editorialising was quickly removed by WhatamIdoing but the initial faulty sentence was retained. QuackGuru then added another statement:

"Cost information may be included in the lead, but this is done on a case-by-case basis"

The reason for this would appear to be to defend Doc James addition of detailed price information in the lead of mainly the WHO essential drugs. It isn't supported by any policy, guideline or consensus of editors on Wikipedia. Indeed, the RFC of two years ago (not 10 as claimed above) rejected the general inclusion of price information in articles. James has found a source that lists a handful of prices from a limited small set of suppliers of a given drug back in 2014 and has extrapolated that to be "The wholesale cost in the developing world" (e.g. Mebendazole). This is a statement that is simply untrue. Nobody has published "The wholesale cost in the developing world". There is a database that contains specific wholesale prices from a very small collection of suppliers who mostly target the developing world. They have got the price information from whoever makes that price information available to them: they don't claim this is actually representative of world-wide pricing or developing-world pricing. There could be huge gaps in that data due to price being withheld or simply hard to retrieve. Perhaps it is likely to be in the approximate ballpark, but no more than that.

Because the source does not perform any statistical analysis on the price, neither can James. But he chooses the smallest and largest price in the database of six suppliers. Picking the smallest and largest value in a dataset is not a typical method used to statistically analyse data and present it to people, because it is prone to giving undue weight to outliers. Performing analysis on a tiny dataset is prone to drawing incorrect conclusions because information is missing. Some of the suppliers in this small dataset could be orders of magnitude larger (and thus more significant) than others, yet we treat them all equally by ignoring all except the cheapest and dearest. That is simply statistical and economic incompetence. The data is certainly now out-of-date as the website has not been maintained for five years. We are told the price in the developing world is between XX and YY. That means we claim to have a reliable source that says the cheapest price in the whole developing world is XX and the most expensive price is YY. But in fact we don't. We have a source that merely gives a price from six suppliers who do not cover the entire developing world. And a price that is 5 years old. And let's not even start on a definition of "Developing country" which may or may not include South Africa, India, China and Russia depending on which measure you use.

James directly quotes these arbitrary prices in dollars and cents. We have in our lead the claim "between USD 0.004 and 0.04 per dose". Let's return to WP:WEIGHT. Do our reliable sources give weight to the lowest and highest prices in the IMP Price Guide of 2014 for this drug? Do our reliable sources quote those prices in tenths of a cent precision? Do they think that the price, to tenths of a cent, charged by German charity Action Medeor International Healthcare to one African country, Tanzania, is important enough to mention in the lead? I hope you agree that is ridiculous. The only reason our lead, as written by one Wikipedian with an agenda, mentions the price in Tanzania by a German charity, is because that's all that James found convenient when web surfing. Not a single one of our reliable sources on Mebendazole mention the price, in tenths of a cent in Tanzania in 2014, in their leads. So neither should we.

Then there is the question of what we even mean by "dose". The article doesn't explain or link. The source gives the price per tablet and also gives a defined daily dose. For example, Mebendazole, the lowest price is $0.0035 per tablet of 100mg and a "defined daily dose" of 200mg. Our article just says "$0.004... per dose". What our article neglects to mention, and most of our reliable sources do mention, is that mebendazole might only need a single dose of 100mg to treat threadworms, though a second dose after two weeks is used if reinfection occurs (which is common). For some other infections, the treatment is 100mg twice a day for three days. Add to that, the advice is often to treat the whole family. Now, we generally don't go into specifics about mg dosage, but it is encyclopaedic that this is a very short and quick treatment (compare e.g. fungal nail infection which might take 6-12 months of treatment daily). Which figure is useful to our readers? The single dose taken three times a day, the 3x combined daily dose, the "defined daily dose" standard "used to standardize the comparison of drug usage between different drugs or between different health care environments" (which is not the therapeutic dose), or the total amount needed to fully treat the condition in the patient, or the amount for the whole family? So, just like wholesale vs retail pricing, we aren't really defining what we mean by dose. Should the lowest price in the lead be $0.0035 tablet in Tanzania, 0.007 per defined daily dose, or give different values for threadworms vs other infections, or total treatment cost?

The article also claims the price in the UK is around $5, citing an FT article I can't read. The BNF says 100mg tablets from Janssen-Cilag Ltd costs £1.34 for a pack of 6 tablets, which is wildly different from $5. The UK retail price is £8 or £9 for a pack of four from two pharmacies I found, which would likely require two packs (£16-18) to treat the family of four with two doses two weeks apart. Why are we mentioning $5 in the UK, when that bears no resemblance to any price whatsoever? It is just wrong.

There is notable commentary in reliable sources about the US price hike. Also this article from 2011 mentions dodgy accounting by reputable charities wrt the price of this medicine.

In summary, we are making patently false and unsupported claims in our leads: nearly every fact presented is just plain false and our wording ambiguous. We are giving way too much weight and precision to prices from an arbitrary and tiny collection of sources, and our major source for such prices has not updated their data for 5 years. And to add to that, our Mebendazole article doesn't mention far more relevant encyclopaedic information such as that just two doses of the tablet, taken two weeks apart, will clear a threadworm infection, or that family members are encouraged to be treated even if they don't show symptoms.

The claim that such information is supported by reliable sources takes a deliberately naive view of how sourcing should influence our article content. I'm sure everyone here knows that people turn up on Wikipedia with claims made by some mouse experiment in such and such a journal. They are perplexed that although we agree that journal is a reliable source for what the experiment did, it isn't a reliable source from which we can make health claims in humans. Similarly, a 5-year-old neglected database of prices from six random organisations is not a reliable source from which we can present the very complex business of worldwide drug pricing to our readers.

Our lead should contain a summary of only the most vital and robust information about the article topic. It is, after all, what is read out to us by our electronic devices and what is translated to other Wikis. Instead we simply have nonsense written by incompetent amateurs who are misinterpreting their inadequate and out-of-date sources. It couldn't really, be much worse. -- Colin°Talk 21:32, 23 November 2019 (UTC)

Well, there is some precedent for this. It depends on whether you think that the price of a product or service is more like a birth date or more like a criminal conviction. We routinely include bare facts in the name of comprehensiveness even when almost no source mentions them. In biographies, that means that we include people's middle names, birth dates, parents' names, cities of birth, high schools, etc., even when no source actually cares about those facts. In articles about chemicals, we include bare facts such as molecular weight, even though we know that nobody's actually going write an article about how it's so interesting that the molar mass of chlorine trifluoride is 92 instead of some other number. We do this in medical articles, too: if there's no known treatment, then we would usually have a section called ==Treatment== that makes a brief statement that there's no known treatment. We just feel that these articles would be incomplete without this kind of information, and the difference in terms of WEIGHT is whether you mention these things in passing, or if you put a lot of emphasis on them – not whether it's mentioned at all. WhatamIdoing (talk) 02:39, 24 November 2019 (UTC)
Much more like a criminal conviction. A birth date or molar mass is a pure, bare statement of fact. The molar mass of chlorine triflouride isn't one thing in San Diego and something very different in Rio de Janiero; it's just 92, everywhere, every time. If you were born on 1 January 1970, that's your birthday, period, everywhere. The price of a given medicine very well might be different everywhere, and even different depending on how it's being procured or purchased. That's not something we can just slap a number on; it would require contextualization and discussion by a reliable source, not just grabbing a number out of a database. There are cases where reliable sources have extensively discussed the prices of certain medications or treatments, and I'm all for including such well-sourced information in the article, but it shouldn't be a routine practice to add to all or most articles. Seraphimblade Talk to me 04:19, 24 November 2019 (UTC)
Agree with Seraphimblade. Someone's middle name, DoB, etc are the sort of information one expects in an encyclopaedia and are either correct or not. The price of a drug is just about the most nebulous thing we can consider: where, when, how much, how delivered all cause huge fluctuations. Consider the price of a mango:
  • Standard whole fresh fruit. 69p
  • Giant whole fresh fruit. £1.80
  • Twin pack wrapped in plastic and ripened ready to eat. £1.89
  • Dried mango slices, 200g. £4.00
  • Mango slices in syrup in tin can, 425g. £1 or 3 for £2.50
  • Mango pulp in tin can, 850g. £1.75
And that's just from one major supermarket in the UK in November. Now imagine that I wrote "The price of mango in the developed world is between £0.69 and £4.00" and inserted that into the lead, and edit warred to retain it there and then went and edit warred at WP:MOSFRUIT to ensure the price of mangos was required. The database James found is just a collection of data values from limited sources, just as MySupermarket.co.uk is a collection of data values from limited sources (it doesn't include the price at markets and small shops, just large chain supermarkets). James has engaged in WP:OR to make claims that this dataset represents the lowest and highest prices in the developing world and presented it in a way we don't really know what the price is for.
The price of mebendazole is not notable except in the case of the US price hike. Without that hike, the price follows the pattern of any cheap-to-produce out-of-patent pill around the world. Further, it is remarkable for being effective with just a very short treatment regime -- something our article does not mention. Time and again I see WP:MED arguing to include factoids or awful videos but not even getting the basics of a text encyclopaedia right. These embarrassingly bad price statements are present on dozens, hundreds, of our key drug articles, and they are all, every one of them, as wrong in so many ways as a Trump tweet. -- Colin°Talk 12:34, 24 November 2019 (UTC)
And yet if I added the annual average FOB (~wholesale) prices for imported mangoes according to the USDA to Mango, I doubt that anyone would tell me that was not encyclopedic information and that readers would be mad to discover that they weren't paying US$4.60 to 5.74 per box in the US during 2018 and therefore all information about prices urgently needed to be removed to stop people from thinking that the wholesale price was their local retail price, or that it was terribly confusing to let the experts at the USDA weight conventional and organic, and different sizes of mangoes, and different ports of entry, and different seasons, and the different countries of origin into a single price range all by themselves. I might get complaints that it's only the US (which is fair; 8 of the top 10 mango producers in the world don't export mangoes to the US in any considerable quantity, but that suggests expanding it with a more prices for different countries, not removing the prices that we do have), and I might have to point out that the US grows very little mango domestically, so the import price basically is "the" price, but I don't think that anyone would say that the overall economics of a popular fruit crop are unimportant unless I'm citing a source that talks about those prices in great detail. WhatamIdoing (talk) 16:57, 24 November 2019 (UTC)
  • I just set up Wikipedia:Defining data. I intend this to be the start of some guidance about what sort of information should go into articles, seemingly in circumvention of the guidance at WP:NOT, WP:RS, WP:WEIGHT, WP:V, WP:LEAD, and the rest. WhatamIdoing, I think this is what you were mentioning above. Colin, I think you are saying that price is not "defining data", and I am saying that some information about price is defining. Again, I agree - we have great difficulty actually getting this data, but if we had it in sufficiently high quality, then I would like to present it as fundamental. I see price to be as fundamental to drugs as concepts like nationality and century of birth are to a biography. Obviously we need more discussion. Blue Rasberry (talk) 15:20, 24 November 2019 (UTC)
    • Bluerasberry the concept of "Defining data" is interesting. I don't think that wiki projects have always been sensible about what to include in info boxes. For example, WP:MED for years put lots of external database links in info boxes, and I see that is now demoted to a collection at the bottom of the article. Even now it tries to include some rather complex concepts in the infobox. Perhaps a good characteristic of the data we should include in info boxes is whether that data describes something simple and straightforward rather than nuanced and complex. Does it need a lot of explanation? Even so, it can be simple for some subjects and hard for others. The infobox for Measles is fairly straightforward. The Tuberculosis one breaks down a bit with its Frequency datapoint being hard to explain and the value given is probably not what most people would consider (numbers with illness). Infoboxes become downright silly with Cancer and Epilepsy which are really groups of diseases that may have very little in common.
But surely "defining data" is information about a subject that if it were significantly different, it would be a different subject. Someone's DoB, parents, partners, children, occupations and year of death all go to define a person. The defining attributes of mebendazole include its chemical and therapeutic properties. It may also include the company that invented it and may still hold a patent on it. Whether it is licensed in Europe or US. In the UK we might be concerned if it is available only by prescription (PoM) or only from a licensed pharmacy (PM) or widely (GSL). In the US you care much more about price than we do in the UK. Depending on your circumstances and age, prescribed medicines are either free of charge or require a fixed £9 charge for about 1 month's supply. Whether a bottle of pills has a wholesale cost of £1 or £2 or £10 is really only the concern of my pharmacist.
The recent astronomical drug price hikes in the US (and occasionally in the UK) demonstrates in fact that price is very much not a defining characteristic of a medicine. In fact, it is more a representation of the economic systems of countries, and the price regimes their governments do or do not enforce. Has drug company mergers led to decreased competition, have monopoly regulators been asleep? These are not, in fact, questions about a drug at all. The drug price is just a pawn in a bigger game. -- Colin°Talk 17:10, 24 November 2019 (UTC)
Don't you think that whether the wholesale cost is £2 or £200 is also a matter of interest for taxpayers and policy makers? Perhaps just as it doesn't actually matter whether some politician's birthday is January 1st or January 2nd (although, for better or worse, we do report that level of detail), the exact details aren't that important, but the overall order of magnitude could be important to the healthcare system and its users (not to mention students writing about the drug for school).
Which brings me to the other point: Prices matter to the business end of healthcare. Any individual consumer or taxpayer might not care about the general price range for a given drug, especially since so many of them are generic drugs, but investors should and do care. Are we actually writing comprehensive articles if we're omitting key information about that POV? I notice that s:1911 Encyclopædia Britannica/Tea mentions the early price of tea in England and the then-present estimated capital value of those big businesses, so I think it's fair to assume that some business matters are a proper subject for encyclopedias. WhatamIdoing (talk) 17:24, 24 November 2019 (UTC)
And that's great, and there are places those can be found in many cases (and if it can't be found, well, then it's unverifiable so we couldn't include anything about it either). It's useful is explicitly not an argument for inclusion in an article. Bus, train, and flight timetables and schedules are indisputably useful and of interest to many people, but we still do not include them, both because they're a greater level of detail than is generally encyclopedic, and they often change and would quickly become out of date. That's the same reason we do not generally include prices. Now, with something like the treatment mentioned above, which reliable sources have extensively noted is the most expensive in existence, sure, then we should mention that, in the same context the sources do—its extreme price is part of what's notable about it. But we shouldn't do that for run-of-the-mill drugs or treatments where their prices are not of any particular note. So, while there are exceptions, the general rule of other articles about products should also apply here: No pricing. Seraphimblade Talk to me 18:40, 24 November 2019 (UTC)
WAID, I just noticed your comment above. My mango comment wasn't made to get side-tracked in to how to price mangoes, but to demonstrate the silliness that I'd found a website quoting a handful of mango prices and had done OR on that to make claims that really aren't true. WAID, we don't have a source for the cheapest price of mebendazole in developing nations. We don't have a source for the dearest price of mebendazole in developing nations. Our source doesn't even claim to represent "developing nations", whatever definition you might use. Nor do we state what the price is actually for. Surely you agree that "what is this price for" is crucial? All our important drug articles make price claims that are simply not true and not supported by the source given. Are you not concerned about that? Consider Ibuprofen:
We can normally take two 200mg ibuprofen tablets up to 4 times a day. The defined daily dose of ibuprofen is 1.2g. So by "dose" do we mean a 200mg tablet, 400mg individual dose, 1200 defined daily dose, or 1600mg maximum normal daily dose?
  • Ibuprofen says "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose"
  • The source has "Lowest Price: 0.0046/tab-cap", "Highest Price: 0.0080/tab-cap" (and explains that lowest/highest is just among their 9 suppliers, not "the developing world"). The tablet "Strength: 200 mg". The "Defined Daily Dose: 1.2 G".
  • If our dose is a 200mg tablet then "$0.005 to $0.008"
  • If our dose is two tablets, 400mg, then "$0.009 to $0.016" (or $0.01 to $0.02 if we round).
  • If our dose is defined daily dose, 1200mg, then "$0.028 to $0.048" (or $0.03 to $0.05 if we round)
  • If our dose is maximum daily dose, 1600mg, then "$0.037 to $0.064" to $ (or $0.04 to $0.06 if we round).
None of these ranges bear any resemblance to what our article claims. The article also claims "In the United States, it costs about US$0.05 per dose." without saying if that is a wholesale price. The source does not, AFAICS, include any price information whatsoever. Consider the epilepsy drug valproate.
  • Valproate says "The wholesale cost in the developing world is between US$0.14 and US$0.52 per day" added by Doc James in 2015.
  • The source has one supplier price (IDA) at $0.1486 per 500mg tablet. If we might multiply by 3 to get the 1.5g defined daily dose, it comes to $0.45 per day. The other two prices it lists are buyer prices, for Peru and Sudan. These work out at $0.23 per day (Peru) and $0.79 per day (Sudan). There isn't AFAICS any price or price calculation that gives "between US$0.14 and US$0.52 per day". The price quoted in our article is similar but not exactly the same as two 500mg tablets, but the defined daily dose would be three 500mg tablets.
Honestly, the more examples you look at, the more one wonders if this is some experiment in trolling Wikipedia to include random false facts and see how long the community takes to notice. You ask if our readers might get confused. No, I think here we actually have an editor who is confused, who really hasn't a clue about what they are writing, and has spread their confusion over all our major drug articles. Are any of our drug prices correct? -- Colin°Talk 18:43, 24 November 2019 (UTC)
Like WAID said, "Don't you think that whether the wholesale cost is £2 or £200 is also a matter of interest for taxpayers and policy makers?" I agree that reporting bus fare and the difference between $0.03 to $0.05 is silly, so let's leave that out because that was never the point. Now you guys say something about uncertainty about whether something is approximately $2 or $200, which is the difference between accessible and inaccessible. Why not communicate this? We seem likely to be able to do so soon with good data and technical capacity, and if people want to experiment with the process now, then why not? Blue Rasberry (talk) 22:21, 24 November 2019 (UTC)
That's...not how any of this works. Wikipedia follows sources. They, not us, should be doing the experimenting. If they decide to go into a lot of detail on the price of certain medication, we follow their lead and do the same. If they don't, we don't either. If we will one day have sufficient data, we will use that data after it exists, not presume one day it will. Wikipedia follows, never leads. Seraphimblade Talk to me 00:45, 25 November 2019 (UTC)
I think we have two different conversations going on. WhatamIdoing and Bluerasberry are still trying to make the case for including a cost figure and not really engaging on the problem that (a) we don't have sources for what they want and (b) all our major drug articles have incorrect drug cost figures. WhatamIdoing, there are several parties who care about price. You say "prices matter to the business end of healthcare". I return to the question "the price for what?" Clearly our current description of pricing is not actually stating "what" because "dose" is not defined and can be defined in many ways (though defined daily dose would appear to be the relevant one for financial comparison). But for a drug company, that only has meaning when combined with how many doses a person takes (or keeps taking for life), how many people get ill and need the drug, whether other drugs compete or will compete, what competition there is in the market for this drug (either on patent or off patent but nobody else making it), what agreements they already have with healthcare providers that control prices, etc, and whether you want to spend eternity in hell for charging $400 rather than $0.04 for a pill to get rid of threadworms. Our cost information will not help anyone make that decision, and frankly, nor can they be trusted. I think if you want to make the case that cost information should be presented in our drug articles, and describe how it should be presented, then you need to start with an example of best practice. MEDMOS shouldn't be based only on speculative information we might not be able to reliably get, nor based on the very bad practice described above, but on best practice. -- Colin°Talk 15:08, 25 November 2019 (UTC)
Seraphimblade, the idea of "experimenting" is to experiment with different approaches for reporting what the sources say. For example, would editors be happier with a properly sourced sentence that says "The wholesale cost of generic ibuprofen tablets is usually about a penny each", and leave it to the reader to guess that we're talking about the standard 200mg pill, or would we be happier with a properly sourced bot-controlled item in an infobox that gives the price for an extremely specific definition, such as the "US Average wholesale price (pharmaceuticals) for the Defined daily dose on the specified date"? US AWPs have been reported for four decades for drugs that require long-term treatment, and DDDs are standardized worldwide (that's the point of the DDD, although for ibuprofen, it'd require more than one entry, because there's more than one DDD), so we can source those prices. The "experimental" part here would be deciding whether we'd prefer something approximate in prose vs something precise (we can get the AWP for ibuprofen down to the thousandths of a cent on any specific day, which might be overkill) vs some other approach (e.g., list price at time of introduction, in the ==History== section for a drug). WhatamIdoing (talk) 16:47, 25 November 2019 (UTC)
WhatamIdoing, that's a false dichotomy. The general rule for products is that we do not include pricing at all, and I see no reason to deviate from that here. Exceptions are made in which the price, in and of itself, is one of the notable features of the product, but if it's just something that's sold and no sources talk about the price being of particular note, it should be omitted entirely. So, in most cases, I would prefer to see neither. Seraphimblade Talk to me 19:38, 25 November 2019 (UTC)
That is simple not correct. We have whole pages on prices.Price of oil Doc James (talk · contribs · email) 21:28, 25 November 2019 (UTC)
Seraphimblade, I'm not sure that the former general rule has as much widespread support as it used to, either in this subject area or in any other. Template:Infobox camera has a parameter for the list price. Template:Infobox battery includes a measure of cost-effectiveness (the amount of energy compared to the price paid by end users). Template:Infobox aircraft type includes the price of an aircraft, and they've clearly put some effort into deciding exactly which price they want to include. Looking at some FAs, Sega Genesis lists the initial or list price for multiple products, plus the clearance price for one of them. 32X has the initial price in three currencies in the infobox. Science Fantasy (magazine) reports the cover price. NeXT names the price for one of their products in the lead. House (TV series)#Distribution tells you how much iTunes is currently charging to buy an episode. If editors genuinely didn't want prices in articles, then I don't think that we'd find it in infoboxes and Featured Articles. WhatamIdoing (talk) 22:00, 25 November 2019 (UTC)

Featured article samples

My apologies for just now getting to this, as I was traveling during Thanksgiving. IMO, these FA samples are not indicative that support for NOTPRICE has necessarily changed. And even FAs are subject to WP:OTHERSTUFFEXISTS-- even more so with the decline in participation at FAC and FAR since 2011. Even in the best of FAC times, it was near impossible to get top reviewers to engage with gaming articles, so one gaming article passing FA with prices in the infobox doesn't indicate much to me about Wikiwide trends.

Looking at each of these examples:

  1. Sega Genesis promoted Dec 2013 seems to discuss historical prices in keeping with the restrictions and exceptions of WP:NOTPRICE, with reasoning for the price information being significant put in context relative to history, competition, marketing, etc. and from independent sources. This is different than using database drug prices with no context for why we are reporting the price.
  2. 32X promoted March 2014 is a great example of WP:OTHERSTUFFEXISTS even in FAs. It is a very weak, WikiCup FAC, with clear fan support (which was a typical issue during WikiCup) and limited independent or expert review, such that it would not have been promoted FA in the days of more active participation at FAC without more strenuous review. (FAC delegates who followed my tenure had to deal with declining reviews, and have less help in confronting effects of fan supports from inexperienced reviewers such as occur during WikiCup.) This weak gaming article is not a strong indication of anything Wiki-wide.
  3. Science Fantasy (magazine) promoted May 2011 by me was written by a top-notch FA writer and reviewer, Mike Christie, who writes about (among other things) historical science fiction magazines, generally no longer in print. Pricing issues mentioned have to do with the historical context of business decisions made, and are in keeping with WP:NOTPRICE. Again, this appears to a policy-compliant use of pricing data, very different from what is being proposed with adding current prices to drug articles without context or significant sourcing.
  4. NeXT was promoted in 2008 (before the time when the bot indicated who promoted it, but it was most likely promoted by me), and the FA promoted version did not have that information in the lead. The article has not been reviewed at FAR (most FAs these days have not been reviewed at FAR, because FAR is dead, which is why most FAs today are not actually FAs at all). In other words, this is a very old, unreviewed FA ... and the mention in the lead, added since its FA promotion, appears to comply with WP:NOTPRICE because of the context given. Again, this is very different from what is being proposed with drug prices.
  5. That House (TV series)#Distribution "tells you how much iTunes is currently charging to buy an episode" (in 2008) shows just what an embarrassing debacle that editor's FACs were. This FA was written by an editor who was later revealed to be a prolific sockmaster, with supports for promotion of his articles coming from his socks. It was promoted by me in June 2009, that editor never wrote an FA I thought was up to standard, but my job was to determine consensus among reviewers, so he had numerous FAs passed based on his sock supports. I thought his work was dreadful, and this article passed FAC based on his own sock supports. That this editor was one of my main FAC frustrations during my tenure is obvious in that I let a line that talks about a "current" price get by me, and that inaccurate line (since it does not specify an "as of" date) is still in the article. This is another great, but embarrassing for me, example of WP:OTHERSTUFFEXISTS, and not an indication of anything Wiki-wide. Its editor was most unpleasant to deal with and his work eventually could attract very few reviewers, so he socked and did his own.
In summary, I am not seeing strong evidence of a trend here. SandyGeorgia (Talk) 06:34, 8 December 2019 (UTC)
Sandy, another good WP:OTHERSTUFFEXISTS example: In the edit war at ivermectin, a "fourth opinion" editor cited the fact that it appears most of our drug articles have prices as evidence that Wikipedia supports prices in drug articles. But since all (as far as I can see) prices in drug articles were added and occasionally maintained by a single editor and that editor was slow edit-warring to retain the price in that article (which suggests strongly that further examples could be found) one must be careful not to use bad practice as evidence that it is accepted. WhatamIdoing, I would be far more impressed if WP:MED could bring a dozen drug articles (not trival ones, but ones like have been discussed here and at WP:MED, which are typical rather than edge-cases) bring them up to first-class standard wrt source->text compliance with policy. If that was done then the matter of "do we include drug prices" would be a much simpler matter of consensus opinion. I really think "first show us how you'd do it" is the obvious step. And you are welcome to try. -- Colin°Talk 10:59, 8 December 2019 (UTC)
What's found in the articles about the Genesis (and some comparable articles, such as Nintendo Switch) is the original list price. I can see why that might be included, as generally the introductory price of a game console will indeed be discussed pretty heavily by sources immediately preceding and following the launch. I would be amazed to find there isn't a bunch of source material about it. On the other hand, something like the iPhone 11 lacks it (though ironically, it had a spam link for it I cleaned up), because it doesn't even really have a particular price. Sure, there's a list price if you just buy one, but very few people outright purchase at that price; most get them through some kind of deal with their carrier. Additionally, the Genesis article doesn't go into what getting hold of a Genesis would cost you today, so that is a static figure that won't become outdated. Those are all very different circumstances than medications. They're much more similar to the phone—ask ten people what they paid for them, and you'll likely get ten different answers. The same is true of oil prices; that in itself is a notable subject. The subject of medication pricing might also be worth an article if we don't have one; I suspect we could find a fair bit of sourcing for that. But that's different than trying to put current prices in individual articles. Seraphimblade Talk to me 22:32, 25 November 2019 (UTC)
You said "On the other hand, something like the iPhone 11 lacks it (though ironically, it had a spam link for it I cleaned up), because it doesn't even really have a particular price." The reason it lacks pricing in the article is because you deleted it from the article.[2] QuackGuru (talk) 00:38, 26 November 2019 (UTC)
QuackGuru, you attempted to put it in after I made that comment. Check the timestamps, please. Seraphimblade Talk to me 07:39, 26 November 2019 (UTC)
Seraphimblade, the original list price for every model of iPhone is listed in the large table at iPhone#History and availability (i.e., the first section). Also, I believe that "deal from your carrier" thing really only happens in the US. Most of the world pays cash on the barrel head for their phones.
It would be trivial to find equivalent sources for the original (US) list price of any new drug. WhatamIdoing (talk) 16:00, 26 November 2019 (UTC)
They certainly are, and with no sourcing to indicate any type of significance. I don't have time to clean up such a large table right now, but I'll correct that once I do. In any case, the correct venue for this, ultimately, is an RfC at WT:NOT. If there's consensus that WP:NOPRICE should be changed or removed, well, then such as it is and we'll go forward with that. But until and unless that happens, that is the global policy on prices and articles, and that rule is "With a few exceptions, don't include them." A local consensus here couldn't change that in any case, so this really is just an academic discussion. If you'd like to work on formulating an RfC, I'd be happy to give my thoughts on wording. Seraphimblade Talk to me 17:25, 26 November 2019 (UTC)
WhatamIdoing with an iphone the original list price doesn't require any maths or judgement about doses or indications. You say it would be trivial to find original US list prices of any new drug. But all you can find is the price for a product with a bar code: a particular pack size of a particular strength in a particular formulation of a particular brand. I can't just transform that into "price for a drug". Also, "any new drug" doesn't help us with the 99% of our drug articles that are existing drugs. -- Colin°Talk 17:33, 26 November 2019 (UTC)
The original list price of a MiniMed artificial pancreas doesn't require any maths or judgment about doses or indications, either. It's just an electronic device, complete with bar code. What we recommend about how to describe prices should encompass more than just drugs. WhatamIdoing (talk) 00:20, 27 November 2019 (UTC)
That's very nice, but what we recommend now is "Don't include them". If you'd like to see that change, the aforementioned RfC is still needed. Seraphimblade Talk to me 04:00, 27 November 2019 (UTC)
See WP:NOPRICE: "An article should not include product pricing or availability information unless there is an independent source and a justified reason for the mention." It does not conclude never include pricing. There is independent sources and there is a justified reason to include the pricing for the IPhone 11 but I believe you violated WP:NOPRICE.[3] There is significant independent press coverage. That is a justified reason. QuackGuru (talk) 12:32, 27 November 2019 (UTC)

International Medical Products Price Guide

The source for most "developing country" prices in our drug articles is the "International Medical Products Price Guide". This is typically expressed in our articles by a statement like "The wholesale cost in the developing world is between US$0.14 and US$0.52 per day." or "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose" which appears in both the lead and a Cost section. Although cost statements are widespread in our drug articles, particularly the WHO Essential Medicines, only one user, Doc James, added them, mostly in 2015.

The guide "provides a spectrum of prices from nonprofit suppliers and commercial procurement agencies, based on their current catalogs or price lists. It also contains prices obtained from international development agencies and from government agencies." Its purpose is "to improve procurement of medicines of assured quality for the lowest possible price". It claims "comparative price information helps in price negotiations, in locating new supply sources, and in assessing efficiency of local procurement systems". It cautions "This Guide is intended as a comparative reference only", "The vendors included are not intended to be a comprehensive list of potential suppliers".

There are two types of price in the guide. Buyer and Supplier prices. Buyers are usually government agencies and the price is what they obtained after negotiation, competitive bidding, etc and are not available to anyone else. The buyer prices are indicative only, and can't be used for comparison or reference purposes. Suppliers are organisations who have a warehouse and supply to customers. The prices are further complicated as there are other cost factors to consider when ordering medicines: insurance, transport (air, sea, inland), import duties, unloading costs, handling fee, currency conversion fees, minimum order.

The guide is aimed at the person ordering medicines for their hospital, charity or government department, who is being offered various prices and wants to know if that is reasonable ("for comparative purposes only"). If much higher than the guide prices, then there may be room to negotiate a better price. If much cheaper than the guide prices, questions should be raised about quality.

What the guide does not claim to be is a comprehensive reference of prices in the developing world. It doesn't weigh the prices offered by a huge international supplier as any more significant than the prices offered by a small regional one. For some drugs, formulations or sizes, there may be no price at all, or only one Buyer or one Supplier price.

The guide gives a unit price: per tablet or capsule, or per ml if liquid, or per g if a cream, etc. Each of these will correspond to a particular amount of medicine (often in mg). The actual dose taken by a patient is likely to be unrelated to the unit price: they may take two tablets, measure out 20ml or spread less than a gram on their skin. The guide makes no attempt to related this unit cost to the actual cost incurred by a therapeutic dose or to fully treat a condition. Instead it offers the defined daily dose. This is the "average maintenance dose per day for a medicine used for its main indication in adults". This could therefore be quite different to the dose actually used or when used for another indication. Its usefulness here is in estimating the likely quantity of a medicine that a hospital may go through each day, and thus need to regularly procure. Do I need to buy 10,000 100mg pills a month or 20,000. For some medicines and formulations, the defined daily dose is not available. It also doesn't indicate the duration of treatment.

The guide mostly presents raw data: the package size (100 tablets), the package price ($5.85), the unit price ($0.0585/tablet), the tablet strength (200mg) and the defined daily dose (1.5g). It does perform a little statistical analysis if several prices are available. It gives the lowest unit price, highest unit price, median unit price and high/low ratio. The guide explains that the most useful indicative value is the median unit price. It explains that a simple average (mean) is too influenced by outlier prices. Similarly the lowest / highest prices are themselves most likely to be outliers. The high/low ratio gives an indication of how close or spread the price is. So the most useful price, if we were to present it to readers, would be this median. And it would serve merely as an indication of a typical wholesale unit price available to purchase on the international market.

But a unit price (per tablet or per ml) isn't very useful to our readers, particularly as we don't tell them how many tablets or ml to take. Instead our current wording usually talks of "dose". This is defined to be the amount of drug administered at one time. For ibuprofen, the BNF tells me: "Initially 300–400 mg 3–4 times a day; increased if necessary up to 600 mg 4 times a day; maintenance 200–400 mg 3 times a day, may be adequate". For children, it gives different dose values for seven different age groups. And that's just for the most common indication. Some pills are available in an extended-release formulation, which requires only a once-a-day dose. Other medicines may be designed to take effect rapidly (e.g., buccal midazolam). There are endless permutations of single-dose, three times a day for two weeks, once a day for the rest of your life. As noted above, there is a defined daily dose but that is intended to be used to help the purchasing team know how many pills to order for the hospital, not how much of a medicine a reader might actually use.

Since the guide is based on purchasing a particular form of a medicine, it may actually have many entries for the one drug. For example Valproate.

So we have 150mg, 200mg or 500mg tablets, and 25mg/5ml or 200mg/5ml syrups. We even have enteric coated and normal tablets. Which of these is "the price"? Most of them have data only for one supplier or one buyer. The drug is used for three main conditions: epilepsy, migraine and bipolar disorder. The recommended initial dose, typical dose and maximum dose is different for each condition. There is no formula to convert unit price to dose.

The International Medical Products Price Guide has not been updated since 2015. I have emailed them to ask why and if there is likely to be a future update. Perhaps they are changing to update less frequently than annually. While the guide provides a lot of data, turning this into information that our readers might use would require a lot of original research and there is no clear algorithm that would take these data values and offer a simple number to the reader. Our source does not give any indication of "dose", nor could it, so we can't quote a price per dose. Although our source does perhaps give a "per day" price (by multiplying the unit price to achieve the defined daily dose), it very much cautions against using this metric for any purpose other than stock control. And our source offers multiple prices for different formulations and no way to choose which is relevant. We could give the median unit price for a 200mg tablet in 2014, but our reader would have no way to interpret what a "200mg tablet" represented for a given condition being treated.

So I think that while this guide is invaluable for its intended purpose, I cannot see any way we can use the raw data it presents in a way that is meaningful to present to our readers of a drug article. It is clear that current use of this source is

  • misrepresenting it as a comprehensive reference of developing nation prices
  • giving too much weight to the two potentially-outlier values of lowest and highest price
  • using the ambiguous term "dose" or "per day", which is not supported by the source, which only gives unit prices
  • simply incorrect -- I cannot find examples where the quote prices match the source under any interpretation of the data values

I propose that all use of this source to present the prices in developing nations be removed from Wikipedia. Statements like "The wholesale cost in the developing world is between US$0.01 and US$0.04 per dose" are factually incorrect and misleading in every single word and every single number.

I see similar problems with our prices per dose for the US. These link to Drugs.com monographs which do not actually contain prices. Instead the price is on another page (e.g. diazepam, valproic acid). Those are retail prices "based on using the Drugs.com discount card which is accepted at most U.S. pharmacies." So our lead juxtaposes "developing nation" wholesale prices with US retail prices for a certain discount card holder. And again we have the whole mess of different formulations, different conditions being treated and different pill sizes or syrup strengths and no way to define one and only one "dose". So it seems likely that we should need to remove the US dose prices as well. -- Colin°Talk 18:22, 25 November 2019 (UTC)

Websites frequently shuffle stuff around. It was present on the page in question when the source was used.[4]
Here we have the "National Average Drug Acquisition Cost"[5] Doc James (talk · contribs · email) 18:33, 25 November 2019 (UTC)
Colin, this particular comment feels like it's more about the application of MEDRS to a specific source than about what general advice to give in MEDMOS. Maybe this particular discussion should be moved to RSN or WT:MED instead? WhatamIdoing (talk) 18:51, 25 November 2019 (UTC)
We could change it to "The wholesale cost in the developing world is about US$0.40 per defined daily dose as of 2015."
But in my opinion "The wholesale cost in the developing world is about US$0.40 per day as of 2015." is reasonable.
So these values are published in a book by WHO. The specific number for valproate comes from the International Dispensary Association Foundation an NGO founded in 1972 that specializes in the delivery of medications to the developing world, so it applies to that area of the world. Doc James (talk · contribs · email) 20:30, 25 November 2019 (UTC)

WhatamIdoing, the problem is that MEDMOS is currently pushing the non-consensus of price/cost into drug articles, whereas MEDRS does not. Price is an economic factor, and not really MEDRS's concern. Perhaps that's why WP:MED is making such a hash of it -- we need some accountants, statisticians and basic maths ability here.

Doc James defined daily dose is explicitly not a useful figure for our readers. See WHO Definition. It exists merely to indicate consumption that affects hospital supplies or consumption in a population. "The DDD is sometimes a dose that is rarely or never prescribed because it is an average of two or more commonly used doses." and "DDDs are not established for all medicines" and "DDD is a unit of measurement and does not necessarily correspond to the recommended or Prescribed Daily Dose". It also corresponds to the "its main indication in adults" -- I have no idea whether the main indication in adults for valproate is epilepsy or manic depression and nor are our readers told. The source doesn't say.

I don't think you are carefully reading what the source does and doesn't include. James continues to talk of a "wholesale cost" as though that is one thing for one drug (e.g. valproate). It isn't. The source gives, for example, an indicative cost of buying bottles of 100 500mg tablets in bulk. There's a different price for 150mg tablets and these are approx 2x more expensive per mg than the 500g tablets. The syrup is even more expensive per mg of drug. These aren't reference prices for "the developing world" but really just indicators from whatever sources they could manage to get that year. So the 200 mg/5 ml syrup only has the Buyer price for South Africa. This isn't a price anyone can buy -- they didn't find any Suppliers with public prices at all. It is the unique price the SA government negotiated and reflects only their contract. And the syrup formulation of valproate is important, as that is the primary form used to treat epilepsy in children.

James, you claim now you used the defined-daily-dose to work out the prices in articles. But as I worked out above, the price you included in the articles does not match that figure. I see you have updated the valproate article. You haven't explained why the figures were wrong before.

What is happening here is a gross over simplification of the issue of "cost". It isn't "developing world" it isn't "one cost" it isn't "lowest" or "highest" and it isn't "per dose" or "per day" in any actual patient and it isn't "one indication/use". A drug has multiples of all these and you are trying to find one number.

WhatamIdoing, this reminds me of Brexit -- all things come back to Brexit in the UK. A slim majority of the UK voted to leave the EU but nobody was told what kind of Brexit we would get. Some promised we would stay in the customs union or the single market, others said we could leave both and still get all the benefits for free. It now turns out that there is no form of Brexit that a majority would be happy with, though it seems like we are heading for a no-deal Brexit in December 2020. You might all wish to include "the cost" as a nice simple figure for readers. But when you actually look at the sources at the fundamentals of how drugs are delivered, indicated, dosed and used, there isn't one cost. -- Colin°Talk 09:32, 26 November 2019 (UTC)

Colin, we're kind of spread all over the map here, and it might be useful to get a little more precise. Whether any specific source is being accurately represented in a specific article isn't really MEDMOS's concern. MEDMOS, like all manuals of style, is primarily concerned with how you say something, after you've already decided that it belongs in the article in the first place.
The things that should be covered by MEDMOS are things like these:
  • Is this something that should normally be included (e.g., similar to side effects; we do not require any sort of extraordinary coverage of side effects to put those in a drug article), or only under certain circumstances (which should be listed)?
  • How much precision should be included? The US Medicaid program says that the National Average Drug Acquisition Cost for 250 mg capsules of valproic acid is running between 15 and 25 cents per pill right now, but it reports it to the one-thousandths of a cent. How much do you think is appropriate for most circumstances? Dollars, dimes, pennies, fractions of pennies? The first two significant figures?
  • Should costs be presented in monetary terms, or in real-world terms? For example, do you say "Recovery costs US$n in lost wages" or "People can usually return to work after n weeks"? This question is not at all specific to drug costs. Assuming that ideal sources exist, would we rather say that:
    • depression represents a drag of US$n trillions on the global economy, or that
    • it results in US$x in medical costs, US$y in increased morbidity, and US$z in excess mortality, or that
    • depression causes in US$x in medical costs, worsens the quality of life, and shaves y years off the lifespan of affected people?
  • When you describe a product or service with significant variations, how do you describe it? There's a lot more variation in laptops than there is in valproic acid, but Laptop#Market Share lists an average sales price. How could we do better?
  • Which prices belong where? Presumably, if you're on this kind of list, then it goes in the lead plus elsewhere in the article. But in other cases, where's the best place? Does it depend upon the nature of the information, e.g., with recent prices going in ===Economics=== and original list prices going in ==History==?
  • How should prices be described? Consider this business source. WP:MEDPOP approves of the popular press for financial information, and that's what this is: Prosecutors said that Fagron's subsidiary "typically sold the fluticasone propionate for about $160 per gram". It's a generic drug, the daily dose is the same for (almost) everyone, the price only applies to the US, and it's specific to one company. What else would you need to know to describe that price? What about the price for Asfotase alfa, which is only made by one company and whose dose varies by weight [6]? There are a lot of ways to calculate prices for drugs in particular, so which one would you use, and how would you explain it, if you needed to include that information in an article?
Completely apart from the question of whether prices should be in many articles or few, I think we could do some good work here in describing how to write about them. WhatamIdoing (talk) 00:12, 27 November 2019 (UTC)
WhatamIdoing, the discussion on the MSHPriceGuide is ongoing at WP:MED so perhaps that's a better place for it than here. The questions you ask may be valid. But I'm interested in the text in actual articles right now being very very wrong, rather than specifying about how may digits of precision some hypothetical price might be. We have an awful lot of basic "text is not supported by the source" problems at the moment, such as a claim that the price offered by a supplier of Ethosuximide only in the Democratic Republic of Congo is "the price in the developing world". It is all well and good speculating about how we might describe prices, if we can't actually do that given the sources we have. -- Colin°Talk 09:01, 27 November 2019 (UTC)

James claims the "The wholesale cost in the developing world is about US$0.40 per day as of 2015." But that is just for enteric coated 500mg tablets from IDA Foundation. The price in any other form or for other suppliers or in some developing nations is wildly different, even if we misuse the DDD in the way James wants (which is Original Research). -- Colin°Talk 09:47, 26 November 2019 (UTC)

Introduction to Drug Utilization Research 2.6 Drug costs lists several cost metrics:

  • total drug costs
  • cost per prescription
  • cost per treatment day, month or year
  • cost per defined daily dose (DDD)
  • cost per prescribed daily dose (PDD)
  • cost as a proportion of gross national product
  • cost as a proportion of total health costs
  • cost as a proportion of average income
  • net cost per health outcome (cost-effectiveness ratio)
  • net cost per quality adjusted life-year (cost-utility-ratio)

These are all the possible costs we might consider in the encyclopaedia. James has (apparently, though not supported by any evidence) chosen cost per DDD. Here's what WHO say about that: "For example, the cost per DDD can usually be used to compare the costs of two formulations of the same drug. However, it is usually inappropriate to use this metric to compare the costs of different drugs or drug groups as the relationship between DDD and PDD may vary."'. So my above table showing that 500mg is the cheapest per DDD and the 250mg/5ml syrup is the dearest is a valid usage. But it is only for one formulation and it would be wrong of us to use it to indicate "the cost of valproate" compared with eg. "the cost of carbamazipine".

Problems using the defined daily dose (DDD) as statistical basis for drug pricing and reimbursement: "The defined daily dose is an artificially and arbitrarily created statistical measurement used for research purposes in comparing the utilization of drugs.", "in most cases, the DDD differs greatly from the typical PDD of the drug in question. In some cases, this gap may be exacerbated by the fact that a drug may be prescribed in two vastly different dosages and the DDD represents the average of those outliers." The paper goes on to explain why cost per DDD is a misuse of the metric.

From these and other sources, it seems that presenting a cost per DDD in the lead of our drug articles would be exactly the misuse of DDD that WHO and others caution against. It is an arbitrary unit for doing research into consumption. Presenting this information in our leads is a clear example of why we do not allow original research. -- Colin°Talk 10:13, 26 November 2019 (UTC)

IDA Foundation sells at these rates to more than 130 LMIC countries. So this represents the approximate wholesale cost in a lot of places.
Per WHO Defined daily dose "is the assumed average maintenance dose per day for a drug used for its main indication in adults."[7]
So this is the approximate or typical dose in adults.
Additionally $0.40 per day is approximately the same as $0.82 per day.
What it is NOT the same as is $638 per dose.[8]
If these approximates both you I would be happy to switch to "generally less than US$1 per day", "$US1 to 2" "2 to 5, 5 to 10, 10 to 20, 20 to 50, 50 to 100, 100 to 500, 500 to 1000" etc.
Sure there are other numbers we could also provide. And I would welcome you finding sources and doing so. Doc James (talk · contribs · email) 14:42, 26 November 2019 (UTC)
James, you keep mentioning IDA foundation as though somehow that makes your numbers right. IDA isn't even listed as a supplier for all formulations. If you want to just use a single source for pricing, then say "IDA Foundation sell ..." rather than misleading our readers that this represents developing world prices. If the IDA foundation price was all that mattered, then there would be no need for the MSH Price guide.
You are just plain wrong about DDD. Please read what I wrote and linked to.
No, $0.40 is not the same approximately as $0.82 per day. Remember these are wholesale prices. The average daily wage in India is $5.50. So a tablet at $0.40 and a syrup at $1.18 + retail markup are hugely different prices to someone in the developing world. You are just waving your hands about at the moment with this "these are pocket change numbers" comment. It very much looks like you don't care about the actual numbers... perhaps that's why every example I looked at is wrong.
I've no idea what you linked to that was $638 per dose. The link doesn't indicate any drug name. I think we can agree cents are not the same as hundreds of dollars, but if you want to make a point about a treatment being cheap vs extortionate you need a source making that point.
I have looked at lots of sources. James, you are missing the big point. When you say, for example, "IDA Foundation sells at these rates" you miss out the "sells what". What are they selling? They are selling 500mg tablets in bottles of 100 in bulk orders. You have arbitrarily chosen just one formulation of this drug and quote that price. It is as daft as me saying "The price of strawberry yoghurt is 50p". I didn't say how much yoghurt and there is no one size for yoghurt and there are all kinds from Greek style to French set to Organic.
What you want to say "The wholesale price of valproate in 2014 was $...." cannot be said. It makes no sense. You are conducting original research in an area well beyond your competence, even if we allowed editors to do so.
James, there is a reason why none of our sources, in their leads, say things like "The wholesale price of valproate in the developing world is $0.40". They don't say it because it is nonsense. Instead, we have some sources that give raw data for prices for specific product items with bar codes.
I think perhaps we need to open this discussion up to parts of Wikipedia with mathematical competence, rather than just WP:MED. -- Colin°Talk 15:35, 26 November 2019 (UTC)
When you are interested in having a civil discussion let me know. Otherwise please do not ping me. Doc James (talk · contribs · email) 15:53, 26 November 2019 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Restart

This has become unwieldy. Let's see if a clean start below helps. Best, Barkeep49 (talk) 21:27, 23 December 2019 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Discussion above seems to have gone into the weeds a bit. I believe the best way forward might be to clearly understand the purpose and context of including pricing. For example:

  • Increases in the wholesale price of insulin, noted above, are clearly a pressing public health concern and noted by large numbers of sources.
  • Variations in costs of drugs between nations may be a matter of public of concern (e.g. salbutamol / albuterol at $30 - $60 per inhaler according to various sources whereas the NHS estimates the annual cost of treatment to the NHS at about £22, with a per-item prescription fee payable by working-age adults of £9 per item). I would suggest this is most relevant on drugs on the WHO essentials list, where there is a notably wide disparity as reflected in independent sources such as news articles.
  • The price itself, if it is not exceptional, would not seem to be relevant or encyclopaedic, unless it is specifically called out as extraordinary (e.g. a new cancer drug that costs hundreds of thousands of dollars for a course of treatment but does not yet have a sufficient body of evidence to unambiguously support use - a perennial problem in the UK - or perhaps aspirin, which is pretty much the cheapest drug in the world and is constantly finding new uses).

In other words, it seems to me that we need to show a reason to care about cost: either it's sharply increased or decreased, or there are large and important disparities between nations, or it's called out as extraordinarily high or low. In all cases the key is that sources we normally trust - serious mainstream newspapers, in-depth journalistic reporting by fact-finding TV shows, published journal articles - note the price as being a significant matter. What would people think of using a benchmark expectation of three or more serious sources mentioning the price or one devoted entirely to the price as a significant concern or factor? Guy (help!) 11:43, 5 December 2019 (UTC)

Addendum The questions I think I'd like to be able to answer are:

  1. Should we include pricing as a matter of course, or only where it's been discussed as a significant issue by RS?
  2. If as a matter of course, based on what part of WP:5P?
  3. If we should include as a matter of course, how should we manage the problem of global variation?
  4. If we should include only based on RS commentary, what level of commentary do we need?
  5. If we include pricing, should it be only from secondary sources or is a primary source acceptable or (because it may be more accurate) preferred?

I think if we have confident answers to these questions, everything else comes out in the wash. Guy (help!) 12:38, 5 December 2019 (UTC)

Guy I fully agree with you that only exceptional pricing is encyclopaedic. This is also WP:NOT policy. And also WP:WEIGHT policy -- if nobody is commenting on the wholesale price of a drug in DRC vs the retail price in the US, then neither should we. James, unfortunately, contests that WP:WEIGHT is supported by database records of raw price data, or by the general concern about drug pricing means it has WEIGHT in every drug article.
However, the only reason this conversation was on this page was because pricing was very recently added to MEDMOS and QuackGuru started edit warring to support James's position of putting detailed prices in the lead. WhatAmIDoing suggested another forum might be better. So it continued at Wikipedia talk:WikiProject Medicine. If you want to start a new RFC on drug pricing, then that may be useful. WP:MED ignored the result of the last one, though. The issue of an RFC is similar to Brexit. The last RFC contained a lot of untruths about pricing sources and wishful thinking about prices being useful, and rather ignored the problem that we don't have sources of anything but raw data, and so require original research to make any statement. As folk say about damn lies and statistics, it is unfortunately a problem that one can make any point one likes about drug pricing merely by selecting the right database records to use. -- Colin°Talk 12:18, 5 December 2019 (UTC)
Colin, You were doing so well right up to the point that you started attacking other editors again. Could you maybe not do that please? Guy (help!) 12:25, 5 December 2019 (UTC)
Guy I've given an example of behavioural problems and open agenda pushing in a sequence of diffs over at ANI.- Colin°Talk 13:21, 5 December 2019 (UTC)
Basically, we're just saying to follow existing policy: NOT, POV, OR. --Ronz (talk) 19:17, 5 December 2019 (UTC)
NOT does not apply to medical content. QuackGuru (talk) 19:53, 5 December 2019 (UTC)
I'm afraid you'll find no consensus for that, or are you just trolling? --Ronz (talk) 20:11, 5 December 2019 (UTC)
I'm dying to know what that assertion is based on. – Levivich 21:37, 5 December 2019 (UTC)
@QuackGuru:, could you please explain the comment that "NOT does not apply to medical content"? SandyGeorgia (Talk) 18:04, 6 December 2019 (UTC)
Where does Not mention drug prices? One size does not fit all. QuackGuru (talk) 18:15, 6 December 2019 (UTC)
Your statement was "NOT does not apply to medical content"; your answer does not address that. Also, almost all of NOT specifically addresses topics just like drug pricing, and the idea that the page should list every one of those items explicitly is faulty. SandyGeorgia (Talk) 18:22, 6 December 2019 (UTC)
Where does the WP:NOT policy mention prices? Seriously? At WP:NOTPRICES:

An article should not include product pricing or availability information unless there is an independent source and a justified reason for the mention. Encyclopedic significance may be indicated if mainstream media sources (not just product reviews) provide commentary on these details instead of just passing mention. Prices and product availability can vary widely from place to place and over time. Wikipedia is not a price comparison service to compare the prices of competing products, or the prices and availability of a single product from different vendors or retailers.

– Levivich 18:37, 6 December 2019 (UTC)
And I agree with that. All the sources being used to discuss prices are independent. And we have sources that justify the importance of pricing information. Doc James (talk · contribs · email) 20:40, 6 December 2019 (UTC)
The problem with the wording above at WP:NOTPRICES is that there is no mention of WP:WEIGHT. We don't typically add information to Wikipedia based on one source. That guideline needs attention, although I suggest that the subsequent sentence does use a plural and clarifies the singular in the first sentence. The not a price comparison service seems to a definition of where WPMED practice currently falls. SandyGeorgia (Talk) 15:38, 7 December 2019 (UTC)
I fully agree with Guy and Colin on this. Exceptional pricing that attracted media coverage should be mentioned, but otherwise Wikipedia is not a drug sales catalogue. As I mentioned elsewhere, pharmaceuticals pricing is among the most opaque ones in the world – a medication will have a list price, a reference price, a benchmarking price, the effective price per payer/country, an out-of-bundle price, and many others. They also vary significantly – a drug with a US list price of $1000 and a regional reference price of $800 may be sold for an effective price of $200 or less after volume discounts, risk sharing agreements and/on bundling or capping arrangements. For this reason list prices are irrelevant countries where drugs are purchased by public payers. Listing US retail prices will usually be confusing to the international reader. Again, I'd like to remind Doc James that there is an entire world outside of the simple buy/sell reality of the US! — kashmīrī TALK 22:52, 5 December 2019 (UTC)

The World Health Organization states "Data on drug costs will always be important in managing policy related to drug supply, pricing and use."[1] They list a few price metrics including cost per treatment day, month or year and cost per defined daily dose (DDD).[1] Doc James (talk · contribs · email) 03:21, 6 December 2019 (UTC)

WHO doesn't dictate Wikipedia's content. It may be important, but it may not be important enough for what's been done or proposed for product pricing. Nor does it mean it is suitable for presentation in an encyclopedia article.
Best focus on Guy's initial comment and addendum. --Ronz (talk) 04:21, 6 December 2019 (UTC)
It is best to oppose Guy's proposal that is against WHO's position.
WHO confirms drug prices are important. Drug prices is basic information readers are seeking. QuackGuru (talk) 04:33, 6 December 2019 (UTC)
No one is against WHO's position. --Ronz (talk) 04:48, 6 December 2019 (UTC)
Excluding drug prices except in rare cases is saying drug prices are not important. WHO says drug prices are important. QuackGuru (talk) 05:10, 6 December 2019 (UTC)
QuackGuru, no it's not, it's just saying that Wikipedia does not play a role in drug policy. Guy (help!) 09:10, 6 December 2019 (UTC)
Yes, it is saying that drug prices is unimportant. Essential information is drug pricing according to WHO. Excluding essential information is against WHO's position. QuackGuru (talk) 12:56, 6 December 2019 (UTC)
QuackGuru, no it's not. It is saying that drug pricing is important to some purpose other than an encyclopaedic description of the drug. Guy (help!) 14:22, 6 December 2019 (UTC)
  • (from ANI) Re Guy's proposal, I think include the pricing only where it's been discussed as a significant issue by RS. The level of commentary should be "discussed as a significant issue by multiple RS"–more than just a mention, and more than just one source treating it in depth. Maybe 2+ RSes giving it significant coverage. The RSes supporting pricing should be secondary, but could be complemented by primary sources in appropriate circumstances. For example, 2+ secondary sources discussing the increase in insulin prices in order to include it in the article, but we might supplement with a primary source to give the most up-to-date price. – Levivich 05:03, 6 December 2019 (UTC)
Being mentioned in a single reliable source is sufficient IMO. There is discussion of prices for every medication more or less but some is easier or harder to find.
For example lenalidomide and bortezomib this WHO text says "Monthly pharmacy costs included in the total monthly cost in the unadjusted analysis were 4101 $US (SD 1931) and 4855 $US (SD 2431) for lenalidomide and bortezomib, respectively."[9] as part of a 6 page discussion of costs.
This is done for all medications added to the WHO Essential medicines list. Doc James (talk · contribs · email) 06:14, 6 December 2019 (UTC)
Doc James, If the price is the primary focus of an independent RS, then I would agree. If it's mentioned as a side issue, then I'd look for more than one. I don't think that's unreasonable, do you? Guy (help!) 09:11, 6 December 2019 (UTC)

User:JzG my thoughts on your questions:

We have plenty of sources that discuss prices.[10][11] Doc James (talk · contribs · email) 20:44, 6 December 2019 (UTC)
  1. Should we include pricing as a matter of course, or only where it's been discussed as a significant issue by RS?
    • WHO states prices are an important aspect of medications and thus IMO we should generally include content on this when covered by a reliable source.
  2. If as a matter of course, based on what part of WP:5P?
    • The content is encyclopedic. The cost of a medication is a core part of cost benefit analysis and a key aspect of public health.
  3. If we should include as a matter of course, how should we manage the problem of global variation?
  4. If we include pricing, should it be only from secondary sources or is a primary source acceptable or (because it may be more accurate) preferred?
    • We have textbooks that list these as well as WHO and the US government. Doc James (talk · contribs · email) 06:22, 6 December 2019 (UTC)
      Doc James, Cost-benefit analysis is not part of Wikipedia's remit, though. We're an encyclopaedia not a textbook or prescribing guide. You and I agree that drugs such as insulin, where costs and cost disparities are identified as a pressing public health concern, should unquestionably be included. But I'm not persuaded that just being on the WHO register qualifies to discuss cost unless other sources do too. Normally I would not expect it to be particularly difficult to find sources discussing cost disparities for essential medicines, given the prices paid in the US compared with other countries. If there's evidence that price, specifically, is a focus of RS coverage for a substantial proportion of the WHO list then I'd be more inclined to support your position. Can you show that? Not price books or anything else, but actual RS coverage showing that costs of WHO essential drugs are considered independently significant? I suspect this does exist and we may be able to word a consensus position for the WHO essential list. Guy (help!) 09:18, 6 December 2019 (UTC)
    user:JzG As I mentioned above we have a 6 page discussion in a WHO document of the price of lenalidomide and bortezomib among others. Would you consider that to be suitable for showing notability? Easier to get this information for newly added medications to the list as documents for when older medications were added are harder to find. If we limit cost discussion to only those picked up by the popular press we will end up with primarily US prices for new medications. And our content will become more a US perspective. Doc James (talk · contribs · email) 20:38, 6 December 2019 (UTC)
    we have a 6 page discussion Can you please link these discussions? --Ronz (talk) 20:45, 6 December 2019 (UTC)
Here starting on page 51[13] Doc James (talk · contribs · email) 02:49, 9 December 2019 (UTC)
Guy, James, this 'discussion' is an application for three drugs to be included in the list of WHO essential medicines (which they were this year). It is in WHO's archives and not AFAICS published elsewhere. I believe Guy asked for independent RS, not internal WHO documents in their archive. Naturally cost and cost-effectiveness is going to feature in that decision. And it is a hard decision because cost is hard thing to consider and weighing that cost against sickness is a hard thing to do. The document is 76 dense pages long and includes discussion of complex pricing models. Cost changes throughout the many years of treatment, there's relapse and some patients are refractory to treatment. The drugs are taken in combination. The descriptions in that document blow my mind and I can't begin to think how to explain the cost of treatment with multiple therapies simultaneously.
Yet on our article Lenalidomide we say "In the United States it costs about US$16,000 to US$21,000 per month". But what we don't tell the reader is that's a retail price, and most other drugs we've looked at cite a wholesale "cost". The Drugs.com article explains all sorts of discounts on offer, possibly as low as $25 a prescription if your insurance is good enough (though I may be misreading as I'm not familiar with US drug pricing).
And we juxtapose that in the lead with "In the United Kingdom this amount costs the NHS about £3,400 to 4,400." This cites the BNF which gives an "Drug tariff price". But hold on a second, shouldn't we be comparing retail prices? Firstly cancer patients in the UK get their prescriptions free (normally £9 for a month's supply). So the true juxtaposition of the US retail price is £0!! Yay for socialist healthcare. Secondly, the UK has a special arrangement price for this drug. These arrangements vary from year to year, but this article from 2014 notes that "The manufacturer proposed the patient access scheme, which provides a price discount: the NHS receives the drug for free for any patients remaining on the drug after 26 cycles.". But when bidding for the drug to be a first line treatment (rather than third), the patient access scheme discount was part of the reason negotiations took so long. This article in 2019 says "The drug’s list price in England is up to £4,368 a month, or over £52,400 a year per patient, with the confidential PAS discount cutting this substantially.".
Lenalidomide is one of those drugs that is on the borderline for affordability even in the rich UK and the high price issue is discussed in the body of literature. However, the statement "costs the NHS about £3,400 to 4,400 per month" was derived from original research reading simplistic drug pricing in the BNF. If a professional expert was writing about the cost of lenalidomide to the UK's NHS, they'd have known about the patient access scheme and explained that any official list price is not the true cost to the NHS, which in this case is confidential. Our wiki article currently has a very incorrect wholesale list price for the UK juxtaposed with a US retail list price, which I really don't know what proportion of the population are actually paying. Honestly, even in articles where price is notable, we still go and ruin it by doing original research on raw data sources. We slip up every time. -- Colin°Talk 19:18, 9 December 2019 (UTC)
This may be the place to introduce Colin to US drug pricing. First, every different insurance company negotiates different deals, to such an extent that if one has a chronic condition, one may look to a specific insurance company whose policy is best depending on that condition. Independent insurance brokers gain clients by studying the pharmaceutical pricing structure of each company so they are in good position to offer recommendations. Ditto with Medicare supplement plans (our social security does not cover all medical costs, so many buy supplemental insurance to cover drugs). We personally pay $ 000000 (ZERO) for many prescriptions, because we chose the right plan.

Second, almost EVERY pharmacy I know of has what is called a $4 plan ... that is, regardless of the retail, wholesale, or whatever price, they offer a 30-day prescription to many common drugs for only $4. This is marketing-- to get customers in the door. They similarly offer 90 days for $10 for many drugs, and their plans are typically posted right next to where you submit your prescription.

Third, coupons are available all over the internet from specific pharmacies for specific drugs-- again, marketing to get clients in the door. (My husband saved hundreds of dollars recently by simply googling the drug he needed.)

And finally everybody with a phone uses the app, GoodRX, which finds the cheapest source of a given drug in any geographic location, and even coughs up the coupons. And by everybody, I mean ... I volunteer as a Spanish-language interpreter in a free clinic that serves hispanic migrants without insurance, and everyone with a phone knows about GoodRX-- even those who don't speak English. To help the uninsured save money, when a physician writes a scrip in the clinic, they first pull up the $4 list and if they don't find something there, they next go to GoodRX.

In short, prices are extremely variable, subject to many factors, and very few people pay retail. Having said that, epipen is expensive for everyone, which is why I brought up that example (had to do with shortages or something.) SandyGeorgia (Talk) 20:18, 9 December 2019 (UTC)

Further adding to the complexity is the variation between insurance plans for drug formularies. Insurance may make one drug free while requiring people to pay out of pocket for another. And I know from personal experience that the classification of a single drug can vary not only from one insurance provider to another, but between different insurance plans from a single provider (bronze, silver, gold, platinum). --Tryptofish (talk) 22:28, 9 December 2019 (UTC)
  • The price of all things is important to somebody. I can find sources or indeed highly authoritative reports from respected bodies that say the price of ____ is important. Food, heating, housing, travel, clothing, broadband, alcohol, pet food, kitchens and bathrooms... I don't think the argument that the WHO say prices are important or even that some authority considers the WHO List of essential medicine prices to be important is actually a valid argument for including in an encyclopaedia. Someone mentioned Wikipedia's place in the information<-->knowledge spectrum. The problem with James's sources are they are raw database records. They tell you that a pack of 60 5mg wonderpam-sodium enteric-coated tablets from DrugMaker corp on the week ending 06/12/2019 cost $xx [insert some country-specific definition of an arbitrary "wholesale" price here]. What James has tried to do and claims we can do, is give 'a price' for 'a drug'. We can't. Above it is claimed "Drug prices is basic information readers are seeking" but read kashmīrī's comment above about all the different concepts of "price" in the US. James claims we can use the BNF for UK pricing. It is no better. It gives several prices for each specific thing with a bar code you might get from a pharmacy. It lists two prices, not one, one of which is indicative of what a pharmacy might pay, and the other an indication of what the NHS would refund the pharmacist -- but that explanation hugely hugely oversimplifies the actual pricing. These prices depend on whether the drug has several generic alternatives or only one supplier, whether the GP wrote the brand name on the prescription or not, they vary depending on supplier shortages, which are not uncommon. They vary in dose size, formulation, coatings, etc. I'm afraid the claims "we have ...." sources that would provide an easy "the price for wonderpam is $3 a day in the US" do not stack up to scrutiny. We have raw data and way way too much original research needed to even begin to simplify the price, and no one concept of "price" from country to country. -- Colin°Talk 11:37, 6 December 2019 (UTC)
    Colin, it's a great argument for an article series on medical costs in $COUNTRY, though. Guy (help!) 14:25, 6 December 2019 (UTC)
    I love this argument. Why don't we add national prices to articles on cement, timber, steel, copper, gas, electricity, agricultural land, university tuition fees, public transport, bread, housing, postal services etc.? For sure there are UN sources that list many of these prices as highly important in some area of human activity. Why stopping at drugs? Let's make Wikipedia a registry of Very Important Prices! — kashmīrī TALK 17:49, 6 December 2019 (UTC)
  • Reliable sources do say it is important and reasons have been given for including them. Prices of the same drug vary in different countries.[14] Drug price monopolies raise drug prices.[15] QuackGuru (talk) 12:56, 6 December 2019 (UTC)
    QuackGuru, also true of cars. You are framing opinion as fact and not citing policy to support it. Guy (help!) 14:19, 6 December 2019 (UTC)
    I cited sources that support it is important. That is framing a fact as a fact. QuackGuru (talk) 14:22, 6 December 2019 (UTC)
    QuackGuru, see WP:ITSIMPORTANT. You cited sources that say it's important to drug policy. Wikipedia has no role in that. Guy (help!) 14:26, 6 December 2019 (UTC)
    Drug prices fluctuate and there is a controversy over the increasing drug prices. QuackGuru (talk) 14:30, 6 December 2019 (UTC)
    QuackGuru, which is either significant or not depending on the importance of the drug, the amount of the increase, and other factors. And we establish that from reliable secondary sources about the drug specifically. What we don't do is to decide that because prescription drug prices in the United States is an important topic, so the price in the United States is important to every drug. Guy (help!) 14:50, 6 December 2019 (UTC)
    Drug prices vary globally for the same drug. This has led to a controversy. Therefore, it is important to include the differences in drug prices. QuackGuru (talk) 14:56, 6 December 2019 (UTC)
    QuackGuru, Your logical fallacy is: begging the question. Guy (help!) 11:28, 7 December 2019 (UTC)
  • (Saw this on ANI). Not my area of expertise, but I just wanted to challenge the argument that because the WHO considers that "Data on drug costs will always be important in managing policy related to drug supply, pricing and use" it does not follow that "Data on drug costs will always be important in informing a reader of an encyclopaedia article about a drug" which is our remit. We might decide it is, we might decide it's not, but as (hopefully) our articles will not be used to manage policy re drug supply, pricing and use, we should consider things by our own requirements. Scribolt (talk) 13:58, 6 December 2019 (UTC)
    • Indeed it is framed that "WHO say X therefore Y". For example James says "[WHO] list a few price metrics including cost per treatment day, month or year and cost per defined daily dose". Yes they do list them, but that doesn't mean they recommend them, or that one has to be very careful when using one. For example the "cost per DDD" is extremely contentious and WHO warn that DDD is essentially an arbitrary number for population studies. They say "the cost per DDD can usually be used to compare the costs of two formulations of the same drug. However, it is usually inappropriate to use this metric to compare the costs of different drugs or drug groups as the relationship between DDD and PDD may vary." (PDD is the prescribed daily dose for an individual with a particular condition). So you can use the arbitrary DDD to do some maths to compare that a person needing 100mg DDD with two 50mg doses a day would in the UK cost the same as someone taking 100mg once a day (because the few UK prices I've seen seem proportional to dose -- not claiming that as a fact) whereas in the US it would likely be twice as expensive (again, based on my experience with US prices being the same for various sizes, though with some odd values here and there). Of course all I just did is original research and just because WHO say an expert performing Drug Utilization Research (the source James cites) could do that sort of maths, doesn't mean that Wikipedians can or should. -- Colin°Talk 16:28, 6 December 2019 (UTC)
  • Guy, you might be starting to realise that there are two sides to this who are completely talking past one another. One side is trying to frame an argument based on polices, on what individual drug articles might warrant, and what they might reasonably expect editors to be able to say based on very limited sources, and an appreciation of just how complex the issue of cost is. The other side is framing their argument based on a political agenda against Big Pharma secrecy, price extortion, etc, based on feelings around drug costs in general, and never budges from considering that cost is a simple basic thing all readers want to know. It is like placing a Brexit party MEP in the same room as a Lib Dem MEP. Each have totally separate value-systems when it comes to emotion/truth/fairness/equalty/etc,
Further, as we know from Brexit, the Devil has the best lines. A punchy dishonest/misleading soundbite zings whereas you need a five-paragraph BBC Fact Check web page explaining why, in fact, it isn't as simple as that and no there never were going to be 40 hospitals or 20,000 police officers. The Truth is complex and messy whereas Lies are simple and easy. I don't have a solution for that.
If we keep talking past each other, as has happened with me and James, and with Ronz and James, then we will not achieve consensus, and any poll in the end will just becomes a numbers game. It may result in another no-consensus result, and it may once again result in a problem where one side says things the other side regard as patently false or obviously irrelevant. Further, a poll merely asking "should all our drug articles have prices", still doesn't solve the problem of which prices and how do we get them without original research or relying on now defunct websites or over-simplifying. When you have editors who are totally happy with juxtaposing wholesale prices from 2015 with retail prices from 2019, the problem is bigger than "should we have prices". Like politics, it is all well and good if a majority want the moon on a stick, we need some agreement on what those drug prices might look like, and acceptance that the methodology used to get them passes our fundamental policies. I don't think that has been demonstrated yet, not even close. -- Colin°Talk 16:28, 6 December 2019 (UTC)
I think there are sound (and source supported) arguments in both sides, that's why this issue is so difficult to solve. We need to be careful and open in our approach here, depending on how we work it out it may be beneficial to the encyclopedia, if only just to create new articles dedicated to the subject. Signimu (talk) 20:11, 9 December 2019 (UTC)

Regarding the importance of pricing: Where is it covered in depth in Wikipedia? Medication costs, Prescription drug prices in the United States, Prescription charges. I expect there are more. There seems to be some disorganization and outdated links currently in articles. Regardless, we clearly think pricing is important and are treating it so. --Ronz (talk) 17:58, 6 December 2019 (UTC)

Thanks for the restart, JzG. I went away over the US Thanksgiving holiday, and lost track of the discussion here (will not try to review the content added since I last weighed in, and appreciate the restart). Colin, henceforth please let the bigger problems speak for themselves (ala ROPE); I believe they are self-evident. I agree with JzG, Colin, Levivich, Ronz, and Scribolt (because their positions address Wikipedia policy). I also agree with Kashmiri (an unusual position for me). I disagree with Doc James and QuackGuru (because those positions do not encompass policy). I also submit that with so much dangerously defective information on Wikipedia, WPMED should get back to working collaboratively to improve content rather than focusing on adding information that is not policy compliant. SandyGeorgia (Talk) 18:13, 6 December 2019 (UTC)

MEDRS is not policy. It is a guideline like this page. QuackGuru (talk) 18:17, 6 December 2019 (UTC)
QuackGuru, it has much greater weight than this page. Guy (help!) 11:28, 7 December 2019 (UTC)
QuackGuru, the relationship is that guidelines explain how to apply policy. Guidelines cannot/should not deviate from policy. SandyGeorgia (Talk) 15:33, 7 December 2019 (UTC)

We have an overview on the relationship between the different price estimates used in the United States here. Indicates as we all agree there is more than one way to present a price. Even those prices for medications are rough figures, they are still useful. $US10 is very different from $US10,000 or $US2.1 million. Doc James (talk · contribs · email) 20:49, 6 December 2019 (UTC)

Try to register that drugs are mostly available for free in much of Europe and parts of Asia and Africa and have entirely different retail prices in most of the developing world. Why would anyone want to stick your American deliberations on Medicaid, etc., into a global encyclopaedia? If a resident of the US, India or Ghana is interested in a particular drug price, he or she can always look up an online pharmacy in own country or ask in real world around the corner. The differences in drug prices between countries can be even 1000-fold. — kashmīrī TALK 23:51, 6 December 2019 (UTC)
In our struggle to GLOBALIZE, are we ignoring how drugs/medical products are typically supplied in the US? Insurance companies negotiate drug and medical product prices and the differences within the US can be significant depending on insurance company. How many people pay retail price in the US? SandyGeorgia (Talk) 15:44, 7 December 2019 (UTC)

"up to 90% of the population in low- and middle-income countries must pay for medicines out of pocket"[2] For the United States out of pocket costs are about 48 to 67%.[16] Doc James (talk · contribs · email) 02:46, 9 December 2019 (UTC)

Thanks for this, Doc, but ... First, the best I can tell, that is 1998 data, which is useless post-Obamacare. Second, the full text of the cited source about US out of pocket costs reveals another layer of complexity: age (not surprising considering how insurance plans work). Annual average out-of-pocket prescription drug expenditures for all adults are $177, but people age 65 and older pay much more for their medications. People age 65 to 79 pay $456 out-of-pocket. People age 80 and older pay even more (see Figure 4). Adults pay almost half — 48 percent — of their expenses for prescription drugs out-of-pocket, but persons age 65 to 79 pay 56 percent and those age 80 and older pay 67 percent of their total drug expenditures out-of-pocket. Doc, with this much variation in drug prices in the US, and this new layer of complexity, could you please take one clear example and plainly explain what numbers you are using, how they can be representative of anything, and how they are not original research? As you are the person seeking to make this change to policy, it would help all of us and expedite matters if you would clearly explain how the data you want to use relates in any way to prices people actually pay. SandyGeorgia (Talk) 03:21, 9 December 2019 (UTC)
Sure. The DDD for valproate is 1.5 grams. The external reference price is 0.1339 per 500 mg tab. 90 tabs per month at this tablet size. So about 12 USD per month at DDD.[17] Doc James (talk · contribs · email) 23:50, 9 December 2019 (UTC)
You didn't answer Sandy's question, which asked you to explain how what you are doing is not original research and why you picked numbers that represent anything useful. You picked the 500mg tablet which has only one supplier. The 200mg tablet has five suppliers, which suggests to me you picked the wrong one. Doses above 250mg should be taken as divided doses according to Drugs.com. A 500mg enteric coated tablet cannot be split, so isn't very useful in practice while titrating dose from an initial 700mg, say. You've been misled by the 1.5 DDD into thinking that is nicely 3 tablets, when real living patients might be on any dose from 700mg to 2.5g and so require smaller tablets. And 500mg is essentially useless for paediatric epilepsy, which is a major use for this drug. You claim "the external reference price". That's just not true, James. The MSH can be a source of external reference prices, plural, and that 500mg tablet is weak in terms of data strength, in only having one supplier. It is "an external reference price", chosen arbitrarily, and the difference is huge and to say "the" is wrong. There are other prices, for syrups and other tablet sizes, but only the 200mg tablet seems to be widely available in the developing world. The 200mg tablet doesn't divide perfectly into 1.5, which itself is just an arbitrary average figure. Lets assume 1.4 for a real patient, which is 7 tablets. That's $0.4865 per day according to your original-research method, and $14.60 per month (or $16.68 for 8 tablets/day). And you don't have 2018 or 2019 data, because the MSH price guide appears to be discontinued. -- Colin°Talk 09:46, 10 December 2019 (UTC)
"The MSH can be a source of external reference prices, plural" I think there is a misunderstanding. When sources or I guess editors here refer to "MSH reference price" singular, they often/always refer to the median supplier price, which is the recommended methodology by the WHO/HAI.[3]: 215  An example is the Lancet paper.[4] The MSH median supplier prices are the basis that was used to build the subsequent WHO/HAI medprices database which uses them to compare drug prices internationally and between generics and original brand packaging.[5] The use of MSH median supplier prices as the basis for surveys, analyses or policies[6] is widely accepted, essentially any document mentioning the use of the "WHO/HAI methodology" is doing that [18][19]. --Signimu (talk) 13:36, 10 December 2019 (UTC)
Signimu, to pick the valproate example, you are right we can have an external reference price for 200mg tablet using the median of the five supplier prices. This is what the MSH guide suggest to use. We could, in negotiations with a supplier, also cite each of five records as examples of external reference prices. I have said this multiple times, and am pleased you say so because so far I've been a lone voice in the wilderness on that one. Our articles mostly use a lowest and a highest price taken from both the supplier and buyer price lists. Gasp with me in statistical horror. Sometimes there are no suppliers for a particular drug size (or at all) and only buyer prices. Sometimes there is only one supplier or one buyer. Sometimes the supplier is international but sometimes they only supply one country (e.g. DRC). We have consistently cited MSH as an authority of "developing world cost" even in this very suboptimal cases. But what the MSH can't give us, is an external reference price, singular, on valproate. I am not aware of any methodology approved by WHO that allows us to select which pill or syrup, or whether to use sodium valproate or valproic acid, or whether crushable or enteric coated. These all have their own price. It has been repeatedly claimed that the MSH data can give us an international reference price or an external reference price or a price in the developing world for "a drug". And of course, the next step after that is to conduct original research to get a treatment price. -- Colin°Talk 16:54, 10 December 2019 (UTC)
I agree, using the range of min-max without providing the median is meaningless, the median is more informative (and more commonly accepted metric) than simply the min-max, although the min is also used in policies (for obvious reasons of trying to pressure to drive prices down). Yes, mixing supplier and buyer price should not be done, but it's unclear which one we should prefer for an encyclopedia, maybe both have their place if it's clearly described along with the pricing. About I am not aware of any methodology approved by WHO that allows us to select which pill or syrup, or whether to use sodium valproate or valproic acid, or whether crushable or enteric coated, there is one, which is to cluster biosimilar compounds according to the ATC classification. The WHO/HAI database, and most databases and studies in fact, use the ATC or a similar classification to cluster products. Comparing prices between biosimilar ATC (what I mean by biosimilar is explained here[7]: 200 ) is often used by studies and policies alike to compare generics vs originator brands, and sometimes to compare prices internationally, and there is evidence that clustering biosimilar products allows to actually have a better representation of prices (so it is methodologically advised to do so),[8] so this may be a possible solution. Signimu (talk) 17:14, 10 December 2019 (UTC)
Signimu, I'll reply in your other post mentioning HAI. I would really really have preferred if wiki could have a bog standard source->text discussion like you and I are doing, where novel sources are offered, and advice about statistical approach given, and an agreement reached. We haven't had that to date, but it would help narrow down the choices of acceptable prices before any RFC on them. -- Colin°Talk 18:37, 10 December 2019 (UTC)
Extended content
  1. ^ a b "Introduction to Drug Utilization Research: Chapter 2: Types of drug use information: 2.6 Drug costs". apps.who.int. Retrieved 27 November 2019.
  2. ^ Measuring medicine prices, availability, affordability and price components (PDF). World Health Organization. 2008. p. 1.
  3. ^ Raju, Priyanka Konduru Subramani (2019). "Chapter 6.2 - WHO/HAI Methodology for Measuring Medicine Prices, Availability and Affordability, and Price Components". Medicine Price Surveys, Analyses and Comparisons (Monograph). Academic Press. pp. 209–228. ISBN 978-0-12-813166-4. Retrieved 27 November 2019.
  4. ^ Wirtz, VJ; Hogerzeil, HV; Gray, AL; Bigdeli, M; de Joncheere, CP; Ewen, MA; Gyansa-Lutterodt, M; Jing, S; Luiza, VL; Mbindyo, RM; Möller, H; Moucheraud, C; Pécoul, B; Rägo, L; Rashidian, A; Ross-Degnan, D; Stephens, PN; Teerawattananon, Y; 't Hoen, EF; Wagner, AK; Yadav, P; Reich, MR (28 January 2017). "Essential medicines for universal health coverage". Lancet (London, England). 389 (10067): 403–476. doi:10.1016/S0140-6736(16)31599-9. PMID 27832874.
  5. ^ http://www.haiweb.org/MedPriceDatabase/
  6. ^ Toumi, Mondher; Rémuzat, Cécile; Vataire, Anne-Lise; Urbinati, Duccio (2014). "External reference pricing of medicinal products: simulation based considerations for cross-country coordination" (PDF). European Commission. Retrieved 27 November 2019.
  7. ^ Vogler, Sabine (2019). "Pharmaceutical Pricing Policies". Encyclopedia of Pharmacy Practice and Clinical Pharmacy. Academic Press. pp. 188–201. ISBN 978-0-12-812736-0.
  8. ^ Danzon, PM; Kim, JD (1998). "International price comparisons for pharmaceuticals. Measurement and policy issues". PharmacoEconomics (Review). 14 Suppl 1: 115–28. doi:10.2165/00019053-199814001-00014. PMID 10186473.
Wrt the 90% figure "out of pocket". Firstly, when you were doing your Black Friday Christmas sales shopping, and the store said "Up to 50% off all prices", what does the "up to" bit mean? It's a get-out-of-jail-free statistic. Let's look at WHO's source, which is another WHO document that says "In many low-income countries in particular, private out-of-pocket spending accounts for 50%-90% of pharmaceutical sales.". So that would be 50-90% then. And these figures are more than 10 years old.
Secondly, "out of pocket" is a retail price. The WHO source says "Duties, taxes, mark-ups, distribution costs and dispensing fees are often high, regularly constituting between 30 to 45% of retail prices, but occasionally up to 80% or more of the total" But all our developing world costs for drugs are given as a wholesale price (and not even a well-defined one, see MSH for the two kinds of prices they have). If we are concerned to supply detailed dollars and cents "out of pocket" prices for drugs then here's what WHO says about MSH: "Medicine price indicator guides [they cite MSH] provide the sales prices from large wholesalers of generically equivalent medicines to governments. However, they do not give the price patients must pay in either the public or private sectors and often do not include new, essential but patented medicines."
That WHO document also says "Prices of the same medicines frequently vary between countries" which makes it hard to justify citing a price in Democratic Republic of Congo in 2014 or the price in Costa Rica in 2015 and claiming this is "the price in the developing world" (which we do). Wrt MSH it has been claimed "what we are using is an accepted international reference price" (singular). But MSH provide "international reference prices" plural. These are multiple prices that then require expert care to use properly. They explicitly state that the Buyer prices "should not be used as international reference prices", yet a Buyer price is frequently given in our articles. So MSH is a source of IR prices, plural, if you know what you are doing, that has some value to someone -- mainly the folk who cite them (plural) in order to negotiate a better price for their government. The closest we might get to "an international reference price" for a 100mg tablet, is to cite the median of the Supplier prices (which is what MSH recommend) but that only has value if the sample size is reasonable (perhaps half a dozen suppliers, and not just one). And that's just the price of a 100mg tablet, not "the drug" in general. The dose-cost ratio of higher and lower dose tablets varies considerably between countries.
Returning to the WHO document, and "out of pocket" prices. WHO's 2007 survey concluded "medicine prices are high, especially in the private sector (e.g. over 80 times an international reference price); availability can be low, particularly in the public sector (including no stocks of essential medicines); government procurement can be inefficient (e.g. buying expensive originator brands as well as cheaper generics); mark-ups in the distribution chain can be excessive; and numerous taxes and duties are being applied to medicines."
WHO give two examples:
  • "Salbutamol inhaler – an important medicine used to treat asthma – is virtually unavailable in the public sector of many countries (where medicines are generally cheaper or even free) and when purchased from the private sector, can cost the lowest-paid, unskilled government worker several days’ wages."
  • "The price of originator brand atenolol 50 mg tablets is over 20 times the international reference price in all the countries except India (where it is still high at 5 times the reference price) and Kazakhstan. Even the lowest-priced generic is very expensive in all countries, and there are some huge brand premiums, e.g. in Uganda the originator brand is about 13 times the price of the generic."
I'm not getting a good feeling that the IRPs bear anything other than a tenuous "less than" relationship with the "out of pocket price" that "50-90%" of people in low income might pay. -- Colin°Talk 09:02, 9 December 2019 (UTC)

References

Agreeing on WP:NOT

On the AN/I discussion, User:WhatamIdoing suggest that perhaps community feeling about WP:NOPRICES might have changed. However, look at the above discussion for the quote box in pastel shade. User:Levivich quotes the existing NOPRICES text and James replies "And I agree with that. All the sources being used to discuss prices are independent. And we have sources that justify the importance of pricing information". This is why I said above that it seems that two sides are talking past each other, and have different value systems that are framing their reading of policy. James thinks that raw database records in an MSH database (independent source) combined with general WHO statements that "drug pricing is important" (justified reason for the mention) together satisfy existing policy at WP:NOT for every drug article on Wikipedia. Most others here take the view that WP:NOT currently requires significant commentary about an individual drug, for example, price extortion due to limited generic suppliers in the US, or the high price of a barely effective cancer treatment not satisfying NICE that it is justified for use in the UK's NHS. So while we each continue to actually interpret current policy differently, it seems premature to suggest we have a vote to change policy. -- Colin°Talk 11:57, 7 December 2019 (UTC)

No evidence has been offered for statements about the community being divided on the application of NOT to pricing, and I think all RfCs show strong support for NOT. --Ronz (talk) 15:54, 7 December 2019 (UTC)
This, I think, is a question that could be settled through an RFC. Is it enough to have multiple sources that justify the importance of drug pricing in general, or do we need a source that justifies the importance of pricing for this specific drug?
There are broadly three possibilities here:
  • People think that sources about the importance of drug prices in general justify some prices for most/any drugs.
  • People think that sources about the importance of a class of drugs justify prices for most/any drugs in that class. For example, if sources say that the price of antibiotics is particularly important, then we add information to articles about antibiotics, but not to other drugs.
  • People think that (only) sources about the importance of prices of specific drugs justify prices only for the individual drugs named in the sources. For example, if the source says that the end-user price of aspirin is important, then we would write about end-user prices in the article on aspirin.
WhatamIdoing (talk) 07:39, 12 December 2019 (UTC)
WhatamIdoing, this is similar to the Q1 below. Firstly I don't think there is anything unique to medicines in this. Take the example I gave below about food. The price of food is hugely important both in terms of feeding one's family but also in production as it is an important source of income for much of the world. Indeed, for some foods, the price is commoditised and traded and speculated in markets. And the price of some foods is also of special interest. In Scotland, there is a minimum price for alcohol at retail, and a ban on multi-buy offers to try to address a drink problem. The price of healthy fresh fruit and vegetables vs junk food is also frequently discussed. In other countries, the price of the staple carb is key. The welfare and environmental issues surrounding meat are associated with its historically low price. And if one wants to, with Google, one can probably find documents discussing the price of any individual thing from the cost of Sainsbury's premium mince pies to Aldi's bargain Scotch. I don't see anyone campaigning that every food article should have "a price".
For example, every single drug in the WHO list of essential medicines will go through a proposal/approval process that involves cost effectiveness which involves some mind blowingly complex statistics to estimate the cost of treatment. James linked to one in the WHO archives for a group of three recent drugs; older drugs are probably not online. And every single new drug is assessed by the NHS and a price negotiated and the cost-effectiveness weighed in terms of whether it should be used vs older drugs. Is the notability of price in our articles dependent solely then on the happenchance that these memos and reports and applications are in an archive that Google can find and permitted for public access?
An RFC gauging opinions on all these thing may end up being largely pointless if simply subverted as seen in the above "I agree" to the existing WP:NOTPRICES. And I can't see your 2nd bullet "class" discussion being constrained to the format of an RFC if there are so many classes of drugs each with their own importance to someone. Elsewhere it was suggested for example, that a discussion of cheap/expensive drugs is likely more appropriate in an article on the drug class, and could then easily be sourced to secondary/tertiary sources. For example, at Anticoagulant: "NOACs are a lot more expensive than warfarin, after having taken into consideration the cost of frequent blood testing associated with warfarin" (unsourced BTW). This is a heck of a lot more useful than original research producing dollars and cents prices of each NOAC and warfarin and expecting our patients to compare for themselves, without the caveat that warfarin has hidden costs, and the efficacy and tolerability both also factor into cost effectiveness. This is why Wikipedia medical articles are built on top of expert knowledge and opinion, not our own editors doing their own research.
We have so greatly abused the raw-database online resources, simply because they are accessible and give an illusion of comprehensiveness, and this has led to the idea that we can put a price on any drug article. It is a mirage. Price is too complex for that, and this approach always requires original research. We need to take it a level up, stop the original research and presentation of essentially random numbers "accurate" to the cent to our readers, and offer them expert-sourced higher-level cost information. --Colin°Talk 08:56, 12 December 2019 (UTC)
Colin, your response is one that a debate teacher would admire – it hits all the points in contention, so that none can be struck as out of play for the next round – but if we want to settle things, I really do think that we need several small RFCs, instead of a huge, multi-part RFC. One of the separable assertions is that "Sources: Drug prices are important!" is enough to "justify" (in the sense of NOTPRICES) prices in articles about individual drugs. Other editors say that's not the case, and that you need to have "Sources: The price of aspirin is important!" to add the price to Aspirin. (There is the possibility of some middle ground, and since RFCs these days tend to be mere votes, I think it's worth calling that out, by way of encouraging people to identify where they land on the spectrum of possibilities.) My point is: even if you're 100% right about the other aspects, they don't have direct bearing on this question. We could actually settle the dispute over this question. WhatamIdoing (talk) 16:00, 12 December 2019 (UTC)
WhatamIdoing, well maybe that set of questions is approaching things from the wrong angle. Most people already think WP:NOTPRICES is your option 3. We already have WP:LEAD, WP:WEIGHT and WP:NOR all of which firmly disallow what is being done in our leads of drug articles. For some reason, these policies are being chucked in the bin. By asking 1 and 2, you are just setting fire to the bin. A better way is to look at what people are trying to say in articles and examine if that is appropriate per these non-contentious policies. You raise the question of what to ask, and this is being discussed far far below, so perhaps this conversation is better done there. There is an open request at the NOR noticeboard for input on the NOR issue in particular. Currently not a lot of input on that, which is odd since examining source->text issues is basic wiki stuff. --Colin°Talk 16:23, 12 December 2019 (UTC)
I think WAID proposition of asking this question (and with the proposed formulation) makes sense. We should separate this issue from NOR, as I will later and in a separate section show that it's possible to use each of these kinds of sources with 0 OR.
I can also provide some sources for each proposition:
  1. Drug prices are important in general: [20][21][22][23]
  2. Prices for classes of medicines: [24] review is quite comprehensive although not exhaustive, it covers generics vs originator, high-priced/orphan medicines, medicines for specific indications (such as HIV/AIDS, cancer, hepatitis C, cardiovascular, Alzheimer, etc), hospital medicines, non-prescription medicines.
  3. Prices for specific drugs: either use databases to simply get a price, or papers to get an interpretation (eg, comparison between countries, evolution over time, etc), this chapter and the following ones provide lists of several studies per geographical regions[25]. Signimu (talk) 16:42, 12 December 2019 (UTC)
PS: in my mind, if WAID's proposition is integrated in the RfC, it should of course be accompanied anyway with concrete examples, else it's just too broad and impractical. --Signimu (talk) 16:47, 12 December 2019 (UTC)
Well we are repeating the discussion below, where I already suggested any classification of rules really needs to include examples of articles that would be excluded/included by that proposal. I think you will find editors willing to argue any drug meets any "rules" you care to invent. I still have a huge problem with the concept that a drug has a price, in the general case -- our sources clearly suggest not. So would be very interested in how you escape the NOR issues to date. Btw, I can't read the book you link and it seems from the abstracts I read to only confirm that claims such as "the price in the developing world is..." are going to be rubbish. --Colin°Talk 17:00, 12 December 2019 (UTC)
Yes my PS was a callback to the discussion below I will go back there, I have finished reading new material to try to answer some of the questions you and others have pertinently raised. You can access all chapters using a website which name starts with sc Alternatively I can send it to you, it's really a great source material, we can mostly use only this source as a reference to help us make sense of the different possibilities, it contains pretty much everything about drug prices and is up to date (2019). --Signimu (talk) 17:22, 12 December 2019 (UTC)

Proposed questions for central RfC

Archived ANI leading to RFC

There appears to be consensus for a central RfC to settle this issue, with the RfC questions decided here.

please focus on the questions to be asked, not the answers you think the RfC should return.

Opening proposal

Q1. In articles on medicines, should pricing be included:

  1. Only where cost has been the focus of significant discussion in reliable independent sources (excluding trade press);
  2. For all drugs on the WHO essential medicines list;
  3. For all drugs.

Q2. Where pricing is included, should it be referenced to:

  1. Authoritative primary sources;
  2. Secondary sources only.

Q3. Where pricing is included, should it be:

  1. In the info boxes;
  2. In the info boxes only where global variation can be documented, otherwise in narrative text;
  3. Only in narrative text;
  4. Via templated external links.

Discussion

Please suggest additional questions and options for the questions, or modifications to the questions or options. Again,. please focus on the questions we should be asking the community, not the answers you think the community should return. Guy (help!) 11:40, 7 December 2019 (UTC)

  • As I note above, I think you need to clarify your first question to make it clear the "significant discussion" is about that drug and that drug alone. James appears to have interpreted the existing WP:NOPRICES requirement for justification and commentary on these details as being already satisfied in the general case for all drugs.
  • For the primary/secondary sources question, again I think this needs clarified. Is the MSH international price guide, the National Average Drug Acquisition Cost (NADAC) weekly reference data, the BNF online, and Drugs.com (retail) a primary or secondary source? None of them are the actual supplier/pharmacy website. So in that sense, their information is secondary. Some of them have processed the data (the NADAC is an average) and some haven't (the MSH and BNF list each supplier, though MHS provide a median, and BNF an NHS reference price). The Tarascon Pocket Pharmacopoeia price in $, $, $, etc symbols is perhaps even a tertiary source, though in my view, mass reproducing their symbolic "rough guide only" price (whether symbolicly or by writing $50-100), is likely to foul copyright on databases as these are not hard facts. James is now claiming Tarascon's prices are "inflated" so he doesn't want to use it any more. So I'm not sure we have any examples of what many of us would regard as secondary sources that could be directly inserted into articles with just paraphrasing and no original research.
  • The location question doesn't include "in the lead". Nor explain what detail. Perhaps some examples for each option would help. For example, many people might be very happy with the lead summarising economic information in the form of the words "cheap" or "expensive" or "very affordable" inserted into the description of the drug. They may however, not think that citing several $xx.xx prices in the lead is appropriate.
  • I don't understand the second infobox question. Perhaps examples would help.

Overall, however, I think an RFC is premature though certainly required at some point. Surely before the RFC we could get 6-10 articles (including, say, the drugs already discussed here and at WT:MED) to a standard that meets our policy (OR, V, NPOV, etc) and sourcing guidelines (MEDRS if claiming treatment costs). Those articles could then be used as examples of "Is this what you want?" If those wanting prices can't even come up with example text or mock infoboxes that satisfy the community on the basics, then we are wasting our time having a divisive vote on the issue. Perhaps we could invite people who are currently respected in reviewing such policy and article texts to examine the proposed pricing lead, body and infobox, and for us to reach a point that says "OK, if Wikipedia is going to include drug pricing in every article, this is the high-quality example you might follow". After all, policy and guideline should be based on best-practice, not wishful thinking. -- Colin°Talk 12:22, 7 December 2019 (UTC)

My suggestions:

Q4. Where pricing is included, what should be the geographical area covered?
  • Any country that sources can be found for
  • Only major economies, notably .... [to be agreed on]
  • Only countries or areas where prices of particular drugs have become a matter of significant public interest (e.g., due to controversies)
Q5. Where pricing is included, what prices should be used?
  • product cost (e.g., per package)
  • treatment cost (e.g., per day/month/year/treatment course)
Q6. If per-package pricing is included, what prices should be used?
  • Ex-factory prices (manufacturer's list prices)
  • Wholesale (distributor) prices averaged using the following method: ... [please propose]
  • Retail prices averaged using the following method: ... [please propose]
  • Effective prices per package or per treatment (i.e., the money effectively received by the manufacturer per drug package sold, after all rebates, discounts, free packages, risk-sharing arrangements, etc.)
  • Where effective prices cannot be calculated (e.g., due to bundle deals, budget capping or other RSS methods), should this be mentioned if confirmed by sources?
Q7. If US prices are primarily used, should they be accompanied by a note stating that US prices are not usually representative of prices in most other countries (as is the case according to many RS)?
  • Yes, always
  • Yes, if differences are indeed observed in pricing data
  • No

— kashmīrī TALK 12:28, 7 December 2019 (UTC)

  • The product/treatment choice needs to be a concrete one with examples that meet our policy. For example, if price per package (or per pill), then how does one select which pill and package size if a source offers many? James has, for example, claimed he chose the cheapest pill-dose to meet the Defined Daily Dose, but this also requires knowledge of how often the dose is taken per day, which also requires knowledge of which indication we are talking about, and requires agreement that the DDD is a reasonable measure to use [its an arbitrary value, contentious, and not available for all drugs]. In one example we had, the 100mg DDD would require a hard-to-get and oddly expensive 100mg pill in the US, whereas two 50mg pills would be much cheaper, provided you accept it is typically taken twice a day. None of this methodology is currently documented in the article -- why a particular dose of pill was chosen -- and there are strong grounds for arguing that this is original research. It gets worse of course when you go onto treatment cost. Again, none of our current prices-per-treatment (or monthly cost for ongoing therapy) indicate how this was chosen in the article/footnote and require original research to achieve it. A drug may be used for depression, anxiety and neuropathic pain, and we give a cost per month currently without explaining to the reader which indication that cost is for [there are no sources I have found, that indicate which condition the DDD is supposed to represent, and "the most common indication" may well differ from US to UK to developing world].
So offering the community a choice: Would you like cost-per-month for longterm use & cost per treatment for short use, as appropriate [again requiring OR to pick] begs the question: how do we achieve it. So I really think all of these question options have to be grounded in an in-article example, where some respected and knowlegable independent wikipedians, agree that it is reasonable to do this. Our RFC cannot rewrite the rules on original research or misuse of raw primary data, even if folk wanted to. -- Colin°Talk 14:23, 7 December 2019 (UTC)
  • Related to Q6 and Q7, in the US, insurance companies negotiate prices and pricing varies considerably depending on insurance. How relevant are retail prices in the US anyway, and if prices are included for the US, how do we factor in the price paid by most people per the largest insurance companies? How many consumers pay retail for medical products/drugs in the US? Let's take prostate cancer as an example. It occurs generally in older populations, so we get into medicare pricing and medicare supplement plans, and the issue that most elderly men with prostate cancer will opt for a medicare insurance supplement that covers cancer drugs/treatment. The variation in those plans is considerable. So what price do we use in this instance for the US? SandyGeorgia (Talk) 16:04, 7 December 2019 (UTC)
  • Related to Q6. Just as a heads up, there is an ongoing switch to a pay-for-results pricing model (also known as outcome-based risk sharing), especially in new super expensive medical technologies (including gene and cell therapies) – i.e., payments are due only relative to the health benefit observed in each treated patient. Simple "let's add some price to each drug article" will certainly not capture this; and by not capturing, it will falsify the real treatment cost. I am mentioning this to show how complicated the area of drug pricing is once one starts looking beyond common medications. That's also why I oppose the simplistic arguments on some "WHO publication" coming from Doc James et al. — kashmīrī TALK 17:18, 7 December 2019 (UTC)
    • See the "WP:OR at Diclofenac"/"Diclofenac in more detail" at WT:MED, where James is starting to appreciate there is more than one "wholesale cost" and has now decided the US list cost is "inflated" so would prefer to use a price after discounts/rebates. However, on the MSH "developing world" source, the Supplier prices James cites as "international reference prices" are supplier list prices, not prices negotiated after discount/rebates, etc, and MSH reckon buyers should add 10% to those list prices to cover transport and other costs. The UK's BNF lists two prices but I'm not expert enough to describe what they mean, and a given pill doesn't always have both prices in the BNF. So a question of "Do you want A or B or C cost figures" might not be achievable for all of e.g. US, UK, developing world. This is why I really feel it is up to those wanting to include costs in articles, to demonstrate that the options they want to ask us about can actually be achieved with the rules of policy. -- Colin°Talk 17:53, 7 December 2019 (UTC)
      • Yeah, I didn't really check all the sources, but from my own area, 10 x 2g oral tablets of salbutamol cost $55.47 in the US[27] and INR1.08 = $0.015 in India[28] - a 3,700-fold difference. Of course such differences would need to be captured, but do we really want Wikipedia to serve as a price comparison website? — kashmīrī TALK 18:05, 7 December 2019 (UTC)
  • Developing world the only source I've seen used so far for "developing world" is the MSH Price Guide. It is a respected source of raw primary data. Though its data comes from a number of suppliers, any given drug or drug pill size, might have few or no supplier prices. We have examples given where "the price in the developing world" was actually just the price paid by the government of Costa Rica in 2015, or the price of a supplier who only supplies war-torn Democratic Republic of Congo in 2014. But bigger than that is that it was updated yearly but has not been updated since 2015. I have tried multiple ways of contacting them and GMail has informed me that the contact email address given on their website is not responding. Perhaps someone in the US can phone them to enquire if there is likely to be a future update. Maybe they can't afford yearly updates and are moving to 5-yearly. I'm just guessing. As we enter 2020, I don't think Wikipedia should be committing to sourcing prices from "the developing world" to a source that will be already 5 years out of date and stale. -- Colin°Talk 17:52, 7 December 2019 (UTC)
It takes time to gather data. When I first began working with the source the data was from 2014. They have since updated to 2015. Doc James (talk · contribs · email) 02:52, 9 December 2019 (UTC)
The 2015 price guide was published in 2016, as one would expect, and "it takes time to gather data" does not appear to be the case. And it had been published annually since 1986. There has been no guide published since 2015. Something is up. -- Colin°Talk 10:03, 9 December 2019 (UTC)
From my readings, I understand that the WHO was tasked in 2001[29]: 3  to monitor more actively pharmaceutical prices and provide new methods to do so on a consistent basis, which led them to finance the MSH Price Guide in 2005[30] and to the WHO/HAI methodology and set of survey tools in 2008[31], geared mostly towards low-and-moderate-income countries as prices and availability infos are not centralized there[32]: 210 . For high income countries, the WHO rather pushes for them to create their own centralized systems, such as Euripid in Europe[33], which contains all drug prices of EU members but is inaccessible to the public unfortunately. Also of note, 2014 is the time when EU began to implement drug pricing policies systematically following this report: [34], and 2009 the first financial contribution of Euripid (following the first WHO/HAI report). So it may be that simply this phase of price monitoring by the WHO is ending, and is progressively shifting towards the countries themselves using these provided tools to make their own platforms, as is happening in EU. Signimu (talk) 13:55, 9 December 2019 (UTC)
Coming back to the fact that it's not updated since 2015, the WHO now lists the MSH update frequency as "regular" instead of "annually" before, and WHO/HAI database as "sporadically" updated. However, I don't think this is really an issue, sure it would be better to have more recent info, but it's not uncommon for economical metrics to not be updated annually, see for example the OECD pharmaceutical expenditures[35], the latest data for several countries dates back to 2015 and possibly before (didn't check all countries), and yet figures are made to present the expenditure "at 2018 or latest data", hence mixing up data from 2018 for some countries with earlier data for other countries. Signimu (talk) 18:32, 10 December 2019 (UTC)
  • Colin, the historical data is interesting enough as an external link IMO. See http://mshpriceguide.org/en/single-drug-information/?DMFId=38&searchYear=2015 (I use this for palliative treatment of C7 radiculopathy). The sharp trending increase is interesting to note. Guy (help!) 08:53, 9 December 2019 (UTC)
    Guy this is exactly why we don't allow original research and I wonder if MSH is really suitable for "the general reader". Looking at the chart on the link you give and the price sharply increases between 2014 and 2015. See 2014 data and 2015 data. Firstly both records only have "Buyer" prices. See MSH Price Sources for an explanation of Buyer vs Supplier -- Buyer prices are less interesting/useful. In 2014 we only have the price the South African Department of Health agreed to pay, which was $0.0226 per 10mg tablet when bought as a bottle of 100. In 2015, the same South African Department of Health agreed to pay $0.0197 per 10mg tablet in a bottle of 100. The price went down 13%. However the 2015 database also has a record for the price the Sudan National Medical Supplies Fund will pay. Strangely the record is for a single tablet, not a bottle of 100 and the price is $0.1330 per 10mg tablet. The median price (which MSH recommend we use, and they use for the chart) breaks down if your dataset is as tiny as this. It ends up being the mean of the two prices. So the "sharp trending increase [that] is interesting to note" is fake, an anomaly of the dataset and artefact of limited sampling. And in my experience a 1 tablet price record is kinda weird for a medicine taken daily for continuous therapy. So I wonder if there's actually a mistake with the data.
    That $0.13 a tablet price for 10mg is far higher than any price for 50mg and 25mg. In fact the 25mg record is far far healthier from a statistical POV because it has loads of supplier prices (and a handful of buyer prices). The median supplier price for 25mg is 0.0084 which is 9x less than the median buyer price for 10mg in 2015. Having records with limited data is quite typical of the MSH. If one cares about statistics, only the 25mg record is respectable. The 10mg record is clearly prone to sampling error due to limited data (and it has no supplier prices at all, which suggests it really isn't generally available). I hope you can see that doing original research on the MSH dataset is exquisitely sensitive to whatever arbitrary tablet size you use, and extremely prone to misuse (such as using both buyer and suppliers and presenting highs and lows which can be outliers).
    But Amitriptyline is interesting for a second reason. My UK GP explained to me that this is an old tricyclic antidepressant that they no longer use much for that, because although effective it is really nasty in overdose, and sedating. Instead it has found a new use, in low dose, for neuropathic pain, and there are newer safer drugs for depression preferred in the UK/US. The Defined Daily Dose on the MSH site is 75mg which corresponds approximately to the maintenance dose for depression listed in Drugs.com. In the article we give a cost [with many mistakes] "per dose", but we don't tell the reader "for what?". You and I might be able to reverse-engineer the price by searching Drugs.com/BNF for doses for various indications and working out that this dose is for depression, and not for neuropathic pain. I have not found any source that states which indication a DDD was calculated for. So I'm not sure there is any general way of using MSH as a source and giving a price "per dose" with the reader informed about what indication (illness) that dose is actually for. The "most common indication" may well vary from country to country. In the UK, amitriptyline is licensed for neuropathic pain. In the US, this is off-label. In the developing world, its very cheap price might encourage greater use for depression vs newer drugs than is the case in the UK. -- Colin°Talk 10:03, 9 December 2019 (UTC)
    Colin, all excellent points. But probably for the RfC. Why don't you and Doc James come up with a list of arguments for and against the three elements of Q1? This would certainly be an argument against any inclusion absent significant coverage in multiple RS. Since a Pro/Con list would not need you to agree on the underlying question, I would imagine you could do that without too much friction. Guy (help!) 10:22, 10 December 2019 (UTC)
    Guy, to be honest, even when I looked at Lenalidomide, which already meets NOTPRICES in having significant interest in its high price, the article conducts original research and gets the NHS wholesale price totally wrong and then juxtaposes that with a US retail price, which Sandy explained, is not actually price most people pay. Asking "do you want pricing" rather assumes it can be done, can be offered simply, and the choice is merely a matter of preference. James below seems to be willing to change his approach to how prices are presented. I suggest we tackle that, and then we can have examples to go with the questions. -- Colin°Talk 10:54, 10 December 2019 (UTC)
  • Threshold questions? Q1 as written above appears to be clearly answered by NOTPRICES already, so I'd suggest two substitutes for Q1:
    • Q1.1: "Does NOTPRICES require one source or more than one source?"
    • Q1.2: "Is the WHO source a sufficient source for NOTPRICES?".
    If the answer to Q1.1 was "more than one" and the answer to Q1.2 was "no", would that kill this entire dispute, rendering the other questions moot? Or would there still be stuff relating to prices in dispute? It seems like the answers to Q4–Q7 above are all going to be "follow the sources", and so the answers will entirely depend on what we deem are acceptable sources (Q1 and Q2), and Q3 (infoboxes, etc.) would have to wait until all the others were answered. – Levivich 20:12, 7 December 2019 (UTC)
I'm not sure I understand either of your questions. Number of sources isn't particularly relevant. What seems to be a problem is the incorrect interpretation of NOTPRICES that interest in drug prices in general is sufficient to warrant detailed drug price information in every article. It wouldn't matter if there are 2 or 100 articles on drug-pricing-in-general. Nor sure what WHO source you mean. I think we all agree that drug pricing is a matter of public interest, but that can be covered in, well, an article on drug pricing. -- Colin°Talk 21:08, 7 December 2019 (UTC)
James wrote above Being mentioned in a single reliable source is sufficient IMO. and pointed to the WHO essential medicines list. – Levivich 21:48, 7 December 2019 (UTC)
Yes, but we shouldn't necessarily frame questions round one idiosyncratic interpretation of policy, or read too much into terse replies. WP:WEIGHT of course requires editors to be familiar with the body of literature on the subject. This is something that requires one to study a subject, rather than flit from one article to another adding factoids. But WP:WEIGHT doesn't require editors to cite all that literature, indeed when editors do find themselves adding multiple sources it is usually a bad sign of contentious editing. To use the US insulin price increase example, this is well known to anyone who follows medical issues, esp. in the US, and only needs one to cite one great secondary source. However, if someone questioned whether this was indeed a story worthy of XX amount of article prose, one could, on the talk page, justify that with more references. And if none were found, then perhaps indeed, it was just one journalist and not notable enough. Using a primary source for insulin, as with all sources, depends on what you do with it. In a simple case, then giving a price "per 1,000 iu of NPH insulin" might be straightforward from primary sources. But it isn't simple, because the source has 10iu, 40iu and 100iu vials in different records, and buyer and supplier prices, and multiple prices that must somehow be extracted to something simpler. That requires a degreee of medical, statistical and economic literacy that is clearly lacking, and why we forbid original research. The more important point appears to be the incorrect thinking NOTPRICES is simply sastisfied by a general interest in drug prices, and never once citing a single source discussing the price, because database records cannot discuss anything. -- Colin°Talk 10:50, 8 December 2019 (UTC)
  • Type of page. I just came here from the ANI discussion, and I've read through the discussions so far. One additional issue that occurs to me is the relative desirability of including prices on individual drug pages versus on pages about drug classes. For example, it seems likely that there are some specific drug products, such as some of the high-priced specific-use biopharmaceuticals, where there has been a lot of independent coverage in terms of pricing. On the other hand, discussion of comparative pricing differences on drug class pages (example: H1 antagonist) might, perhaps, be more encyclopedic that listing individual prices on every drug page (examples: Diphenhydramine or Loratadine). I'd be interested in finding out what the community thinks about that. --Tryptofish (talk) 20:32, 7 December 2019 (UTC)
    • I don't think there has ever been any question that inclusion of issues that have been discussed widely and independently in sources can be encyclopaedic, whether at the drug level or drug class level. The contentious issue (and the discussion continues at WT:MED not just here) has been the egregious original research from raw database records in order to supply drug pricing in every single article, irresepective of whether our sources are actually finding the price of that drug a matter of interest or not. -- Colin°Talk 21:08, 7 December 2019 (UTC)
    • I think it's an interesting idea to explore – maybe it would indeed make our life easier to list prices in articles on drug classes if feasible. But this still won't get us past the issue that in real world, prices are set for commercial products whereas Wikipedia articles are predominantly about chemical molecules (that's also why we use INN as article title). I can't fathom Wikipedia listing all doses and formulations (tablet, ointment, inhaler, IV, etc.) and corresponsing prices across the world. That's not only impossible to achieve but also utterly pointless. Oh, and what about drugs that contain more than one active ingredient, like for example Fourderm [36]? Where do we stick in its price? — kashmīrī TALK 11:01, 8 December 2019 (UTC)
  • I agree with you entirely. I think there needs to be some decision-making about the most plausible options for those things, and then have the community decide among those in the RfC. And without that, it's an exercise in futility. --Tryptofish (talk) 20:22, 8 December 2019 (UTC)
  • Agree it is an interesting idea to explore, and one of the reason wiki prefers discussion to polling -- polling should be to confirm the consensus you feel is already established and based on clear examples. If people want "prices on drug class" or perhaps "prices on disease treatment section", then before we ask, I think there should be some valid examples to look at. For example, Wikipedia policy requires that if folk are comparing X's then we need sources that are comparing X's -- it isn't acceptable for us to do original resarch from raw data and present prices for readers to compare or even worse, for us to declare which is cheapest or best value [that's would require we take efficacy, tolerance, etc into account too]. Take the DDD, currently being used for original research dose prices. WHO explains how to use DDD and really it is not intended for this purpose and they explicitly state it should not be used to compare drugs within a class. This is why such comparisons should be left to experts and we cite the experts, not database records. So, if we have good secondary sources for drug-class price comparison, and someone can make a few typical-case examples, then it is a good question to ask. If we can't make valid examples, we shouldn't ask. -- Colin°Talk 12:11, 8 December 2019 (UTC)
    The problem with that approach, which usually is based on comparing drugs in a similar ATC class and is usually called internal reference pricing, is that it can lead to some biases whether we use per dose or per gram price, and is always unfit for international comparisons (only unfit when using DDD it seems) [37]. I'm not saying we should not explore such approaches, it could be very interesting, but we should first ensure we use the proper metrics for the proper purpose, I plan to post what I find after the discussion on prices encyclopedic pertinence. Signimu (talk) 22:07, 9 December 2019 (UTC)
  • For Q3 "Where pricing is included..." I'd like a question option something like 'Should the price in the lead be detailed specific dollars and cents for a variety of countries/areas or be simplified to e.g. "cheap", "inexpensive", "very expensive", "affordable"'. IMO this is more in keeping with WP:LEAD and is what a professional article on a drug might say, and I suspect we are only including the detailed figures because we only have primary data records for sources and and to simplify translation efforts that are of no concern to this encyclopeadia. I suspect the community, if asked, would much prefer or even insist on the simplified lead text. Can this actually be achieved in the general case? -- Colin°Talk 12:44, 8 December 2019 (UTC)
    Colin, that would be good, but where do we source it? Guy (help!) 08:54, 9 December 2019 (UTC)
    Guy I suspect a simplified statement like "wonderpam is a cheap antidepressant ..." is only valid for one country. So a US or UK publication could well say this. Whether we can say that is harder to say, particularly as our concept of "cheap" is distorted by people in the world earning less than a dollar a day. I think there are likely tertiary country-specific sources that say this sort of "knowledge", but may be hard to find a source that does this routinely. Our biggest problem with prices IMO is not just what we would love to say to our readers, but the extremely limited things we can say to our readers based on very primitive raw datasets. Already all the examples in articles require horrendous amount of original research. That's why I throw down the gauntlet on those who want to push for a "prices in all our drug articles" to be up front about what kind of prices, use a meaningful and consistent definition, and demonstrate you can do this routinely while obeying policy. As with WP:V, it is up to those wanting to insert text to justify their inclusion. -- Colin°Talk 10:21, 9 December 2019 (UTC)
    Such a concise qualificative summary would be nice, but I don't think possible. This looks like trying to qualify 'accessibility', which is not a directly observable factor, as it is influenced by prices, volume (availability) and socio economic factors (such as wage and general 'richness' of a country and its citizens), and probably others. The closest thing that I've seen is the 'affordability' metric by the WHO/HAI[38], which is the number of wage days the lowest paid unskilled government worker has to pay to "purchase 7 days’ supply of a medicine to treat an acute condition, and 30 days for a chronic condition, based on standard treatment regimens". Eg, for Metformin 500 mg cap/tab (sorry they don't provide direct links to specific entries), the affordability ranges from 0.2 day of wage to 14.6 days. This can be used for comparisons, either between generics and originators, or between countries (since affordability is always based on a local wage, no adjustment is necessary). However, I did not find any source that defined what is unaffordable, although the WHO/HAI manual says that 2 days is already considered unaffordable "for many low-income patients" [39]: 127 . Signimu (talk) 21:57, 10 December 2019 (UTC)
    Signimu, Tryptofish See post 22:07, 9 December 2019 by James: Example text. The link opens at "valproate", which is a drug we discussed. This is a UK book and the book says "Valproate is inexpensive (around £9 for a 100-pack of 500mg tablets). However, cost increases with the complexity of the formulation. So the book qualifies "inexpensive" with a sample price and gives a caveat. In a global encyclopaedia, we can't just say "Valproate is an inexpensive anticonvulsant" and cite a UK-centric book that has caveats. If you can't read the whole book, google the title and you may find a handy PDF of the 2015 edition, which I'm not going to link. Looking at the other "cost" examples in the book, it reads very much like a doctor->doctor advice on saving the NHS a few bob and, in the 2015 book, I felt the commentary on the cost at "warfarin" vs newer drugs was outdated (the authors are pharmacology registrars, not cardiologists or neurologists) This, cough, is one reason why I keep banging on about there being a problem with MSH database ending in 2015. Basically, if you look at the book (which only covers 100 most subscribed drugs in the UK), there are only a small number of cases where the "cost" is described neatly in one word. It is never as simple as claimed.... -- Colin°Talk 22:36, 10 December 2019 (UTC)
    If we explicitly mention this qualification is for UK only, then I think why not, but if not the problem I would have with that kind of qualification is that it is 1) not systematic since there is no explicit definition, 2) not generalizable to other drugs or countries (because of lack of explicit definition). The WHO/HAI database is certainly not perfect as you say, it's not complete and not updated as frequently as we would like, but at least they explicit (and validated) a standard method for each metric they provide. I'm not saying we should use it, I'm just describing possibilities Signimu (talk) 23:58, 10 December 2019 (UTC)
  • What kind of price stage and measure There are multiple stages of prices in the supply chain that can be used, and not all have the same availability, see external reference pricing which is the first article of a serie we can make about pharmaceutical price monitoring and policies. Then there are also multiple price measures, as pointed out elsewhere by Collin. There are unfortunately statistical intricacies, such as the cost per DDD being suitable for estimating consumption but NOT for estimating price nor doing cost comparisons, except maybe for internal reference pricing, see this excellent literature review[40]. And not all of these price measures, and stages in the supply chain, are available publicly. If we are to accept prices, we need to have a consensus on what stage and what measure we should use. We can choose multiple stages, but then it should be clarified in text. I think we still should research sources and discuss further about what measures of price we should use (the ERP article is a first attempt at that), so I think doing a RfC at this point would be premature. --Signimu (talk) 13:02, 9 December 2019 (UTC)
  • Current summary needed It feels like it might be helpful for someone to summarize current thinking on what the RfC questions are. If need be I will do this but I am hopeful given the talent of the editors involved that this won't be needed and we can see if there is consensus to launch the RfC. Best, Barkeep49 (talk) 02:53, 10 December 2019 (UTC)
    • Barkeep49, several editors have suggested an RFC is premature if the encyclopaedic text offered by the question cannot actually be achieved in the general case without original research and essentially arbitrary choices. The price for a drug can change by a factor of 15x or more depending on which database record is cited and what method used. There is even the more basic question that we are claiming a drug has a price (in US, developing nations, UK, etc) and, other than exceptional cases, it clearly very much doesn't. So really, what is the point of asking the community "should pricing be included ..." when the examples already discussed fail fundamental non-contentious policies. it really doesn't seem unreasonable to ask those wanting prices to demonstrate it can be done, and then we can have the community-consensus-opinion of whether we want them. If necessary, I can create a section that lists the sources currently being used, and how they are being misused in articles. Perhaps then the community can come to some consensus on a legal valid source/wording of prices that would be permissible if we are convinced they should be included. -- Colin°Talk 10:16, 10 December 2019 (UTC)
      Colin, I understand that is the contention of several editors. The idea that prices are original research has been a contention of those opposed to to pricing information all along. If editors who have been in favor of including pricing information and editors who have been opposed feel like headway is being made towards consensus, well great. Don't let me stand in the way. However, that's not quite how I'm reading recent discussions, and I don't think it's how Guy, who is another uninvovled sysop trying to help this along, is reading it either. I really do mean it that if editors on both sides feel like progress is being made and not just old arguments rehashed, then great. But absent that - and so far I admit I don't see this feeling from both sides - I think a focus on coming to agreement on how to structure what will be a fairly complicated RfC could be the best use of editor efforts and energy. Best, Barkeep49 (talk) 14:48, 10 December 2019 (UTC)
      Although I do not agree with Colin's behavior focusing on specific editors, I do agree that the RfC is premature as it would be too vague if we don't discuss/prepare first some concrete options of what kind of prices/pricing can be used, in other words to discuss a bit more about the methodology. I think we made a tiny bit of progress with Ronz's price/pricing concise distinction below, with pricing seemingly being more acceptable. Now more specifically I think we should discuss what kind of pricing is acceptable, I'm expecting everyone will have a different opinion ofc, but I plan to provide some concrete examples I hope will further our discussion and give us material to properly formulate an adequate RfC. Signimu (talk) 14:59, 10 December 2019 (UTC)
  • We would need to define "excluding trade press". I imagine you would be referring to medical sources here, like textbooks, review articles, and government websites? Doc James (talk · contribs · email) 19:21, 10 December 2019 (UTC)
    • Guy is your Q1.1 a reformulation of the existing WP:NOTPRICES? If not, what are you trying to achieve with any difference? Immediately above is a query on "excluding trade press". I should point out that at 18:37, 6 December User:Levivich quoted WP:NOTPRICES, to which the response was "And I agree with that. All the sources being used to discuss prices are independent. And we have sources that justify the importance of pricing information". Multiple editors disagree that WP:NOTPRICES permits the insertion of prices across all drug articles, though their opinion on the matter is dismissed. Your Q1.2 and Q1.3 rather assume your Q1.1 restricts the insertion of prices to a smaller or distinct set of articles. Yet Q1.1 does not specify whether the "cost" is the cost of that drug or drugs in general or that class or grouping of drugs. This matters because "a justified reason for the mention" has been argued to be satisfied merely by global interest in drug pricing. My point is that WP:NOTPRICES has been ineffective at preventing the addition of prices to all drug articles, and that if one's attempt is merely to restate it, then frankly what is the point? If you are trying to improve or clarify WP:NOTPRICES, then I think you really need to give some examples of what you think it would include/exclude, and get agreement from those have so far dismissed that WP:NOTPRICES is a restrictive policy wrt drug pricing, to agree that it will make a significant difference and agree with your examples of in/out articles. Without that agreement, we are simply wasting all our time. -- Colin°Talk 21:57, 10 December 2019 (UTC)
      Colin, I am trying to forestall attempts to override consensus by claiming that articles in Chemist and Druggist discussing the price of Pharmaco's new generic wonderpam don't get used to try to crowbar pricing in when it's not demonstrably a matter of significant public discussion. Guy (help!) 22:36, 10 December 2019 (UTC)
      • I have not seen trade press (by which I presume you mean pharmacy magazines) used to justify costs in drug articles. I have however seen the general interest in drug prices by WHO. I have seen the suppression of drug prices in US commercials. I have seen the claim big pharma want to suppress and censor drug prices. All these are used as reasons to justify NOTPRICES is satisfied. I don't think merely excluding "Chemist and Druggist " is going to work. -- Colin°Talk 22:43, 10 December 2019 (UTC)

Sample article please ?

This discussion could go nowhere fast if we don't have concrete examples to focus on. We will get dizzy from different dosages, generic and brand names, prices in different currencies and countries, and so on. And for example, in Trypto's post just above this, I did not realize he was proposing a question rather than asking for discussion, so the format here could get muddled.

If we had a mockup to work from, editor commentary here could be more focused and specific. I have suggested that an article like Epipen could be worked up to include pricing information that is fully compliant with WP:OR, WP:WEIGHT, WP:LEAD, WP:NOT, and WP:RS. Why do we not work that article up as one example, and then pick a drug from those that are currently contested (that is, based only on a database source), and work up the best possible mockup of what is proposed? If we have two samples, we could then analyze, side-by-side, what each of us suggests as questions based on those mockups, and it would be much more clear to the wider community when launched.

The idea of coming up with a question list is revealing that we are essentially planning to ask the community whether pricing data should be exempt from policy (NOT, WEIGHT, OR) and guideline (LEAD, MEDRS). Should we just ask that question outright? Putting forward an RFC to ask the community to basically exempt pricing from policy (or change policy to accommodate pricing) is going to be complex: we could at least present a well-thought out sample, showing our best effort. My impression from the text I've read so far is that we don't have a good sample that doesn't involve OR. Could we see the best shot at what is proposed? I believe that will make discussion much easier ... particularly when we take this to the wider community. SandyGeorgia (Talk) 03:30, 8 December 2019 (UTC)

Sorry that it was confusing. I was going by what Guy posted at the top of that discussion section. --Tryptofish (talk) 20:23, 8 December 2019 (UTC)
Tryptofish, the problem I am seeing is that it is unproductive for us to be suggesting questions until/unless we thoroughly discuss each issue and have sample sourced text to reference-- that is, I am concerned that Guy's format is not going to work here. Your suggestion provides a great example: we should explore this issue before proposing it as a question to the wider community. Again, I think we need a working sample. Colin has done utterly boatloads of work to uncover all of the problems here, and yet we still do not have answers to those issues, or a GOOD sample of a best practice proposal from which to work that uses sources correctly. JzG, I fear we will need to reboot the reboot, and set up individual sections to discuss each question, but do that only after those wanting to add prices can show us a sample where they overcome the voluminous issues Colin has uncovered. As of now, I fear we are not working towards something that can be presented to the community with a straight face. We are asking to change policy, but we haven't produced one good example to support that, and our sources do not support the text added in apparently any case (at least, one has not been offered as a sample).

I am also very concerned about the number of RFCs that fail because they ask too many questions. At some point, we have to gel this down. I suggest a new format is needed for this page, the facilitates discussion of each issue, and only when it can be based on side-by-side examples showing the goals and the issues. How, with a straight face, are we going to the community to ask for a policy change or interpretation without even presenting one article where cost is well added ? SandyGeorgia (Talk) 17:05, 9 December 2019 (UTC)

I think those are all good points. As a newcomer here, I was just trying to follow what Guy said, but I also did not suggest actual wording for an RfC question, because I'm unsure about those things too. I agree with you that we need to better understand what it is that we should ask, before we will be able to ask it. I also agree with Guy that, as we begin to workshop possible questions, we need to focus on the questions themselves, rather than on editors' opinions about what would be the right answers. We might, perhaps, want to consider a simpler RfC to come first, evaluating community sentiment on whether we should include pricing at all – sort of like just Q1 and Q2 above at #Opening proposal, probably with further revisions. (Although I'm sensitive to the issue that editors won't be able to answer broad questions about pricing until they understand what is being proposed about how the pricing would look on the page, I think that those two basic questions can be answered from the start.) Then, if the community puts boundaries on what is or is not acceptable, we could work on a follow-up RfC that explores the more detailed options. --Tryptofish (talk) 20:45, 9 December 2019 (UTC)
I would be strongly opposed to running "a simpler RfC to come first, evaluating community sentiment on whether we should include pricing at all" because essentially we had that in 2016 and it contained the same emotive pleas that "pricing is important so our drug articles should contain prices" and the same unfounded claims "we have great sources". Voters were asked to give an opinion on prices without a clear idea of what price. I mean, a lot of us might support "XXX is a cheap anticonvulsant" or "YYY is an expensive cancer drug" [assuming sources for that, and worldwide applicability of that claim] but have a big problem with the dollars and cents approach adopted so far. There seems to be an emerging consensus on this page that the pricing => prices logic is not justified and that we should encourage the improvement and expansion of articles on the issues of drug pricing. I hope I have managed to convince folk here that the sources are unable to supply the sort of prices some want to offer our readers, and that the concept of "a" cost or price for "a drug" is, in the general case, a mirage. So let's not repeat past mistakes. Our policy and guidelines are best built upon a foundation of best practice. Where is the "best practice" for drug prices in articles? I see plenty "awful practice; don't do this folks" examples. Learn from the mistake of Brexit: there was a slim majority for "Yes" but then it turned out there was no consensus of what kind of "Yes" folk actually wanted or could be achieved in practice.
Since the suggestion of an RFC was raised, I have only seen further tangential pleas about the general importance of "drug pricing", and no effort whatsoever to solve the problems I've uncovered already with "drug prices". I suspect strongly it is because the whole thing is busted and they know it. -- Colin°Talk 22:08, 9 December 2019 (UTC)
I agree we should not ask questions that appear to offer a get-out-of-jail-free card to exempt editors from established non-contentious policy. ALL the current examples looked at fail abysmally in terms of WP:OR, WP:V, WP:WEIGHT, WP:LEAD. There was an attempt at WT:MED to get editors to give examples for valproate but framed as "How should we summarize this source" and cited a database record of one particular pill size that conveniently happened to have only one supplier (a reasonably international one at that). Instead how should the cost of valproate be described to our readers, and how would you achieve while obeying policy? The problem with Epipen is that we can probably find lots of discussion on the price and secondary sources for the price and it already meeds WP:NOPRICE in terms of the notability of its price (in the US anyway). The article economic details are also a bit of a mess and I think that will only complicate any discussion about it. The contentious issue is whether we can do this for all/most drugs. The examples looked at so far are:
  • Ethosuximide -- the leading treatment for childhood absence epilepsy.
  • Diazepam (aka Valium) -- leading benzodiazepine for anxiety, alcohol withdrawal syndrome, muscle spasms, seizures, trouble sleeping, and restless legs syndrome.
  • Diclofenac (aka Voltaren, etc) -- NSAID for pain and inflammatory disease. Also has controversial veterinary use.
  • Carbamazepine (aka Tegretol) -- anticonvulsant and neuropathic pain.
  • Valproate (aka Depakote, Epilim) -- for epilepsy and bipolar disorder and to prevent migraine headaches. One of the most broadly effective anticonvulsants.
  • Mebendazole -- parasitic worm infestations. This one does have potential for notability on its US price and only its US price. This is simpler than epipen. If, for example, we wish to compare US vs UK prices, how would we do it under policy rules.
Perhaps other editors have other examples they'd like to see. -- Colin°Talk 12:38, 8 December 2019 (UTC)
Another example is Amphotericin B. Signimu (talk) 23:41, 8 December 2019 (UTC)
I will try to express myself more clearly. We cannot expect and should not ask RFC or talk page participants to scroll through long discussions and sort out a multitude of issues. We need to present concrete examples demonstrating the issues and summarizing them in a digestible format so that a new reader to this page can easily view, review and check vs. sources.

When we discuss OR, WEIGHT, etc, we have to discuss them relative to sources, what the text says, what the source says, etc, in a way the new participant can easily engage. Colin has thoroughly analyzed a number of articles and found numerous problems that are already detailed on this page: one of those as a sample needs to be summarized in a visually digestible format such as a table. We have already established there are many: we just need one that is typical and can be summarized.

Colin, could you pick an example and do something like what I did in sandbox so that any reader coming to this page can clearly see the issues? Could Doc James or QuackGuru pick a typical drug that you believe is not affected by the issues Colin mentions and do the same? Once you settle on a table or format, then we can give maybe six examples or something like that. Then, for example, Trypto could give an example of what typical text for a drug class would look like in the same visual format. And we could see a sample of what a policy-compliant drug cost (eg insulin or epipen, for which the sources do the work for us) would look like in the same format. SandyGeorgia (Talk) 04:56, 9 December 2019 (UTC)

Moving my sandbox content to here, so it can be deleted from my userspace. SandyGeorgia (Talk) 20:02, 22 December 2019 (UTC)
Extended content

Valproate [41]—(Depakote, Epilim)—for epilepsy and bipolar disorder and to prevent migraine headaches. One of the most broadly effective anticonvulsants.

Article text Source Source text Notes
The wholesale cost in the developing world is about US$0.40 per defined daily dose as of 2015. "Sodium Valproate | International Medical Products Price Guide". Put here exactly what the source says Put explanation here of concerns about OR, WEIGHT, etc. and limitations such as those discussed on talk (as in, only one data point supplied for one country, and so on.
In the United States, the National Average Drug Acquisition Cost is roughly US$1.30 per day for the short acting formulation as of 2019. "NADAC as of 2019-11-27 | Data.Medicaid.gov". Centers for Medicare and Medicaid Services. Put here limitations related to US insurance, etc.
The price to a consumer in the United States for this amount is about $US2.30 per day as of 2019. "Valproic acid Prices, Coupons & Patient Assistance Programs". Drugs.com.</ref>
The long acting formulations are more expensive. "NADAC as of 2019-11-27 | Data.Medicaid.gov". Centers for Medicare and Medicaid Services.
In the European Union, end-user costs are less than 0.60 EUR for an average daily dose in Germany. Regular pharmacy price, including all taxes, et cetera: less than 34,43 EUR for 200 controlled release pills with 500mg each; date: 2016-11-30[citation needed] No source is given.
Sandy, wrt your sandbox, I think you underestimate how much can be wrong with so few words and numbers in article text. I've taken apart several examples and examined several sources and explained their limitations, and I'm not sure there is a way to do that in just a few words. At this point I feel strongly that every single example of drug pricing we have examined breaks so many rules, whether wiki policy or elementary statistics, that I think the ball is in the court of those who want to include prices. Over to you guys. Mock up these example of how you think it should and can be done, and let the community judge. If you can do this, and satisfy the community, then we can vote on whether we want it or not. I have serious doubts that MSH, Drugs.com, BNF, US NADAC databases of raw price data can be used to present prices in a way that has currently been done, or that people want, without breaking WP:NOR and without being plain incorrect to our readers about our claims. If people really do want a single place to see the criticism of a source and how it is used in an article, then I could do that in a sub-page, say. But if e.g. we start to agree that MSH is a non-starter, then I don't want to waste my and your time discussing how badly it was used. If we still don't agree that MSH is causing OR on a massive scale, then sure, I could write up something. -- Colin°Talk 13:43, 9 December 2019 (UTC)

One of the questions that could be usefully asked is what an encyclopedic presentation of pricing/prices/costs would look like. Imagine that the sourcing question is solved (which it isn't, but that's a separable question). Which of these sounds most like a Wikipedia article?

  • Compared to other treatments for the same condition, Wonderpam is inexpensive.[1][2]
  • Wonderpam's high price makes it unaffordable to about half the people in the world.[3][4]
  • The worldwide median National Average Drug Acquisition Cost (or some other specifically named price) of Wonderpam was approximately US$5 for a 30-day supply in 2015.[1][2]
  • Wonderpam's manufacturer shipped 30 million pills and reported a total net revenue of US$5 billion in 2015.[5][6]

(Yes, these statements are contradictory. They're meant as examples of writing style, not as examples of facts.)

Assuming that you had the perfect sources in hand, and they were saying exactly this (only in slightly different words, because copyvio), and that nobody disputed the idea of including something about the financial aspects of Wonderpam in a Wikipedia article, which of these is the kind of thing that you would want to write?

I think that editors who aren't willing to go through the complicated stuff about how the pharma world works could still share a valid opinion about whether the best practice (assuming ideal sources, etc.) focuses on relative costs (it's inexpensive), access (it's unaffordable), wholesale costs ($5), manufacturer's revenue, or something else. WhatamIdoing (talk) 23:43, 14 December 2019 (UTC)

WhatamIdoing, I don't think your first, second or third sample sentences are problematic wrt should-we-shouldn't-we, although none of us can override WP:WEIGHT (though there has been an attempt to wrt global-interest in pricing => prices in each article). If our sources think the relative cheapness of wonderpam compared to its peers is relevant, of course we can include that. But that doesn't justify for example, us additionally doing OR and inserting random numbers into our articles based on our own made-up-and-usually-wrong treatment costs. The first three samples all make global statements, which can be problematic. While a relative-cost statement might be ok, I think any absolute claim that X is inexpensive is likely to be country-specific. There is plenty evidence from WHO that drugs we might consider inexpensive simply aren't available though national health services in the developing world, leading to extremely high private retail price. This makes a nonsense of the claim that a portion of the world are paying out-of-pocket => we should include wholesale list prices in our articles. I don't think we can have all discussion purely on an "assuming perfect sources, saying exactly this" basis because we do live in the real world. The fourth bullet point is likely to come from a press release so we must be cautious not to repeat self-reported claims. I assume figures that come from a company annual report to shareholders have a legal obligation to be correct, so that might be better. I haven't really looked to see if that information is generally available. It probably isn't for drugs that don't affect company bottom line, and likely to be a commercial secret. The third point... well that's the problem because we simply don't have sources doing that because it doesn't make sense to do it (in the general case) without being much more specific. I think there has been a naivety / deliberate over-simplification to suggest that in the general case drugs have prices. They really don't. -- Colin°Talk 13:05, 15 December 2019 (UTC)

Disputed

Guideline protected. Guy (help!) 10:23, 10 December 2019 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

What is currently in WP:MEDPRICE[42] is disputed. The current text was edit warred in, and does not reflect consensus. I don't edit war, and don't want to revert to older consensus version, but we should not leave the impression that this text has consensus. It seems the best interim solution, as we wait to formulate an RFC, is to add a disputed tag to that section. Do we now have neutral admins following this page, and is it acceptable for me to tag that section? We cannot leave the impression that this is consensus text during RFC formulation and execution. SandyGeorgia (Talk) 19:44, 8 December 2019 (UTC)

I think the best option is to blank it. It was added in this edit in October by QuackGuru, a diff that I very much encourage others to read as it explains the agenda being pushed. It was blank before the article edit-warring that sparked this renewed debate began, and in the interests of neutrality, should be blank again. -- Colin°Talk 21:58, 8 December 2019 (UTC)
The text was: When including content about product pricing an independent source should be used. The pharmaceutical industry has tried to conceal medication prices because of their continuing legal cases in the United States. There is a lack of transparency regarding medication and vaccine prices among non-governmental organizations. SandyGeorgia (Talk) 02:24, 9 December 2019 (UTC)
So there's no discussion, and a clear BATTLE behind it? Unless there's some non-disputed version along the way (I'm not seeing one), blank it. --Ronz (talk) 22:08, 8 December 2019 (UTC)
Yep, there has been no discussion beforehand. QuackGuru amended the manual first and only then posted on Talk. That's already bad. Another issue is the text itself: suggesting that patients' "lost wages" are a component of drug price (sic!). Man, this guy should start reading before writing for an encyclopaedia. — kashmīrī TALK 22:45, 8 December 2019 (UTC)
There was a discussion back in October. Changes made without a RfC may be against the closure at AN/I. There was a consensus among editors back in October for a rewrite. QuackGuru (talk) 22:49, 8 December 2019 (UTC)
It wasn't WP:CONSENSUS as you can see - the discussion is right above and there were several voices who did not agree with it. That's why we are having this RfC by the way - because there was no consensus. The closure at ANI prohibited adding drug prices to articles. — kashmīrī TALK 22:54, 8 December 2019 (UTC)
The main discussion, which this is a subsection, was started about adding pricing information to this guideline. (Wow, we've been here a while.) The RfC we're now forming is to determine what will be included. Starting with a clearly disputed section is probably a very bad idea. --Ronz (talk) 23:17, 8 December 2019 (UTC)

Noting that I have fully protected this MOS page pending the completion of the RfC. All are also reminded that the community, including I believe nearly every editor in this subsection, expressed broad consensus for The above debates will be subject to civility restrictions with strict enforcement of WP:AGF, WP:CIV, no WP:BLUDGEONing and no rehashing of grievances. (emphasis added). Best, Barkeep49 (talk) 23:32, 8 December 2019 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Proposal for resolution

These discussions have been going on for several months, and although we are now actually discussing and trying to develop consensus, the issues revealed are increasingly complex, and we still don't have a best practice working example. So, I want to float an alternate proposal in the hopes we can find a less burdersome way to resolve the entire matter.

In some cases (samples Epipen and insulin), we have no problem discussing costs and pricing without breaching policy and guideline, as the MEDRS-compliant sources do the work for us.

In other cases, we are relying on database-type information, and having to do the math which involves a multitude of problematic factors, discussed at length above and which will be quite cumbersome for the wider community to digest.

What if we were to simply provide an External link template for each of the data sources that have been employed so far, and agree to add them to External links on drug articles, and let the reader do their own original research? Then we would also need to agree that only compliant text would be added to the article, that is, subject to WP:V, WP:MEDRS, WP:WEIGHT, WP:LEAD, WP:NOT, etc.

And if we were to come to consensus to do this instead, would that satisfy the ANI closers although avoiding an RFC? We are short on medical editors, and all of us could be more "gainfully employed" here improving content, so my hope is that we can find an alternate solution that gives our readers more access to price information, while not breaching Wikipedia policy. SandyGeorgia (Talk) 17:40, 9 December 2019 (UTC)

SandyGeorgia, that's one of my options above. I am OK with proposing just this (include where primary subject of multiple sources specifically about that drug; external link for WHO essentials) rather than multiple options. Guy (help!) 18:37, 9 December 2019 (UTC)
For the databases, the HAI did the work for us: multi-countries databases list[43] (+ the HAI one: [44]) and national databases list[45]. But I don't think this will fix the issues. There are real methodological issues, even if we just take the prices straight from the databases, they are not necessarily comparable because they may originate from different stages in the supply chain (see external reference pricing where this is clarified). The HAI somewhat simplify this issue as they specify the supply chain stage for each price database. I think we should further study prices before launching a RfC, the creation of the external reference pricing article is a first attempt at doing that, but we can do more with the references I've lately found. Signimu (talk) 19:05, 9 December 2019 (UTC)
I think we need to have an RfC in any case (in other words, to not avoid having one). There really needs to be a community consensus about some of the most basic questions about whether or not to include pricing on all drug pages. --Tryptofish (talk) 20:50, 9 December 2019 (UTC)
With respect to prices for LMIC, I consider the sources we are currently using to comply with WP:V, WP:MEDRS (which they do not necessarily have to as this is not a health claim), WP:NOT, and WP:WEIGHT. With respect to having these details covered both in the body and in the lead that is easy to comply with, and I am happy to do so.
With respect to the wording around how to summarize this source I am more happy to discuss.[46]
Currently we have "The wholesale cost in the developing world is about US$0.40 per day as of 2015."
If people feel this is overly summarized we could go with
A external reference pricing in the developing world is about US$0.40 for a typical maintenance dose in adults as of 2015."
Would be interested in hearing other peoples suggestions.
We also have sources such as this one we list it as "inexpensive".[47] Would also be happy with simple that in the lead with the rest of the details going in the body.
Not supportive of simple adding an external link.
Agree we have a lot of other stuff that needs doing such as updating epidemiology in a few hundred articles in 2017 data.[48] Doc James (talk · contribs · email) 22:07, 9 December 2019 (UTC)
Tryptofish, yes, absolutely. That's the consensus from the ANI thread and is also the best way to definitively resolve the underlying dispute. Guy (help!) 10:18, 10 December 2019 (UTC)
Well this is the first time James has accepted that his approach to listing detailed prices in the lead should end. Yay for WP:LEAD compliance. He cites a UK textbook that interprets "inexpensive" in terms of NHS drug pricing and socialist healthcare, which, as we have seen many times, is not reflected in US let alone developing nation pricing -- we have quotes above from WHO that drugs with low wholesale prices may be unavailable through public health and only on private prescription charges, which can be many many multiples of the wholesale price and most certainly not "inexpensive". Unfortunately the various suggested article texts above are simply (a) not supported by the source and (b) still require original research. And currently I'm not seeing suggested texts by anyone other than one editor, yet several people have previously expressed support for prices. This should ring alarm bells. We only have text offered by one editor that has already multiple times shown to break even the most basic non-contentious policy. Wikipedia is a collaborative editing project, which means I expect to see a consensus of editors offering examples of their wonderful drug prices in articles they wrote. I don't. At the moment, a premature RFC would effectively giving carte blanche to one editor to do as they wish. That doesn't sound like how a collaborative editing project works to me. -- Colin°Talk 11:57, 10 December 2019 (UTC)
"And currently I'm not seeing suggested texts by anyone other than one editor, yet several people have previously expressed support for prices." That's not entirely true, other editors (such as me) may be working on the issue in other ways, such as by reviewing the literature and expanding articles such as external reference pricing and median price ratio. For an example of a possible basis of prices, see my reply above [49], although please note that I plan to provide more content for other editors to evaluate as I wrote elsewhere. Signimu (talk) 13:52, 10 December 2019 (UTC)
Colin, As noted by Barkeep49, please restrict your comments to the specific issue of what questions should be asked at RfC and do not comment on other editors or you may be sanctioned. Guy (help!) 14:17, 10 December 2019 (UTC)
Guy, Barkeep49, Any questions in the RFC must contain examples of what the proposer means: article text + sources, or a worked example. And if you want to ask a question about "Authoritative primary sources" or "Secondary sources only" then you'll need to give examples of those because editors may disagree on what you mean. And of course we can't discuss sources without an example of what you propose is drawn from them. We know quite well in MEDRS and elsewhere that whether a source is permitted is not just the subject-domain but what exactly you are saying about it. So it isn't fair to ask us to say primary/secondary without an example of what you are using the primary source to say or what you are using the secondary source to say. Same for product cost and treatment cost. I don't really understand how, in good faith, an admin on Wikipedia can propose to ask the community if they want to see treatment costs in articles, say, without a working example of an article with treatment cost and an agreement that this was achieved without breaking other policy that it is not within your power or remit to change. I see it as a bit futile to run an RFC with a question + example that will instantly be rejected because of obvious shortcomings with the example. So I really really would like to see more than one editor participate in doing that. Perhaps you want to post a notice at WP:MED asking for further contributions to examples of great article price text. -- Colin°Talk 16:37, 10 December 2019 (UTC)
I can give a few examples of primary/secondary sources issues:
  • The most primary level of sources for prices would be databases, but there are some particularities. The MSH International Medicines Price Guide for instance allows to write things like 'the median supplier price is $XXX'[50]. The WHO/HAI database[51] builds on the MSH guide and provide higher-level price metrics, such as the median price ratio or affordability. Of note, some of these databases are accompanied with manuals that also provide some analyses (eg, the WHO/HAI methodology and database manual[52], see figure 1.1 and table 1.1).
  • Papers analyzing the databases, such as those using the WHO/HAI methodology[53] or MSH price guide[54]. We then get higher level infos, ranging from "prices of originator medicines and generics are respectively 21.33 and 11.53 times the international reference price in Boston area"[55] to the availability and prices of cardiovascular medicines in 36 countries[56][57].
  • Reviews such as [58][59][60][61], but at this level they usually only pertain to the effect of policies on prices, but sometimes we get some specific infos when it's used to illustrate a point such as "Public sector data from 13 LMICs showed that the mean public sector price of human insulin (100 IU isophane/regular 70/30) represented 0·7 to 6·2 days’ minimum public sector wages for 1 month of treatment" or "Private pharmacy prices for four commonly used cardiovascular medicines in 18 countries showed that they were potentially unaffordable for at least some patients in every country [...] 0·14% of households in HICs, 25% in upper-MICs, 33% in lower-MICs, as much as 60% in LICs (excluding India), and 59% in India"[62].
So can we consider papers studying databases as enough for citing prices? If not, we are left with review sources, which are mostly focused on a high, policy level view.
Worth mentioning is this excerpt from a chapter from a specialist on how to choose an appropriate source for medicines prices: "In general, literature does not qualify as appropriate data source for price surveys, analyses, and comparisons. As the price studies in Part 1 also indicated, primary price data were published in only few papers, and they were usually focused on a low number of countries and medicines. Despite its limitation as data provider, scientific literature can prove to be useful in offering additional (background) data for the analysis and interpretation of prices (e.g., a global price survey on sofosbuvir and ledipasvir/sofosbuvir was based on the collection of medicine price data from national sources, but a discount mentioned in literature was considered for the calculation of discounted prices)."[1]: 257  In other words, if we use the databases directly, we have a more uniform and standardized data source, whereas with literature we get more interpretation but less data. That's a trade-off we see with pretty much any other metric in fact. --Signimu (talk) 18:15, 10 December 2019 (UTC)
Remember that advice for academic research and publishing does not necessarily align with Wikipedia sourcing best practice. Academics are taught to cite the original research paper whereas we are taught to cite the review. Of the sources used so far, MSH, BNF, Drugs.com all operate at the level of low-level data. Whether they are strictly "primary sources" is an academic debate (clearly they aren't the drug companies or suppliers own list prices, so their data is second hand). I'm only just now looking at HAI and so tried searching for some of the drugs that we've discussed. For example "valproic acid" and "amitriptyline". For the former I get 500mg "Latin America, lower-middle income December 2009" or for 200mg I get "Morocco April 2004". For the latter I get 25mg in loads of regions. I tried "Sudan March 2012" and got nothing for "Public sector procurement prices ", nothing for "Public sector patient prices " but data for "Private sector patient prices" And then lots more numbers and percentages I don't yet understand. I'm seeing two things about this (and the MSH it builds on) that (a) the data is often very very limited and (b) doesn't go past 2015 and (c) is still for a particular tablet at a particular mg. So while the HAI may have some additional metrics, it still doesn't solve the problem of original research to pick which pill-size. Neither let us write "the median supplier price is", we have to say "of a 500mg tablet is.." And then we might need to be honest with the reader if that isn't a median of 5+ records, say, but is one record from one supplier who is only supplying one country that was in civil war in 2014. All the pricing so far suffers from gross over-extension in terms of claim vs what the source actually honestly tell us.
"Papers analyzing the databases" have names like "A Survey of Medicine Prices Availability, Affordability and Price Components in Shanghai, China Using the WHO/HAI Methodology". WP:MED readers will immediately recognise such as primary research papers. Looking at a couple gives me the shivers. I don't think they are useful on drug articles or even on drug-pricing articles. But let's pause a second and consider that this "primary research" is exactly what has been happening with all our drug articles to date. We have been doing original research on the raw data.
The reviews you cite seem to deal mostly with public health and I agree that any specific information they contain is likely to be a side-issue wrt the conclusions/focus of the paper, and filled with caveats.
The more I see of this sort of thing, the more respect I have for a field of researchers trying to make sense of an extremely complex topic and using their university degrees and PhDs and extensive professional experience in this domain to produce good research that moves health onwards. It is unfortunate that none of them seem to be editing wikipedia and could advise us the errors of our ways.
I don't think we can use the databases nor the research papers that investigate those databases. That's just too low level and requires original research. Maybe you could search more to see if you can find higher level sources that are closer to the sort of thing some people want to include in our articles. While there may be academic advice of methodologies, that doesn't mean editors here are allowed to do it. -- Colin°Talk 19:17, 10 December 2019 (UTC)
A focused RfC is certainly preferable to an open ended RfC. An RfC where people clearly understand the options is obviously preferable to one which can be misinterpreted by participants and the closer. That's all great. However, people opposed to this change cannot say "find me an example I agree with before we offer it as a choice". The RfC is about what should be allowed. It is complex and the people who wish to include the information have the burden of convincing the whole community that what they propose is either compliant with other policies and guidelines or else that as part of the change they wish to amend those policies and guidelines. However, the people that need to be convinced of this is the community as a whole, not to the editors who have thoughtfully considered this issue and decided it's not appropriate. Best, Barkeep49 (talk) 19:25, 10 December 2019 (UTC)
Yes Barkeep49, that's exactly what I'm trying to do, to document the issue, so that others can evaluate and decide with more info. I agree for the need of a RfC, and as you say, with adequate and precise questions (and possibly examples), it would be much better I think. I'll try to summarize how the recent progress may pertain to the formulation of the future RfC. Colin: yes for the methodology, it's not our job, and the documents mostly pertain to conducting price studies, but it can help us navigate in the various methods available and avoid us from making bad choices (I hope). Signimu (talk) 21:22, 10 December 2019 (UTC)
I just saw this as a result of Signimu's subsequent post: However, people opposed to this change cannot say "find me an example I agree with before we offer it as a choice". Barkeep49, I fear there is a serious misunderstanding. I continue to ask for a sample not because I expect to agree with it (the methodology), rather to be fair in what sample we use to discuss the problems once the RFC is advanced !!! We should not pick apart one of the weakest drug article's pricing text; I want to know which is a drug article for which the sources used don't have as many of the problems as have been highlighted, in the interest of fairness. Do all of them have all of the problems Colin has highlighted? Do any of them have only a few of the problems? Is there any one that is better than the norm that will serve as a best-practice example?. We should use the best example the editors advocating for those sources have to put forward, so as not to be accused of having cherry-picked one of the worst articles as an example to highlight, and to avoid going through four, six or eight examples of all the problems. Bracket the problem by showing one of the best, and one of the worst when arguments are advanced in the RFC.

Had you not made that statement above, I would not realize that perhaps intent was being misunderstood here. (I also note that it was Signimu's closed/archived section below to led to a breakthrough in price/pricing, so again, let discussion run … we are making progress.) SandyGeorgia (Talk) 21:33, 10 December 2019 (UTC)

I am catching up here at the bottom of the last post re several items earlier on the page. I have to attend an unexpected and heartbreaking wake and funeral, and may be tied up for a few days, but will continue to weigh in as much as I can.

First, I do believe we are making progress, albeit of the "slowly but surely" type, perhaps not as quickly as some may want. While the adminning is greatly needed and very helpful, please consider not shutting down discussion that is finally happening after so many years of seeing !voting rather than discussing on this page. It is a complex issue, and that people are finally really engaging is, IMO, a very positive sign. A cultural switch to discussing rather than !voting is unlikely to happen in just a few days.

As can be seen on this page in the last few days, not all of the very involved parties have even had a chance to weigh in here, so we should remember that an RFC does not have to happen tomorrow. Hearing all voices, and getting all issues out on the table to give the RFC the best possible chance of reaching a useful conclusion is much more important than being in a hurry to put something up, that could be a debacle because of unexplored issues. While it seems like this conflict has been going on for a very long time, I suggest the real timeline started only a week ago, when everyone started engaging.

Another thing that can be seen on this page is that there is voluminous information to digest, and different editors are digesting that information (or not) at different rates. Yes, it is frustrating that this results in lengthy repeat posts, but that is part of the problem we must deal with. So again, no hurry. That we are seeing some progress in getting editors (not yet all, as I am concerned about those that have gone silent) to actually read the page, and digest and respond to concerns, is a very good sign. A better sign would be to begin to see less terse answers to the concerns being raised, to help avoid repeatedly asking the same question(s).

As to getting back to organizing the actual RFC, I am more concerned than ever that we not launch an RFC fraught with the kinds of issues that often crater RFCs. The idea of advancing specific question proposals was a good one, but the organization of that on this page is less than optimal. We need a name for the RFC and to begin to draft it in a subpage, with the idea of eventually moving it to a Wikipedia page. Next, we are risking having too many questions, which could be a killer; once we get to work, we can try to gel them gelled down. Also, the questions we are asking is only recently becoming clear even to many of us-- without an example of the problem that brought us here, it seems to me that the average editor coming to the RFC will miss the point, and respond with, well of course you can discuss prices following the restrictions and guidelines in NOPRICE. I continue to think we need an example that facilitates discussion of WHY we are asking what may appear to be a no-brainer question. We don't want to end up with a conclusion based on people not understanding what the dilemma is or what we are asking.

And finally, MANY of the questions we've advanced above are, considering the complexities here, oversimplified. All of the questions have been put forward in single sections, with lots of mixed-up discussion in each section. If we were to begin work on the actual text in a sub-page, we will begin to see what concerns remain and how the wording is coming. But I do not believe we are to the point of doing that yet: I do believe we are making progress. SandyGeorgia (Talk) 20:17, 10 December 2019 (UTC)

SandyGeorgia, my deepest condolences to you and your relatives, please take the time necessary for your mourning, that's certainly more important than whatever happens in Wikipedia... (and I agree about the slow but sure move forward) Signimu (talk) 21:25, 10 December 2019 (UTC)
Sandy, likewise from me. My deepest condolences and my best wishes. --Tryptofish (talk) 21:52, 10 December 2019 (UTC)
Thank you both very much. Not a family member, buta very close family friend; there is a bit more detail at Barkeep's talk. Thanks again, I will be following best I can here, SandyGeorgia (Talk) 22:01, 10 December 2019 (UTC)

References

  1. ^ Vogler, Sabine; Schneider, Peter (1 January 2019). "Chapter 8 - Medicine Price Data Sources". Medicine Price Surveys, Analyses and Comparisons. Academic Press: 247–268. doi:10.1016/B978-0-12-813166-4.00014-0.
Likewise, sending lots of good thoughts your way, @SandyGeorgia. — kashmīrī TALK 01:09, 11 December 2019 (UTC)
Thanks, Kashmiri; much appreciated. SandyGeorgia (Talk) 02:48, 11 December 2019 (UTC)

An alternative suggestion

I think it's very clear that, at some point in the future when editors here are comfortable with it, an RfC is going to be necessary. It's not optional.

After following this discussion for several days, I'm frankly very pessimistic that editors who are active here are ever going to agree on a series of questions (ie, should we use primary or secondary sources?) for the RfC structure. It's hard for me to envision a scenario in which there will be wide agreement about what any given choice, among multiple-choice options for a single question, even means.

So I want to suggest an alternative approach. Present the RfC respondents with a series of proposals, and ask them to choose among those proposals. For example, Proposal 1 might be to not list prices on every page, but instead discuss prices only when there has been a lot of independent discussion in sources – accompanied by reasons why editors think alternative proposals are no good. Proposal 2 could be to have prices in the infoboxes of every drug page, sourced to the WHO – accompanied by reasons why editors think this is a good idea. Or something roughly like that. And Proposal 3 would be something else, and Proposal 4 would be something else, and so on.

I think that most editors here either are able now, or expect to be able later on after more discussion, to say what they think the right solution is, and why. But they might not agree with what some other editors want to do, although they will be able to explain why they oppose those alternatives. Some editors might support Proposal 1, and oppose all other proposals; others will support multiple proposals while opposing just a few. I think it's feasible to develop such proposals, in whatever amount of time that will take, and maybe agree on some basic points about fairness regarding how each proposal is presented.

So the RfC would have the proposals listed by numbers, and there would be a !vote section below where each responding member of the community could indicate what they support or oppose and why, followed by a discussion section for more extended discussion. I think it would be good to have 3 uninvolved administrators determine the consensus at the end. Something like this worked surprisingly well at WP:GMORFC, in a subject area that was actually a lot more contentious than this one is. --Tryptofish (talk) 22:12, 10 December 2019 (UTC)

That sounds like a very reasonable and concrete way to formulate this RfC, plus it would allow to avoid the issue with presenting a combination of options that is impossible to follow (eg, lacking data). Please allow me to document a bit more what I read so that we can maybe weed out some more impossible choices. Signimu (talk) 22:32, 10 December 2019 (UTC)
And this is another of the reasons I am after a good sample drug: we can show under each of the proposed scenarios for that best sample drug what would be allowable. The complexity here may to be too much for average RFC (non-pharm) respondents to digest. And we are risking getting a non-meaningful result with all the complex questions. The other thing I like about this proposal is that it won't make it look like we are asking the community to endorse or reject what is already in NOPRICE. SandyGeorgia (Talk) 22:53, 10 December 2019 (UTC)
Ok, for example I would argue that generally there isn't "one price" for "a drug". For some there is. For some drugs that are highly notable for price issues, our best tertiary sources may simplify pricing and give "one price" for "the drug" (often as an example or rule-of-thumb or ballpark price). We could use those prices from those sources, as long as we caveat them as much as the source does. Pretty much all the existing examples of prices in our drug articles are based on cherry-picking raw database records from inappropriate sources, conducting extensive original research involving statistically/economically/medically unsound methods, and then presenting a statement to the reader. That statement either over-simplifies the facts or adds artefacts that misrepresent or is simply just totally unsupported by the source. We then compound this by juxtaposing incompatible prices without even informing the reader.
I suspect the response to my argument might be that the sources are the WHO and are offical international reference prices or respectable sources like the UK's BNF, and any mistakes that have been made can be fixed and would be more than happy to do so. Which, you know, is just 100% BS but that never stopped 100% BS from winning votes. The proposal 2 "prices in the infoboxes of every drug page, sourced to the WHO", is frankly making me lose the will to live. -- Colin°Talk 23:27, 10 December 2019 (UTC)
Sandy is right that another potentially useful aspect of this approach is that it pretty much forces everyone to make it specific. I had forgotten to mention that, but I agree. And I certainly don't want to make anyone lose the will to live – but of course I only wrote those things as examples, not as what an actual proposal would be. I was the filing party in the GMO ArbCom case, and I was worried that the RfC would succumb to BS arguments, but what actually happened was that all "sides" got discussed and a rational result (or at least the proposal that I had written!) won out after 30 days. There were quite a few individual editor comments that made me want to facepalm, but overall the three uninvolved admins were readily able to determine a proper consensus. (And the GMO dispute makes this one look like a love-fest! I've lost count of the number of users who ended up topic-banned or site-banned.) --Tryptofish (talk) 23:48, 10 December 2019 (UTC)
Need to check out that one, too. Must be lots of fun to edit there! — kashmīrī TALK 01:12, 11 December 2019 (UTC)
Colin here is why I see this proposal working. The proposed questions we were presenting earlier boil down to asking the responders to deal with policy questions as if NOPRICE, NOR, WEIGHT, LEAD, etc didn't already exist (making it unlikely we would get a meaningful result because of the confusion of what we are asking). Then, each question we came up with was oversimplified or leading, with a complexity of issues that would be very hard to present (without specific examples of how the sources were used, and how that relates to WEIGHT, V, NOT, etc). Even if we were to finally decide on questions, I could not imagine what kind of neutral preamble we would come up with to introduce the conflict.

In this proposal, I envision a neutral preamble that begins with "Here's what NOPRICE, WEIGHT, LEAD and NOR say, there is a conflict about how to use certain sources for text about drug pricing". How do you see using these kinds of sources, and then offer the various options for the sources we are conflicted about.

  • Preamble … conflict over how to interpret NOTPRICE, V, OR, LEAD for drug pricing in articles
    • Can a source like A be used to (and then provide Trypto's various option idea):
      • Option 1
      • Option 2
      • Option 3
      • Option 4
        • with the options for a sample drug summarized in a table that shows what text looks like for a sample drug for each disputed source. This gives a more natural format for one source, one drug, for us to lay out pros and cons in our responses.
    • Can a source like B be used to:

… and so on.

The heart of the dispute here is not whether drug articles can include prices: they can, per NOPRICE and subject to NOPRICE, WEIGHT, LEAD, etc. The core question is what kinds of sources can be used for what kind of text.

I haven't put nearly the amount of work you and Signimu have in to looking at the sources and the problems, so please feel free to tell me I'm wrong. I just think this is a more productive direction to head than the list of questions we have above.

Separately, I don't think we should be asking about prices in the LEAD at all in this RFC … there is already a guideline on LEAD, the question here is complex enough, and if WP:MEDLEAD persists in being a non-consensus section extending beyond and breaching the Wikipedia-wide WP:LEAD, we are likely to end up with another community-wide RFC to resolve that similar problem. I am hoping it will not have to come that, as this RFC will become a model for better consensus discussions occurring at WPMED. SandyGeorgia (Talk) 02:46, 11 December 2019 (UTC)

The thing is the "Can a source like A be used for ----insert article text here ----" is really a simple question of the sort Wikipedians deal with every day. If folk want, I can write up a very simple question of the "Can ... source be used to say ...." variety and we encourage some community discussion on it. It might resolve things much quicker and with less pain than an opinionated RFC about the emotive subject of pricing and Big Pharma censorship. -- Colin°Talk 11:36, 11 December 2019 (UTC)
I think I understand the proponents argument for using databases directly is akin to using WHO ICD codes, we don't need a reliable source to mention each and every one of the ICD codes to be able to add them to entries. So in theory, if we had such a well established and complete database of medicines prices, I think there would be no argument against using that. The problem here is that we do have well established and wide databases of prices such as WHO/HAI or the MSH guide, but they aren't complete, as they are particularly missing high income countries (of course, their purpose is to improve accessibility, and accessibility is most often, but not always, an issue in low to moderate income countries). So are these databases enough to be used systematically for pricing, akin to WHO ICD codes? I'm not sure, and that may be a question for the RfC. It could be implicitly formulated through an example (and explicited in a description of this option). Signimu (talk) 13:03, 11 December 2019 (UTC)
No, Signimu; there is no original research in using an ICD code with citing it, and that practice is fully compliant with all policy, practically akin to not having to cite "the sky is blue". Not only do we not need a citation for common information, if we wanted one for ICD codes, we can find them in legions of MEDRS-compliant sources that present none of the kinds of issues these database-type sources do. And there is only one DSM and one ICD directory-- not multiple different interpretations using different rubrics. The OR that is being used to price drugs in contested articles is not available-- at least in any case presented so far-- in any source other than database types. (If that data were available in a source that doesn't require OR, we could use that source and avoid these problems.)

I think this discussion has also shown that there can be no such thing/never will be any such thing as a "well established and complete database of medicine prices", because that is not how medicine pricing works (see my description of drug prices in the US). I think the comparison of WHO/HAI/MSH to WHO ICD codes will get us off on the wrong track-- ICD codes have very little variability (allowing for the simple descriptions of where they differ from DSM).

I agree that we should be asking if WHO/HAI/MSH can be used as they are being used: I disagree that there is any similarity to ICD codes, and am reluctantly seeing that Colin may be correct in asserting that we are not engaging the core dilemma here.

Rephrasing: basically the approach that Tryptofish and I are advocating asks the simple question about whether these kinds of sources can be used as they are, by giving the RFC respondent a choice of how to use the source. That is the correct question. Thus far into the RFC formulation, we may not have yet fully engaged the core issue, and I fear that equating ICD codes to the dilemma with drug prices may mean that we may still have more work to do to make sure we all understand the problems. SandyGeorgia (Talk) 14:57, 11 December 2019 (UTC)

SandyGeorgia, please allow me to have a slightly different perspective, and opinion. Until recently, I did not know about WHO ICD codes. Maybe for you this is the de facto, and unique, standard, but that's not the case, there are several other diseases codification schemes. Diseases codification is of course less dynamic than prices since there is less impact from the market (the only impact probably being R&D allowing to discover and treat new diseases), but it's still dynamic and flexible, it's a human mind construct. There are different codifications depending on the interpretation and purposes, and even controversies as are well known for the DSM. That said, I'm not saying that OR is not an issue, obviously it is and we should provide examples that do not fall in this issue. I can give you an example of a pricing example straight out from the WHO/HAI database involving 0 OR, in the two figures of: median price ratio. Again, it's not to say we should do that, but although OR is of course a concern to have, using databases does not necessarily involves doing any level of OR. --Signimu (talk) 15:01, 11 December 2019 (UTC)
PS: to be clear, I'm not saying I agree with this view, as I said before I see merits in a lot of arguments from the various sides. I'm trying to understand all points so I can provide adequate documentation to help us make sense of this issue. My point is that yes OR is a concern, and databases may be at higher risk of misusage, but databases != OR if used properly. --Signimu (talk) 15:15, 11 December 2019 (UTC)
My sincerest apologies, Signimu for overstepping in my response. I believe your perspective and hard work here has been invaluable, and I do see your broader point about how/when we can apply database info. I suppose I reacted strongly because even though I dislike infoboxes because they often oversimplify nuanced information, even I am not opposed to having ICD codes in infoboxes, even without citation, as they are so standard and widely accepted (controversies you mention notwithstanding).

Also, because discussion on this alternate proposal is advancing, I am hoping @WhatamIdoing: will provide feedback. SandyGeorgia (Talk) 15:09, 11 December 2019 (UTC)

No need for apologies, you did nothing wrong Ah the whole issue around infoboxes is... a separate issue to say the least. I'm not even sure that's possible. That's why I agree Tryptofish's suggestion is great, it's better to provide an example of a 'good' or at least 'possible' option rather than propose something that may be in fact impossible for us to do. --Signimu (talk) 15:20, 11 December 2019 (UTC)
Thanks everyone for this useful feedback. I feel cautiously optimistic about how this part of the discussion is going. I like the idea of asking about whether "this source can be used for this text" or not. One option (certainly not the only one), if we ask RfC respondents to support or oppose various proposals, is to have something like Proposal 6 be "Source X may be used for example text", and editors could support or oppose that statement along with the rest. Such "proposals" would be very brief, so they wouldn't be much to read.
I do encourage editors who have not yet expressed their opinions of the alternative suggestion to do so. And if we continue to have support for it, I think it would be fine for anyone who wants to (and no hurry at all about this!) to start a new section of this talk page called "Proposal 1" and start workshopping possible text for it. I think the fairest, most neutral, way to number the proposals is that the first one to start composition here is #1, the next one #2, and so on, with the short statements about source acceptability coming after. --Tryptofish (talk) 18:44, 11 December 2019 (UTC)

I think that this might be an interesting source to consider, because it doesn't have many of the usual complexities:

At a price of more than $600,000 for a 25kg patient on a WAC basis and with 8% of all DMD patients amenable to exon 53 skipping, Vyondys has an annual U.S. total addressable market opportunity north of $500 million, Credit Suisse analyst Martin Auster said in a note.

The drug only exists in one country. It (so far) only has one price (per amount of drug, not per patient). What, if anything, could you use that for? WhatamIdoing (talk) 02:18, 15 December 2019 (UTC)

Firstly, the article is I suspect little more than a reformatted press release, and as such, somewhat dodgy wrt reliable sources. It wasn't entirely intelligible, and I don't know whether that's because the subject is very complex, the writer didn't actually understand the subject, or I haven't had my second cup of tea yet today. This kind of article (assuming we had one written by a business editor of a professional business publication, rather than a press release) seems more appropriate for the economic area of the body of a drug, and if this is a key drug for the company, then the company article. Hard to know what if any of that belongs in the lead, other than perhaps something about approval. I found Understanding Drug Pricing interesting wrt "on a WAC basis" and their comment "The wholesale acquisition cost (WAC) is an estimate of the manufacturer’s list price for a drug to wholesalers or direct purchasers, but does not include discounts or rebates.3 Without including rebates and other incentives provided by manufacturers, it is hard to estimate the actual cost of the drug.". We see the same in the UK with some new cancer treatments that are very expensive having an official price but then a "patient access scheme" to discount. This scheme can be very complex and can also at times be confidential. So is the "At a price of more than $600,000 for a 25kg patient" little more than a boast to encourage the market to buy shares of the company (I assume this is completing treatment, though drug cost will only be part of the treatment cost). As such, the press release does not need to care that the $$ figure is inaccurate and merely suggestive of "extremely large". We would, for example, be very wrong to write "It costs $600,000 to treat a child with DMD using the drug Vyondys. But then, being "very wrong" hasn't seemed to matter wrt drug prices. -- Colin°Talk 11:05, 15 December 2019 (UTC)
This is not a press release; it's a business article written by a professional journalist, who quotes a business analyst to estimate potential future revenues for the manufacturer. The price given is for the (alleged) drug costs, not the full cost of the treatment (e.g., the cost of doing the IV infusion).
Would you accept a statement like "One business analyst estimated the manufacturer's initial list price at approximately US$600,000 per 25 kg patient"? WhatamIdoing (talk) 03:12, 18 December 2019 (UTC)
WhatamIdoing, here is the press release. I can find only tiny portions that do not come directly from the press release, with mixed competence in paraphrasing vs straightforward copy/paste. This is Churnalism: reformatted press release. I have no idea if "a note" from Martin Auster is considered a reliable source. Seems he gives ephemeral buy/sell market tips. I'm actually struggling to get the $600,000 number independently. The press release states it is "priced at parity to Exondys 51, the price of which has not increased since its launch in 2016". This article claims that net annual cost is "$300,000 per patient, less than Wall Street analysts had thought". Presumably those "Wall Street analysts" include the likes of Auster. The article also notes that 25kg is a 7-8 year-old boy, and the price will increase with age, perhaps to $450,000. And this is per year (for life?) whereas I read the statement as a total treatment cost. Further, the efficacy of both drugs and lack of good quality trial data seems to be highly questionable, raising concerns about whether we could claim "to treat" while meeting MEDRS (probably not, yet). It seems the jury is out on that matter, and that they are going ahead with approval on basis of parent's request, and to seek evidence as it becomes used. So while I can find sources suggesting $300,000 parity price, the only sources for $600,000 cite Auster. Perhaps we should give stock analyst market tips as miss when it comes to this. It seems terribly hard to get it right. -- Colin°Talk 13:19, 19 December 2019 (UTC)
IMO "getting it right" can be accomplished by making the claim small enough. One analyst said $600K for the specified size of child, but perhaps we should give more WP:WEIGHT to the $300K source. (Of course it's cost per year; that's typical for discussing the cost of DMD drugs.) WhatamIdoing (talk) 16:27, 23 December 2019 (UTC)
NOR sample discussion

User:Signimu your investigations and research into pricing sources, etc is very helpful and progressing things, but I think you are looking at this from a global-health researcher POV rather than a Wikipedian POV. Btw, MSH guide is incomplete for more than just high income countries but also most modern drugs and even for the drugs it does include, it very often has few, one or no suppliers for pill sizes, and of course, has no prices since 2015. Our opinion on how complete/good/etc the MSH data is and what WHO recommended methods are used to extract this into meaningful results for public health research is one thing and perhaps lets us judge if tertiary sources are really doing a sound job. But from the Wikipedian point of view, it is really quite simple. Let me cite some policy from NOR.

  • "Wikipedia articles must not contain original research...facts...for which no reliable, published sources exist. This includes any analysis or synthesis of published material that serves to reach or imply a conclusion not stated by the sources. To demonstrate that you are not adding OR, you must be able to cite reliable, published sources that are directly related to the topic of the article, and directly support the material being presented.
  • "Take care not to go beyond what is expressed in the sources....The only way you can show your edit is not original research is to cite a reliable published source that contains the same material....Drawing conclusions not evident in the reference is original research regardless of the type of source.
  • "All analyses and interpretive or synthetic claims about primary sources must be referenced to a secondary or tertiary source, and must not be an original analysis of the primary-source material by Wikipedia editors"
  • "A primary source may be used on Wikipedia only to make straightforward, descriptive statements of facts that can be verified by any educated person with access to the primary source but without further, specialized knowledge."
  • "Do not analyze, evaluate, interpret, or synthesize material found in a primary source yourself; instead, refer to reliable secondary sources that do so"
  • "Do not combine material from multiple sources to reach or imply a conclusion not explicitly stated by any of the sources. Similarly, do not combine different parts of one source to reach or imply a conclusion not explicitly stated by the source. If one reliable source says A, and another reliable source says B, do not join A and B together to imply a conclusion C that is not mentioned by either of the sources"
  • "Routine calculations do not count as original research, provided there is consensus among editors that the result of the calculation is obvious, correct, and a meaningful reflection of the sources. Basic arithmetic, such as adding numbers, converting units, or calculating a person's age are some examples of routine calculations"

Take that and examine the following article texts:

  • Ethosuximide: The wholesale cost in the developing world is about US$27.77 per month as of 2014[63]
    • The number $27.77 does not appear in the source. This particular drug has one pill size in MSH for 2014 and only one supplier. However, the supplier ASRAMES only supplies the north of Democratic Republic of Congo using their own small fleet of vehicles [i.e. not "the developing world"]. The maths done to reach the price in the article is
      • Use Defined Daily Dose as the "daily dose". This is contentious as the DDD is an artificial metric created for research into population usage of drugs, and not drug costing. It is explicitly not a therapeutic dose and may in fact not be related to any therapeutic dose used in practice. This drug is mainly used in paediatrics.
      • Use 30 as a month supply. In the UK the package size is 56 (for 28-day months). In the source, the price is for a bottle of 100 tablets. In the UK and US oral syrup is also available.
      • ( Defined Daily Dose [1250mg] / Strength [250mg] ) * Unit Price [$0.1845] * Month [30] = $27.675. Round to 2 decimal places.
  • Carbamazepine: The wholesale cost in the developing world is about US$0.07 to US$0.24 per day as of 2015[64]
    • The numbers $0.07 and $0.24 do not appear in the source. Searching for Carbamazepine in 2015 produces 5 results though 2 are to the same URL (just different therapeutic categories). The 100mg/5ml syrups has no suppliers and five buyers.The 200mg sustained release pill has no suppliers and two buyers. The 400mg sustained release pill has no suppliers and one buyer. The 200mg standard tablet has 12 suppliers and 4 buyers. The maths done in the article is
      • Pick the 200mg standard tablet. The database does not indicate which to pick, nor is there an established/official method for picking one in any literature.
      • Take the lowest unit price [0.0138 from the Suppliers] and the highest unit price [0.0480 from the Buyers].
      • ( Defined Daily Dose [1000mg] / Strength [200] ) * Unit Price [$0.0138] = 0.069. Round to 2 decimal places.
      • ( Defined Daily Dose [1000mg] / Strength [200] ) * Unit Price [$0.0480] = 0.24
      • Although we give a min and max price, we have taken the min and max from different kinds of price (supplier/buyer) but not from different strengths or formulations, which may be cheaper or dearer.
  • Mebendazole: The wholesale cost in the developing world is between USD 0.004 and 0.04 per dose.[65]
    • The numbers 0.004 and 0.04 do not appear in the source. Searching for Mebendazole 2014 produces 5 results. The 100mg standard tablet has 6 suppliers and 5 buyers. The 100mg chewable tablets have 4 suppliers and no buyers. The 100mg/5ml syrup has 6 suppliers and 3 buyers. The 500mg standard tablet has one supplier and one buyer. The 500mg chewable tablet has 3 suppliers and no buyers. The maths done in the article is
      • Pick the 100mg standard tablet. The drug is used for lots of different parasitic infections. Drugs.com generally indicates the chewable tablets and either the 100mg as a single dose with another perhaps in two weeks, 100mg twice a day for 3 consecutive days, or the 500mg tablet once as a single dose. Some require 200-400 three times a day for three days, then 400-500mg orally 3 times a day for 10 days. This is not a drug taken long term, but in short treatment duration that is quite variable in strength/length depending on indication.
      • Assume the "dose" is 100mg. The source does not indicate what a "dose" is. It does give a "defined daily dose" of 200mg but see above for that.
      • Round the lowest price from the suppliers [$0.0035] to three decimal places [$0.004] and round the highest price from the buyers [$0.0393] to two decimal places [$0.04]. As above, the min and max come from different kinds of price (supplier/buyer) but not from different strengths or formulations (e.g. chewable or 500mg).
      • The article does not give a treatment cost even though the length of treatment is knowable depending on indication (from other sources). The article does not state what treatment that dose cost is for.

I don't think it is rocket science to see this is original research of the most obvious kind. The NOR point "can be verified by any educated person with access to the primary source but without further, specialized knowledge" is particularly useful here. There is nothing in the source to suggest which record to choose. There is some advice on the help pages about buyer and supplier prices and about perhaps using a median, but nothing concrete enough to say what our educated person might actually do, and our articles have generally ignored the advice, and statistical common sense about avoiding outliers, buy choosing the min and max. A careful reader might question if buyer and supplier prices can really be combined, as they have in our articles. The help pages also have some advice about DDD, which is generally off-putting in terms of thinking it should be used for pricing a daily treatment cost (e.g. " the ATC/DDD system by itself is not suitable for guiding decisions about reimbursement, pricing, and therapeutic substitution."'). There is nothing in the source about whether a daily dose is divided (usually in two) or single. Nothing to indicate starting dose and maximum dose and titration increments which strongly influence a choice of pill size the doctor prescribes. Nothing to explain for example that an enteric-coated pill or a capsule cannot be split in two half doses, but that dose variation is simple with a syrup (though watch out for the per 5ml gotcha). Nothing to suggest picking chewable or enteric-coated vs standard. Nothing to suggest e.g. that paediatric usage might suggest a syrup. Nothing about treatment duration (continuous, pulsed, short-term, one-off). Nothing to say what indication the DDD was calculated for. Nothing to suggest additional costs such as blood level monitoring or follow up appointments. Nothing to explain what kind of wholesale cost a supplier price is, or the many caveats to use when considering the relevance (if any) of a buyer price. -- Colin°Talk 16:23, 11 December 2019 (UTC)

Here's a very brief summary of the other database-level sources we have used:
  • NADAC for "US wholesale cost ... per month". This starts off bad with a database result for a historical week but no selection on drug at all (25,194 records returned). If we enter a drug name in the "Find in this dataset" box we again get a mix of syrup and tablets of various kinds and mg sizes. Might also get variants on the drug itself such as different salts or in combination with other drugs. The NADAC does have an average price, so we don't have the problem of multiple prices per supplier/manufacturer. There is no dose information or notes on what therapeutic purpose the drug might be used. So no way to use this site to calculate a "per dose" or "per day" or "per month".
  • Drugs.com for "In the United States, it costs about US____ per dose". Website return various sizes and formulations (e.g. 10mg, 25mg, 50mg, 75mg, 100mg, 150mg), syrups, capsules, etc. Webpage does not explain e.g. low dose is for neuropathic pain off label in US, higher doses for mental illnesses licenced indication. No information on this page about typical dose or how many tablets to take per day or how to divide doses. Price-per-pill reduces considerably the more pills you buy in one go.
  • BNF for UK prices (might not be accessible outside of UK). Again multiple size tablets, syrups and capsules, and different package sizes. Multiple manufacturers, some with different "NHS Indicative Price". Most records have "Drug tariff price". No indication on website which price to pick or what they mean in terms a normal "educated person" might grasp. No indication if drug is subject to special discounting (due to e.g. patient assistance program) or currently has a higher price (due to shortage). Although other pages in the BNF give dosage advice (for professionals) this of course varies per indication or patient age and health. So no way to determine one "daily dose" or "individual dose".
All these are highly respected sources of raw data about prices. Nobody is questioning the accuracy of the data. -- Colin°Talk 18:42, 11 December 2019 (UTC)
As several of us have now pointed out that this is the core issue affecting a potential RFC, I have posted a query for further feedback to the NOR noticeboard. I suggest we cannot put together an RFC that is likely to yield a useful conclusion without getting to the heart of the NOR matter. SandyGeorgia (Talk) 17:58, 11 December 2019 (UTC)
Well, I see that noticeboard is just hopping with activity. Perhaps by next week we need to start putting up some concrete proposals of format for the RFC. I still think we need a name for the RFC, and a draft page for coordinating work. SandyGeorgia (Talk) 15:11, 13 December 2019 (UTC)
I will do, hopefully today but more likely on sunday, hard to manage such a big issue along with work :-/ Signimu (talk) 19:45, 13 December 2019 (UTC)
  • This is going to end up at ArbCom. The task is to formulate a question for a central RfC. Nobody's doing that, all that's happening is endless rehashing of the answer you think the RfC should return. Guy (help!) 23:22, 15 December 2019 (UTC)
  • JzG, then I must be misreading or missing something, because my impression is that both WAID and Signimu have indicated that they are working on it. From my end, it's holiday season, and hashing out wording isn't the spirit I'm in. I have said multiple times that we first need a name and a draft space: it is my understanding that Signimu has said they are on it. SandyGeorgia (Talk) 00:13, 16 December 2019 (UTC)
  • It's difficult to create an RfC when the content policy seems clear, and those asserting exceptions for policy are almost totally silent.
  • For the RfC, we need to distinguish questions about pricing from questions about specific prices. --Ronz (talk) 16:00, 16 December 2019 (UTC)
  • About the problem of some editors seemingly being silent, I have a suggestion on how to jump-start it. Just go ahead and start formulating RfC questions that represent the proposals that you would likely support. At some point, it will become pretty obvious that the RfC is going to happen, and when previously silent editors realize that their perspective might not be presented, that will have a tendency to focus the mind. And if that doesn't work, then propose something for them, and make them realize that they would be better off revising it so that it doesn't look bad. And if that doesn't work, then let the RfC turn out overwhelmingly to support what you are hoping for. --Tryptofish (talk) 20:03, 16 December 2019 (UTC)

Are pharmaceutical prices encyclopedic information?

Please focus narrowly on the question for RfC, you can make arguments for your preferred outcome in the RfC itself. Guy (help!) 10:16, 10 December 2019 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Now that I caught up with the latest developments (sorry I could barely connect the last week due to IRL things), I would like to suggest to discuss what I think is one of the most essential issues: whether pharmaceutical prices are encyclopedic information. There are other crucial questions, such as the methodology (what prices to use?), editorial (where to place prices - lede, infobox, article's body, wikidata? what sources are acceptable?) and others, but I think this issue precedes the others, so I'll start here with that and later write about the methodology in a separate section. Since finding sources is something I like to do, and I like to learn new things, I will here show the results of my findings on this question, which may help further our collective understanding and discussion. Before going further, please note I had no in-depth knowledge of the whole issue around medicines prices before the new discussion started at WT:MED#MEDPRICE, and I here applied the same WP:MEDRS methodology we use for finding reliable sources.

  • WHO stance on prices (emphasis mine): "Access to essential medicines is part of the fulfilment of the right to the highest attainable standard of health (in short: the right to health). So why do millions of people across the globe go without the treatments they need? The reasons are now becoming clearer – and the price and availability of medicines to those who need them are crucial factors."[1]
  • United Nations stance: "Greater transparency, fairer prices for medicines ‘a global human rights issue’, says UN health agency"[2]
  • The latest World Health Assembly approved a resolution on price transparency: "Seriously concerned about high prices for some health products, and inequitable access to such products within and among Member States, as well as the financial hardships associated with high prices which impede progress towards achieving universal health coverage; Recognizing that the types of information publicly available on data across the value chain of health products, including prices effectively paid by different actors and costs, vary among Member States and that the availability of comparable price information may facilitate efforts towards affordable and equitable access to health products; Seeking to enhance the publicly available information on the prices applied in different sectors, in different countries and the access to and use of this information, while recognizing different national and regional legal frameworks and contexts and acknowledging the importance of differential pricing;"[3][4]
  • Vogler, researcher at WHO (emphasis mine): "Ensuring affordable access to essential medicines (i.e. those medicines that satisfy the priority health needs of the population) is a major policy objective globally. One-third of the world's population; however, is estimated to have limited or no access to essential medicines. One of the key barriers is the high price of medicines. [...] Pharmaceutical pricing policies are key because medicines are no normal goods and health care, including pharmaceutical, systems are not normal competitive consumer markets."[5]
  • A Lancet commissioned review paper: "Governments, national health systems, and the pharmaceutical industry must promote transparency by sharing health and medicines information."[6]

To summarize: the sources above, and virtually all sources I have, have noted the specificity of pharmaceutical prices, which directly impacts medicines affordability and thus efficiency. All sources agree price is not a side parameter, but a "crucial" or "key" factor. It's not like the prices of other consumables, where the price is mostly/only an economic factor: medicines prices are also and foremost a health issue.

Now let's go back to Wikipedia's policies and RfCs:

  • WP:NOPRICES states: "Sales catalogues. An article should not include product pricing or availability information unless there is an independent source and a justified reason for the mention. Encyclopedic significance may be indicated if mainstream media sources (not just product reviews) provide commentary on these details instead of just passing mention. Prices and product availability can vary widely from place to place and over time. Wikipedia is not a price comparison service to compare the prices of competing products, or the prices and availability of a single product from different vendors or retailers."
  • 2016 RfC Wikipedia_talk:WikiProject_Medicine/Archive_84#Price_of_medications (emphasis not mine): "Except in the cases where the sources note the significance of the pricing (which did have consensus), there is no consensus to add the pricing to the articles. There were several ways to present this information proposed in the discussion (such as wikidata). Please feel free to start a follow-up discussion regarding that."

We have several sources above that independently assert the importance of prices, for all medications. I have yet to find any reliable source that would qualify pharmaceutical prices as irrelevant. Not to mention that a wide array of countries are actively implicated and/or implementing medicines price policies, external reference pricing being just one of them. Since virtually all sources seem to agree on the importance of medicines prices as a significant (even "crucial") health factor (contrary to prices of other consumables), and the worldwide consideration of various entities such as health organizations, national and transnational regulatory bodies, and researchers, it looks to me that medicines prices may be considered encyclopedic information per se, just like we consider epidemiological information as encyclopedic without requiring each source to specify why epidemiology is important for a particular disease, the epidemiology is simply a factual bit of knowledge that stands by itself, and furthers our knowledge on the topic (eg, the prevalence of an illness is important by itself, we don't need to put this in perspective with other diseases).

What do you guys think? --Signimu (talk) 18:46, 9 December 2019 (UTC)

PS: to be clear, even if we reach here an agreement that prices may be encyclopedic in theory, we would still need to resolve other issues as stated above by me and others, such as the methodology (eg, if we can't find a common ground with a good methodology to base the pricing, we can't add prices anyway). Signimu (talk) 18:53, 9 December 2019 (UTC)

References

  1. ^ Measuring medicine prices, availability, affordability and price components. World Health Organization & Health Action International. May 2016. pp. 123, 289. Mirror: 2008 original publication by WHO.
  2. ^ "Greater transparency, fairer prices for medicines 'a global human rights issue', says UN health agency". UN News. 14 April 2019.
  3. ^ Fletcher, Elaine Ruth (28 May 2019). "World Health Assembly Approves Milestone Resolution On Price Transparency". Health Policy Watch.
  4. ^ "Improving the transparency of markets for medicines, vaccines, and other health products". WHO. 2019World Health Assembly, 72{{cite web}}: CS1 maint: postscript (link)
  5. ^ Vogler, Sabine (2019). "Pharmaceutical Pricing Policies". Encyclopedia of Pharmacy Practice and Clinical Pharmacy. Academic Press. pp. 188–201. ISBN 978-0-12-812736-0.
  6. ^ Wirtz, VJ; Hogerzeil, HV; Gray, AL; Bigdeli, M; de Joncheere, CP; Ewen, MA; Gyansa-Lutterodt, M; Jing, S; Luiza, VL; Mbindyo, RM; Möller, H; Moucheraud, C; Pécoul, B; Rägo, L; Rashidian, A; Ross-Degnan, D; Stephens, PN; Teerawattananon, Y; 't Hoen, EF; Wagner, AK; Yadav, P; Reich, MR (28 January 2017). "Essential medicines for universal health coverage". Lancet (London, England). 389 (10067): 403–476. doi:10.1016/S0140-6736(16)31599-9. PMID 27832874.
  • Thank you Signimu for your thoughts. Just a minor although I think important comment from my side. When the sources you quoted mention prices, they talk about them in relative terms: prices matter only to the extent as they hinder access to medicines. Prices are discussed as relative to the wealth of a given society, to its per capita GDP, etc. But here on WP the entire discussion has been about absolute pricing – i.e., disconnected from any socioeconomic context and simply presented in crude monetary units.
In my view, it absolutely is encyclopaedic to add information that a cost of a specific therapy has prevented access to it in specific countries. I even have no problem with listing countries where a given drug is largely not available due to cost.
But dumping raw pricing data into all the drug articles on Wikipedia - no, I do not think this is the type of information that any encyclopaedia should provide. — kashmīrī TALK 19:08, 9 December 2019 (UTC)
Thank you Kashmiri, to be accurate, yes and no, let me explain: prices are mostly studied in relation to essential medicines accessibility, because that's the goal the UN and WHO were tasked (or tasked themselves) to do. Usually, pricing and accessibility issues arise in low and moderate income countries, but they also happen in wealthy countries: it was reported that hepatitis C medicines, which are now part of the essential medicines list, are too expensive even for wealthy countries, and following the 2008 global crisis, european union countries which were affected reported having restricted access to essential medicines.[1] That's why this same source writes: "Affordability of medicines has become a key issue for governments, as well as public and private payers for health care, regardless of a country’s income level."[1] I however agree with you that a dump of raw prices is not encyclopedic, but we should keep that for the discussion around the methodology of what kind of measure we can use (which I am still researching). A preliminary example of what is possible can be seen at median price ratio, although I don't think this can be generalized to multiple articles as this is limited to LMICs countries. Signimu (talk) 19:17, 9 December 2019 (UTC)
  • (after edit conflict) Signimu, I suspect the answers to your well-positioned query (thanks for the analysis) are buried in the lengthy discussions above, so at the risk of repetition (no bludgeoning intended :), my summary response is that:
  • Yes, MEDRS-compliant sources address the importance of pricing, and that information would be relevant to a broad, general article on drug pricing; but,
  • No, we don't have any indication that the sources put forward so far are adequate to add the content that has been added to many/most drug articles, and we don't even have a best-case example to go on. As we've discussed, we can specifically discuss pricing in some cases (sample, insulin and epipen) because the sources do the work for us, but we are using database-style sources that carry a multitude of problems discussed above to add pricing info as of now.
Overall, while we possibly all agree that prices are important per MEDRS, the specifics applied so far to most drug articles are non-compliant and even misleading. That is the dilemma. I feel we are spinning our wheels until we get one, good, compliant example that does not involve breaching NOT, WEIGHT, V, LEAD, NOR and MEDRS. Short answer: for specific drugs the importance and relevance of pricing/cost must be based on good sources for and about that specific drug, which is not the case that is being advanced now. Pricing content for an individual drug is encyclopedic, when specific sources, compliant with policy, are used to add the relevant information, subject to policy and guideline. SandyGeorgia (Talk) 19:14, 9 December 2019 (UTC)
  • Signimu I think your initial quotes about the importance of drug pricing is sufficient to justify a set of articles on the issues and methodologies and projects worldwide. But they don't seem convincing, as Kashmiri notes, to dump raw dollars and cents prices into our articles. Think of the basics for life: a home, heating and electricity, food, water, education, health (not just drugs), etc. These are all important and countless words have been written about affordability of food [In the UK, we have a scandal of being one of the richest nations on earth and yet those working in the poorest jobs, which can include nursing and teaching, can find themselves reliant on food banks]. But the affordability of food as a world issue, is not being used to justify prices of mince pies and others into every single food article. That's what we see here: a plea to emotion.
See other posts I've made today. Even for drugs that are notable for their high price like Lenalidomide, Sandy I'm sorry but we don't even get those right. The article juxtaposes a retail US price (without saying it is retail) with a wholesale UK list price and claims that is the price it costs the NHS. But the BNF book does note that this eyewateringly expensive drug is subject to a "patient access scheme" that heavily discounts the price to the NHS. The discount for some of those schemes is public and very complex and for some confidential.
Read also my post above about "out of pocket" prices paid by those in developing worlds. The WHO document cited basically busts to smithereens the idea that prices in the MSH price guide bear any resemblance to the often private prices paid by patients in such countries, if they have access at all. The misuse of MSH data is basically an example of how many 1001 ways can you completely screw up when doing original research. Guy's "price sharply increases" observation is another good example. Perhaps, when this is over, someone can use it as a textbook example of why original research is a bad idea. -- Colin°Talk 19:40, 9 December 2019 (UTC)
Please let's keep methodological issues, which are real, for another discussion, else if it's mingled with the issue of pertinence I fear we will go in circle without reaching any agreement. Your point and analogy with food is very interesting, thank you. Signimu (talk) 19:59, 9 December 2019 (UTC)
Pricing in general is clearly important, and I earlier pointed out Medication costs, Prescription drug prices in the United States, Prescription charges [66].
We generally don't use references that don't mention a specific item to determine weight within an article about a specific item. (That's a horribly convoluted sentence, but I hope it makes sense.) So, while prices are important, that doesn't mean we've met NOTPRICE criteria for an article about a specific medicine unless we have a source clearly showing the importance of prices for that specific medicine.
Prices may not exist in a form that is suitable for encyclopedia articles at all, as all the discussions are indicating.
To answer the question of this subsection: No. Encyclopedic prices don't appear to exist.
Meanwhile, I think a good case could be made for giving extra weight to pricing information when it's available: Information about the ranges in pricing and the many factors involved. --Ronz (talk) 20:09, 9 December 2019 (UTC)
I agree very much with Ronz's distinction between prices and pricing. --Tryptofish (talk) 20:52, 9 December 2019 (UTC)
Yep, Ronz gets the Succinct Prize for the day, with the price/pricing distinction. SandyGeorgia (Talk) 20:58, 9 December 2019 (UTC)
Your pricing/price distinction captures what I was qualifying as "prices may be encyclopedic in theory", I should have written "pricing". Signimu (talk) 21:49, 9 December 2019 (UTC)
I like your pricing/price distinction, but pricing needs price metrics, which complicates things (because it requires discussing the methodology of 'what price metrics to use'). I agree however that the sources could be interpreted that way (the pricing, or affordability/accessibility, of medications is a health issue, not the price per se). --Signimu (talk) 21:56, 9 December 2019 (UTC)
And that seems to me to be a good reason to treat the matter as something to be discussed in the text, based on independent secondary sources, as opposed to just listing numbers. (Of course, other editors will disagree.) --Tryptofish (talk) 22:03, 9 December 2019 (UTC)

References

  1. ^ a b Cite error: The named reference Wirtz2017 was invoked but never defined (see the help page).
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

While I agree that we don't need to have the discussion right now, this section heading is a potentially important question. If editors believe that some kind of information about money is normally appropriate to include (subject to the limits of the available sources, just like everything else), then we need to know that. WhatamIdoing (talk) 01:06, 15 December 2019 (UTC)

Wikipedia is full of prices of products

Please focus narrowly on the question for RfC, you can make arguments for your preferred outcome in the RfC itself. Guy (help!) 10:16, 10 December 2019 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

We have an entire article on the price of oil, with multiple sub pages. We have a section on Gasoline#Use_and_pricing in various parts of the world. And than an entire subpage on the topic.

For other commodities we see the price of wheat, milk, and rice. For consumer products we have the price of the Nintendo_Switch (both in the infobox and body of the article in various currencies), Nintendo_Switch_Lite, Apple Watch Series 3, Honda Legend, Acura RL, and Pentax K20D.

Now if we just want to look at just featured articles, we include prices in Sega Genesis, Wii (in the infobox and body), Sega Saturn (in the lead and body), NeXT, Macintosh Classic, and PowerBook 100 (in the lead, infobox, and body).

We include prices for paintings The Triumph of Cleopatra and Streatham portrait. We include prices for elements such as Xenon and Germanium. We include prices for universities such as University of Michigan and Shimer College#Tuition and fees. We include prices of Durian and Maple syrup. We include prices for water fluoridation.

I am happy for us to continue including all of these things. Including the prices of medications is not out of line with current practice both generally and for featured articles. Doc James (talk · contribs · email) 01:41, 10 December 2019 (UTC)

Given the discussions we've had, it's clear that determining a price for a medication that we can use in an encyclopedia article is nothing like any of these examples. --Ronz (talk) 01:46, 10 December 2019 (UTC)
Doc James, have you read this section? You are repeating arguments that have already been addressed in several sections on this page. WP:OTHERSTUFFEXISTS, and yes, there are many instances of pricing that comply with the restrictions of WP:NOTPRICE.

What is needed next for us to be able to advance towards an RFC is for you, Quack Guru, or anyone else to provide a sample of your strongest, tightest, most accurate drug article on pricing, that does not have the multitude of problems that have been outlined on this page, so that we can advance RFC questions based on what you consider to be your best practice example. Many people here have expended considerable effort in analyzing and discussing the issues to be resolved: please give us what you consider to be the best-sourced drug article from the sources you want to use, so that we can begin to address things with specificity. Reading this section for a list of drug articles that have already been well discussed here may help. SandyGeorgia (Talk) 02:16, 10 December 2019 (UTC)

Sure I have tried here.[67]
And happy to continue if others are interested. This at Xenon "approximate prices for the purchase of small quantities in Europe in 1999 were 10 €/L for xenon, 1 €/L for krypton, and 0.20 €/L for neon" is what we do sometimes do for medications. Here we have a very specialized source used to support it.[68] One could summarize this table Wii_launch#Release_dates_and_pricing by just giving the highest and lowest numbers. But meh.
Including prices of commodities and products is common practice. Just noticed you went through a number of them above. Doc James (talk · contribs · email) 02:33, 10 December 2019 (UTC)
I'm not sure any of you are going to be able to convince each other of anything at the moment. Not for lack of knowledge or skill, but simply because you have discussed this so much and still have a good faith disagreement. I would suggest that your energy might be better spent coming to consensus on the RfC questions as that, in my estimation, is your best way out of this morass. However, if you all feel like you're making progress please ignore me and continue your discussion :). Best, Barkeep49 (talk) 02:48, 10 December 2019 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

WhatamIdoing's summary

I've been taking notes, and I've left comments elsewhere, but some themes come up in multiple sections, so I'm skipping down here to tell you what this looks like to me at this point.

First, I agree with User:SandyGeorgia that we're really making progress. This section has some 300 comments and about 50,000 words on how to address these questions. I know that long discussions on complicated subjects aren't everyone's cup of tea, but I like them. Hang in there if you can, and don't worry if you can't.

Some things that have really been helpful include:

  • User:Signimu and User:Colin in particular have done a really amazing job in identifying the complexity of "the price" for "a drug". I hope that drug pricing will be improved while people still have this information in mind and sources at hand, because I strongly suspect that some editors will be relying on our articles to figure out their answers to the RFC.
  • SandyGeorgia proposed providing ==External links== to a drug price database. This kind of out-of-the-box thinking is really helpful, and the particular idea is something that we might want to do no matter what else is decided.
  • User:Ronz had a particularly good comment on the difference between "prices" and "pricing". We need to be clear about which one(s) we want in which circumstances.
  • Several people have identified specific details that future editors will want help with, such as User:Tryptofish's point about whether prices/pricing is better suited to articles about individual drugs (e.g., Diphenhydramine) or articles about a class of drugs (e.g., H1 antagonist or Antihistamine). The only point that I didn't see mentioned recently was the effect of patent status on drug pricing, but that was touched on in the pre-ANI part of the discussion.

Next, it seems to me that there's no way to write this RFC until we have settled the question about how to present the decision to editors. That means that we have to decide whether we're asking questions that identify relevant principles or proposing guideline text. I can go either way with this. If people answer our questions, then there are several people in this discussion who can write guideline text to align with their answers; if editors instead pick a proposed addition to the guideline, then we'll know the answers to our questions. This is not a decision that will change what editors say, but it is necessary to decide this now. (Feel free to actually vote, straw poll-style, for your preference. I can work with any outcome on that point.)

Among the open questions, the main one seems to be whether (other) editors think that drug prices should normally be included or normally be excluded. Yes, NOTPRICE says to normally exclude, but some editors are uncertain whether that applies as strictly to drugs as it does to T-shirts. One way to put this is, are we talking about a subject that is mostly a business matter or a healthcare matter? If it's mostly a business matter, then we'd mostly leave them out, because prices aren't encyclopedic except when there is significant discussion (e.g., a price-gouging scandal). If it's mostly a healthcare matter, then we'd mostly include something (as much as we could based upon sources, which might not be much).

Among the other open questions, my notes include:

  • What constitutes "discussion" of prices in a source for NOTPRICE? Is it enough to have a source that says Drug Prices Always Matter, and then we add prices to all drugs? Or do we need something saying that the price of a specific drug, or a specific class of drug, matters? Is it enough to have solid numbers (which are hard to come by for many – but not all – drugs), or do we need solid numbers plus an explanation of whether and why the number matters for that particular drug?
  • Is it the price, or the affordability? Price matters for business reasons. Affordability matters for medical reasons.
  • Do we actually care about anything other than drugs? Does anyone care about the cost of surgeries, hospitalizations, time-consuming therapies, durable medical equipment, or consumable supplies?

Finally, there was a point when it seemed like people were saying that we can't get a good source for all drugs, so we should all give up and not do anything. I don't think we should write this guideline on that basis. We have, in the past, had people tell WPMED folks that they were going to write a review article, and was there anything in particular that we needed a decent source for? It's likely that drug pricing will get more transparent over time, and the ideal sources will emerge. That time might be years from now, but it will likely happen. I think that we should take User:Bluerasberry's advice above and talk about that magical future, and what we consider to be the right way to address that question. A worldwide average? A range? A number for wealthy countries, and another number for low and middle income countries? A description (such as "affordable" or "expensive")? Historical prices? List prices? Something else? Would it be one sentence or one paragraph or one whole section? There are a lot of options, and I think that people ought to be considering the options. WhatamIdoing (talk) 02:14, 15 December 2019 (UTC)

Wrt healthcare vs business matter. Hmm. I think for most drugs, they don't have significant business WP:WEIGHT. Any sales figures are likely to be commercial secret unless they hugely affect share price. None of us dispute that there are drugs that are important for company share price or important for high-cost scandals where there is a business aspect. Again this doesn't justify the original research that's been going on. There is also a political aspect and a retail aspect. I'm really not convinced that a healthcare issue suggests we include a price [there are so many to choose from, only only a few available]. While some drugs have become notable for their healthcare burden, is the price of valproate relevant to healthcare -- that drug is chosen for its effectiveness in treating epilepsy. Another newer on-patent drug might be notable wrt healthcare, if it was for example, not recommended by NICE as a first-line therapy largely because it is more expensive than existing treatments which work much of the time. And we'd have sources that say that information and our reader cares that "expensive -> not-first-line choice" rather than $45.04 median price per month.
I strongly disagree with the magical wishful thinking approach. It doesn't in fact seem that "drug pricing will get more transparent over time". Already there are doubts if MSH price guide is being maintained since 2015 (I really would like an answer to that). The UK seems to be doing more secret patient access pricing for expensive new drugs, leaving behind the BNF to show a mythical wholesale list price that is misleading. Brexit and a US trade deal is likely to lead to the UK being pushed towards US drug pricing. But I think the fact that the sources we have used so far all give prices down at a basic level "A 100-pill bottle of 50mg tablets of enteric-coated sodium valproate" is important. The BNF even gives separate prices per brand for some drugs, the NADAC gives an averaged price, MSH gives prices for each supplier and likely-to-be-misleading buyer prices. Drugs.com gives retail list prices that very much depend on how many pills you order, and that Sandy has explained nobody pays. None of these give us a "therapeutic dose" and none of them give a price per day or month -- doing that is ALWAYS original research. In a regulated drug price market like the UK, I've seen prices per MG relatively constant for different MG size tablets. But in the US, it seems many drugs are the same price per tablet regardless of how many MG, though there are exceptions. The MSH prices are even more variable for pricing which size tablet. Our sources operate at this very low level because the real world is complicated and can't be simplified. Sure, a high-level source could suggest a rule-of-thumb price (like one of the pocket pharmacy sources did with their $$ symbol system, but we abused that source too.
We are surrounded by experienced editors here. If this was simple to do, we'd have done it. -- Colin°Talk 13:28, 15 December 2019 (UTC)
Appreciate the thoughtful consideration. Worried about work progressing during busy holidays; my IRL calendar is jam packed. SandyGeorgia (Talk) 13:20, 15 December 2019 (UTC)
I am not sure we need an RfC at this time. I see other relatively more simple directions we can develop to make progress.
  1. Develop Wikipedia:Experiment to present guidance on how to run and interpret live experiments. We have thousands of drug articles and some subset of them should be experimental. We can get information from including anything new, prices or side effects or any new data, in these. We still do not have standards for running experiments. In the case of prices being yes/no, we need a safe and sane path to do limited experiments with measurable outcomes and to collect feedback on live articles.
  2. Develop WP:Infobox for the sake of prices and structured data. Suppose that we had perfect price information on all drugs, and we had consensus to include it in some form. What would our pipeline be and how would we manage it?
  3. Develop wiki process of original research for aggregating data to present scales. Right now I have a draft for this at Wikipedia:Defining data. If we included price information, it may not be exact numbers, as what we actually want to communicate is low/medium/high price. I asset that it is absolutely fundamental to knowing about a drug if it is inaccessible by price. I think for any product shown in Wikipedia, communicating whether a product is luxury or affordable is also general reference encyclopedic content. Wikipedia claims to some interpretation, but actually we have a backlog of undocumented examples of original research which we designate as acceptable. It is taboo to talk about this because of respect for the simplicity of WP:OR, but we need an evaluation method for identifying and considering when some original research and primary data interpretation is necessary context for understanding a topic.
  4. We need to develop English Wikipedia's relationship with Wikidata. The Wikidata community is light on its editorial policy and ethical standards and right now, I expect that project would draw heavily from the precedents of various language Wikipedias, if Wikipedias wanted to recommend policy. It seems apparent to me that Wikidata is attracting government, nonprofit, and corporate investment at a pace and scale which Wikipedias never have. It does not take too much money to massively disrupt community projects, and if English Wikipedia does not build out the social infrastructure to have discussions about its relationship to general reference data in Wikidata, then the cost of integrating Wikipedia and Wikidata will rise as Wikidata develops with corporate capture. Some imagination about the future of Wikidata and new technology would be helpful here. English Wikipedia could draft the first Wikidata policy on drug prices, which is a privilege which someday companies might pay money to do. English Wikipedia does not have to do this but doing something seems worth considering.

If we had more developed positions on these things then answering the question about what to do with prices on English Wikipedia would be easier.

Blue Rasberry (talk) 20:33, 15 December 2019 (UTC)
Those all sound like interesting projects, but in the meantime, we have to deal with original research in articles. Since there has been no response from the NOR noticeboard, it looks like an RFC is unavoidable. Like Colin, I am not a fan of the idea that these problems will go away with wish lists; we need to deal with problems in articles now, based on current policy, not as the Wikipedia wheels turn on new ideas. SandyGeorgia (Talk) 20:40, 15 December 2019 (UTC)

Break on WAID's summary

Two points:

  • The consensus coming out of the ANI discussion means that there must be an RfC. It isn't optional. I see nothing wrong with thinking things over until after the holidays, but by then, we will have to go forward with it.
  • I still think that the best approach to the organization of the RfC is to have "proposals" in the form of how prices should or should not be included, and how they should be sourced, and have editors who respond either support or oppose each of the various proposals. But I don't think we need to propose guideline language for adoption. Please remember that editors said at ANI that there needs to be a widely-announced RfC, and the basic issues at hand are whether or not we should include prices, and if so, how they should be presented and formatted. I agree with Guy's comment higher up on the talk page, that this will end up at ArbCom if editors don't come up with a workable RfC. --Tryptofish (talk) 00:34, 16 December 2019 (UTC)
    • Tryptofish, Guy. I have an idea for a question then. There are some here who believe that a drug has a price that can be expressed in dollars and cents. It has been claimed this is "defining data" and thus appropriate for leads, info boxes and wikidata. None of our sources used to-date have "a price" for "a drug". Those who believe in this one-drug-one-price approach have stated that either the raw database sources and OR required to produce a price is just fine (e.g., James), or that we should run an RFC based on a "magical future" that at some point sources will emerge that directly state a price for a drug (e.g., WAID, BR). There are a bunch of proposed questions about leads and info boxes and the old primary/secondary chestnut, but many of them presume that a price for a drug is a sensible and encyclopaedic concept and all we need is an opinion about whether or where to put it. And this is before we even get into thinking about which kind of price we might present (many different wholesale and retail prices, most of which are "list" prices that fail to present a true cost/price anyone or any organisation actually pays). And btw, simplifying a price to an adjective has been suggested, but to date no reliable comprehensive source of such drug-adjectives has been offered, and any source is likely to be very country-specific -- so unless someone finds such sources, that's a non-starter.
So perhaps the question we first need to ask is
"Do you think that a pharmaceutical drug has a one price that can be expressed in dollars and cents, for each region such as US, UK, developing world, and this is information that Wikipedia drug articles should include and can routinely be sourced by editors without original research."
Something like that. This question could be presented with example texts, example sources, and perhaps opening statements by both sides. Perhaps we need to have little *clarification that we are talking about a drug, such as ibuprofen and not a packet of 16 Nurofen Express 200mg Liquid Capsules from Boots the Chemists in London Liverpool St Station. -- Colin°Talk 13:15, 16 December 2019 (UTC)
Pinging previous participants in this conversation SandyGeorgia Bluerasberry, Signimu, WhatamIdoing, QuackGuru, Doc James, Kashmiri, Levivich. Barkeep49 (talk) 16:11, 16 December 2019 (UTC)
​Speaking for myself that is really two seperate questions and I think RfCs are generally more successful (as in process not outcome) when there is not a compound ask in the initial question. If people feel that's generally on the right track I would simply recommend making it a two question RfC. Best, Barkeep49 (talk) 16:13, 16 December 2019 (UTC)
I like the idea of working on a series of questions that incrementally address the various aspects of the content disputes. Colin's question would be a good start.
Yes, this will make for a problematic RfC, but that seems to be a result of the complexity of the topic itself. --Ronz (talk) 16:17, 16 December 2019 (UTC)
Barkeep49, there are sort of three questions, so I'm not clear which you think is "two". I don't see any of them as independent, so unclear how a two or three question RFC would make sense. If people reject the idea there is a price for a drug, then really it is back to square one for those promoting prices to rethink some alternative approaches to the whole matter: no point then in asking the other two. And if people think it is a valid concept and want it on wiki, then we must absolutely require them to examine carefully whether their wishes can be achieved with the suggested sources and without OR. The policy requirements are clear: it is up to those adding information to Wikipedia to demonstrate that this can be verified by reliable sources and does not require original research to calculate, synthesise or select records from a database. I wish that clarity, in my mind, was a no brainer, but it seems we have to ask -- nobody responded to the noticeboard request to examine the OR. --Colin°Talk 17:10, 16 December 2019 (UTC)
Of course there are subsequent questions for a later RFC if that is required, such as what should go in the lead, or infobox or body, and which prices we might show. Again the latter is in danger of offering choices we simply don't have. It seems we have different sorts of hard-to-explain "wholesale list price" for each region, never mind the abomination of juxtaposing wholesale and retail list prices. We only seem to have retail prices for the US, and those are "list prices" of dubious worth. So plenty meaty questions could follow on, assuming the first one is accepted. --Colin°Talk 17:14, 16 December 2019 (UTC)
Colin, in taking another look you're right it's really three questions (expressing in dollars and cents, should include, sourcing). That only amplifies my concern about it - such questions are much harder to get clear community consensus and input on. Best, Barkeep49 (talk) 20:09, 16 December 2019 (UTC)
  • I endorse that Colin's basic question or questions must come first. I don't mind that we then go on to ask other questions, even if the answers may be obvious per the first. (I have seen this done in an 'if-then format on other RFCs.) I also don't mind if we do one RFC, starting with Colin's question or questions, and later do another, if that will lend greater clarity.

    I reject pressure to get the wording finalized right away: I am hosting a holiday party this Friday with 200 invitees, and I have done no Christmas shopping. As others advance on the wording, I am following every post, but I cannot initiate work during this stage of December. SandyGeorgia (Talk) 17:25, 16 December 2019 (UTC)

  • I'm not convinced this dispute requires any changes to policy pages, or even necessarily a full-blown RfC. On the "pricing v. prices" issue, I'd venture to say that nobody would disagree that Aspirin should say that aspirin is an inexpensive drug; the question is whether Aspirin should say, as it does now, that aspirin costs two- to two-and-a-half cents "per dose" in "the developing world", sourced to a pricing catalog (a primary source). Why not go to an article like Aspirin and just have a talk page thread about whether the price should included or excluded? And then start a thread here about whether a sentence should be added saying "default include" or "default exclude". I wonder if those threads would end up with quick and clear consensus without even needing an RfC tag. If opinion is divided, then an RfC tag can be applied. So I guess those are the "two questions" I see: a specific example, in or out; and a "default" to apply that specific example to all articles, probably on the MOS page (or maybe elsewhere). To the extent this has already been tried in the past, I'd say try it again, given there are more editors paying attention now. – Levivich 19:42, 16 December 2019 (UTC)
    • Again, there must be an RfC, and it's not optional. (But I see nothing wrong with waiting until after the holidays.) I think that Colin's question(s) do/does get at the basic issue at hand. I can imagine that editors who come to the RfC without being familiar with the discussions here would initially react to that question by wondering how they could answer it: do I think that such prices exist in that way? I'm not sure, and that question sounds like a loaded question that is challenging me to conclude that there must be some kind of problem with sourcing, but I'm not sure what that problem is. For that reason, I'm leaning towards an RfC format that presents editors with (1) an assertion that such prices can be included, and specific reasons why, and (2) an assertion that such prices should not be included, and specific reasons why. And then, editors can support one of those and oppose the other. In that way, the reasons for and against will be easy to see, and editors can weigh which argument is the strongest. Also, a proposal or assertion of this type is less restrictive in terms of having a single question or two questions. --Tryptofish (talk) 20:18, 16 December 2019 (UTC)
Sure we could start with "do you think the approximate price of a medication can be estimated for various regions of the world?" And "Should Wikipedia articles contain information about the cost of medications?" Doc James (talk · contribs · email) 23:50, 16 December 2019 (UTC)
I think that the first of those two questions is problematic, because it fails to capture all of the issues that play into the disagreements here. If reworded, the second question does seem more apt, although I don't think that "cost" is the correct word to use. But in any case, the more I think about it, the less I like the idea of asking yes/no questions. First of all, there must be some explanation of why a particular editorial approach is either good or a problem. It's not nearly enough to frame things in terms of "approximate" and "estimated", because one editor's "estimate" might be another editor's "original research". For those reasons and more, I think it would be much better to present responding editors with specific examples of how one might write about drug prices, with explanations, and ask them to support or oppose those proposed approaches. If we don't have specificity, there will be no agreement about how to implement the RfC results. And if there is going to be specificity, there might as well be one example that reflects the position of some of the editors here, another example that reflects a different position held by other editors, and so on – and then ask the community to endorse one or more of those concrete formulations and oppose one or more of the others. --Tryptofish (talk) 21:22, 17 December 2019 (UTC)
Tryptofish, I would like to see one or two such proposals from you. If you write one, others might be able to write similar ones, to show their idea of the best approach. WhatamIdoing (talk) 05:27, 18 December 2019 (UTC)
That's a good suggestion, thanks. I had actually been starting to think the same thing. On the other hand, I think there are several other editors who have been working on these problems a lot longer than I have, and understand the issues a lot better than I do. So I'm thinking that I can create some examples of the formatting, but leave it to others to fill in the details once they can see what I'm talking about. I was traveling all day yesterday and I'll have a very unreliable internet connection through the weekend, but I'll try to get on it soon. --Tryptofish (talk) 16:11, 19 December 2019 (UTC)
James's questions are problematic because a "medication" could be viewed as a packet of pills with a bar code: we need to be clear the scope is the article scope, because that's the scope that has been done several hundred times already. "the wholesale cost is" is article-scope. We claim this entire drug has a price in dollars and cents, and also apparently, has a dose. One. Not many. The "can be estimated" is asking people to speculate and is answered true if you can do it for even one drug. And when people think of "estimate" they think of rounded numbers or even adjectives, not of "$43.77" or even "from $38.34 to $99.55 per dose" which isn't estimating. We mustn't ask for permission for something that is ambiguous or tangential to the problem with the current approach to drug prices. The second question is clearly broken. Clearly some of our articles should (though we still need to avoid OR and incorrect use of sources), and that's permitted by WP:NOTPRICES. Below, I see an argument between two editors where one wants every sentence and sub-clause and fact in the lead to be followed by at least one inline citation number. And the other wants far fewer. The claim is "No guideline prohibits addition citations to leads", which is used to suggest that citation proliferation is permitted. We have the same problem with this second question: it would be taken as permission to add them to every drug, to add them to (often only) the lead, and to do it in a way that misleads our users and breaks OR. I don't think either of those questions are acceptable or will resolve the problem. I think the one I proposed (clarified to avoid the misunderstanding below) is necessary. -- Colin°Talk 15:20, 18 December 2019 (UTC)
User:Colin, the question that you wrote needs some tweaking, or perhaps should just be replaced by a statement of the facts. Do I think that a pharmaceutical drug has a price that can be expressed in dollars and cents, for regions such as US? Yes. In fact, I think that a drug has multiple prices that can be expressed in dollars and cents in the US: AMP and WAC and retail price so forth. I think it is possible for us to write, without violating WP:NOR, a sentence or two about one or more of these prices for most drugs. It would have to be specific, e.g., "Shortly after its original approval in 2010, Medicaid's US average wholesale price for a 60-mg prefilled syringe of denosumab was reported at US$990 per dose, with two doses expected each year to treat osteoporosis.[[[PMC (identifier)|PMC]] 2957751]". Is that "the only" price? No. But it is "a price", and I do not believe that it violates WP:NOR. WhatamIdoing (talk) 06:01, 18 December 2019 (UTC)
WhatamIdoing, where is a difference between "a" and "multiple" that is key to my question, and I agree we need to be extremely specific: but being specific is not what has been done in hundreds of articles where we have been misleadingly vague, and is impractical for most drugs. The stubbornness that we can simplify basic database record queries down to single prices for the US, UK & developing world, is really why we are here facing an RFC. So perhaps the question needs rephrased and I think the question would also be clearer with examples and/or opening statement that summarises the problem.
Most medicines are available in multiple formulations, strengths, package size and brand. There are a minority of typically new and expensive medicines where their cost is well documented and not really for debate here (other than to consider OR or cherry picking examples, etc). Many medicines are used for multiple indications. The dose and duration of treatment varies with indication and patient characteristics and simple trial-end-error. There are some medicines with one simple dose regime (e.g. contraceptives) but even then there are brand choices and price choices. Look at ethinylestradiol with desogestrel. A pack of 63 20mg/150mg formulation costs £5.04 (Bimizza) or £5.08 (Gedarel) or £8.44 (Mercilon) or £5.07 (Munalea). The 30mg/150mg formulation costs £6.13 (Alenvona) or £3.80 (Cimizt) or £4.19 (Gedarel) or £7.10 (Marvelon) or £4.18 (Munalea). And of course, that cost is only borne by the NHS, because contraceptive prescriptions are free in the UK. Which specific price should our article include?
You tried to pick an example where the price was fairly simple. But according to the BNF, denosumab has several dose regimes. There is "60 mg every 6 months" for osteoporosis. Then there is "120 mg every 4 weeks" for bone metastases. Or there is "120 mg every 4 weeks, give additional dose on days 8 and 15 of the first month of treatment only" for giant cell tumour of bone. Try sticking that in an infobox. There are also associated-costs because the treatment has risks that suggest dental examination and treatment are done prior to taking them, and plasma-calcium concentration monitoring required ongoing. While the 60 mg vial costs £183, the 120mg vial only costs £309.86 (a £56 saving on two vials). Those are list prices that the NHS "pays". The people getting this probably pay nothing in the UK (due to age or cancer exemption). But here's the problem: the drug was only approved on a "patient access scheme" with confidential discount. You can read the proposal here -- look at page 6 and you will find black boxes over the discount offered. So the "list price" is simply a fantasy "recommended retail price" that the NHS has agreed to display as part of negotiations. In the US, according to the official website there are discounts on what anyone pays, all they way to $0 for those who are uninsured and unable.
If you do think that being very specific about our prices is an approach to consider, then I don't think even denosumab has information we could explain in the lead, in an infobox, and for the UK we'd have to admit that we really have no idea what price the NHS pays. See how that approach works for valproate -- the most commonly prescribed anticonvulsant and available as generic worldwide. -- Colin°Talk 15:05, 18 December 2019 (UTC)
"Prices vary according to dose and other factors, but Medicaid's US average wholesale price for a <size> pill of valproate was reported in <source> as being US$0.xx on <date>". It might be a WP:DUE violation, but it would not be a NOR violation.
But do you know what I might rather do? I might rather say that PMID 29564159 in the BMJ has estimated the production costs for the active pharmaceutical ingredient to be about US$26/kg. This is likely to be fairly stable and to apply wherever the drugs are made.
People who care about the public-policy-and-patients side more than the chemicals-and-manufacturer's side of that drug will note that the same lines of that table give an estimated generic price of 1.5¢ each for a 100mg pill, 1.9¢ for a 200mg pill, and 2.9¢ for a 500mg pill. I don't think that would violate WP:NOR, although we'd probably need to explain exactly what their phrase "estimated generic price" means. WhatamIdoing (talk) 17:50, 18 December 2019 (UTC)
I think we are getting a little off track, and you missed my joke at WT:MED: "Maybe we should list the wholesale price per kilogramme of the pharmaceutical intermediate [of valproate] in 200kg drums at Alibaba". The paper you cite is primary research, and I'm not sure a bunch of medics are qualified to estimate production costs of pharmaceuticals, and if doing this at a 10% profit was good business, it begs the question why it isn't being done. One thing I did find interesting in the paper was
"Most of the high-priced medicines in India were found only in the private market price source, and not the (Tamil Nadu) government tender list, suggesting a lack of availability in public facilities. Over 75% of health expenditure is out-of-pocket in India, of which the majority is spent on medicines. While we found Indian prices to be below our estimated generic price in many cases, Indian prices were mostly government tender prices, which are likely to be significantly lower than the private market prices more often experienced by those needing medicines in India. Further analysis of the Indian market would be necessary to determine prices available to various facilities, provinces and patient groups"
There has been an argument that we are doing our readers a favour by informing them of the price that Big Pharma want to hide, or because patients in some countries pay out of pocket. The irony is, by quoting wholesale list prices or retail list prices, we are in fact publishing the mythical prices that Big Pharma is happy to display, but which neither government, pharmacies nor patients actually pay. Government is usually paying much less than list. In the US the patients are paying much less than list. In the UK they may often be paying nothing at all or a fixed £9. In India they are paying through the nose because of lack of government-supplied medicine availability, and we have no data on what is actually paid in the developing world, assuming the drug is even reliably available. So we are actually doing Big Pharma a favour, and not actually educating our readers of anything meaningful at all. -- Colin°Talk 19:11, 18 December 2019 (UTC)
  • I've made a tweak to the above proposed question to try to address WhatamIdoing's concern that there may be ambiguity about the scope of "a" meaning "one for the whole article (in a region)" or "one of many possible". Open to other ideas about how to word that. More extensive tweaks or suggestions will probably require to be stated again, rather than done in-place. I think sometimes here (and below) the discussion is a bit deviating towards imagining/thinking about various potential prices (e.g. industrial) or other articles being in scope for discussion at this RFC. I don't think we should expand that far right now. We aren't here because of that. We need to remain grounded in the edits that have been made to hundreds of drug articles, and to MEDMOS, and the intractable nature of the discussions about those edits (article/guideline text + sources). That's why we reached ANI and prompted an RFC, so that is the article text + sources that should be considered. We can all imagine other kinds of price-related edits or subjects to discuss another day. And as Sandy warns below, we need to guard the RFC against permitting all sorts of tangents to be created during the RFC. -- Colin°Talk 08:34, 19 December 2019 (UTC)
    • Colin, if editors actually voted that yes, they did think that there was one single price for common drugs in each area, and that it both could and should be included routinely, what would you do with that (I mean, after the part in which we all groan and wonder whether humanity is doomed)? I am not convinced that this question is going to produce actionable results. WhatamIdoing (talk) 16:33, 23 December 2019 (UTC)
      • WhatamIdoing, what would I do with that? I would have to conclude I was wrong. Other people clearly have different value systems when it comes to Wikipedia. I haven't seen any indication from you or indeed mostly anyone else, that this would be a groan-worthy and humanity-doomed scenario. We've even had the opinion expressed and codified in proposed RFC text below that we might go further and include these "simple" prices in our info boxes, sourced to wikidata. WhatamIdoing, we advertised for comments about whether the source->text broke NOR and not a single person responded. So I'm totally expecting James to be vindicated and if folk think the text in those articles is just dandy, then that's me put in my place. I'm quite used to election results defying rational sense, so it would not be at all unexpected. -- Colin°Talk 18:44, 23 December 2019 (UTC)

Timing

Does anyone urgently need this RFC to start in the next week or two? I propose that – unless there are significant objections – we plan to start the RFC no sooner than Thursday, January 2nd. Does anyone object to that? WhatamIdoing (talk) 05:30, 18 December 2019 (UTC)

That should work for me. SandyGeorgia (Talk) 14:29, 18 December 2019 (UTC)
Certainly not during the holiday period. -- Colin°Talk 08:36, 19 December 2019 (UTC)
Yes, perfect for me. I am sorry for falling silent recently. Had a couple of conferences, international travels, etc., over the last weeks which taken together consumed most of my energy. Now let's have a peaceful Christmas break and get back to work on the 2nd. — kashmīrī TALK 14:35, 19 December 2019 (UTC)
I, too, agree. I think it should definitely not be any earlier than that, and may even need to be a bit later. --Tryptofish (talk) 16:24, 19 December 2019 (UTC)
"A bit later" is perfectly fine with me, too. WhatamIdoing (talk) 16:35, 23 December 2019 (UTC)

Name the RFC

It does not appear that anyone is taking the lead on this; it was clearly stated on the ANI that Colin should not be the one to formulate the RFC, but no one else has. I have repetitiously suggested ad nauseum above that we need a name and a draft space to start work. Please pick a sub-page so that we can use it, with an Under construction tag until ready for launch. The situation is complex, so how to name it is a problem; the name should not mislead. Let's not obsess on this; the name can be changed as we work (via a page move), but we need a starting place, that can be labeled as "under construction". SandyGeorgia (Talk) 17:23, 17 December 2019 (UTC)

  1. RFC on NOPRICE and pharmaceutical drugs
  2. RFC on pharmaceutical drug content
  3. RFC on pharmaceutical drug pricing and sources
  4. RFC on pharmaceutical drug prices and sources

Please pick one, or add suggestions. SandyGeorgia (Talk) 17:27, 17 December 2019 (UTC)

  • All of them work for me. I prefer 3, because application of NOPRICE is not at the core of this dispute-- the proper use of sources relative to NOR is. We also have the Ronz notion of prices/pricing. I believe 3 to be a correct formulation of the essence of the dispute. SandyGeorgia (Talk) 17:34, 17 December 2019 (UTC)
    Option 4 works for me as well; in either case, we have to define our terms. SandyGeorgia (Talk)
  • Support 3. I agree with you, and I think that 3 is simply the most informative and neutral of those options. And thank you for moving this along. --Tryptofish (talk) 21:07, 17 December 2019 (UTC)
    • Update: Support 4 as first choice and 3 as a close second choice.. Either of those is OK with me, but I think that 4 puts the spotlight on the center of the disagreement that needs to be resolved. --Tryptofish (talk) 16:15, 19 December 2019 (UTC)
  • Support 3. The issues are broader than NOPRICE, while certainly more specific than all pharmaceutical drug content. --Ronz (talk) 23:26, 17 December 2019 (UTC)
  • Do we only care about pharmaceutical drug prices? Should this RFC cover other types of prices/pricing/costs? WhatamIdoing (talk) 05:26, 18 December 2019 (UTC)
    • I remain very concerned that if we try to do too much in this RFC, we will not get a usable result. The disputed matter right now really boils down to how a few sources are being used, by not very many editors, in very defined ways. I suspect that by addressing that issue first, we will solve any disputes not already addressed by NOPRICE. I really appreciate the work you've done to show that we have similar potential problems with medical devices, etc, but I feel those are bridges we should cross separately, so as not to muddy an already complex issue. SandyGeorgia (Talk) 14:33, 18 December 2019 (UTC)
      • I think any single RFC is going to be difficult. The first question could probably be more generic, however. There's a line of argument that says money matters more in healthcare industry than it does in, say, the tech industry, so NOTPRICES applies to mobile phones but not to drugs. That could be addressed in the general case, inclusive of all medicine-related financial content. NB that I'm not saying that the first question needs to be more generic; if you all would prefer to talk only about drugs, that's okay with me. WhatamIdoing (talk) 17:24, 18 December 2019 (UTC)
      • I'm a bit confused that #3 says "pricing" especially given Ronz distinction of pricing being unarguably important but "Prices may not exist in a form that is suitable for encyclopedia articles at all, as all the discussions are indicating". So I think the question is "drug prices" not "drug pricing". -- Colin°Talk 19:15, 18 December 2019 (UTC)
        • I thought the preamble of the RFC would define the Ronz concept of price/pricing early on. If we use "prices" in the RFC title, people may automatically go straight to NOPRICE and not engage the distinction. IF we use "pricing", I am hoping they will be more likely to digest the distinction we are making in the discussions here. SandyGeorgia (Talk) 19:46, 18 December 2019 (UTC)

User:Barkeep49, do you have some examples of RFCs (in the same approximate area of content) you think worked well, and perhaps some you think were bad because of the question(s) asked. -- Colin°Talk 20:38, 18 December 2019 (UTC)

Colin, good question. Give me a couple days to see what I can dig up. Best, Barkeep49 (talk) 20:53, 18 December 2019 (UTC)
Well, here's an epic fail, for which I am (only partly) responsible, and providing examples of all the things we are trying to avoid here: Wikipedia:WikiProject Medicine/RFC on medical disclaimer. It started out with a few choices, to show how it might be done, and then as the RFC was running, other editors kept tacking on other sample ideas, to the point that we ended up with nothing. Tryptofish, have a look at this in terms of what can happen when you offer samples ... that the participants will grow to their own idea of what can be done. The RFC became so confusing and diluted that it became a nothing-burger. It should have been a yes-no question: does Wikipedia need a medical dislaimer. There's lots more wrong with that RFC, so it should serve as a good example of what to avoid. SandyGeorgia (Talk) 03:44, 19 December 2019 (UTC)
Thanks Sandy. As I noted above, I'm traveling and will have an antique internet connection for the next few days, but I will definitely look into that. I agree that we must ensure that there will be a usable outcome. (Perhaps one solution would be to disallow adding new options after the RfC has opened, or perhaps not. I think there are significant dangers of inconclusive results with almost any RfC format, and it's essential that we figure that out before going live with the RfC, even if it takes a little longer than expected. Better to get it right than to get it fast.) --Tryptofish (talk) 16:21, 19 December 2019 (UTC)
Thanks for the link to that earlier RfC. I've studied it, and I think I need to take a drug now. I fully agree with you that we need to avoid that. I don't claim to have all the answers, but something that stood out to me right away is that editors were presented with a list of choices for the disclaimer, of which to approve or disapprove, and then were presented with views by individual editors about which choices were good or bad, and ultimately had to support or oppose those secondary views. That clearly did not work. Also, there was a lot of background information to work through. I think it might have been better to ask editors to support or oppose each of the disclaimer versions, as opposed to supporting or opposing what other editors thought about those disclaimer versions. Again, I don't have all the answers, but I'm leaning towards giving editors a limited number of specific choices, giving them succinct reasons for and against each of those choices, and asking each editor to support or oppose the various choices directly (as opposed to supporting what somebody else said about it). (At WP:GMORFC, there were a whopping 23 different options, and yet it was quite clear at the end of the month what the community consensus was.) --Tryptofish (talk) 16:33, 22 December 2019 (UTC)
Re, "I think I need to take a drug now", pass the prozac. I think I can afford it (although I notice the a price is mentioned in the lead, with no mention whatsoever in the body of the article, WP:LEAD problem). I was partly responsible for the dismal structure, but did not foresee it would get so out of hand with alternate options tacked on. SandyGeorgia (Talk) 16:52, 22 December 2019 (UTC)

Structure of the RfC

I've created User:Tryptofish/Drug prices RfC draft. Some editors here suggested that I should show what it might look like if we structured the RfC around propositions to support or oppose, as opposed to questions to answer yes/no. So that's what I've tried to do there. I don't think it's anywhere near to being ready for prime time. It's just an example to start some discussion. What I'm trying to do is to (1) pin down what actually might appear on pages (as opposed to generalized descriptions that might mean different things to different people), and (2) get responding editors to say clearly that some options are OK and others are not. --Tryptofish (talk) 16:41, 22 December 2019 (UTC)

Considering my epic fail discussed just above this, I will wait for others to weigh in. SandyGeorgia (Talk) 16:53, 22 December 2019 (UTC)

I really really don't like the 6 propositions options (individually and the idea of having 6). Can't see how that wouldn't be a total disaster. Can we try to be consistently clear about our terminology. "Drug pricing" is not the same as "drug prices". The former is the whole social/economic/policical/health issue of how to set the price of drugs, and the consequences of those decisions or methods. Drug prices are numbers. We really aren't discussing whether articles should mention "drug pricing". The question "a simple listing of a price or price range may be cited to a primary source" is loaded unfairly -- if there is one continuing theme about the discussion over all these pages is that (a) the person putting up the price thought and still thinks it was simple but (b) it turned out to be anything but. I don't think we should ask questions that subvert WP:NOR, but instead ask people to judge whether it can be done while obeying policy. For example, here's what WHO say in their essential medicines review report: "The MSH (Management Sciences for Health) International Medical Products Price Guide 2015 reports a median supplier unit price for tramadol hydrochloride 50-mg tablet/capsule of US$ 0.0427.". Not once has any price in our articles even approached that level of direct reporting of a primary source, and unlike the WHO researchers, we can't arbitrarily pick the 50mg tablet vs the 100 mg/ml solution, 100 mg/ml ampoule, 50 mg/ml ampoule, or the controlled-release capsules in 50, 100, 150, 200, 300, 400mg sizes.

I think we need to stick to a simpler one question design, and forget for now the fanciful ideas about info boxes and external links. We are here because of the seriously problematic prices in hundreds of articles leads, let's stick to discussing and resolving that problem first. Tryptofish, if you think, after all this discussion about the problems, that this will be resolved by someone voting "There's nothing wrong with sourcing a drug price to a primary source, and readers need to know this information" then we really have all been wasting our time and have failed to explain the issue.

I'll repeat my suggestion below:

"Do you think that a pharmaceutical drug has one price that can be expressed in dollars and cents, for each region such as US, UK, developing world, and this is information that Wikipedia drug articles should include and can routinely be sourced by editors without original research."

I think this works because it requires the participant to explain why (or at least, agree with someone else's explanation). And the above should include examples from articles, and the good-idea/problematic sections like your proposal so people get some info up-front. -- Colin°Talk 09:27, 23 December 2019 (UTC)

I'm fine with using whatever format for the RfC that gets consensus from editors here. But I do want to emphasize that the examples that I gave are not actual proposals for how the RfC would be worded. I'm only trying to show a possible format. And I've said explicitly that the number of questions could be very different than what I put there. And I certainly wasn't trying to illustrate the best possible RfC responses: those were just some rather crude examples and nothing more. So let's please not get bogged down in the way that sentence was written on Tryptofish's draft page is bad, so we have to reject the entire approach shown there.
Since you do appear to be proposing the specific wording for your single question, I will repeat, in turn, my opinion that the question as you wrote it is not really neutrally worded, because it subtly implies that anyone who would answer "yes" must be making a mistake. Also, it's hard for me to envision a way that your single-question format would allow us to show responding editors what it would actually look like on a given page. Let's see if we can combine the better parts of your approach with the better parts of mine.
I do agree with you that the issue contained in your proposed question is the most essential one here. And I'm happy to agree that other stuff (infoboxes, external links, and so forth) can be omitted, assuming that other editors here will agree with that. And it sounds like we both believe that it's useful to provide "for" and "against" reasons.
So, how about this:
Proposition: A pharmaceutical drug has one price that can be expressed in dollars and cents, for each region such as US, UK, developing world, and this is information that Wikipedia drug articles should include and can routinely be sourced by editors without original research.
Supporting arguments:
  1. reason
  2. reason
  3. etc.
Example of specific page content that is written assuming the proposition is true.[1][2]
Opposing arguments:
  1. reason
  2. reason
  3. etc.
Example of specific page content that is written assuming the proposition is false, and showing how to write it correctly based on that assumption.[3][4]

References

  1. ^ cite1
  2. ^ cite2
  3. ^ cite3
  4. ^ cite4
That would be in effect a single question, and all I did was omit the first few words of your question. But it presents the issue as two alternatives, one of which could be supported and the other opposed. If that works for you, I can happily support it, although I anticipate that other editors will want to discuss the exact wording. --Tryptofish (talk) 15:12, 23 December 2019 (UTC)
Tryptofish, there isn't an "example" of text that assumes the proposition is false. The conclusion would be the current approach to listing prices is wrong and needs removed. If people then want to finesse the cases where prices might appear and what sort of sources to use, that's a question for another day. The "blockage" the RFC is trying to resolve is whether the approach on hundreds of existing articles is acceptable at all, not to merely tweak at the wording. -- Colin°Talk 19:14, 23 December 2019 (UTC)
The problem with that proposition is that we're asking editors to vote on whether certain verifiable facts (e.g., it's usually more complicated than "one price") are actually facts, which is not an appropriate subject for voting. What if they all vote in favor of the non-factual side? WhatamIdoing (talk) 16:50, 23 December 2019 (UTC)
WhatamIdoing, it may be your opinion that it is usually more complicated than "one price", and I agree with you, but we have hundreds of articles clearly stating one price for "developing countries", one price for US, etc. And despite me complaining about that for a month, virtually nobody has backed me up on this. I've repeatedly pointed out multiple pill sizes and formulations and nobody cared. I've repeatedly pointed out multiple indications and other factors mean there is not one "dose" but nobody seemed concerned. I've repeatedly pointed out that we are doing original research and again nobody seems to mind. I've found mistakes in nearly every price and nobody even shrugged. We advertised for people to comment on whether the current text broke NOR and nobody came. If they vote in favour, then Wiki gets to keep these prices in the lead of hundreds of articles. The "facts" haven't mattered so far. -- Colin°Talk 19:02, 23 December 2019 (UTC)

What is the reasoning behind adding to articles prices of drugs in developing vs developed nations? Is it meant to imply or inform readers that drug companies are overcharging people in developed nations?--Literaturegeek | T@1k? 20:05, 23 December 2019 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Manual_of_Style/Medicine-related_articles/Archive_11&oldid=943063316"