Talk:Abortion/Archive 36

Archive 30

←

Archive 34

Archive 35

Archive 36

Archive 37

Archive 38

→

Archive 40

Reference to settle the debate

This ref states [1]

Death following legal abortion induced in appropriately equipped and staffed medical settings is very rare, with rates ranging from zero to two deaths per 100,000 procedures in the 13 countries for which accurate statistics are available (3). The aggregate mortality rate for these countries is 0.6 deaths per 100,000 legal abortions; this rate is lower than that of tonsillectomy and makes induced abortion about ten times safer than pregnancy carried to term.

I propose summarizing this as: "In 13 countries with proper facilities and legally allowed abortion the average mortality was 0.6 per 100,000 procedures which was ten fold safer than a live birth." Doc James (talk · contribs · email) 21:25, 8 June 2010 (UTC)

Why would you leave out the words "for which accurate statistics are available"? They are significant. Does the source tell us which countries? HiLo48 (talk) 21:35, 8 June 2010 (UTC)

We would not be quoting it if these statistics were not accurate so in the context of Wikipedia it seems a little redundant.Doc James (talk · contribs · email) 22:20, 8 June 2010 (UTC)

That source is pretty dated and in any case only pertains to the developed world, which grossly misses the point. Pregnancy, childbirth and illegal abortion are all vastly more dangerous for poor women in developing countries. They are also vastly more common there. LeadSongDog come howl! 22:32, 8 June 2010 (UTC)

This page is about both legal and illegal abortions. No one disagrees that illegal abortions are way more dangerous.Doc James (talk · contribs · email) 22:52, 8 June 2010 (UTC)

I assume that James is proposing we use this language/ref to address safe abortion. We would still have separate language addressing unsafe abortion (the majority of which takes place in the developing world), I presume. As LeadSongDog notes (and I don't think anyone disagrees), the risks of unsafe/illegal abortion are exponentially greater. MastCell ^Talk 23:00, 8 June 2010 (UTC)

Please note that there was already a long discussion above about whether to get into a comparison to childbirth. Several editors have objected to doing that. I initially had no objection (and even inserted it into the article myself), but the arguments the other way have persuaded me that it's best not to include it. This article needs to stay on topic, and not digress into tonsillectomies or childbirth, or what have you. It's fine to give the death rate for abortion, but giving it for certain other procedures is problematic. We already refer readers to Family_planning#Risk_of_death which gets into the risks not just of childbirth but of other family planning methods (e.g. contraceptives). So, let's please stick with the earlier consensus according to which we don't compare to childbirth in this article, until we reach consensus to include it. Thanks.Anythingyouwant (talk) 23:03, 8 June 2010 (UTC)

(ECx2)We already refer readers to Family_planning#Risk_of_death Ok, let's refer to a section of another article that was created today (which, BTW, I still oppose and would still like to remove) for the purposes of bolstering an argument to exclude content here. If that section "Risk of death" is deleted from that article, what then... Just saying the argument seems a bit circular (or the conditions cleverly orchestrated to support one side)-Andrew c ^[talk] 23:12, 8 June 2010 (UTC)

Ah, I see that a certain admin has just edit-warred this back into the article.[2] Not surprising. Per WP:BRD, "If an issue is already under discussion or was recently discussed, people may take offense if you boldly ignore the discussion, especially if you make a change away from a version arrived at through consensus, to an earlier or suggested non-consensual version. Ignoring earlier consensus is generally not a wise approach!" So, if a revert cycle is to be broken, it is best to break toward the earlier consensus. The edit-warring admin has not said one word at this talk page about staying on topic or the family planning article, both of which I've mentioned repeatedly above (and in edit summaries).Anythingyouwant (talk) 23:09, 8 June 2010 (UTC)

The comparison to childbirth is explicitly made by numerous expert, reliable sources as a key element of contextualizing the risk of abortion. It is thus "on topic". The objection to our mention of it seems to be based on personal opinions and viewpoints of anonymous Wikipedians. Perhaps you could elaborate on a policy-based reason for excluding this comparison, which numerous expert sources deem appropriate and relevant? MastCell ^Talk 23:16, 8 June 2010 (UTC)

It is redundant to what is at the family planning article. It also omits the comparison to tonsillectomy (at the top of this talk page section), and omits the risk of contraceptives, presumably to advance the POV that women should opt for abortion instead of childbirth. More importantly, you're edit-warring, MastCell. You slap a bad faith label on everyone who disagrees with you about this, and disparage us as "anonymous" (as if your identity is known to all). I cited Google search results above that indicate only a very small fraction of books about abortion compare its risks to those of childbirth. The Wikipedia guideline about staying on topic instructs us to avoid redundancy between articles. This has nothing to do with personal opinions and viewpoints, at least at my end. You blasted me a few days ago for including the comparison to childbirth in this article, and now you're blasting me (and other editors) for taking it out; this seems unfair.Anythingyouwant (talk) 23:22, 8 June 2010 (UTC)

Correction: Andrew c has now removed all the information about risk from the family planning article.[3] I must say, that's not surprising either.Anythingyouwant (talk) 23:28, 8 June 2010 (UTC)

I could easily find 20 references the makes the comparison between child birth, abortion and maternal mortality. How many review articles does one need? This point is mentioned in the exact context of this articles subject in top quality recent reviews, governmental sources, and textbooks. It is not controversial within the field of obstetrics and gynecology. I do not understand why it is controversial here. None of the reasons for removal make any sense. It does not matter if it is discussed elsewhere it should still be discussed here.Doc James (talk · contribs · email) 23:34, 8 June 2010 (UTC)

There is a HUGE amount of information about abortion in reliable sources. We have to be somewhat selective here. Just because something is in 20 references does not mean it belongs here in this article. In fact, if it's only in 20 out of the many thousands of reliable sources on this subject, then that might be a very good reason for not including it. In any event, I was for including it in another more appropriate article, and linking from this article. I'm not going to argue about this further today, since arguments are not addressed, and material is jammed into this article regardless.Anythingyouwant (talk) 23:41, 8 June 2010 (UTC)

I do not see any good arguement to exclude it. Uptodate makes the comparison so there is no reason why we should not. Providing numbers without context is close to meaningless. We need to provide context.Doc James (talk · contribs · email) 23:47, 8 June 2010 (UTC)

The Geneva Foundation for Medical Education and Research (1993) source predates widespread access to mifepristone in many of the developed countries. If that point does need making can we at least find something more current to get more accurate numbers? LeadSongDog come howl! 23:55, 8 June 2010 (UTC)

We have it referenced with the article from the US in 2006. The The Geneva Foundation for Medical Education and Research is part of the World Health Organization. I have added a 2009 medical text.Doc James (talk · contribs · email) 23:57, 8 June 2010 (UTC)

I've a problem with this source, I can't find a pubmed or ISBN for it, looks like self-published

• "Induced abortion". Geneva Foundation for Medical Education and Research World Health Organization.

Doc James, GFMER doesn't belong to the WHO, it's a non-profit organisation focused in "Sexual and reproductive health" that collaborates with the WHO, it's alike of "The Guttmacher Institute"--Nutriveg (talk) 03:50, 9 June 2010 (UTC)

Mastcell reverted my edit saying the WHO published that document, please provide the source supporting that information.--Nutriveg (talk) 04:10, 9 June 2010 (UTC)

You removed it as a "self-published sources", which is clearly inappropriate. Material published by the WHO, or by GFMER, is not "self-published" in the Wikipedia sense, just as scholarly material from the National Cancer Institute, the American Cancer Society, or the US Surgeon General is not "self-published". GFMER is a non-profit associated with Geneva University, and works in close partnership with the WHO. The authors of that particular source are from the WHO. It appears to be a reasonably expert body, and certainly qualifies under WP:MEDRS. MastCell ^Talk 04:18, 9 June 2010 (UTC)

Make yourself clear, the authors work for the WHO or it was published by the WHO? If the later is true please provide a source (like a PMID or ISBN) supporting that information. The only source supporting it was ever published is the GFMER website itself.--Nutriveg (talk) 04:24, 9 June 2010 (UTC)

The publication series is listed as being in the NLM (NLMID 9602354) OCLC 34525562 and a few other libraries, but not a great many. Perhaps someone with ready access to NLM, BL, or CISTI can check the details agree with the gfmer website's version of the title page.LeadSongDog come howl! 05:20, 9 June 2010 (UTC)

I think that doesn't matter any more, that data is too old, from 1990, and so doesn't respect WP:MEDRS#Use up-to-date evidence.--Nutriveg (talk) 05:26, 9 June 2010 (UTC)

Yes it does look like it is just a collaboration of the WHO rather than the WHO my mistake.Doc James (talk · contribs · email) 06:21, 9 June 2010 (UTC)

The rate hasn't changed at all, really. It's been remarkably consistent from the 1990 ref through the mid-to-late 2000's, so I don't see that we're providing out-of-date information. We are using up-to-date evidence, by the way - note the other 3 cites attached to that sentence. I'm a bit confused as to why we seem to be throwing everything in MEDRS against the wall to see what sticks, all in the name of keeping out a piece of information that reliable expert sources clearly deem relevant. The spirit of MEDRS is to help us create high-quality, informative medical coverage that accurately reflects scholarly knowledge. For all the citations of the guideline, I think we're actually moving away from its intended goal. I'll leave it to others to comment further, because I'm finding Nutriveg's objections increasingly arbitrary and unreasonable. MastCell ^Talk 06:25, 9 June 2010 (UTC)

The other sources don't support the previous text. So far this is just an old study self-published by a NGO. It's not my problem if respected epidemiologists don't make such poor comparisons as the one that has been tried to be included in the article and the only worldwide source you can find is an old one, so old that it doesn't comply with WP:MEDRS#Use up-to-date evidence.--Nutriveg (talk) 12:37, 9 June 2010 (UTC)

You mean like UpToDate (which directly compares abortion risk to childbirth risk)? The cited reliable sources make the comparison, so I'm not sure why you believe that respectable sources don't make it. You've rejected Grimes 2004 as "US-centric" (which is a reason to balance our coverage with more worldwide perspectives, not a reason to simply delete an otherwise informative and reliable source). You've rejected Grimes & Creinin because... why, exactly? You've rejected Beckmann 2009 (the standard OB/GYN textbook)... why, exactly? (It draws the comparison on p. 150). You've rejected the GFMER source because it's "self-published" (incorrect), and because it's "out of date" (although the rate has not changed since the paper was published, so its information is not in fact out of date). Perhaps you could clarify, because it's a waste of time to keep finding good sources in this sort of environment. MastCell ^Talk 16:59, 9 June 2010 (UTC)

As pointed before, Uptodate uses US numbers when comparing those two: 0.567, so it's not a worldwide source, the same problem of Grimes.

Doc James presented Beckmann as a US source "We should than include a comparision such as in the US this compares favorable with normal pregancy which has a mortality of 7.7". Later, when I didn't realize that was the same source being used, I couldn't find the citation in Beckmann since I was searching for the word "safer" which in fact it doesn't have. Now I have found the text but it doesn't say it's safer, it only present the numbers. I doesn't say where that 7.7 number come is from, but that whole section only references US data and facts, the is book published by the " American College of Obstetricians and Gynecologists" and this 7.7 number has been reported as a US one "To discern possible trends in maternal mortality, data were divided into two 5-year periods (1987-1991 and 1992-1996). The national MMR was 7.7 for each time period."

So repeating sources that were already debunked as US centric won't resolve this issue doing original research to apply US numbers to the world or to making comparison where it wasn't made.

GFMER is a 17 old source when WP:MEDRS#Use up-to-date evidence says "Look for reviews published in the last five years or so, preferably in the last two or three years" which is an obvious thing since a review of still older studies is useless. The publishing status of that source can not be checked as it is not properly formatted as such

Stop searching for data you can't find, no respectable epidemiologist would make such poor comparison and deaths caused by abortion are usually counted in general maternal death, by the WHO/ICD definition of maternal death, so that number will always be higher and incomparable by respected scientists.--Nutriveg (talk) 18:45, 9 June 2010 (UTC)

It's a bit frustrating to have you repeatedly assert things that are clearly incorrect, and to persist in making untrue statements despite correction. Numerous reliable and scholarly sources make this comparison. It is untrue to say that no respectable source would do so, unless you're trying to illustrate the No true Scotsman fallacy.

Also, calling a source "US-centric" doesn't "debunk" it, and I'm not sure where you got that idea. It does mean we should try to balance it with more global sources, but the fact that a study was performed in the US does not render the source's content untrue or "debunked". MastCell ^Talk 22:15, 9 June 2010 (UTC)

I didn't insist in anything after evidence were provided if that looks different is because discussion of a single source is often broken between different sections of this talk page, while you often insist in using sources without addressing their reported problems. As I've said numerous times my problem with US centric sources is mainly the one of applying them in a worldwide context.--Nutriveg (talk) 22:53, 9 June 2010 (UTC)

You keep claiming that "no respectable scientist" would compare abortion risk to childbirth risk. You've been provided with nearly a dozen examples in which respectable scientists do exactly that (and there are more). Yet you keep making the same assertion, as recently as two posts above this one. Do you see what I mean? MastCell ^Talk 23:17, 9 June 2010 (UTC)

To make myself clear that was mortality risk, and (mainly) epidemiologist. I pointed that because I have not seen so far data to support such comparison in a worldwide level since, maybe due to how ICD defines it, maternal death data isn't so stratified to allow such epidemiological study, so what most are doing is directing comparing which is also clear by their presentation of those numbers. I do think those who may do such direct comparison do a poor job, but I don't oppose citing them I just don't believe their work is easily available in editorially reviewed publications.--Nutriveg (talk) 23:37, 9 June 2010 (UTC)

(undent) I have provided another 4 references for people perusal.Doc James (talk · contribs · email) 17:24, 9 June 2010 (UTC)

Thanks, I'll comment those in the section you created.--Nutriveg (talk) 19:15, 9 June 2010 (UTC)

I think the above source is fine for use on Wikipedia. We also have many other great sources which confirm these conclusion.Doc James (talk · contribs · email) 19:44, 9 June 2010 (UTC)

I see you have a personal opinion, maybe you should first change WP:MEDRS to reflect the use of unreferenced phrases (and rounded numbers) from two decades old, poorly cited (no pmid or isbn) source with no signs of editorial review.--Nutriveg (talk) 19:53, 9 June 2010 (UTC)

(Sigh) Let see the GFER has a peer reviewed journal

'Reproductive Health' - the official journal of GFMER As a peer reviewed, online, open-access journal, 'Reproductive Health' offers a platform for scientists to publish results of good quality research in human reproduction. The journal is supported by an international editorial board experienced in reproductive health and peer review thus ensuring the quality of the journal. Published articles are freely and universally accessible online and are indexed in PubMed. Information about submitting a manuscript can be found at 'instructions for authors'.

[4] The problem with pubmed is it is very US centric. Much of the rest of the world is not on pubmed as it is run by the US NIH.

The rest of the title says

Special Programme of Research, Development and Research Training in Human Reproduction, World Health Organization, 1211 Geneva 27, Switzerland

It mentioned the World Health Organization in the title. Doc James (talk · contribs · email) 19:58, 9 June 2010 (UTC)

I got that wrong at first as well, James - the attribution of WHO is actually to the authors, both of whom work for WHO. --RexxS (talk) 20:31, 9 June 2010 (UTC)

Misrepresenting sources

I've removed a sentence on contraindications from the article, and wanted to register a concern here. Please compare the actual source to our representation of it:

• Our article: Women who have uterine anomalies, leiomyomas or had previous difficult first-trimester abortion are contraindicated to undertake surgical abortion.
• Source: Abortion should not be undertaken for women who have known uterine anomalies or leiomyomas or who have previously had difficult first-trimester abortions, unless ultrasonography is immediately available and the surgeon is experienced in its intraoperative use. ([5], emphasis mine)

This is a real problem, because we've misrepresented the source pretty badly. It does not list "contraindications" to abortion. It lists factors which may complicate the procedure, and which may prompt a referral to an experienced surgeon, which is quite different. Can I put in a plea for everyone to spend a few extra minutes double-checking the actual content of the sources they cite? MastCell ^Talk 23:35, 8 June 2010 (UTC)

I did not insert that material, and I have no idea whether there was an innocent mistake by another editor, or an intentional misrepresentation according to your accusation of bad faith. However, I would urge you, MastCell, to include something about contraindications in this article. You've repeatedly removed all such information over the past several days. Surely you can find some way to include info about contraindications.Anythingyouwant (talk) 23:38, 8 June 2010 (UTC)

I assumed it was an unintentional misrepresentation. That's why I encouraged greater care going forward. If I thought it was intentional, I'd take a different approach. I don't remember removing any material on contraindications previously; perhaps you could supply a diff to clarify your concern? I do think that a contraindication is distinct from a "health risk". After all, if a procedure is contraindicated, then it will not be performed, and thus carries no health risk whatsoever. MastCell ^Talk 00:03, 9 June 2010 (UTC)

The word "misrepresent" connotes lying.[6] You would do well to use another word if you don't want to piss people off. As far as removing mention of contraindications, you did that here for example. Good night, MastCell.Anythingyouwant (talk) 00:11, 9 June 2010 (UTC)

Misrepresenting means "representing badly" (def. #2), as in "we have represented this source badly". In any case, since that was my meaning, please accept this as a post hoc correction if you prefer. I'm not sure I can apologize for removing an unsourced lead-in about contraindications 10 days ago, but I remain open to a discussion of them that involves concrete sources. MastCell ^Talk 00:20, 9 June 2010 (UTC)

Stop misrepresenting. The erased material was sourced concretely: "For women without [[contraindication]]s,<ref name=Trupin>Trupin, Suzanne. [http://emedicine.medscape.com/article/252560-overview "Elective Abortion"], Medscape (updated 2010-05-27).</ref>...." Sheesh.Anythingyouwant (talk) 00:29, 9 June 2010 (UTC)

(edit conflict) I can't see the Google books ref, but that obviously doesn't disqualify it. Trupin gives:[7]

• "Absolute contraindications are virtually unknown ... Surgical abortion is contraindicated in patients with hemodynamic instability, profound anemia, and/or profound thrombocytopenia. The conditions should be managed and the context of pregnancy continuation must be considered."

PubMed finds 35 records for a search on contraindication+surgical+abortion but I've found nothing so far discussing contraindications to surgical abortion there.

For contraindications to medical abortion, may I suggest:

• "Medical management of abortion". The National Guideline Clearinghouse. October 2005. Retrieved 9 June 2010.

as a reliable secondary source? --RexxS (talk) 00:41, 9 June 2010 (UTC)

Trupin also describes contraindications for medical abortion (not just surgical abortion): "Medical abortion is contraindicated in patients with clotting disorders, severe liver disease, renal disease, cardiac disease, and chronic steroid use. Medical abortion is also contraindicated in women with no access to emergency services and no partners or family to be with the patient during the heaviest bleeding times." May I suggest that the subject does not warrant elaborate treatment, and that it would be more than adequate to simply mention the word "contraindications" with a footnote to Trupin for the interested reader?Anythingyouwant (talk) 01:45, 9 June 2010 (UTC)

I suspect that not everyone will agree that Trupin is the best source to use if we have alternatives. To some extent, I'm inclined to that view, since she summarises without giving us the leads to her own sources. If you look at the NGC page I cited above, they lay out their sources and methodology clearly, which is why I offered it as an potentially unimpeachable source for the medical abortion part. --RexxS (talk) 01:54, 9 June 2010 (UTC)

MastCell already recommended the Trupin source for other purposes.[8] If another good source is used instead, I have no objection.Anythingyouwant (talk) 02:07, 9 June 2010 (UTC)

The emedicine ref mentioned by anything says

Absolute contraindications are virtually unknown. If abortion presents a medical risk to the patient, then continuation of the pregnancy presents an even greater risk. The type and timing of an abortion procedure or method may be contraindicated based on the medical, surgical, or psychiatric condition of the patient.

Doc James (talk · contribs · email) 00:44, 9 June 2010 (UTC)

I have no serious issue with it being used.Doc James (talk · contribs · email) 04:07, 9 June 2010 (UTC)

The source is fine so long as it is accurately represented; that's been the sticking point, repeatedly. As James notes, the source emphasizes that contraindications are "virtually unknown", so an accurate representation of the source would probably convey that. The source also emphasizes that for any given risky situation, abortion is safer than the alternative of proceeding to childbirth. Again, an accurate discussion of contraindications, or an accurate representation of the source, would convey that.

My view is that since a) contraindications are "virtually unknown", and b) contraindications are distinct from "health risks", we should not discuss them in the "Health risks" section of this article. They might be more appropriately discussed in articles on specific procedures (vacuum aspiration, dilation and extraction, etc). MastCell ^Talk 04:11, 9 June 2010 (UTC)

You'll have to correct my understanding, but I thought that 'relative contraindications' described factors that significantly increased risk, but not enough to completely rule out the drug/procedure, so that judgement is required in making a decision (or the relative contraindications eliminated before treatment). If that's so, I would expect expect to find some mention of contraindications somewhere in the article. Again, I'll gladly be corrected, but my reading of the literature gives me the impression that surgical abortion is effectively free of any contraindications of any kind, while medical abortion has relative contraindications for a number of factors relating to interactions of the two drugs used. As the article grows, is there a point at which it should be switched to summary style, spinning off a daughter article something like 'Abortion (medical procedure)' which would describe in more detail surgical and medical abortion in the same way as other medical procedures/treatments/drugs (I'm thinking of Brachytherapy (GA) and Sertraline (FA) as templates for the layout)? Or are the sociological aspects too closely tied into this topic for that to be realistic? --RexxS (talk) 11:36, 9 June 2010 (UTC)

There's already a Medical abortion article we can create a Surgical abortion one. The original source sounded like a relative contraindication for me so I don't understand why so much noise, it would be easier simply add the word "relative" or complement that citation with the mitigative measures.--Nutriveg (talk) 13:29, 9 June 2010 (UTC)

Ok. The Medical abortion article is only start-class and would benefit from expansion (we have a lot of sources here that could be used) as well as a copyedit to transform much of the list-style information into prose. It would benefit from a structure similar to Sertraline, imho. There's no point in condensing this article until the daughter articles are at least as complete, so I'll put my suggestion hold for now. --RexxS (talk) 14:11, 9 June 2010 (UTC)

After the lesson on "Misrepresenting sources", do we now need a lesson on using your own words, plagiarism/copyvio, and using quotes when you are copying verbatim text? -Andrew c ^[talk] 14:45, 9 June 2010 (UTC)

No, we need helpful people who can fix those minor mistakes instead of creating catch-22 discussions: "misrepresentation" or "too much alike".-Nutriveg (talk) 15:04, 9 June 2010 (UTC)

Can someone propose a change to avoid the copyvio? Then we can make an editprotected request.-Andrew c ^[talk] 23:19, 11 June 2010 (UTC)

Use of MEDRS

WP:MEDRS, by consensus, has the force of WP:RS when considering sources that make medical claims. It imposes a higher standard of reliability for those claims than does WP:RS. However, it has to be read in full and used carefully, as a superficial reading can lead to reliable sources being removed because of a misunderstanding of the guidance it gives.

In the case of WP:MEDRS#Use up-to-date evidence, it contains a "rule-of-thumb" to help us decide between multiple sources. If we don't have recent secondary reviews on the topic concerned, we should not use this section as a tool to delete sources simply on the grounds of age, until more recent sources are found.

In the case of WP:MEDRS#Respect secondary sources, it contains the following points: We cannot use primary studies to contradict reliable secondary sources such as reviews. This does not mean that primary sources are forbidden. Indeed, they may be the only sources available and may be used, albeit carefully, mentioning the study in the text and avoiding any analysis. However, once reviews or other reliable secondary sources are published, those should be substituted for the primary source. The only occasion where removal of a source is clearly indicated is when reviews in the area have been published which ignore a particular primary source – in that case the study is deemed "unimportant" and should be removed. quotes from WP:MEDRS removed per suggestion

I hope editors here will subscribe to my summaries and abide by what WP:MEDRS actually says. In particular, I'm asking that the likely debate on the current addition and removal of sources confines itself to discussing how the actual principles of MEDRS should apply here. --RexxS (talk) 13:57, 9 June 2010 (UTC)

It's better to discuss problematic sources individually in the sections of the talk page where they were brought to and are currently discussed. Text from WP:MEDRS is available for every user so copying it here brings no help, use links as everybody else, pasting large amounts of text here doesn't improve your argument. Create an essay if you want to express your personal views about MEDRS this is not the appropriate place for that.--Nutriveg (talk) 14:24, 9 June 2010 (UTC)

Thanks for taking the time to discuss this issue. As you obviously prefer to be dismissive, let me explain that I placed it here to try to help you could understand how you have misused the guidelines and policy.

The source that you removed here (edit summary: Self published source) is not a self-published source:

• "Induced abortion". Geneva Foundation for Medical Education and Research. World Health Organization.

The publisher is the World Health Organization: per WP:RS a "reliable, third-party, published sources with a reputation for fact-checking and accuracy". Take it to WP:RS if you don't believe me.

In this diff (citing the original source f), you change the text which was based on a secondary source to reflect a primary source - exactly what WP:MEDRS#Respect secondary sources tells you not to do.

In this diff (despite being self-published this is too old and so doesn't respect WP:MEDRS#Use up-to-date evidence), you remove a secondary source that is available in two forms:

• "Induced abortion". Geneva Foundation for Medical Education and Research. World Health Organization. 1993. Retrieved 2010-06-07.
• . Frontiers in Endocrinology. Vol. 2. Rome: Ares Serono Symposia Publications. 1993. OCLC 34525562. {{cite book}}: |work= ignored (help); Missing or empty |title= (help); Unknown parameter |editors= ignored (|editor= suggested) (help)

After tagging the latter with {{failed verification}} despite the fact that a quick Google search shows that it has been cited by others. From {{failed verification}}: "Use this tag only if a source is given, you have checked the source, and found that the source says something other than what is contained in the text, or for whatever reason is illegible or unreadable. Explain in detail on the talk page (my emphasis). MEDRS does not give you the right to just delete sources simply because they are old. You need to be replacing them with newer reviews.

In this diff (removing primary old source by WP:MEDRS#Use up-to-date evidence and WP:MEDRS#]Respect secondary sources), you remove a retrospective chart audit of 1677 medical charts:

• Westfall JM, Sophocles A, Burggraf H, Ellis S (1998). "Manual vacuum aspiration for first-trimester abortion". Arch Fam Med. 7 (6): 559–62. PMID 9821831.{{cite journal}}: CS1 maint: multiple names: authors list (link)

It's a primary source, but you don't suggest why it contradicts a secondary or that it has been excluded from later reviews - the reasons required by MEDRS.

If you are having problems in understanding Wikipedia's sourcing policies and guidelines, please discuss here first, rather than presenting other editors with a fait accompli by removing numerous sources without taking the time to find a consensus for your actions first. You know your removals are contentious, and acting before discussion is inviting others to edit-war. --RexxS (talk) 16:45, 9 June 2010 (UTC)

That first NGO source was discussed in this section where we reached consensus it was not published by the WHO. I described that reference was not found due to lack of a pmid, isbn and results of some basic search, which is still the case since no one was able to check it maybe the template used before was not the best, but it's not being used due to other problems with that source better described in that discussion. Feel free to discuss those issues in the appropriate section now you know where it is.

The other is an old primary source. It's primary, the doctor evaluates the safety of the procedures performed in his own office. Currently it supports no unique information in the text to worth citation, it's too old (1998) to not have been cited in a secondary/tertiary source we could use. And the removal strictly follows WP:MEDRS as an example: "Prefer recent reviews to older primary sources on the same topic".--Nutriveg (talk) 17:17, 9 June 2010 (UTC)

I'm commenting here because of a note left on my User Talk page. Note also that I have previously warned Nutriveg regarding the removal of sources or sourced material using MEDRS as a justification. This ought to be obvious without being explicitly stated, but here it is: MEDRS provides supplementary guidance as to which sources are to be preferred assuming that multiple sources are available. MEDRS should not be used to justify removal of sources or sourced material that, while reliably soured according to WP:RS, is nevertheless not of the very highest standard possible. Again, I note that Nutriveg started removing sources from this article after being told on this talk page that their source was not up to MEDRS standards (though that was not the only problem with their addition) and I reiterate my concern that this course of conduct is disruptive and intended to make a point. SHEFFIELDSTEEL^TALK 18:29, 9 June 2010 (UTC)

(edit conflict) Thanks for the helpful response, Nutriveg. If you don't mind though, I'd rather keep this particular discussion together. I can see the debate about the publisher of the "Induced abortion" source, but I can't see where consensus was reached that it is not a reliable source or is unsuitable for other reasons. Your view seems to be that it's published by the GFMER, rather than the WHO. In that case, why is GFMER not a reliable source? It has articles cited 164 times in Google Scholar, and 7 times in PubMed. I know that's only an indicator (not a guarantee) of a reliable publisher, but what is the rationale suggesting that it isn't RS? Should I make a post at WP:RSN to get some other opinions?

The 1998 primary source was supporting the sentence "Possible complications include hemorrhage, incomplete abortion, uterine or pelvic infection, ongoing intrauterine pregnancy, misdiagnosed/unrecognized ectopic pregnancy, hematometra (in the uterus), uterine perforation and cervical laceration." If you're happy that the Handbook of Obstetric and Gynecologic Emergencies p.258 supports that list, I'll gladly withdraw my objection to the removal of the primary source. I now see that Doc James has been providing more sources, so hopefully much of this will become moot. --RexxS (talk) 18:33, 9 June 2010 (UTC)

SheffieldSteel, you should be aware of your former involvement as an editor in this article. It's not because a source was removed before that it can't ever be removed again, all those changes where justified and explained, stop making unecessary out-of-context warnings on content disputes. I see the message Rexxs left on your talk page but he was the only that made any noise about the removal of a 1998 study which is not used to support any wikipedia article text not supported by the other tertiary sources in the article. His unjustified call for the use of that source is what looks disruptive for me.

RexxS, GFMER is being discussed in that section, there is a whole section about it, I will discuss it there as everybody else. The consensus I pointed was about it not being published by the WHO, you're welcome to join that discussion. I'm glad you finally recognized how useless that 1998 source was.--Nutriveg (talk) 19:06, 9 June 2010 (UTC)

I am quite aware of my involvement in editing this article. My last edit was to revert an edit - by you - which misrepresented its source (a newspaper) in order to push a particular POV. Administrators routinely revert edits which are uncontestably detrimental to Wikipedia; such actions do not render us "involved" to the point where we cannot subsequently use our admin tools. SHEFFIELDSTEEL^TALK 19:52, 9 June 2010 (UTC)

Please refer where is administrative role to get involved in this discussion about content--Nutriveg (talk) 20:03, 9 June 2010 (UTC)

The 1998 primary source probably isn't as useless as you may think: It's been cited at least 60 times :) --RexxS (talk) 20:24, 9 June 2010 (UTC)

It was useful to those articles, don't know in which context (good or bad), but at Wikipedia we don't make science so we prefer secondary/tertiary sources like the one already supporting that same text.--Nutriveg (talk) 20:31, 9 June 2010 (UTC)

Perhaps a little light can be shed on the credibility of the "Induced abortion" source by pointing out that, whoever "published" the book, there were no fewer than eleven WHO contributors listed for various chapters amongst the contributors listed here. The title page shows five editors, one of whom, J. Villar, is listed as "Special Programme of Research, Development and Research Training in Human Reproduction, World Health Organization, 1211 Geneva 27, Switzerland". The specific chapter is listed as being authored by Van Look P (63 pubmed hits) and von Hertzen H (46 pubmed hits, 13 together), both of whom are listed as from the same organization. According to pubmed, their collaboration spans from PMID 8324605 (in 1993) to PMID 20159186 (in 2010), almost entirely focussed on medical abortion and emergency contraception, and including four review articles. GFMER lists their terms of reference as collaborating experts to the WHO. Now does anyone seriously contend this source is a hoax played out by someone claiming to be Aldo Campana, in Geneva, under the very noses of the WHO, simply because they made an online copy available in apparent accord with those terms of reference, or is the sole remaining objection that it is out of date? A refined citation if used, would thus read:

• Van Look PFA, von Hertzen H (1993). "Induced abortion". In Campana A, Dreifuss JJ, Sizonenko P, Vassalli JD, Villar J (eds.) (ed.). Frontiers in Endocrinology. Vol. 2. Reproductive Health. Rome: Ares Serono Symposia Publications. ISBN 9788885974128. OCLC 34525562. Retrieved 2010-06-07. {{cite book}}: |editor= has generic name (help); Missing or empty |title= (help)CS1 maint: multiple names: editors list (link)

The NLM ID 9606196 maps to a longer malformed title, "Reproductive health : postgraduate course for training in reproductive medicine and reproductive biology / Faculty of Medicine, University of Geneva and Special Programme of Research, Development, and Research Training in Human Reproduction, World Health Organization ; editor, A. Campana ; co-editors, J.J. Dreifuss ... [et al.]." LeadSongDog come howl! 20:06, 9 June 2010 (UTC)

Thanks for your effort now we have checked that reference and solved that specific problem. It's not unusual for people which work for an organization to write articles not reviewed by such organizations while still using their work correspondence address that's why I raised that point. I needed to check if it was actually published to confirm if that 1993 book was editorial reviewed.--Nutriveg (talk) 20:22, 9 June 2010 (UTC)

Sources comparing other risks

Nutriveg, you have looked at a lot more sources about this issue than I have. Have you found any sources --- either US centric or not --- that put the risk of death from abortion in PROPER context by comparing it to other risks, or explaning that the risk of death from childbirth is also very low? We do need context in this Wikipedia article, but IMHO the bare comparison to risk of death during childbirth is insufficient and potentially misleading.Anythingyouwant (talk) 19:44, 11 June 2010 (UTC)

There are a boatload of reliable sources listed on this talkpage, which makes "proper context" obvious. When you say "PROPER context", you have in mind not the context provided by experts in the field or by scholarly sources, but the context you personally think is best. You are asking Nutriveg to cherry-pick available sources to find one that agrees with your personal viewpoint. I would strongly suggest (not for the first time) that we follow where the sources lead instead. MastCell ^Talk 20:11, 11 June 2010 (UTC)

No, I am asking if Nutriveg has come across a contextualization of the abortion death risk that HE THINKS would be proper to cite somewhere at Wilipedia according to what he perceives to be Wikipedia policy. Your constant stream of accusations against me is most tiresome MastCell.Anythingyouwant (talk) 20:27, 11 June 2010 (UTC)

Anythingyouwant, worldwide (Lloyd, 2005) does say risk from childbirth is also "low". What other risks are you talking about? There a lot of US centric sources out there but I didn't mind to check because I'm not challenging that information despite the low quality of/access to (pregnancy) health care for poor women in the United States and how those studies usually ignore that when comparing both. Sources presented in Abortion in the United States#Maternal death may be a start for you.

MastCell, assuming bad faith: " You are asking Nutriveg to cherry-pick available sources to find one that agrees with your personal viewpoint" doesn't help. And so what if he finds the sources he's searching? Isn't you the one saying he should follow what sources say? Let's let him find such sources then.--Nutriveg (talk) 20:36, 11 June 2010 (UTC)

I'm not "assuming" anything. We have dozens of high-quality sources on this talk page alone. Yet Anythingyouwant has declared that these sources don't put things in "PROPER context", further mentioning his personal opinion that the sourced comparisons are "insufficient and potentially misleading." It remains a source of deep concern to me that discussion on this talk page revolves around whether certain editors personally agree with the comparisons used by a given expert reliable source. MastCell ^Talk 21:03, 11 June 2010 (UTC)

I explained what I meant by "proper" above. You can ignore that if you want MastCell. And yes, to the extent that this article mentions the risk of childbirth without mentioning that it is a "low" risk, that is potentially misleading. Reliable sources confirm that it is a "low" risk.Anythingyouwant (talk) 21:15, 11 June 2010 (UTC)

Way back, I had proposed including the absolute risks of both abortion and childbirth in addition to the relative risks. This approach has several benefits: a) it makes clear that both risks are very low in an absolute sense in the developed world, b) it does not editorially tell the reader what to think (one can see that the risks are low for oneself, rather than being editorially informed that they are "low"), and most importantly c) it is the approach used by many, if not most, expert scholarly sources. Should we reconsider that approach? MastCell ^Talk 21:26, 11 June 2010 (UTC)

This article currently says that abortion has one of the lowest risks of any medical procedure. I don't think that's editorializing any more than is the statement that the risk of dying in childbirth is also low. You're suggesting that we keep the first but omit the second?Anythingyouwant (talk) 23:03, 11 June 2010 (UTC)

The first is a direct quote from a recent review article appearing in a major medical journal. Thus it's not us editorializing. If you want to describe the risk of death from childbirth as "low", I don't really have a problem with it. I just think that using the actual numbers to illustrate the low risk is more consistent with the content of actual reliable sources, but whatever. MastCell ^Talk 23:37, 11 June 2010 (UTC)

Ok, so if we mention the low risk of childbirth, citing Lloyd 2005, then we may have a deal. I prefer not overwhelming the reader with numbers, and would therefore prefer a summary in words (perhaps with the numbers in a footnote), but whatever.Anythingyouwant (talk) 01:58, 12 June 2010 (UTC)

And later you complain the discussion process isn't evolving... Why didn't you opine on the text I suggested in the section above instead of going back in the process to propose the very same text, with the same problems (US centric) as explained before? I also need to remember the main issue here so far is comparing mortality with childbirth so I extracted that text:

"In the US, the risk of death from abortion is 0.567 per 100,000 procedures, making abortion approximately 14 times safer than childbirth (7.06 maternal deaths per 100,000 live births).(primary source about US) The risk of abortion-related mortality increases with increasing gestational age, but remains lower than that of childbirth through at least 21 weeks' gestation.(Other primary source about US) and ( Suzanne R Trupin, emedicine source).

At first look that emedicine source the following useful information: "At every gestational age, elective abortion is safer for the mother than carrying a pregnancy to term." But, it doesn't cite references for those claims, the author owns a clinic that provides abortion services which advertises as safe since 1973 and doesn't require parental consent and that reference doesn't add up anything to what is already said by (Lloyd, 2005), least it has the benefit you mentioned: "a) it makes clear that both risks are very low in an absolute sense in the developed world," .--Nutriveg (talk) 22:22, 11 June 2010 (UTC)

Your proposed text doesn't work for me, because it incorrectly singles out the Guttmacher Institute as claiming that abortion is safer than childbirth. In fact, this view is universally held by experts in the field, as demonstrated by a plethora of sources, and is in no way unique to Guttmacher. MastCell ^Talk 23:51, 11 June 2010 (UTC)

So the only problem with it is the Guttmacher attribution?--Nutriveg (talk) 00:25, 12 June 2010 (UTC)

Keep in mind who Guttmacher is and where their money comes from. We should probably mention it in the article. - Schrandit (talk) 02:13, 12 June 2010 (UTC)

(undent) I came across a comparison to taking antibiotics and to appendectomy. It was safer than both of these if I remember correctly.Doc James (talk · contribs · email) 22:30, 13 June 2010 (UTC)

Health risks rewrite

I've taken a shot at rewriting the section on health risks. The previous section had several shortcomings (IMHO):

• Numerous one-sentence paragraphs
• Spotty overview of health risks, with focus on isolated factoids and no clear summary
• Written in a somewhat opaque style
• Did not address the difference in risks between "safe" and "unsafe" abortions, which is of great importance to a worldwide (as opposed to Western) overview

I'll leave it open for comments and changes. I think the new version is a bit better (obviously), in that it clearly delineates the difference in risk between "safe" and "unsafe" abortion. It also provides a more organized overview of risk - both the absolute risks and relative risks for abortion and childbirth are given, and they are juxtaposed (as they are in virtually all scholarly sources on the topic) rather than presented in isolation. I think I've emphasized that there are actual risks associated with abortion, to address the charge (above) that the article made it sound like abortion was entirely risk-free as opposed to merely very safe.

I've tried to avoid sourcing material from websites of organizations like Planned Parenthood and Guttmacher. While these websites have ample and high-quality information, their association with the pro-choice viewpoint in the US always renders them problematic. I've tried to go for review articles, WHO material, etc supported by some key primary sources which illustrate the conclusions of secondary sources.

There are still some areas that need work. For example, the specific health risks of medical abortion (at least those that differ from surgical abortion) remain to be spelled out, pending a good source. Additionally, the "mental health" section should be condensed and probably conflated with other non-risks like breast cancer. I'd welcome any updates, new sources, or comments. MastCell ^Talk 23:47, 1 June 2010 (UTC)

Looks pretty good and thorough and sourced. I thought you had devoted a good amount of space to medical abortion. Thanks for taking the time and effort. Kodus. That said, I found it a little awkward to have the parenthetical phrase (requiring surgical abortion) twice in the same paragraph. And now we have two sections in the article called "unsafe abortion" so direct linking will break (and should we be discussing the health risks of unsafe abortion before discussing the topic of unsafe abortion more generally?) I may have some more comments later, but overall, an improvement for sure. -Andrew c ^[talk] 01:30, 2 June 2010 (UTC)

I haven't got time right now for a detailed critique, so this will be brief.

I do not see any need for this article to contain two separate subsections titled “unsafe abortion”. Much of this newly overhauled section is redundant. For example, the new material says: "Medical abortion with mifepristone and misoprostol is effective through 49 days of gestational age." But the article already says, "When used within 49 days gestation, approximately 92% of women undergoing medical abortion with a combined regimen completed it without surgical intervention." Why so much redundancy?

I do not see any reason to completely exclude information about contraindications, especially for medical (i.e. non-surgical) abortion. We have previously discussed a source that details the contraindications for both medical and surgical abortion.[9]

I support the statement that for most women abortion is safer than childbirth, when performed according to recommended protocols. But why now omit that the risk of maternal death is slight even for childbirth? This hugely slants the article. We've discussed this at this talk page repeatedly, so should we conclude that the omission from the article is deliberate? All of this material was in the NPOV (but reverted) edit I made to this section of the article, including the statement that abortion is safer than childbirth (not just early abortion or surgical abortion). Now that material is omitted. This is not just a matter of emphasis, but of complete omission.

Additionally, we're not writing a textbook here, so I would think you would be able to find accurate sources that are available on the internet for free, via Google Books or Google News Archive, or Google Scholar, so that other editors can look at them. Did you look and were unable to find them? If so, doesn't that say something about the non-notability of the information being cited?

When saying that abortion is safer than childbirth, I would prefer if we hedge a little bit, as our frequently cited author (Grimes) does. He says: "For most women, fertility regulation by contraception, sterilization, or legal abortion is substantially safer than childbirth" in his article "The morbidity and mortality of pregnancy: still risky business", Am. J. Obstet. Gynecol., Vol. 170, pages 1489–1494 (1994), PMID 8178896. Is there some reason to depart from what Grimes says here?

I object that this article devotes so much space to rebutting notions that could easily be rebutted in much less space (health risks, breast cancer, fetal pain, effect on crime rate), while not addressing a primary reason why women may decide not to get an abortion, namely the alleged indicia of humanity in the embryo or fetus.Anythingyouwant (talk) 01:43, 2 June 2010 (UTC)

You guys both identified the duplication of "unsafe abortion", which I have to admit I missed. I agree that we should streamline and combine those two sections, though I'm not sure at a glance how best to do so. I do think that a discussion of health risks needs to detail the distinction between "safe" and "unsafe", but perhaps we can shorten/merge somehow.

Anythingyouwant identified duplication of the timeframe for medical abortion. I admit that I missed that redundancy as well; it's not essential to repeat the timeframe for a discussion of health risks, so it could probably be excised from the "Health risks" section.

Contraindications are not the same as health risks. In fact, there is no health risk to someone with a contraindication, because they won't be eligible to have an abortion. That's not to say that there is no room for a discussion of contraindications to various methods of abortion, but it didn't seem to fit properly in "health risks", since contraindications to the procedure are clearly distinct from health risks caused by the procedure.

I don't think I have omitted information about the small risk of maternal death from childbirth. In fact, I gave precise and properly sourced figures for the exact magnitude of the risk, just as I did for the risk of abortion. The reader can see (rather than simply take our word for it) that both absolute risks are slight. If anything, I've emphasized the safety of both childbirth and abortion by providing absolute risk estimates, while also mentioning the relative risks of the two (as reliable sources do).

Quality of information is not synonymous with its free availability. Any overview worthy of a serious, respectable reference work will necessarily be based on the scholarly literature. In many cases, the scholarly literature is not freely available online - although it is generally freely available at any halfway-decent library, thanks to our tax dollars at work. I looked for the best available sources - as in those in high-quality journals, those that dealt with the topic at hand directly, and those that are frequently cited by reputable authorities - and this is what I came up with.

Most sources don't "hedge" at all when saying that abortion is safer than childbirth - I actually hedged by saying "through 21 weeks", because a lot of expert sources make a blanket statement that abortion is safer at any gestational age. I don't think it is an accurate representation of Grimes' article, or of the scholarly literature in general, to excessively hedge here - if anything, we're understating the degree to which reliable sources hold that abortion is safer than childbirth.

As to space, we go where the sources go. For awhile, there was a lot of literature addressing the (non-)link between abortion and breast cancer. Likewise with the supposed mental-health sequelae of abortion, and fetal pain. I'm not as familiar with available sources on the impact of views on the humanity of the fetus on the decision to have or not have an abortion, but I agree that this would be a relevant topic for this article to discuss. If you have some decent sources in mind, please lead the way. MastCell ^Talk 04:20, 2 June 2010 (UTC)

This edit has several problems.

• There are specific articles for Unsafe abortion and Medical abortion so much of that edit belong to other articles. ::::* Concerning surgical abortion it gives undue balance to mortality risk as a health risk, since that risk is minimal, but the first paragraph only talks about it.
• The statement that complications is rare is not supported by updated sources, complication risk is low but they are common due to the high number of abortions. Botha, Rosanne L.; Bednarek, Paula H.; Kaunitz, Andrew M. (2010). "Complications of Medical and Surgical Abortion". In Guy I Benrubi (ed.). Handbook of Obstetric and Gynecologic Emergencies (4 ed.). Lippincott Williams & Wilkins. p. 258. ISBN 978-1605476667. {{cite book}}: Unknown parameter |coauthors= ignored (|author= suggested) (help)
• Complications are not well represented and the presented solution is deceptive. (Botha, 2010)
• Procedures details, like: description of aspiration, places where they are performed and antibiotics should belong to the procedures section.

So I'm reverting that major change. Please add unrelated text to specific sections/articles and make other changes incrementally.--Nutriveg (talk) 14:21, 2 June 2010 (UTC)

I reverted your major change as well, as MastCell's first 3 bulletted points above discuss the shortfalls of your version. I'm not going to talk down to you and ask you to make changes incrementally. Bold edits are to be encouraged, and I think MastCell's went a long way to improve that section. The section is how it was for quite a long time before the recent edits. It is the longstanding consensus version. There is no consensus for any change (Nutriveg or MastCell's) so we revert to default. What can we do to address everyones concerns and perhaps find a harmony between the two versions?-Andrew c ^[talk] 14:42, 2 June 2010 (UTC)

Or we can continue to edit wildly and pray out edits don't get reverted, in lieu of discussing our differences :Þ -Andrew c ^[talk] 15:23, 2 June 2010 (UTC)

I've addressed the issues raised by MastCell and moved unrelated content elsewhere. That previous version (where I started from today) is closer to that older version where incremental changes were made since then. Beyond that I don't like putting specific numbers to maternal death since that number varies yearly, IIRC it's currently around 0.4 for abortion and 2.5 to live births. I refuse making straight comparisons of maternal death because those are not statistically adjusted (age, risky pregnancies,...), both numbers are very low and comparing near "zeros" is stupid.--Nutriveg (talk) 15:28, 2 June 2010 (UTC)

Andrew c's deletion edit note implies this is going somewhere else, but that doesn't seem to have happened.LeadSongDog come howl! 16:51, 2 June 2010 (UTC)

Abortion and mental health already mentions the Johns Hopkins University's 2008, as is information regarding economic and decision issues. Did I delete any content which isn't covered already at Abortion and mental health in more detail? Or am I missing your point entirely?-Andrew c ^[talk] 17:01, 2 June 2010 (UTC)

Sorry if it was already there in Abortion and mental health that's fine, but the edit comment implied the insertion was yet to be done. LeadSongDog come howl! 19:37, 14 June 2010 (UTC)

← To address Nutriveg's concerns:

• I agree that we should avoid duplication of content. But I don't see how we can accurately discuss the health risks of abortion without drawing some sort of distinction between safe and unsafe procedures (I think there was general agreement on that). Do you have any suggestions about how to cover this distinction in the "health risks" section? That might be more useful than reverting.
• Mortality is the most prominent health risk, both as a matter of common sense and as a matter of weight in reliable sources, so I think our coverage reflects that weight (whatever Nutriveg's personal opinion of it might be).
• If you want to add information on non-lethal complications, feel free. The rate is very low. The source you favor (Botha 2010) states: "Fewer than 0.3% of abortion patients experience a complication that requires hospitalization." If you'd like to include that sort of information, be my guest.
• If you wish to use different language to discuss the frequency of side effects, then propose some rather than reverting the entire edit. For example, it might be more precise to say that the complication rate after abortion is very low (perhaps citing the numbers given by Botha in my bullet point above). We could add that although the rate of complications is low, abortion is a common procedure and therefore it is "not uncommon" for gynecologists to encounter women who have had complications from abortion. Personally I think this is a bit silly - the article should probably be written from the perspective of the individual reader rather than the practicing gynecologist - but whatever.
• I'm not sure what you consider "deceptive"; elaboration might be useful.
• If you want to move specific details of the procedure, please feel free (doing so does not require a blanket revert). I think some minimal level of detail is necessary, since the risks may vary depending on which procedure is used, but I agree that most detail should be in the "procedures" section.
• I'm interested in the comparisons made by reliable sources. Those sources repeatedly and commonly compare the risk of abortion to that of childbirth, and so this is the appropriate comparison for article, regardless of whether Nutriveg personally finds it "stupid".

Thoughts? MastCell ^Talk 18:15, 2 June 2010 (UTC)

Answering

• Your separation of safe vs unsafe was already incorporated to introduce the specific articles.
• Mortality is not the most prominent health risk in this case, it's negligible so there's no reason to be lengthy about that.
• That source said despite the low risk of complications incidence is common due to the high number of abortions not rare as you early said. The current text uses the same terms of that source "low".
• I didn't revert the whole edit, you that rewrote the whole section. I incorporated those relevant changes in this an other specific articles. I'm against citing numbers since they vary from country, year, source and method. Most of the sources, like Botha is US centric.
• Your text mentioned only one trivial medical procedure that would be necessary in case of abortion complications.
• I may not be the most appropriate person to move those procedure descriptions to the appropriate section or article, the article is already big and it doesn't need to repeat itself.
• As I said those comparisons weren't statistically adjusted (for age, risky pregnancies, ...) and dividing anything to zero results in a large number. According to your 0.56 per 100000 maternal death number, abortion would be almost 3 times more deadly than vaginal birth. These comparisons don't carry relevant death risk to worth comparing, less a direct (unadjusted) one.--Nutriveg (talk) 19:29, 2 June 2010 (UTC)

Have you read the article you mention in your last bulletpoint (PMID 18455140)? It notes a maternal death rate of 6.5 per 100,000 births, which is very close to the figure we cite. You must be looking at the rate of maternal death causally related to mode of delivery, which for vaginal birth was 0.2 per 100,000. That statistic reflects only deaths which (in the authors' opinion) could have been avoided by going to C-section. It does not represent the overall maternal death rate. You're comparing apples to oranges.

Do you understand the difference? It's a subtle point, but it nicely points up the reason why WP:MEDRS urges against cherry-picking primary sources. You've gone to a primary source which has nothing to do with abortion and pulled out a number. As it turned out, you've incorrectly interpreted the paper, which is a risk when you employ papers to make a point that the authors themselves don't make. If we didn't double-check your assertion carefully, we might have made a similar error.

That's why I'm citing papers which specifically address the safety of abortion as it relates to childbirth. The numbers are actually quite consistent if you read carefully. But even that issue becomes superfluous if we choose sources which directly address the questions at hand, rather than trying to pull numbers from unrelated publications to buttress our personal viewpoints. MastCell ^Talk 20:56, 2 June 2010 (UTC)

First of all I'm not trying to use that source, I just used it as an example it's not worth to compare such near zero numbers. ::::I just read the article abstract, I don't have access to the whole article but it says "The rate of maternal death causally related to mode of delivery was 0.2 per 100,000 for vaginal birth" not "The rate of preventable maternal death causally related (...)" as you stated above. Sure there are other causes but it's just an example we can't make a direct comparisons (without adjust) since the causes of maternal death are restricted to specific factors. --Nutriveg (talk) 21:11, 2 June 2010 (UTC)

I still don't expect to use comparisons in the article but just as an information see what I've found "After the 19th week of pregnancy the maternal death rate due to abortion is greater than that of childbirth"

Jones, Richard Evan (2006). Human Reproductive Biology (3 ed.). Academic Press. p. 429. ISBN 978-0120884650.--Nutriveg (talk) 22:19, 2 June 2010 (UTC)

You don't need full-text access. See the abstract, first sentence under "Results": Ninety-five maternal deaths occurred in 1,461,270 pregnancies (6.5 per 100,000 pregnancies.) This is not an "example" of anything except the danger of using primary sources without reading them carefully.

As to the Human Reproductive Biology book, it doesn't describe the basis for its claim that abortion is riskier than childbirth after 19 weeks. Nonetheless, if you'd prefer to provide various expert interpretations of the risk at late time points, we could cite it as part of a full discussion, along with sources such as eMedicine and UpToDate, which argue the opposite (that abortion is safer at any gestational age). It seems a bit excessive, since extremely few abortions are performed after 19 weeks, but I'm fine either way. MastCell ^Talk 22:32, 2 June 2010 (UTC)

Please don't read problematic content without discussion, ignoring the problems reported before, like making Division by zero to imply a high risk from maternal death from child birth when it's actually very low, using unadjusted correlation, adding content that belong to other specific articles and making major changes that can't be individually reverted.--Nutriveg (talk) 01:43, 8 June 2010 (UTC)

One of the changes repeatedly made by you is replacing maternal death with mortality rate, please use the correct naming definition by WP:MEDMOS. Abortion is pregnancy management.--Nutriveg (talk) 02:18, 8 June 2010 (UTC)

Since those US statistics were so important to you I moved them to the specific article and created a link for it. I also would like to remember the same was already done for Medical abortion long before.--Nutriveg (talk) 02:52, 8 June 2010 (UTC)

About medical abortions, and the four 2005 deaths due to infection. From my recollection of those events, and their media coverage, the deaths were related to an off label application of misoprostol. Misoprostol was not recommended to be taken vaginally, yet all 4 cases were related to the off label application. I personally feel that if we are to mention the deaths, perhaps we should also mention that it was off label usage? But then again, the RS we are citing doesn't think it is important enough to mention, so maybe I should be quite as I clearly don't know better than published scientists... (quick google search found something like [10])-Andrew c ^[talk] 02:51, 8 June 2010 (UTC)

This is likely a issue for the Medical abortion#Health risks article.--Nutriveg (talk) 03:03, 8 June 2010 (UTC)

More reliable references

I have added more reliable references to support the comparison to child birth in the developed world:

• A Lance review from 1996: Kulczycki A, Potts M, Rosenfield A (1996). "Abortion and fertility regulation". Lancet. 347 (9016): 1663–8. PMID 8642962. {{cite journal}}: Unknown parameter |month= ignored (help)CS1 maint: multiple names: authors list (link)
• A medical text from 2004 Danielle Mazza (2004). Women's health in general practice. Oxford: Butterworth-Heinemann. p. 93. ISBN 0-7506-8773-8.
• Another medical text from 2007 Eric Sokol; Andrew Sokol (2007). General gynecology. St. Louis: Mosby. p. 238. ISBN 0-323-03247-8.{{cite book}}: CS1 maint: multiple names: authors list (link)
• A third medical text from 2005 Lloyd, Cynthia B. (2005). Growing up global: the changing transitions to adulthood in developing countries. Washington, D.C: National Academies Press. p. 215. ISBN 0-309-09528-X.
• Returned a 2009 medical text Douglas W. Laube; Barzansky, Barbara M.; Beckmann, Charles R. B.; Herbert, William G. (2009). Obstetrics and Gynecology. Hagerstwon, MD: Lippincott Williams & Wilkins. p. 150. ISBN 0-7817-8807-2.{{cite book}}: CS1 maint: multiple names: authors list (link)

I hope this addresses the concerns regarding the previous references being old and question about if they were peer reviewed and from a reliable source.Doc James (talk · contribs · email) 17:20, 9 June 2010 (UTC)

It would be helpful if you added the quote to those references, it's problematic to search for that book to find it uses US data or don't make a comparison (safer) as in the Wikipedia text.--Nutriveg (talk) 17:36, 9 June 2010 (UTC)

Just pull it up using google books. If you wish to type out the text though feel free.Doc James (talk · contribs · email) 17:48, 9 June 2010 (UTC)

Yes, the lancet 1996 review does support the text "In developed countries, mortality associated with childbirth is 11 times higher than that for safely performed abortion procedures". It doesn't follow WP:MEDRS#Use up-to-date evidence since it is 14 years old but I will check the other sources first.--Nutriveg (talk) 19:29, 9 June 2010 (UTC)

We have used many other sources of this age. It is sufficiently uptodate and before making claims it is not you should request comments at WP:MEDRS. I am fine with saying 1996. The other references mainly say that child birth is more dangerous than abortion in developed countries. Thus we can either clarify with the date "1996" or the "11 times". I do not have any strong opinion either way. Doc James (talk · contribs · email) 19:42, 9 June 2010 (UTC)

You should try to change WP:MEDRS#Use up-to-date evidence to reflect that.--Nutriveg (talk) 20:06, 9 June 2010 (UTC)

(edit conflict) If it's any help:

• The Kulczycki 1996 review seems to be cited in other texts and supplies data on abortion rates and maternal death from unsafe abortions across the globe;
• Mazza 2004: "Mortality associated with pregnancy is 30 times higher than that associated with abortion prior to 8 weeks gestation" - a comparison made on page 93;
• Sokol 2007: (aside: pages 233- overview of surgical abortion; useful if that article is created?) search for 'mortality' brought up some results in termination section that I wasn't allowed to view;
• Lloyd 2005: "Mortality and morbidity related to pregnancy and childbirth (particularly in sub-Saharan Africa and South Asia, where levels of early childbearing remain high) and as a direct consequence of unsafe abortion across all developing regions remain among the most significant risks to young women's health." (p.5) - factors juxtaposed, but no direct comparison; "Maternal deaths are those that occur during pregnancy and up to 42 days after birth. About 80 percent of these are due directly to maternity; the most common cause is hemorrhage, followed by sepsis and complications of unsafe abortion ..." (p.191); 3 more pages that I can't see;
• Laube 2009: "Risk of death from abortion during the first 2 months of pregnancy is less than 1 per 100,000 procedures, with increasing rates as pregnancy progresses (versus 7.7 maternal deaths per 100,000 live births)." (p.150) - makes a comparison.

It seems clear that some reliable secondary sources make a comparison. There should be enough in the sources there to produce a supportable piece of text. --RexxS (talk) 20:13, 9 June 2010 (UTC)

(edit conflict)Reading what you posted.

• (Kulczycki, 1996) is the one we just cited
• (Mazza, 2004) is just citing (Kulczycki, 1996), using exactly the same phrase so it doesn't qualify as a new review just a copy
• (Sokol, 2005) missing the quote
• (Lyond, 2005) missing a quote comparing both
• (Laube, 2009) it's Beckmann we previously discussed as US data, it just presents both the numbers.--Nutriveg (talk) 20:52, 9 June 2010 (UTC)

Absolutely wrong. You are not qualified to draw conclusions from your own opinion of a reliable source. That's the single biggest problem you bring here. Mazza 2004 is a secondary source and your opinion that "it doesn't qualify as a new review just a copy" is pure nonsense. You have no way of knowing what primary sources were used by Mazza in compiling the book, and it is contrary to all established policy for you to start making your own guesswork analysis of a reliable source. You are also totally wrong to think that Laube 2009 "just presents both the numbers". By using the "versus" phraseology, it is abundantly clear to everyone else that it is a comparison. This is not the first time that it's been pointed out to you that you cannot pick and choose which references you want depending on whether they fit your own preconceptions. --RexxS (talk) 23:00, 9 June 2010 (UTC)

Yeah, I can only opine from what I see if I'm wrong it's up to other more informed people to clear that misunderstanding. The phrases from both articles are the exactly same but I don't know why you're bringing the issue of (Mazza, 2004). Do you expect to use it in a different way than it's currently being used?

Comparing data is not the same thing of presenting both to the reader compare. Compare is using words like "higher", "safer", we usually just cite the source, without advancing it by making such comparisons of different subjects especially when we don't know where those numbers came from... As usual I don't see how you raising that semantic issue will result in a change in the article text.--Nutriveg (talk) 23:14, 9 June 2010 (UTC)

Okay, I can sympathise with the frustration of "knowing that text is wrong" - I often feel the same when editing scuba articles. But I've had to learn that whatever expertise I may have outside Wikipedia is worth zilch inside it, unless a reliable source has already made that point. Even then, sometimes it's not clear-cut and two different views have to be presented; but that's ok as well, as long as I don't give undue weight to my own personal opinion.

I think we'd go mad if we had to track down and analyse the source of every piece of numerical data that shows up in sources. At some point we have to just say "It's a good quality reliable source: I trust it".

Finally, I'm willing to be corrected if I'm wrong, but I thought the issue with the comparison of mortality rates was not "what the text said", but whether the text should even be in the article? I thought that we were trying to settle the point that secondary sources used such a comparison, and therefore it is appropriate for our article to do so. If I've misunderstood, please accept my apologies - it's quite possible that I'm not up-to-date on all of the debates here. --RexxS (talk) 00:05, 10 June 2010 (UTC)

Yanda

• I have found an even better reference. Which I will add Yanda K.; et al. (2003). "Reproductive health and human rights". International journal of gynecology and obstetrics. 82 (3): 275–283. {{cite journal}}: Explicit use of et al. in: |author= (help); Unknown parameter |month= ignored (help). Is 2003 recent enough? Here is the wording they use "Studies of medical abortion since the 1980s in Europe and in 2000 in the United States, in addition to studies in various other countries, have shown the method to be extremely effective and significantly safer than childbirth."Doc James (talk · contribs · email) 20:48, 9 June 2010 (UTC)

It's a newer source but that claim is about Medical abortion so it belongs to that article.--Nutriveg (talk) 20:58, 9 June 2010 (UTC)

Look the answer is no. You must start getting consensus for your deletion of excellently sourced material. Obviously nothing no matter how good a reference will ever be ever good enough for you. This has become silly. Doc James (talk · contribs · email) 21:04, 9 June 2010 (UTC)

The source looks good, but it didn't supported the claim, since the quote is clearly about Medical abortion--Nutriveg (talk) 21:11, 9 June 2010 (UTC)

The answer is still no. The rest of the references support the rest of the claim. Sokol 2007 "In developed countries the mortality rate from induced abortions is less than 1 in 100,000 procedures making the procedure safer than pregnancy and childbirth"Doc James (talk · contribs · email) 21:13, 9 June 2010 (UTC)

And on that note, I will add that this was "a revert too far" in my opinion also. Hence, I have asked an outside admin to take a look at this. SHEFFIELDSTEEL^TALK 21:21, 9 June 2010 (UTC)

DocJames, well you didn't provide the quote to that 2007 article when I requested trying to figure what text it was supporting, and said "Is 2003 recent enough?" as if you had based your last edit on that 2003 source you just added and previously quote here, where the quote indeed reflected your edit. You can use that Sokol,2007 phrase if you think it's better, but the other one is clearly about Medical abortion.--Nutriveg (talk) 21:25, 9 June 2010 (UTC)

SheffieldSteel, what's the ongoing problem you're seeing? This was newly added content, you can't evaluate an edit during the editing process you need to wait few minutes for the outcome.--Nutriveg (talk) 21:30, 9 June 2010 (UTC)

Having most additions of references I have added in an attempt to solve the ongoing debate removed I have posted at WP:3RR [11]. I will stop further editing of the main article / looking for references until this matter is resolved. I see no reason to search the literature in this environment. BTW that sentence is based on 6 excellent sources and the information is in all 6 of the sources. Doc James (talk · contribs · email) 21:40, 9 June 2010 (UTC)

I feel sad about that but good look searching for sources if you think they are not good enough or they directly don't support the text. As opposed of what you're saying most of the references you added are still in the article. If you don't like to have your changes reverted you should discuss them first instead of editing and waiting to see if someone will revert them. Despite of how good you think they are some people may not agree with you.--Nutriveg (talk) 22:05, 9 June 2010 (UTC)

Usually you should see if you can find two people who disagree rather than just one.Doc James (talk · contribs · email) 22:10, 9 June 2010 (UTC)

Nutriveg, take a look at Wikipedia:Ownership of articles#On revert. You've demonstrated nearly all of the examples of inappropriate ownership listed there. MastCell ^Talk 22:11, 9 June 2010 (UTC)

Doc Jones, if you don't care about being reverted I don't care if you do bold edits, but if you do care about being reverted your advice also applies to you: get agreement before making a change.

MastCell, I don't identify with those points since we have added a lot of content so far and the problems we are having is mainly one of inappropriate use of sources, except in what concerns your edits, that remove a lot of sourced content.--Nutriveg (talk) 22:32, 9 June 2010 (UTC)

I agree with Doc (although I don't think it's necessary to include all of the sources he's turned up, since a few should suffice for any reasonable interpretation of verifiability). I think several other editors also agree. The thing is that you're sort of drowning them out, both here on the talk page and by your ready recourse to edit-warring (which I don't really see from other editors). The other thing I see is about 4 or 5 editors working hard to find and vet sources, and one editor (you) simply shooting down that work and reverting (with increasingly odd rationales).

Here's an idea - self-impose 1RR. If people agree with your edits, they'll stick without the need for your rapid-fire reverting. If people don't agree with your edits, it will become apparent once you've given other editors a tiny bit of room to breathe and work. I've done this before myself and it helps (in fact, I decided to limit myself to 1RR here, although I broke my own rule to fix your error about contraindications since it seemed fairly serious). MastCell ^Talk 23:10, 9 June 2010 (UTC)

Cites cleaned up

I've cleaned up the cites Doc James offered above:

• A Lancet review from 1996: Kulczycki A, Potts M, Rosenfield A (1996). "Abortion and fertility regulation". Lancet. 347 (9016): 1663–8. doi:10.1016/S0140-6736(96)91491-9. PMID 8642962. {{cite journal}}: Unknown parameter |month= ignored (help)CS1 maint: multiple names: authors list (link)
• A medical text from 2004 Danielle Mazza (2004). Women's health in general practice. Edinburgh; New York: Butterworth-Heinemann. p. 93. ISBN 9780750687737. OCLC 52515824.
• Another medical text from 2007 Eric R. Sokol; Andrew I. Sokol (2007). General gynecology. St. Louis, Missouri: Elsevier Mosby. p. 238. ISBN 9780323032476. OCLC 70663738.{{cite book}}: CS1 maint: multiple names: authors list (link)
• A third medical text from 2005 National Research Council (U.S.). Panel on Transitions to Adulthood in Developing Countries (2005). Lloyd, Cynthia B. (ed.). Growing up global: the changing transitions to adulthood in developing countries. Washington, DC: National Academies Press. p. 215. ISBN 9780309095280. OCLC 57529038.
• Returned a 2009 medical text Douglas W. Laube; Barzansky, Barbara M.; Beckmann, Charles R. B.; Herbert, William G.; American College of Obstetricians and Gynecologists (2009). Obstetrics and Gynecology (6th ed.). Baltimore, Maryland: Lippincott Williams & Wilkins. p. 150. ISBN 9780781788076. OCLC 567310751.{{cite book}}: CS1 maint: multiple names: authors list (link)

Most of these have google previews available.LeadSongDog come howl! 03:06, 10 June 2010 (UTC)

Thanks LeadSongDog You missed one Yanda K.; et al. (2003). "Reproductive health and human rights". International journal of gynecology and obstetrics. 82 (3): 275–283. {{cite journal}}: Explicit use of et al. in: |author= (help); Unknown parameter |month= ignored (help) and yes all of these were visible by google books from were I come from.Doc James (talk · contribs · email) 03:48, 10 June 2010 (UTC)

(undent) To continue this above paper states "Studies of medical abortion since the 1980s in Europe and in 2000 in the United States, in addition to studies in various other countries, have shown the method to be extremely effective and significantly safer than childbirth." It describes world wide statistics for medical abortions in developed countries. I see that the US does primarily procedural abortions which is different than up north. The other refs provide the comparison for procedural abortions so I think we should changed the wording to "significantly safer than childbirth" and drop the 1996 as there is no evidence that this has in any way changed.Doc James (talk · contribs · email) 05:44, 10 June 2010 (UTC)

I checked the text of those references:

• (Kulczycki,1996) sounded too informal and counseling "When women ask about the safety of abortion it is also worth pointing out that in developed countries, mortality associated with childbirth is 11 times higher than that for safely performed abortion procedures and 30 times higher than for abortions of up to 8 weeks gestation" Not the kind of text I would expect to see in an epidemiological analysis, sounds more like advice for those supporting a specific POV so they can have an argument to get handle of insecure mothers.
• (Mazza, 2004), is in a section called Surgical abortion (not general abortion) as expected it was just a plain citation of (Kulczycki,1996)
• (Sokol, 2007) "In developed countries, the mortality rate from legal, induced abortion in the first trimester is less than 1 in 100,000 procedures, making the procedure safer than pregnancy and childbirth." "First trimester" was an important omission from what Doc James previously quoted.

That book chapter "Pregnancy Loss and termination" was written by Allison Cowett, which is a pro-abortionist and E. Steve Lichtenberg who serves on several committees of the Board of Directors of the National Abortion Federation. NPOV anyone?

• (Lloyd, 2005) is not a medical book so not a good source by WP:MEDRS#Books standards and it says: "In representative developed countries, the risk of dying is no more than 1 in 100,000 procedures lower than the relatively low risks associated with pregnancy and childbirth in these countries (The Alan Guttmacher Institute, 1999)" Even when making poor comparisons they took care to "quantify" the risk. Guttmacher is the one which promote abortion and its legalization: "The Institute works to protect, expand and equalize access to information, services and rights that will enable women (...) exercise the right to choose abortion"
• (Laube, 2009) seems to use US data in the whole section, not worldwide, as previously discussed. It says "Risk of death from abortion during the first 2 months of pregnancy is less than 1 per 100,000 procedures, with increasing rates as pregnancy progresses (versus 7.7 maternal deaths per 100,000 live births)." same 7.7 number of "To discern possible trends in maternal mortality, data were divided into two 5-year periods (1987-1991 and 1992-1996). The national MMR was 7.7 for each time period."
• (Yanda, 2003) says "Studies of medical abortion since the 1980s in Europe and in 2000 in the United States, in addition to studies in various other countries, have shown the method to be extremely effective and significantly safer than childbirth." It's about a specific procedure, medical abortion, which has its own article.

Good luck finding good sources next time.--Nutriveg (talk) 03:17, 11 June 2010 (UTC)

These are all good via WP:MEDRS. I am not sure what you mean by next time as this time I think I have already done very well. You do not get to say "sounded too informal and counseling" as a means to dismiss a reference. This page is about both medical and surgical abortions thus we are not going to exclude everything that does not specifically deal with both.

I recommend we change it to: "Abortions are safer than child birth in developed countries." referenced to the 1996 paper. No this is not too old as nothing new has come along to refute it.Doc James (talk · contribs · email) 03:29, 11 June 2010 (UTC)

That's your interpretation, and since it was you who suggested those sources as reliable I wasn't expecting to hear anything much different.

To you it1s not informal when someone brings a number in a phrase which point was how to address insecure mothers? And that number (11) can't be found (previously) anywhere else so we just don't know if the author took an aesthetic decision to summarize numbers during the editing process since that phrase was directed to (insecure) lay people where an strong (but poor) comparison would sound better to convince those insecure people than presenting individual numbers and letting them do that (poor) "high school math" (as opposing to serious epidemiological analysis) in their head?

Even by bad sources that phrase you suggested is inappropriate. Even those bad sources say "first trimester" and they make clear the risk is low for both procedures.

Using that 14 years old source clearly contradicts WP:MEDRS#Use up-to-date evidence "Look for reviews published in the last five years or so, preferably in the last two or three years." beyond the problems of the context where that argument was used in that old source.--Nutriveg (talk) 03:53, 11 June 2010 (UTC)

Wow so the cut off is 5 years but preferably within the last 2 to 3 years? No reviews before 2005 are sufficiently uptodate for use on Wikipedia.Doc James (talk · contribs · email) 04:05, 11 June 2010 (UTC)

We are not talking about the whole wikipedia, but to support medical claims which WP:MEDRS#Use up-to-date evidence is about.

I wouldn't have a problem if it was just a few years older, but that's not the case here, specially when we think a review is about even older studies. And the problem of the context where that claim was made, that doesn't help neither.--Nutriveg (talk) 04:10, 11 June 2010 (UTC)

"Up to date" implies "not superceded by more recent sources of similar or better quality". Are there any such more recent sources that you would care to suggest? LeadSongDog come howl! 04:40, 11 June 2010 (UTC)

I don't know why you quote text without pointing their source, I pointed the source for quote that mentioned the years. It's not hard to find people that are careless enough to make bold claims without appropriate scientific support that's why the burden of evidence lies with the editor who adds or restores material, not me in this case.--Nutriveg (talk) 05:06, 11 June 2010 (UTC)

(edit conflict) Please try to understand that your characterisation of a source is not an argument. Just because you feel a source is bad doesn't make it so. Your insistence on presenting your own analysis is disrupting the discussion on how best to use these sources.

Your objection to Kulczycki is that it's not the kind of text you'd expect to see?

Your objection to Mazza is that it's a plain cite. It's not any sort of cite. This is a secondary source and you've already been told that none of us know if the source was X,Y or Z unless the author specifically attributes it to another source. Our job as editors is to neutrally report what sources say, not play detective with where the secondary might have got its data from.

Your comment on Sokol that it refers to first trimester is useful and will helps the editors in using that source accurately. However, you completely misunderstand WP:NPOV - that prescribes the way in which editors have to report sources. It is nothing whatsoever to do with authors, publishers, or the words used in a source. The only opinion about the POV of a reliable source is one which is stated in other reliable sources. You cannot dismiss a source on the grounds that you feel it (or its authors or editors) have a POV. It's not our job to do that.

Your objection to Lloyd is that it isn't a medical book. MEDRS says medical books are excellent sources, but you make the logical fallacy of thinking that non-"medical books" are therefore not "excellent sources". I agree that Lloyd certainly isn't a medical book. It's a report on a project undertaken by the National Research Council, drawing from the National Academy of Sciences, the National Academy of Engineering and the Institute of Medicine (as it states on its first page). It's a quality piece of research undertaken by some of the most eminent institutions. It is inconceivable that anyone else could think it was not a reliable source. The fact that it quotes another reliable source is not an objection to its reliability.

There's an essay at WP:TE that gives good advice on how to avoid these sort of problems. You have the capability to make good points – the limitations of sources discussing only part of the issues, for example. But you are drowning that out by insistence on trying to "debunk" all of the sources that don't suit, by repeated arguments that rely only on analyses that are beyond the competence of any editor to make. --RexxS (talk) 04:46, 11 June 2010 (UTC)

I agree with RexxS. A good-faith effort has been made to address Nutriveg's objections, with several editors going above and beyond the call of duty producing a raft of good sources. At this point these are simply not reasonable objections, as RexxS has explained in more detail, and I think it's time to stop wasting effort addressing them. None of us are infinitely patient, and we need to refocus our efforts on actually improving the article with the many sources that we've accumulated, rather than trying to satisfy endless goalpost-moving obstruction. We are at the point where we're going to lose editors who have a lot to contribute (if we haven't already) because of this, so let's not let that happen. MastCell ^Talk 05:04, 11 June 2010 (UTC)

RexxS, I didn't merely characterized the sources, I pointed the problems they had, detailing those issues.

I've said much more about Kulczycki than that out of context word you picked, curiously the same out of context way as that Kulczycki text is trying to be used. It's mainly a problem of taking an unreferenced informal "how to advice" as scientific fact. I've already pointed unreferenced claims by sources saying that abortion could be riskier than childbirth but you didn't see me supporting the use of such text. Do we want to lower the level of the sources used in this article just because some people want to make a comparison? Kulczycki is also older than defined by WP:MEDRS

I later checked the source and Mazza explicitly cites that whole Kulczycki phrase (with the advice part) attributing that claim to him. Don't play detective with me so.

I didn't say WP:NPOV, but NPOV which is short for "neutral point of view", Wikipedia don't have exclusive rights over acronyms. The authors of that text are biased since they are abortion lobbyists and even sue the government for defending that minors should commit abortion without their parents consent. It's appropriate to clarify who was making those (again) unreferenced claims. WP:NPOV applies to text presentation not to WP:Verifiabilty or the selection of reliable sources. Source selection is decided by Wikipedia editors. External criticism about a source would only be required if we decide to include text criticizing that source.

MEDRS in its definition also says "Ideal sources for these aspects include (...) widely recognised standard textbooks ", "specialist textbooks are examples of secondary sources" and "A good secondary source from a reputable publisher will be written by an expert in the field". I initially noticed that problem of not being ideal as that book wasn't categorized as medical, but now, looking further at least some parts of (Lloyd, 2005) seems to have been written by experts, so we can use it.

MastCell, try to really address those issues instead of just saying "Oh, that was already addressed" that kind of participation doesn't help to solve those issues.--Nutriveg (talk) 06:52, 11 June 2010 (UTC)

Having reviewed this, I see no problem with what RexxS, MastCell etc are proposing. Verbal chat 07:48, 11 June 2010 (UTC)

First suggestion by Nutriveg

It looks that Schrandit, by the description he makes, apparently agrees with me about that one source selection: "give the 1 more reliable, recent number".

Verbal, I see you read fast but so far in this current discussion I didn't see what text they (RexxS, MastCell) were proposing or in which source it was mainly based.

I propose we use (Lloyd, 2005) "In representative developed countries, the risk of dying (from abortion) is no more than 1 in 100,000 procedures, lower than the relatively low risks associated with pregnancy and childbirth in these countries (The Alan Guttmacher Institute, 1999)" we may also attribute that claim to their original author Guttmacher Institute, since (Lloyd, 2005) did that attribution and given this is not an "ideal source" as discussed above.

So an early version would look: "According with the Guttmacher Institute maternal death from abortion in developed countries is bellow 1 in 100,000 procedures, lower than the relatively low risks associated with pregnancy and childbirth in these same countries "--Nutriveg (talk) 13:31, 11 June 2010 (UTC)

Arbitrary break 1

Nutriveg, I understand your objections to Kulczycki, but refute them. Your criticisms were "sounded too informal and counseling"; "Not the kind of text I would expect to see ..."; "sounds more like advice for those supporting a specific POV ..." Can you not see that such objections are wholly inappropriate concerning a "review of abortion-related issues", (i.e. a secondary source) published in the The Lancet. It simply doesn't get any better than that.

I was mistaken when I said that Mazza does not cite Kulczycki. I retract that and have struck it from my previous comment. Nevertheless, you still miss the point about Mazza - it's a secondary source; when we use it, we rely on it's own authority. We accept that the author made the judgements and conclusions based on earlier sources – that's what makes it a secondary source – but anyone who wants to contradict those judgements needs to have their contradiction published in a reliable source to have any weight. There is the source you pointed to (Jones, Richard Evan (2006). Human Reproductive Biology (3 ed.). Academic Press. p. 429. ISBN 978-0120884650.) and I simply don't understand why you aren't suggesting a caveat expressing a minority opinion. I'd suggest something like '... although Richard Jones concludes that "[a]fter the 19th week of pregnancy the maternal death rate due to abortion is greater than that of childbirth"(ref)'

Again however, you make the mistake of confusing fact with opinion. The fact is that in developed countries the mortality rate from abortion during at least the first trimester is an order of magnitude lower than the rate of mortality from childbirth. That fact is attested in several reliable sources and you have to understand that the POV of the authors is not our concern. Read again WP:MEDASSESS:

"Assessing evidence quality" means that editors should determine the quality of the type of study. Editors should not perform a detailed academic peer review. Do not reject a high-quality type of study because you personally disagree with the study's inclusion criteria, references, funding sources, or conclusions.

I can't make it any clearer for you than that. --RexxS (talk) 15:27, 11 June 2010

No, that's not my argument, that's part of one of the conclusions (informal/unscientific context) about that reference (the others being unreferenced claim and old source by MEDRS standards) I took from my arguments. So we are going nowhere if those arguments are just ignored and you (again) just pick some words and restrict this discussion as if baseless personal opinion were being discussed. That Kulczycki claim was made under an informal context (how to address doubtful women) where scientific rigor is not expected, during the editing or reviewing process, since that phrase was directed to a lay audience and in that context other features of the message like easiness to read and clarity may outweigh scientific reliability. That claim has no reference in that text and is too old by WP:MEDRS#Use up-to-date evidence standards since it is 14 years old.

I don't want to use and reject the use of (Evan, 2006) since it shares the same problem of making claims without citing the sources. That was just an example of how sources can be problematic in both ways.

I do see the way Mazza cited Kulczycki as a caution measure: This guy, not me, is saying that. It does cite Kulczycki the same way I quote it here, so it doesn't solve those reported problems.

I suggested the text above based on a respectable source, there's no point in further discussing (Kulczycki,1996) (or Mazza which cites exactly that same quote) if those sources aren't being used to support a change in that text or they disagree with it.

That's not a fact, some people here do see it that way and so may show little care about problematic sources supporting that claim. It's likely mortality from induced abortion is lower in many countries and undoubtedly that's the case of the United States. But to make bold claims, especially when mentioning numbers (like you did in "one order of magnitude"), we need good sources. We don't compare apples with oranges, so far I haven't seen worldwide epidemiological reviews, we and others could cite, comparing both cases or data from where those studies could be made: are they more poor? they have less access to good health services? do they have a specific condition? do they decide to continue pregnancy even when advised that carries a death risk? Are those just exceptional cases that generally don't apply? Does it carry statistic relevance when comparing to abortion? Another problem is that around the world women who die from abortion are generally counted as death from pregnancy, by ICD-9/WHO definition, so, in most places, distinction is not done even in the most basic raw data.

But that doesn't stop some people from seeing two different numbers, one higher than other, and saying: "Oh, this number is higher than the other, let's make a simple math", "It's late night and I'm reviewing what I previously wrote, let's make this minor change to make this argument stronger". So when citing those studies we need to make sure they are about real epidemiological analysis and not a case of someone, in a review article/book, presenting original research as if it was a result of such epidemiological analysis that hasn't been done.--Nutriveg (talk) 19:30, 11 June 2010 (UTC)

You've just illustrated the behavior that RexxS' green-colored quote specifically warns against. Again. MastCell ^Talk 23:47, 11 June 2010 (UTC)

Exactly about what source? RexxS said:

"The fact is that in developed countries the mortality rate from abortion during at least the first trimester is an order of magnitude lower than the rate of mortality from childbirth."

I contradicted that argument of him not a source. Right as the opposite, if that was the case I wouldn't have agreed to use (Lloyd, 2005)

Again you show up here to question the editors instead of the content being discussed.---Nutriveg (talk) 23:59, 11 June 2010 (UTC)

There's nothing difficult about this process. Doc James finds six reliable sources. I look for what they said about the comparison between mortality rates for abortion and for childbirth:

Kulczycki is a 1996 secondary review in Lancet, cited 44 times, and as recently as 2009 according to Google Scholar.

Mazza is a 2004 medical text from a quality publisher and says "Mortality associated with pregnancy is 30 times higher than that associated with abortion prior to 8 weeks gestation"

Laube is a 2009 medical text from a quality publisher and says "Risk of death from abortion during the first 2 months of pregnancy is less than 1 per 100,000 procedures, with increasing rates as pregnancy progresses (versus 7.7 maternal deaths per 100,000 live births)"

I put forward the fact that in developed countries the mortality rate from abortion during at least the first trimester is an order of magnitude lower than the rate of mortality from childbirth. I also mention Jones which says that by the 20th week, the mortality rate from abortion has overtaken the rate for childbirth.

Other editors put information from sources that address geographical factors, or the second trimester, or contradict what I assert as a fact, or contradict Jones. Eventually a form of words is agreed and it becomes the article text. That's how collaborative editing reaches a consensus.

Now I'll ask if you can see where this process is being derailed? Instead of looking for sources that add to the process, you have repeatedly sought to analyse the sources put forward, attempting to "debunk" or smear them. That is destructive, not constructive debate. "Kulczycki is too old per MEDS#up-to-date, therefore Mazza isn't acceptable because it uses Kulczycki" – well, read "up-to-date":

Here are some rules of thumb for keeping an article up-to-date while maintaining the more-important goal of reliability. These guidelines are appropriate for actively researched areas with many primary sources and several reviews, and may need to be relaxed in areas where little progress is being made and few reviews are being published.

What are the later reviews in the area and in what way do they differ from what Mazza or Kulczycki say? We have no reason to throw out older secondary sources, in the absence of equally authoritative recent ones that modify their conclusions. What "up-to-date" is telling us, is that we should seek to replace older conclusions as newer ones supersede them. Are you confusing it with "Respect secondary sources"?

If an important scientific result is so new that no reliable reviews have been published on it, it may be helpful to cite the primary source that reported the result ... After enough time has passed for a review to be published in the area, the review should be cited in preference to the primary study. If no review is published in a reasonable amount of time, the primary source should be removed as not reporting an important result.

That tells us we may use a fresh primary source in the absence of a secondary one; that such primary sources should be replaced by secondary ones when available; and that the primary should be removed if no secondary emerges after a time. That's a procedure for discarding primary sources, not secondary.

Where did you get the idea that secondary sources have to reference their claims? A secondary source is either a reliable source or it's not. If it is a reliable source, then we do not attempt to deconstruct it, relying on our own assessment of what it says or who the authors are. And we most certainly do not attempt to "mine-down" to its sources (be they primary or secondary) and quote them instead of the secondary. The authority of a secondary source to support a piece of text rests on that source itself. You merely weaken the support by trying to tie it to the primary sources used by the secondary. I really hope your intention of sourcing your suggested text to Guttmacher 1999, instead of Lloyd 2005 wasn't to deliberately weaken it. I'll make a counter-proposal for some suggested text:

• Maternal death from abortion in developed countries is below 1 per 100,000 procedures when performed during the early stages of pregnancy. The corresponding mortality rate for childbirth is 7.7 per 100,000 live births.(Lloyd 2005) However, Richard Jones in 2006 concluded that "[a]fter the 19th week of pregnancy the maternal death rate due to abortion is greater than that of childbirth."(Jones 2006)

I have no attachment to that or to any other form of words. Feel free to do with it what you will. Now, please review what I'm asking you to do here: to engage in a process of consensus-building, and please do your best to avoid what I clearly believe to be spurious criticism of reliable sources.

We have a reliable sources noticeboard. I've made use of it recently to garner outside opinion on "Geneva Foundation for Medical Education and Research". I'm going to suggest that to enable debate to focus here, when anyone feels that a source is unreliable, they take their objections to RSN (feel free to use my request as a template), and make just a courtesy note here. Perhaps we could try that for a period of two weeks?. Is that an acceptable way to move forward? --RexxS (talk) 02:47, 12 June 2010 (UTC)

Rexx, although you were addressing Nutriveg, I just want to briefly mention that, if the comparison to childbirth is included in this Wikipedia article, then I think it would be essential to say that the risk of death from childbirth is "low", citing Lloyd 2005. Merely reciting numbers is not adequate, because lay readers may not understand that those numbers for childbirth really do reflect a low risk in the big scheme of things. I think MastCell has agreed that we can explicitly say that the risk of dying in childbirth is "low".Anythingyouwant (talk) 03:39, 12 June 2010 (UTC)

That sounds eminently sensible to me, and I follow your rationale. I think LeadSongDog also makes a valid point that will need to be addressed below in Worldwide, so chop away at what I suggested, or suggest something else. Thanks to James, we have enough sources! I have no expertise in this area at all – you're the experts. --RexxS (talk) 04:19, 12 June 2010 (UTC)

There's something difficult about this process, when people defend the use of problematic sources even when those sources won't add anything to the text, just for the sake of doing so. Me, MastCell, Anythingyouwant and Schrandit likely agreed to use (Lloyd, 2005) with the only problem being to attribute or not Guttmacher Institute. But you came back here and say: "We need to use all those six sources" when I see no gain from that.

Except for one point, its age, we aren't analyzing (Kulczycki, 1996) as a whole article but one phrase from that article, so it's pointless to count how many people cited that article if you don't know what or how they cited it. That counting doesn't solve the reported problems with that phrase/source: informal/unscientific context, unreferenced claim and old source by MEDRS standards.

That Mazza whole book also isn't being critized, but the mere citation of (Kulczycki, 1996)

Laube, 2009 is a medical text about US data, not worldwide representative, although, you want to use that number in a worldwide context.

You're trying to put forward original research, we only cite what reliable updated sources say.

Richard Jones is an awful source and I don't believe you'll get support for its use, so I'll wait and see before spending my time.

Collaborative editing reaches consensus when the reported problems are listened and addressed.

I'm not forced to search for sources supporting comparison, that WP:BURDEN lies to those trying to add such text, I did some search before and didn't find anything reliable in a worldwide context. We already have agreed on (Lloyd, 2005) so I have even less motivation to do a search.

If the sources don't respect WP:MEDRS they shouldn't be used to support medical claims, that's my point. If such source doesn't exist I won't blind eye to add that content by anyway, this is simple WP:V compliance, there's nothing about being constructive or destructive. On the other hand insistence in adding such content when the sources are problematic looks like a WP:V and WP:MEDRs violation.

I didn't see your point in citing that MEDRS phrase, there are many recent reviews about abortion, it's an actively researched area, so there's no reason to not follow Look for reviews published in the last five years or so, preferably in the last two or three years, if that information is important it will be mentioned in recent reviews otherwise it became outdated, simple like that. WP:BURDEN lies to those trying to add such text.

I didn't suggest to reference (Guttmacher, 1999) but to attribute Guttmacher as the author cited by (Lloyd, 2005). That was an initial suggestion, we can make some changes from it but I expect to respect what the sources said.

Your counter-proposal has the problem of using a US source "7.7 per 100,000" (Laube, 2009, which you referenced as Lloyd, 2005) a bad source (Jones, 2006) using a text/restriction I don't know where you got "is below 1 per 100,000 procedures when performed during the early stages of pregnancy", and not qualifying the risks involved "low", since the lay reader isn't an specialist in getting "a per 100,000" and qualifying that number by comparing it to other risks that may affect him.

We can go to the RSN if you insist in using a specific (all) sources, but I was expecting to stop when we've got just the necessary ones so we should be able to already have finished this discussion by using Lloyd, 2005.--Nutriveg (talk) 05:21, 12 June 2010 (UTC)

Have you considered this: Wikipedia articles should be based on reliable, published sources, making sure that all majority and significant minority views that have appeared in reliable, published sources are covered? Of course I want you to make use of all the reliable sources that have been brought forward! But there's no way you are going to be accepted as judge and jury of what is a reliable source. All of the sources Doc James found are reliable sources, and all of them need to be considered when writing article text.

Once again, you're not qualified to criticise what a RS says. I've demonstrated that Kulczycki is a RS (published in Lancet) - looking at how many times it has been cited is an indicator of its quality: it's high quality. MEDRS has no definition of "old source", and certainly does not forbid older sources from being used For example, Genetics might mention Darwin's 1859 book On the Origin of Species as part of a discussion supported by recent reviews.

Mazza is a RS. It makes a statement about this issue. Any RS putting forward a view should be included unless it's so much a minority view among the sources that it is not significant. Mazza is not a minority view, so should be included.

Laube is a RS and its view needs to be included for the same reasons. It's US data, so the article needs to say something about US data; we have a RS. You want world-wide data? You go find the sources for word-wide data.

Is Jones a RS? Probably. Is the view expressed a minority view? Probably. Is it such a minority view that it's not significant? I don't know, and I'd welcome debate on that point.

The "1 per 100,000 procedures" is given by Lloyd, Sockol and Laube, as you noted above at 03:17, 11 June 2010 (UTC). So, do the sources support "is below 1 per 100,000 procedures when performed during the early stages of pregnancy" or not?

The point of the MEDRS quote was to give context. "Up-to-date" is a rule of thumb for keeping an article up-to-date, not a "bright line" that automatically disqualifies reliable sources over five years old. Hope that's clearer now.

I agree that it would be good to further contextualise the mortality rates for the lay reader, but not at the expense of the figures that the reliable sources use. Perhaps they should also be characterised as "low", or "very low" - what do the sources say?

I note your accusation that I'm performing original research. I'm not, since I'm only reporting what reliable sources say, but I'd be very interested to hear why you think so. Back it up with some explanation, that will be a help.

Naturally, I agree with your statements concerning WP:BURDEN and WP:V. Naturally, I reject your implied assertion that you should be the judge of when the sources are problematic. Once we've got past the stage where you think you can cherry-pick just the sources you are prepared to accept, I'm hopeful we can start to make some progress with the article. --RexxS (talk) 07:31, 12 June 2010 (UTC)

RexxS, the text you point (without the source, so I can't understand the context) mentions "all (...) views" it doesn't say we should include mention "all the content" or "all the sources". WP:MEDRS let's us find the most approriate sources. (Lloyd, 2005) follows that criteria and represents that view. Your insistence in pushing for the inclusion of all sources, what won't make any significative improvement to the text, is delaying this process we should have already resolved

Kulczycki article is a RS, but that claim was used in a different context from the remaining of the article, so we can't just take that phrase as strict scientific knowledge and use it as so. Yes, it is 14 years old, and so outdated by Look for reviews published in the last five years or so, preferably in the last two or three years. The example you cited was described as exceptional, it was a seminal source, it was a primary source. Reviews, like Kulczycki, aren't supposed to be seminal they just cite other people work, I wonder from where he got that data. Epidemiologic data are representative to a specific population, we can't use old epidemiologic data an present is as current.

As I've said before, Mazza just copy Kulczycki, it doesn't add anything more and doesn't solve it's problems.

About Laube, this is not a US centric article as Abortion in the United States#Maternal death is, so far we've been searching for world representative sources if you didn't notice.

I don't think Jones is a RS. I don't think a source that doesn't provide references to its claims is a RS. If you think otherwise open a separate discussion for that source so it will be clear for you no one is supporting its use.

There's no individual source saying "is below 1 per 100,000 procedures when performed during the early stages of pregnancy" you were making WP:SYNTH by combining sources.

That's not my view of Look for reviews published in the last five years or so, preferably in the last two or three years. when we are talking about a 14 year old epidemilogic review using data from no one knows when.

(Lloyd, 2005) does say "low" did you mind to check my suggestion?

You're doing original sources when you say "the fact that in developed countries the mortality rate from abortion during at least the first trimester is an order of magnitude lower than the rate of mortality from childbirth" There's no updated reliable source saying that and just because a source say something that doesn't mean we should support that as a "fact". We just cite the sources, we don't defend them as true or make further conclusions about what they say.

I only pointed the problems I saw with the sources by WP:MEDRS, while most of those essential problems have been unadressed just because some people personally agree with what those problematic sources say.--Nutriveg (talk) 23:28, 12 June 2010 (UTC)

I'm sorry I forgot to quote the source of the text I drew your attention to. It's in the lead of WP:RS, our fundamental guideline for determining reliable sources. I hope the context is clearer now. You're quite right we don't include "all the content", and I never asked that we should. But WP:RS does insist that we include mention of all the views that reliable sources express. I am, unsurprisingly, pushing for us to consider what all of the sources say when we write the text. What basis do you have for believing that considering all reliable sources will not make a significant improvement to the text? There's no deadline, and taking short-cuts with the sourcing process is contrary to how we work on wikipedia.

Kulczycki is a reliable secondary, agreed. It's also on-topic for the subject of health risks of abortion. So the only question is: have its conclusions been modified by later reliable secondary sources? If not, then its views need to be included. I agree it's inappropriate to write our text to give the impression that old epidemiological data is current. Is there any evidence that the figures or the conclusions have changed significantly in the last 14 years? I'm sorry if I missed that.

The same goes for each of the sources Doc James provided. You're the one who suggested Jones - is it a reliable source you want us to consider, or not?

I had noticed we'd been searching for world-wide data, but not found much. We have found more data relating to the USA and other developed countries, and I see that the sources indicate that there's a significant difference between the data for developed countries and the rest of the world.

I read the three quotes you provided: "In developed countries, the mortality rate from legal, induced abortion in the first trimester is less than 1 in 100,000 procedures" (Sokol 2007); "In representative developed countries, the risk of dying is no more than 1 in 100,000 procedures" (Lloyd, 2005); and "Risk of death from abortion during the first 2 months of pregnancy is less than 1 per 100,000 procedures" (Laube, 2009). I suggested "Maternal death from abortion in developed countries is below 1 per 100,000 procedures when performed during the early stages of pregnancy". Do you seriously believe that is WP:SYNTH?

You propose 'just because a source say something that doesn't mean we should support that as a "fact".' Nobody is asking you support what the source says. But WP:RS, WP:MEDRS and WP:MEDASSESS require you to respect secondary sources; to include all significant views; and not to reject reliable sources because of your personal opinion of them. On Wikipedia, a "fact" is what a reliable source says it is (absent a contradiction from an another reliable source), regardless of whether you think it is wrong or not. Your disagreement with what a source says does not make it "problematical".

RexxS, everyone seems to be satisfied with the Lloyd 2005 source, and from the perspective of expedience and practicality we only need one reliable source to support inclusion of content. Is there any content that is not covered by Lloyd 2005 that we need from other sources regarding the childbirth comparison?Anythingyouwant (talk) 18:55, 12 June 2010 (UTC)

Is expediency a useful factor when writing text? Isn't that coming to the process from the wrong direction? Surely we're not writing text, then finding a source to support it. Anyway, as for your specific concern, if the text you're going to propose includes the views of all the reliable sources (duly weighted of course), then the editors here have done their job. If that text can be cited with just one source, that's a bonus. If it's any help, my feeling is that the following points related to the comparison are reflected in the sources:

1. More data is available for developed countries than for the rest of the world;
2. Both abortion and childbirth carry low risk in developed countries;
3. The risk for both abortion and childbirth is greater in most non-developed countries;
4. The risk for unsafe abortion is much greater than for "safe" (is that the right word?) abortion;
5. The mortality rate for abortion is greater in the later stages of pregnancy that in the earlier stages;
6. In developed countries, in the early stages of pregnancy, the mortality rate for childbirth is around 7 to 11 (an "order of magnitude" if you prefer words to numbers) times greater than the mortality rate for childbirth;
7. In developed countries, in the early stages of pregnancy, the mortality rate for safe abortion is somewhere around 1 per 100,000 procedures;
8. In developed countries, the mortality rate for childbirth is somewhere around 8 per 100,000 live births.

That's not proposed text, just my impression of what I've seen in the sources. Is there anything that you think I've missed, or misrepresented?

The current section on Health risks has an introduction of just four sentences. Are you thinking of replacing it or expanding it? One small point: the subsections that have a level 2 section as parent should be level 3; at present Physical health is level 4. --RexxS (talk) 20:05, 12 June 2010 (UTC)

Nutriveg, do you think that the points just described by RexxS are adequately supported by Lloyd 2005 (and any other sources that you think are Wikipedia-compliant)? If so, perhaps you could start a subsection here at this talk page with some draft content for this article?

RexxS, I usually try not to use the word "misrepresented" due to it's connotations; "mischaracterized" might be a better word, and I'd like to reserve judgment on that until there's some concrete draft language. I think expediency is a worthwhile goal, in the sense that we ought to accomplish what we can agree on now, and discuss the rest later.Anythingyouwant (talk) 22:01, 12 June 2010 (UTC)

Thanks, AYW, I appreciate your sensitivity on the use of terms that could be pejorative, but I hope you'll allow me a little leeway when I'm describing my own contributions. As I've said, I'm have no expertise in this topic. I can only offer what I've learned about sourcing and editing. As a result, I freely acknowledge that I may inadvertently misrepresent/mischaracterise what a source is really saying. The part of the process where editors fine tune text to go in the article is best left to those who understand the background and nuances, so at that point, I'll step aside. --RexxS (talk) 22:38, 12 June 2010 (UTC)

AIW, To my understanding (Lloyd 2005) does represent that view we need to represent. About the points RexxS is concerned:

• The first and second points are addressed by (Lloyd 2005). :The third and fourth are out of the scope of the problem we are discussing "the comparative mortality of safe abortion vs average maternal death in developed countries" We may open that discussion when we finish this one.
• The fifth is also off-topic but since it's simple I'll comment: We can point that maternal death by itself (not associating with anything), is greater in later pregnancy, but I believe that's already represented by "major complication"
• About the sixth: there's no updated data supporting that number, the source that support it are old (so are the numbers), those numbers are presented in an informal/unreferenced context and you can not combine numbers to create another (numeric representation).
• About the seventh: That's original research, we can use terms "early stage" that weren't used by the sources. That's redundant if the average number is the same.
• About the eighth: By what updated source? Those numbers are always changing to use old data or an undated number and I need to remember you that abortion is included in maternal mortality numbers by the WHO/ICD definition. We should present maternal mortality rate in its relevant context: mortality rate from safe abortion in developed countries is lower than the average maternal mortality rate in these same countries.
• The discussion so far is about "the comparative mortality of safe abortion vs average maternal death in developed countries" discuss other issues in another discussion section but a lot of people are expecting this discussion, where you remain the only opposing voice, to end so we can change that problematic text.--Nutriveg (talk) 00:58, 13 June 2010 (UTC)

Have you now decided that we should limit the text that goes into the article to just developed countries? You suggested rejecting Laube less than six hours ago because it was US-centric and you were looking for worldwide sources.

1. If Lloyd addresses "More data is available for developed countries than for the rest of the world", why doesn't your proposed text mention non-developed countries?

3. How can "The risk for both abortion and childbirth is greater in most non-developed countries" be out-of-scope of a comparison between the risks for abortion and childbirth?

4. The risk of unsafe abortion is massively different from safe abortion. How can explaining that any comparison depends on the conditions under which abortion is performed be out-of-scope?

5. If the mortality rate changes with stage of pregnancy, how is that "off-topic" for a comparison using that rate? Where is the phrase "major complication" and how does it address this?

6. Laube (2009) gives 7.7 per 100,000 for US and makes the comparison directly. Looking back in time Clarke (2008) gives 6.5, while CDC (1999) gives "approximately 7.5" and 7.7 for the US. I don't see any significant variation over the time period you object to. Laube is a 2009 source and does the comparison of rates, not me. Do you think WP:SYNTH applies to reliable secondary sources?

7. If one source uses "first trimester", another "first 8 weeks", another "first two months", what is the problem with me summarising that as "the early stages of pregnancy"? I utterly reject the notion that we are disallowed from using terms synonymous to those used by the sources. The whole process of writing for an encyclopedia is to neutrally summarise what the sources say in a manner understandable to our audience. An article written by simply stringing together quotes from sources is unlikely to be much of an article. Please remember that too close a paraphrase of a source is plagiarism.

8. see 6

The sole voice derailing the consensus process is yours. You are the only voice that's proposing a change of scope at this late stage, and the only voice that's proposing to reject all but one source. Everybody else has accepted the other sources as reliable. I still don't see how we can make progress when you unilaterally change the terms of the discussion, and insist on cherry-picking a single source to work from. --RexxS (talk) 04:46, 13 June 2010 (UTC)

1,3,4 The very issue we are discussing here is "comparing a global rate of maternal death from safe abortion with a comparable rate of general maternal death". That's the problematic text, if you came up wanting to discuss unsafe abortion in the same discussion that's off-topic of this discussion, so you should create a separate discussion for that

5, Why you didn't mind to check the article? "rate of major complications (...) varies depending on how far pregnancy has progressed" It's offtopic when that's a isolate point from that discussion above.

6,You said "In developed countries, in the early stages of pregnancy, the mortality rate for childbirth is around 7 to 11" but presented US data to support that

7,Lloyd 2005 doesn't say early stage, if all sources say lower that 1 in 100000, no matter the stage they used, there's no need to differentiate a specific period unless another source puts a higher number for other period.

We have analyzed those sources and the one which was mutually accepted by WP:MEDRS standards was (Lloyd, 2005), (by AIW, Mastcell, Schrandit[2] and me/Nutriveg) which fairly represents the other sources view. So far you're the only one disagreeing with that demanding the use of all sources even knowing they have several problems by WP:MEDRS and don't carry any other consistently or representative information.--Nutriveg (talk) 01:50, 14 June 2010 (UTC)

I made it clear that That's not proposed text, just my impression of what I've seen in the sources.

1,3,4: The exact issue we're discussing is "comparison of the health risk of abortion with that of childbirth", nobody but you has suggested it has to be confined to just developed countries or only safe abortions.

5: That's a statement about complications, and it only refers to surgical abortions. How does it address what we know the sources have to say about mortality from abortions in general?

6: So if I quote the sources that show the US is typical of developed countries, and the MMR is around 7 to 11 for other developed countries, you'll agree that the issue should be included in any text we write?

7 Kulczycki uses the phrase "up to 8 weeks gestation"; Mazza uses the phrase "prior to 8 weeks gestation"; Sokol uses "first trimester"; Laube uses "the first 2 months of pregnancy"; Lloyd doesn't qualify the stage. We're sure of the figures for the early stages, but guessing that they are the same later on is pure speculation.

We don't analyse secondary sources; we report them. Lloyd does not cover the all of the views presented in the other sources. The other secondary sources are no less reliable than Lloyd and the issues they raise cannot be excluded, no matter how much you claim they are old, inconsistent, unrepresentative, unsourced, perform synthesis, cite another source, or any of the other spurious reasons that you've advanced. --RexxS (talk) 02:59, 14 June 2010 (UTC)

Another source

Sorry if this has been examined and rejected earlier, but isn't this the recent, relevant review that we've been looking for? It seems to specifically address the issue we wanted – at least for the USA:

• Christiansen LR, Collins KA (March 2006). "Pregnancy-associated deaths: a 15-year retrospective study and overall review of maternal pathophysiology". American journal of forensic medicine and pathology. 27 (1). National Association of Medical Examiners: 11–9. PMID 16501342.

I can only see the abstract, but it looks promising. Does anyone have Swets or other access to the full text? --RexxS (talk) 05:04, 12 June 2010 (UTC)

I have full access. Anything specific you want me to look for? It doesn't appear to address abortion.-Andrew c ^[talk] 22:10, 12 June 2010 (UTC)

Thank you Andrew. If it doesn't address abortion then there's nothing else needed. It won't be relevant here, other than to confirm a recent estimate for maternal mortality in the USA. I think we already have that from the existing sources. --RexxS (talk) 22:38, 12 June 2010 (UTC)

Rephrasing my earlier suggestion

Earlier I've said:

I propose we use (Lloyd, 2005) "In representative developed countries, the risk of dying (from abortion) is no more than 1 in 100,000 procedures, lower than the relatively low risks associated with pregnancy and childbirth in these countries (The Alan Guttmacher Institute, 1999)" we may also attribute that claim to their original author Guttmacher Institute, since (Lloyd, 2005) did that attribution and given this is not an "ideal source" as discussed above.

So an early version would look: "According with the Guttmacher Institute maternal death from abortion in developed countries is bellow 1 in 100,000 procedures, lower than the relatively low risks associated with pregnancy and childbirth in these same countries"

Likely me(Nutriveg), MastCell, Anythingyouwant and Schrandit agreed with that version. While MastCell expressed a problem in attributing Guttmacher Institute and Schrandit expressed a problem in not attributing. So far RexxS disagreed and proposed another version. Other editors remained silent so I understand they agree or don't care about the outcome.

Expecting to address Mastcell and Schrandit concerns I'll remove the attribution while rephrasing that text. I expect others will maintain their position despite of this change (otherwise we can return to the previous suggestion) and RexxS will make an effort so we can get at least a stable version even if he thinks that's not complete enough.

So this is the new proposed version:

"In developed countries the maternal death rate from abortion is bellow 1 in 100,000 procedures, lower than the general maternal death rate, which is relatively low."

By the linked "maternal death" definition it's clear that's associated with pregnancy and we avoid direct associations that some understand as erroneous or not consistently supported by updated reliable sources.--Nutriveg (talk) 01:57, 13 June 2010 (UTC)

Let's be clear then what you're proposing. You're asking us to discard consideration of all of the sources except the one that doesn't compare the figures for mortality in childbirth, and then use a quote from that source. What part of the current text are you proposing to remove in order to insert that text? --RexxS (talk) 02:52, 13 June 2010 (UTC)

I'm using a source that fairly represents a view and all the editors who had opinion to express agreed with that except for you.

I'm using the scientific technical terms to represent that source, so I'm presenting that data as what it is and in a neutral way, not advancing a position that's WP:SYNTH and not giving more emphasis to it than the one given by the source itself.

The main focus here is to replace the text which was focus of the recent problems "by 1996, mortality from childbirth in developed countries was 11 times greater." The earlier part of the sentence "risk of maternal death is between 0.2-1.2 per 100,000 procedures" should also be replaced since that would became redundant and a range that varies that much doesn't give an idea of the data distribution within that range.--Nutriveg (talk) 18:00, 13 June 2010 (UTC)

I have no problem with writing text that accurately and neutrally reflects the views of all of the reliable sources, including Lloyd, so let's be clear: None of us are objecting to Lloyd - you are the only one objecting to also including what is written in other reliable sources.

So you want to replace the sentence:

• "In such settings, risk of maternal death is between 0.2-1.2 per 100,000 procedures[33][34][35][36] in comparison, by 1996, mortality from childbirth in developed countries was 11 times greater.[37][38][39][40][41][42]"

with:

• "In developed countries the maternal death rate from abortion is bellow 1 in 100,000 procedures, lower than the general maternal death rate, which is relatively low. (Lloyd 2005)" (or is it to be cited to Guttmacher?)

You've not used the figures for maternal death rate, avoiding the comparison that Laube 2009 makes. That is despite the previous lengthy discussions at #Safer than childbirth and #Reference to settle the debate asking for the comparison to be made because it's a comparison made accurately in reliable sources.

You've ignored mention of the fact that the "1 in 100,000 procedures" varies significantly between different stages of pregnancy. Is it not misleading to leave that out? --RexxS (talk) 18:46, 13 June 2010 (UTC)

Yeah (Lloyd,2005) is global, better fits WP:MEDRS, including by being update and fairly represents the other sources in a consistent way, meaning some of these (worse by MEDRS) sources make claims not supported by the others, like using the word "safer", presenting (different) comparative numbers, or not citing which global epidemiological study reached that conclusion. While none of these (worse by MEDRS) sources refute (Lloyd,2005) claims in a relevant way.

(Laube, 2009) is US centric and doesn't detail its numbers (date, source).

The proposed text gives due representation to those issues you mentioned when it compares: "Maternal death rate from abortion is lower than maternal death rate from general causes" where a link to "maternal death" is provided to make sure the reader understand the association with pregnancy.

Adding further terms is an issue first because (Lloyd,2005) is a better but not an ideal source by WP:MEDRS and rely on a single primary source (Guttmacher) for his claims. Secondly because this and the other (worse) sources don't agree with a common clearly understandable medical term "pregnancy", "childbirth", "live birth", "pregnancy or childbirth", "brought to term" like the standardized WHO definition of maternal death. The WHO term is not just standardized, but also is the one statistics are based on and so readily available, where the reader can also further check for himself instead of trusting a claim from a single author or weak sources.

The increased risk of major complications is already expressed in the section bellow, we may repeat it as a separate phrase.--Nutriveg (talk) 20:57, 13 June 2010 (UTC)

For comparison, I'll suggest that the paragraph should look more like:

"Abortion, when legally performed in developed countries is among the safest procedures in medicine.[31][32] (remains the same) Maternal death from abortion in developed countries is below 1 per 100,000 procedures performed during the first eight weeks of pregnancy, with the rate increasing in the later stages; while for childbirth in the USA - itself very low risk - the mortality rate is 7.7 per 100,000 live births.(Laube 2009, p.150)(Lloyd 2005) In the developing world, much higher mortality rates from abortion and childbirth can be found, and they remain some of the most significant risks to young women's health.(Lloyd 2005, p.5) (possible new text) Unsafe abortions (defined by the World Health Organization as those performed by unskilled individuals, with hazardous equipment, or in unsanitary facilities) carry a high risk of maternal death and other complications.[43] For unsafe procedures, the mortality rate has been estimated at 367 per 100,000.[44]" (remains the same)

If you felt it essential to specifically reference the "much higher mortality rates from ..." then Okonofua 2006 (full text, p.7)ought to suffice, but I would have expected the rest of the paragraph would supply the context. Thoughts? --RexxS (talk) 20:31, 13 June 2010 (UTC)

Perhaps there is an over-emphasis on the comparison with childbirth, so the last sentence of the possible text could have 'childbirth' removed, to read:

• Maternal death from abortion in developed countries is below 1 per 100,000 procedures performed during the first eight weeks of pregnancy, with the rate increasing in the later stages; while for childbirth in the USA - itself very low risk - the mortality rate is 7.7 per 100,000 live births.(Laube 2009, p.150)(Lloyd 2005) In the developing world, much higher mortality rates from abortion can be found, and it remains one of the most significant risks to young women's health.(Lloyd 2005, p.5) (possible new text)

It less close to what Lloyd wrote, but may balance concerns of undue weight. --RexxS (talk) 20:55, 13 June 2010 (UTC)

RexxS the very issue we are discussing here is comparing a global rate of maternal death from safe abortion with a comparable rate of general maternal death. We don't compare one world data with a US one, we don't cite data from every country here when they have their own abortion in <named country> article, we don't combine data to advance a position, we don't present inconsistent data between sources, we choose the sources that better fit WP:MEDRS, we don't give more attention to a issue than the sources themselves did in a similar context. Your suggestion suffer from all these problems, while every other editor showed support or had no disagreement with my suggested text. Let's commonly agree on something. Even if you think that's not complete it's better than nothing.

Unsafe abortion is not the issue here, open a different discussion section if you want to this discuss that.---Nutriveg (talk) 21:15, 13 June 2010 (UTC)

If we're back to comparing a global rate of maternal death from safe abortion with a comparable rate of general maternal death, then your proposal doesn't address it. I used a comparison directly from Laube, making it clear that he was using US data. I didn't "combine" the data, Laube did - we don't call reliable sources "inconsistent" on the basis of nothing more than our own opinion, particularly when multiple other sources reach the same conclusion. Unlike you, I have no position to advance, other than to ensure our core standards of respecting secondary sources is complied with. We don't exclude relevant summary from a parent article just because a daughter article exists. We do assemble the views of multiple sources in a neutral manner to produce article text. We don't cherry-pick a single source from many and completely disregard the other equally reliable sources. We don't leave out issues that are covered in multiple reliable sources, by spuriously claiming that they are undue weight - are you claiming that health risks is given too much weight appropriate to its significance to abortion, or that five sentences devoted to data and comparison in a section containing eight paragraphs and two subsections is giving too much weight appropriate to the significance of data to health risks?

Nobody's expressed agreement with your proposed text. Why are you so wed to it, despite its obvious short-comings? Is it simply that Lloyd's phrasing is least inimical to your own point of view?

Unsafe abortion is an issue here. Are you trying to tell us that unsafe abortion is not an issue relevant to the health risk of abortion? You seen a dozen reliable sources that say otherwise. --RexxS (talk) 22:06, 13 June 2010 (UTC)

Why my proposal doesn't address "comparing a global rate of maternal death from safe abortion with a comparable rate of general maternal death"? As far as I'm concerned "developed countries" is not country specific, unfortunately we don't have a mortality from safe abortion rate encompassing more countries.

You compared "Maternal death from abortion in developed countries" with mortality "for childbirth in the USA"

(Laube, 2009) is a US source it says nowhere: "Maternal death from abortion in developed countries is below 1 per 100,000 procedures performed during the first eight weeks"

As I've said other sources are inconsistent because they "make claims not supported by the others, like using the word "safer", presenting different comparative numbers or not citing which global epidemiological study reached that conclusion"" beyond the terminology used. They don't reach the same conclusion but their general view is fairly represented by (Lloyd,2005)

Well I expect to follow WP:MEDRS when writing medical claims but I don't think that's advancing a position since that's exactly MEDRS purpose.

We try to represent a global view not giving more emphasis to a specific country whose statistics is already included in the represented global view, otherwise we would have to include every single country which has such data but this article is already too big. If you think bellow 1 per 100000 is not representative enough we can include a abortion mortality table.

We give WP:DUE to these issues, there's a single, updated source that fits MEDRS, making such claim in a global base (while we have thousands of sources talking about abortion mortality) and it do that shortly compared to the remain of its text. On the other hand we have a small intro section and you want to combine information from several (worse by MEDRS) sources giving more emphasis to this issue than the original source did.

We only have one source fitting WP:MEDRS because we have few sources about this issue! So far the discussion is: "comparing a global rate of maternal death from safe abortion with a comparable rate of general maternal death"

AIW did support, so did Mastcell, as Schrandit[2] and me (Nutriveg). We only have the issue of atributing or not (Guttmacher) since that (Lloyd,2005) information relies and is attributed to that single source, the only problem we are trying to address. The only one disagreeing is you. I'm only trying to find a stable due version supported by updated reliable sources by WP:MEDRS, if it was just for me I wouldn't have accepted Lloyd,2005 since it's not an ideal source, but I made that concession. While you want to add anything that you believe is true, which you call "fact".

I'm only trying to say unsafe abortion is outside of the problem we are trying to solve, so you should discuss it in a new discussion section--Nutriveg (talk) 00:30, 14 June 2010 (UTC)

When did we decide that we were going to only address health risks of safe abortion in developed countries? The vast majority of abortions occur in developing countries, and many of those in unsafe conditions.

One source says 'safer' and another says 'the risk is lower' or compares '1 per 100,000 with 7.7 per 100,000' and you call that inconsistent. Humpty Dumpty would have been proud of that.

It's not your place to reject a reliable secondary source because you don't know where gets its data from; it has the authority to analyse and perform synthesis from its author, its publication, and its peer-review; you don't get to apply your own filtering criteria to reliable secondary sources.

We have half a dozen or more sources that "fit" MEDRS, but as a concession to your demands for stringent application of the rule-of-thumb "Up to date", I only cited the most recent secondaries, although I did weigh what was in the earlier ones, but found no contradiction.

• AIW did support?: "everyone seems to be satisfied with the Lloyd 2005 source ... Is there any content that is not covered by Lloyd 2005 that we need from other sources regarding the childbirth comparison?" - yes, support for including Lloyd. I see no support for your text.
• so did Mastcell?: "Your proposed text doesn't work for me ... " - that's strange support.
• Schrandit[2]?: "Keep in mind who Guttmacher is and where their money comes from. We should probably mention it in the article.", "Over here I've got Guttmacher saying its around 1,100 a year. I also hear that most of those state bans are very, very weak.", "Yeah, intuitively it didn't make sense to me either but I usually hear abortion classified as family "planning". To AYW's earlier point about linking to a more expanded article - I think it would still be worth it to give the 1 more reliable, recent number to give the reader an idea and then link to the main article if they want to know the rest." - Where's the support for your text in that?

It seems more like everybody is questioning your text. On Wikipedia, I call "fact" what can be sourced from reliable sources. Here, I have no beliefs other than that following Wikipedia's policies and guidelines is the only acceptable route to writing an article. Do I need to repeat Wikipedia articles should be based on reliable, published sources, making sure that all majority and significant minority views that have appeared in reliable, published sources are covered?

And I'm only trying to say that we have sources discussing the health risks of abortion in the developing world (the vast majority), and addressing only abortion in developed countries is surely a violation of WP:UNDUE.

Both of seem to be repeating the same objections without persuading the other. Let me make an offer. Get the clear support for your text of the other editors who have contributed heavily to this debate: Andrew, AYW, LeadSongDog, James, Mastcell, Schrandit (have I missed any?) and I will certainly support it as well, despite my reservations. Optionally, let's go to dispute resolution (3O, mediation) or open an RfC to decide. --RexxS (talk) 01:48, 14 June 2010 (UTC)

When we start discussing in a section we should be on topic with the discussion subject of that section, this one being "comparing a global rate of maternal death from safe abortion with a comparable rate of general maternal death" the main contendious point discussion in this article. If you want to discuss other stuff, like unsafe abortion, open a different discussion section!

Yes it's different to say "safer" than presenting something as having a lower mortality rate. "Safer" implies an analysis was performed, a lower data rate is just lower raw data rate. We should present information as by the source. And I always need to rember you Laube is US centric

If a study doesn't attribute the source for some claim, and that claim can't be found in any source, we should threat that claim as primary research.

Whether a source is regarded as primary or secondary in a given context may change, depending upon the present state of knowledge within the field. For example, if a document refers to the contents of a previous but undiscovered letter, that document may be considered "primary", since it is the closest known thing to an original source, but if the letter is later found, it may then be considered "secondary"

If you did not find contradiction beteween the sources then we should base the contendious text in (Lloyd,2005) and that should be enough

We first look for sources and later write text, after having a source we could use, I made my first suggestion based on that source we finally had (Lloyd, 2005)

• AIW clearly points to be satisfied by (Lloyd,2005). Satisfied in the sense a need was fullfilled and it wasn't necessary "other sources regarding the childbirth comparison". He sees everyone the same way, except for you, the only one asking to use other sources
• Mastcell, said why he saw a problem with the text "Your proposed text doesn't work for me, because it incorrectly singles out the Guttmacher" so if he agrees with the text but not with the Guttmacher attribution
• Schrandit after I agreed (Lloyd,2005) was reliable and updated he said: "it would still be worth it to give the 1 more reliable, recent number to give the reader an idea" later he commented about the suggested text saying he supported keeping the Guttmacher attribution in the text, opposing MastCell "Keep in mind who Guttmacher is and where their money comes from. We should probably mention it in the article."

"Fact" has a whole different meaning use the appropriate words to reflect your ideas, in Wikipedia we usually use words like "view" and "information" to represent what you're talking about

Can't you separate the "developing world" "unsafe abortion" discussion from this one? In a diferent discussion section or after we finish this "develop world" issue? That would be helpful since everyone else is having no problem discussing these different topics separatedly.

I'll ask other editors on their talk page to make clear if they are satisfied or not by (Lloyd,2005) and the text derived from it.--Nutriveg (talk) 03:46, 14 June 2010 (UTC)

Health risks rewrite

I've taken a shot at rewriting the section on health risks. The previous section had several shortcomings (IMHO):

• Numerous one-sentence paragraphs
• Spotty overview of health risks, with focus on isolated factoids and no clear summary
• Written in a somewhat opaque style
• Did not address the difference in risks between "safe" and "unsafe" abortions, which is of great importance to a worldwide (as opposed to Western) overview

I'll leave it open for comments and changes. I think the new version is a bit better (obviously), in that it clearly delineates the difference in risk between "safe" and "unsafe" abortion. It also provides a more organized overview of risk - both the absolute risks and relative risks for abortion and childbirth are given, and they are juxtaposed (as they are in virtually all scholarly sources on the topic) rather than presented in isolation. I think I've emphasized that there are actual risks associated with abortion, to address the charge (above) that the article made it sound like abortion was entirely risk-free as opposed to merely very safe.

I've tried to avoid sourcing material from websites of organizations like Planned Parenthood and Guttmacher. While these websites have ample and high-quality information, their association with the pro-choice viewpoint in the US always renders them problematic. I've tried to go for review articles, WHO material, etc supported by some key primary sources which illustrate the conclusions of secondary sources.

There are still some areas that need work. For example, the specific health risks of medical abortion (at least those that differ from surgical abortion) remain to be spelled out, pending a good source. Additionally, the "mental health" section should be condensed and probably conflated with other non-risks like breast cancer. I'd welcome any updates, new sources, or comments. MastCell ^Talk 23:47, 1 June 2010 (UTC)

Looks pretty good and thorough and sourced. I thought you had devoted a good amount of space to medical abortion. Thanks for taking the time and effort. Kodus. That said, I found it a little awkward to have the parenthetical phrase (requiring surgical abortion) twice in the same paragraph. And now we have two sections in the article called "unsafe abortion" so direct linking will break (and should we be discussing the health risks of unsafe abortion before discussing the topic of unsafe abortion more generally?) I may have some more comments later, but overall, an improvement for sure. -Andrew c ^[talk] 01:30, 2 June 2010 (UTC)

I haven't got time right now for a detailed critique, so this will be brief.

I do not see any need for this article to contain two separate subsections titled “unsafe abortion”. Much of this newly overhauled section is redundant. For example, the new material says: "Medical abortion with mifepristone and misoprostol is effective through 49 days of gestational age." But the article already says, "When used within 49 days gestation, approximately 92% of women undergoing medical abortion with a combined regimen completed it without surgical intervention." Why so much redundancy?

I do not see any reason to completely exclude information about contraindications, especially for medical (i.e. non-surgical) abortion. We have previously discussed a source that details the contraindications for both medical and surgical abortion.[12]

I support the statement that for most women abortion is safer than childbirth, when performed according to recommended protocols. But why now omit that the risk of maternal death is slight even for childbirth? This hugely slants the article. We've discussed this at this talk page repeatedly, so should we conclude that the omission from the article is deliberate? All of this material was in the NPOV (but reverted) edit I made to this section of the article, including the statement that abortion is safer than childbirth (not just early abortion or surgical abortion). Now that material is omitted. This is not just a matter of emphasis, but of complete omission.

Additionally, we're not writing a textbook here, so I would think you would be able to find accurate sources that are available on the internet for free, via Google Books or Google News Archive, or Google Scholar, so that other editors can look at them. Did you look and were unable to find them? If so, doesn't that say something about the non-notability of the information being cited?

When saying that abortion is safer than childbirth, I would prefer if we hedge a little bit, as our frequently cited author (Grimes) does. He says: "For most women, fertility regulation by contraception, sterilization, or legal abortion is substantially safer than childbirth" in his article "The morbidity and mortality of pregnancy: still risky business", Am. J. Obstet. Gynecol., Vol. 170, pages 1489–1494 (1994), PMID 8178896. Is there some reason to depart from what Grimes says here?

I object that this article devotes so much space to rebutting notions that could easily be rebutted in much less space (health risks, breast cancer, fetal pain, effect on crime rate), while not addressing a primary reason why women may decide not to get an abortion, namely the alleged indicia of humanity in the embryo or fetus.Anythingyouwant (talk) 01:43, 2 June 2010 (UTC)

You guys both identified the duplication of "unsafe abortion", which I have to admit I missed. I agree that we should streamline and combine those two sections, though I'm not sure at a glance how best to do so. I do think that a discussion of health risks needs to detail the distinction between "safe" and "unsafe", but perhaps we can shorten/merge somehow.

Anythingyouwant identified duplication of the timeframe for medical abortion. I admit that I missed that redundancy as well; it's not essential to repeat the timeframe for a discussion of health risks, so it could probably be excised from the "Health risks" section.

Contraindications are not the same as health risks. In fact, there is no health risk to someone with a contraindication, because they won't be eligible to have an abortion. That's not to say that there is no room for a discussion of contraindications to various methods of abortion, but it didn't seem to fit properly in "health risks", since contraindications to the procedure are clearly distinct from health risks caused by the procedure.

I don't think I have omitted information about the small risk of maternal death from childbirth. In fact, I gave precise and properly sourced figures for the exact magnitude of the risk, just as I did for the risk of abortion. The reader can see (rather than simply take our word for it) that both absolute risks are slight. If anything, I've emphasized the safety of both childbirth and abortion by providing absolute risk estimates, while also mentioning the relative risks of the two (as reliable sources do).

Quality of information is not synonymous with its free availability. Any overview worthy of a serious, respectable reference work will necessarily be based on the scholarly literature. In many cases, the scholarly literature is not freely available online - although it is generally freely available at any halfway-decent library, thanks to our tax dollars at work. I looked for the best available sources - as in those in high-quality journals, those that dealt with the topic at hand directly, and those that are frequently cited by reputable authorities - and this is what I came up with.

Most sources don't "hedge" at all when saying that abortion is safer than childbirth - I actually hedged by saying "through 21 weeks", because a lot of expert sources make a blanket statement that abortion is safer at any gestational age. I don't think it is an accurate representation of Grimes' article, or of the scholarly literature in general, to excessively hedge here - if anything, we're understating the degree to which reliable sources hold that abortion is safer than childbirth.

As to space, we go where the sources go. For awhile, there was a lot of literature addressing the (non-)link between abortion and breast cancer. Likewise with the supposed mental-health sequelae of abortion, and fetal pain. I'm not as familiar with available sources on the impact of views on the humanity of the fetus on the decision to have or not have an abortion, but I agree that this would be a relevant topic for this article to discuss. If you have some decent sources in mind, please lead the way. MastCell ^Talk 04:20, 2 June 2010 (UTC)

This edit has several problems.

• There are specific articles for Unsafe abortion and Medical abortion so much of that edit belong to other articles. ::::* Concerning surgical abortion it gives undue balance to mortality risk as a health risk, since that risk is minimal, but the first paragraph only talks about it.
• The statement that complications is rare is not supported by updated sources, complication risk is low but they are common due to the high number of abortions. Botha, Rosanne L.; Bednarek, Paula H.; Kaunitz, Andrew M. (2010). "Complications of Medical and Surgical Abortion". In Guy I Benrubi (ed.). Handbook of Obstetric and Gynecologic Emergencies (4 ed.). Lippincott Williams & Wilkins. p. 258. ISBN 978-1605476667. {{cite book}}: Unknown parameter |coauthors= ignored (|author= suggested) (help)
• Complications are not well represented and the presented solution is deceptive. (Botha, 2010)
• Procedures details, like: description of aspiration, places where they are performed and antibiotics should belong to the procedures section.

So I'm reverting that major change. Please add unrelated text to specific sections/articles and make other changes incrementally.--Nutriveg (talk) 14:21, 2 June 2010 (UTC)

I reverted your major change as well, as MastCell's first 3 bulletted points above discuss the shortfalls of your version. I'm not going to talk down to you and ask you to make changes incrementally. Bold edits are to be encouraged, and I think MastCell's went a long way to improve that section. The section is how it was for quite a long time before the recent edits. It is the longstanding consensus version. There is no consensus for any change (Nutriveg or MastCell's) so we revert to default. What can we do to address everyones concerns and perhaps find a harmony between the two versions?-Andrew c ^[talk] 14:42, 2 June 2010 (UTC)

Or we can continue to edit wildly and pray out edits don't get reverted, in lieu of discussing our differences :Þ -Andrew c ^[talk] 15:23, 2 June 2010 (UTC)

I've addressed the issues raised by MastCell and moved unrelated content elsewhere. That previous version (where I started from today) is closer to that older version where incremental changes were made since then. Beyond that I don't like putting specific numbers to maternal death since that number varies yearly, IIRC it's currently around 0.4 for abortion and 2.5 to live births. I refuse making straight comparisons of maternal death because those are not statistically adjusted (age, risky pregnancies,...), both numbers are very low and comparing near "zeros" is stupid.--Nutriveg (talk) 15:28, 2 June 2010 (UTC)

Andrew c's deletion edit note implies this is going somewhere else, but that doesn't seem to have happened.LeadSongDog come howl! 16:51, 2 June 2010 (UTC)

Abortion and mental health already mentions the Johns Hopkins University's 2008, as is information regarding economic and decision issues. Did I delete any content which isn't covered already at Abortion and mental health in more detail? Or am I missing your point entirely?-Andrew c ^[talk] 17:01, 2 June 2010 (UTC)

Sorry if it was already there in Abortion and mental health that's fine, but the edit comment implied the insertion was yet to be done. LeadSongDog come howl! 19:37, 14 June 2010 (UTC)

← To address Nutriveg's concerns:

• I agree that we should avoid duplication of content. But I don't see how we can accurately discuss the health risks of abortion without drawing some sort of distinction between safe and unsafe procedures (I think there was general agreement on that). Do you have any suggestions about how to cover this distinction in the "health risks" section? That might be more useful than reverting.
• Mortality is the most prominent health risk, both as a matter of common sense and as a matter of weight in reliable sources, so I think our coverage reflects that weight (whatever Nutriveg's personal opinion of it might be).
• If you want to add information on non-lethal complications, feel free. The rate is very low. The source you favor (Botha 2010) states: "Fewer than 0.3% of abortion patients experience a complication that requires hospitalization." If you'd like to include that sort of information, be my guest.
• If you wish to use different language to discuss the frequency of side effects, then propose some rather than reverting the entire edit. For example, it might be more precise to say that the complication rate after abortion is very low (perhaps citing the numbers given by Botha in my bullet point above). We could add that although the rate of complications is low, abortion is a common procedure and therefore it is "not uncommon" for gynecologists to encounter women who have had complications from abortion. Personally I think this is a bit silly - the article should probably be written from the perspective of the individual reader rather than the practicing gynecologist - but whatever.
• I'm not sure what you consider "deceptive"; elaboration might be useful.
• If you want to move specific details of the procedure, please feel free (doing so does not require a blanket revert). I think some minimal level of detail is necessary, since the risks may vary depending on which procedure is used, but I agree that most detail should be in the "procedures" section.
• I'm interested in the comparisons made by reliable sources. Those sources repeatedly and commonly compare the risk of abortion to that of childbirth, and so this is the appropriate comparison for article, regardless of whether Nutriveg personally finds it "stupid".

Thoughts? MastCell ^Talk 18:15, 2 June 2010 (UTC)

Answering

• Your separation of safe vs unsafe was already incorporated to introduce the specific articles.
• Mortality is not the most prominent health risk in this case, it's negligible so there's no reason to be lengthy about that.
• That source said despite the low risk of complications incidence is common due to the high number of abortions not rare as you early said. The current text uses the same terms of that source "low".
• I didn't revert the whole edit, you that rewrote the whole section. I incorporated those relevant changes in this an other specific articles. I'm against citing numbers since they vary from country, year, source and method. Most of the sources, like Botha is US centric.
• Your text mentioned only one trivial medical procedure that would be necessary in case of abortion complications.
• I may not be the most appropriate person to move those procedure descriptions to the appropriate section or article, the article is already big and it doesn't need to repeat itself.
• As I said those comparisons weren't statistically adjusted (for age, risky pregnancies, ...) and dividing anything to zero results in a large number. According to your 0.56 per 100000 maternal death number, abortion would be almost 3 times more deadly than vaginal birth. These comparisons don't carry relevant death risk to worth comparing, less a direct (unadjusted) one.--Nutriveg (talk) 19:29, 2 June 2010 (UTC)

Have you read the article you mention in your last bulletpoint (PMID 18455140)? It notes a maternal death rate of 6.5 per 100,000 births, which is very close to the figure we cite. You must be looking at the rate of maternal death causally related to mode of delivery, which for vaginal birth was 0.2 per 100,000. That statistic reflects only deaths which (in the authors' opinion) could have been avoided by going to C-section. It does not represent the overall maternal death rate. You're comparing apples to oranges.

Do you understand the difference? It's a subtle point, but it nicely points up the reason why WP:MEDRS urges against cherry-picking primary sources. You've gone to a primary source which has nothing to do with abortion and pulled out a number. As it turned out, you've incorrectly interpreted the paper, which is a risk when you employ papers to make a point that the authors themselves don't make. If we didn't double-check your assertion carefully, we might have made a similar error.

That's why I'm citing papers which specifically address the safety of abortion as it relates to childbirth. The numbers are actually quite consistent if you read carefully. But even that issue becomes superfluous if we choose sources which directly address the questions at hand, rather than trying to pull numbers from unrelated publications to buttress our personal viewpoints. MastCell ^Talk 20:56, 2 June 2010 (UTC)

First of all I'm not trying to use that source, I just used it as an example it's not worth to compare such near zero numbers. ::::I just read the article abstract, I don't have access to the whole article but it says "The rate of maternal death causally related to mode of delivery was 0.2 per 100,000 for vaginal birth" not "The rate of preventable maternal death causally related (...)" as you stated above. Sure there are other causes but it's just an example we can't make a direct comparisons (without adjust) since the causes of maternal death are restricted to specific factors. --Nutriveg (talk) 21:11, 2 June 2010 (UTC)

I still don't expect to use comparisons in the article but just as an information see what I've found "After the 19th week of pregnancy the maternal death rate due to abortion is greater than that of childbirth"

Jones, Richard Evan (2006). Human Reproductive Biology (3 ed.). Academic Press. p. 429. ISBN 978-0120884650.--Nutriveg (talk) 22:19, 2 June 2010 (UTC)

You don't need full-text access. See the abstract, first sentence under "Results": Ninety-five maternal deaths occurred in 1,461,270 pregnancies (6.5 per 100,000 pregnancies.) This is not an "example" of anything except the danger of using primary sources without reading them carefully.

As to the Human Reproductive Biology book, it doesn't describe the basis for its claim that abortion is riskier than childbirth after 19 weeks. Nonetheless, if you'd prefer to provide various expert interpretations of the risk at late time points, we could cite it as part of a full discussion, along with sources such as eMedicine and UpToDate, which argue the opposite (that abortion is safer at any gestational age). It seems a bit excessive, since extremely few abortions are performed after 19 weeks, but I'm fine either way. MastCell ^Talk 22:32, 2 June 2010 (UTC)

Please don't read problematic content without discussion, ignoring the problems reported before, like making Division by zero to imply a high risk from maternal death from child birth when it's actually very low, using unadjusted correlation, adding content that belong to other specific articles and making major changes that can't be individually reverted.--Nutriveg (talk) 01:43, 8 June 2010 (UTC)

One of the changes repeatedly made by you is replacing maternal death with mortality rate, please use the correct naming definition by WP:MEDMOS. Abortion is pregnancy management.--Nutriveg (talk) 02:18, 8 June 2010 (UTC)

Since those US statistics were so important to you I moved them to the specific article and created a link for it. I also would like to remember the same was already done for Medical abortion long before.--Nutriveg (talk) 02:52, 8 June 2010 (UTC)

About medical abortions, and the four 2005 deaths due to infection. From my recollection of those events, and their media coverage, the deaths were related to an off label application of misoprostol. Misoprostol was not recommended to be taken vaginally, yet all 4 cases were related to the off label application. I personally feel that if we are to mention the deaths, perhaps we should also mention that it was off label usage? But then again, the RS we are citing doesn't think it is important enough to mention, so maybe I should be quite as I clearly don't know better than published scientists... (quick google search found something like [13])-Andrew c ^[talk] 02:51, 8 June 2010 (UTC)

This is likely a issue for the Medical abortion#Health risks article.--Nutriveg (talk) 03:03, 8 June 2010 (UTC)