Talk:Thrombophilia/GA1

GA Review
The edit link for this section can be used to add comments to the review.''

Reviewer: WhatamIdoing (talk) 04:40, 18 March 2011 (UTC)

Checklist
GA review – see WP:WIAGA for criteria
 * 1) Is it reasonably well written?
 * A. Prose quality:
 * B. MoS compliance for lead, layout, words to watch, fiction, and lists:
 * 1) Is it factually accurate and verifiable?
 * A. References to sources:
 * B. Citation of reliable sources where necessary:
 * C. No original research:
 * 1) Is it broad in its coverage?
 * A. Major aspects:
 * B. Focused:
 * 1) Is it neutral?
 * Fair representation without bias:
 * 1) Is it stable?
 * No edit wars, etc:
 * 1) Does it contain images to illustrate the topic?
 * A. Images are copyright tagged, and non-free images have fair use rationales:
 * B. Images are provided where possible and appropriate, with suitable captions:
 * 1) Overall:
 * Pass or Fail:

Initial comments
These are mostly content questions that occurred to me while I was reading:
 * 1) What's the relationship to superficial venous thromboses?  Are DVTs even possible in the arms?  (Aren't all the veins in the arm more or less superficial?)  Is Paget-Schroetter syndrome the only form of VTs in arms?  Is it properly thrombophilia?  (It sounds more 'mechanical' than 'problem with consistency of the blood' to me.)
 * 2) Under ==Congenital==:  Are the "type II defects" part of a general classification of congenital defects, or specific to this?  Are all rare forms more severe?
 * 3) Under ==Acquired==:   I don't understand HIT.  You take away all the platelets, and then you somehow have blood clots?
 * 4) Under ==Treatment==:  I thought that ==Acquired== named four kinds that require "particular" treatment, but the first sentence says there is no specific treatment.  The phrase "Apart from the abovementioned forms of thrombophilia..." isn't working for me.  Which forms?  All of them?
 * 5) Under ==Prognosis==:  Is it worth adding something about the prognosis from a thrombotic event, e.g., if a clinically significant clot forms, what's the short-term mortality rate?  I suppose that someone who had this condition would like to know about any overall effects on lifespan.  (Excellent job contextualizing the risks.)
 * 6) Should we address the risk of overdiagnosis (e.g., finding tiny PEs on CT scans for other purposes)?
 * 7) If you've had a DVT/something similar, how likely is it that you have thrombophilia?

These are largely article-writing issues:
 * 1) I think the lead needs to more firmly differentiate between 'having a blood clot, regardless of the reason' and 'having a blood clot because of a problem with the consistency of the blood'.  Can we say something like "The thrombophilias are a group of medical conditions that cause some, but not all, blood clots in veins"?
 * 2) Under ==Signs and symptoms==:  Pregnancy stuff might be more 'Risk factors' or Epidemiology than 'Signs and symptoms'.  Isn't the stuff about Protein C deficiency really a congenital cause rather than a symptom?
 * 3) Under ==Mechanism==:  Probably each of the proteins should be wikilinked once in this section.

I haven't read any of the sources yet; I've only read the article. I don't expect to be able to get back to this in a serious way until at least Saturday, but ping my talk page if you want a quick response or need clarification. WhatamIdoing (talk) 04:56, 18 March 2011 (UTC)

Responses to content questions:
 * 1) The sources said very little about superficial thromboses, and I am not sure if there is much data on a link with thrombophilia. DVTs of the arms are certainly possible (they are called UEDVT or "upper extremity DVT"), but the actual clot tends to be in the subclavian or axillary vein. Whether to use the term "Paget van Schroetter syndrome" for all UEDVT is a matter of nomenclature - when I corresponded with Frits Rosendaal about the subject some time back, he used the terms interchangeably.
 * 2) I'm wondering if I'm overusing the Crowther/Kelton classification, which doesn't seem to have found wide acceptance. They don't seem to classify all forms of thrombophilia into their system. The classification is specific to thrombophilia. Not all rare thrombophilias are severe, as shown later.
 * 3) HIT is fascinating. I wrote the article a few years ago. The number of platelets is diminished (although the number that remains is usually at a level where effective coagulation would be sustained), but the ones that remain are seriously pissed off (they are activated by the antibody reaction), leading to a profound prothrombotic state.
 * 4) This depends on whether you regard PNH, HIT, IBD and nephrotic syndrome as "thrombophilias" in the narrow or the wide sense. The "Kumar4" source and many other sources bunch them together with the "pure" thrombophilias that do not have other repercussions. I'll continue to think of a way to phrase this in a brain-friendly way.
 * 5) I was hoping to bypass the discussion of the prognosis of individual events, because this requires a lot of discussion. For instance, small peripheral pulmonary emboli have a great prognosis, while large pulmonary emboli can cause life-threatening haemodynamic instability; this would require discussion of all the pathophysiological mechanisms. The topic is potentially endless (e.g. the prognosis of other forms of thrombosis, the risk of pulmonary hypertension due to PE).
 * 6) This is again something I would prefer to leave out, as it is more pertinent to the individual events. I do agree that the subject of overdiagnosis requires attention in the pulmonary embolism article (especially in pregnancy, as clearly discussed in 10.1016/S0140-6736(09)60996-X
 * 7) I think we answer this question in a roundabout way. There are probably still undiscovered forms of thrombophilia, so all we can do is list the frequency at which the currently known forms are detected.

Article-writing issues:
 * 1) In many cases, thrombophilia is not the cause but one of a number of contributory factors. If an overweight smoking man develops a DVT after a 4h flight, and is subsequently found to have the factor V Leiden mutation, was the thrombophilia the actual cause? Only in a small proportion of thromboses, a subsequently discovered thrombophilia can legitimately be regarded as the cause.
 * 2) The "pregnancy stuff" was added to "signs and symptoms" because thrombophilic states cause these complications. Similarly, purpura fulminans is a symptom of protein C deficiency.
 * 3) I was being cautious with overlinking, but I will add them back for completeness.

Thanks for your comments so far, and I am looking forward to the rest. JFW &#124; T@lk  01:00, 20 March 2011 (UTC)


 * In case someone is reading this later, I have read and agreed with all of the above responses. WhatamIdoing (talk) 21:26, 26 March 2011 (UTC)

Ref notes
Here are my current notes from reading the refs:


 * Lead
 * In Ref 2, a quick sum of the numbers in Table 8 suggests a prevalence as high as 11%. I don't know if we'd want to rely on that.
 * Signs and symptoms
 * Ref 2 also lists 'sudden death' as a sign
 * Unable to find "Whether thrombophilia also increases the risk of arterial thrombosis (which is the underlying cause of heart attacks and strokes) is less well established" in Ref 2 (did I overlook it?). Ref 24 (Middeldorp in BJH) does cover it, though.
 * Ref 8 technically unnecessary in this section, as Ref 2 includes "recurrent fetal loss"
 * Causes
 * I'm not sure that "Various explanations exist for this link" is the best statement here. Perhaps something about multiple mechanisms being involved might be more faithful to the sources?
 * History
 * The Hughes paper isn't an ideal source for the fact that the condition is named after Hughes.
 * In general, the citations to the original papers seem unnecessary to me. (Keep them if you want; it doesn't matter to me.)

There's a dead link in Ref 4 (both the URL and the doi; it looks like they're rearranging their website).

I haven't yet read Refs 9–13. Of the others, I don't have:
 * Ref 1 (Robbins textbook)
 * Ref 3 (Kyrle in The Lancet, )
 * Ref 21 (Tchaikovski in Thromb Res, )

Off hand, I'd say that there are no significant problems. I hope to get through the remaining five sources within the next 24–48 hours. (Please ping my talk page again if I don't get the notes posted!) WhatamIdoing (talk) 18:52, 25 March 2011 (UTC)

Here's the last few notes in the verification pass:
 * Signs and symptoms
 * "Whether thrombophilia also increases the risk of arterial thrombosis (which is the underlying cause of heart attacks and strokes) is less well established"—Ref 9 (practice guideline, under "Coronary, cerebral and peripheral arterial thrombosis") seems to say that the association has been established, but is just too small (relative to other risk factors) to change the treatment.
 * Causes
 * "...and familial dysfibrinogenemia (an abnormal fibrinogen)." I didn't find this mentioned in Ref 11 (Rosendaal in ASH)
 * Refs
 * For ref 10 (in Annals), you might want to link to the (free) lay summary at http://www.annals.org/content/138/2/I-39.full

As you've indicated that you want to have the article in American English, I will make it that way in a few minutes. WhatamIdoing (talk) 20:40, 26 March 2011 (UTC)

Response to ref notes
I have bunched responses per section, mainly for my own convenience.

Lead:
 * I have rephrased the sentence in question because I was not comfortable with a non-worldwide figure that I could not trace to Heit's paper on re-reading.

Signs and symptoms:
 * I would imagine "cardiac arrest" and "sudden death" are technically synonymous. Even so, completely unheralded death is unusual; often, there have been symptoms of dyspnoea etc
 * The mention of arterial thrombosis is in Table 3 of Heit's article; agree that Middeldorp and also de Moerloose address the question. Reading Middeldorp's conclusions, I'm wondering how Baglin et al can be so certain in view of the methodological issues
 * Heit gives fairly little detail on the miscarriage issue, on which Rai/Regan is an authoritative subject (and the same source is needed elsewhere anyway)

Causes:
 * Rephrased the line on the various explanations. In the end, one of the several mechanisms will probably turn out to be the predominant link.
 * Dysfibrinogenaemia is in table 1 in Rosendaal2005
 * Have added the lay summary

History:
 * Will look for better source wrt Hughes syndrome
 * I like the old citations from a historian's perspective - I don't think there are specific MEDMOS guidelines about them

Thanks for all the comments so far! JFW &#124; T@lk  22:05, 26 March 2011 (UTC)


 * I'm not convinced that Hughes syndrome really requires a citation for GA standards. It's not exactly a controversial sort of statement.  If you're absolutely determined to have one, a relatively weak source like the WhoNamedIt entry would be more than adequate.
 * The high prevalence 'hidden' tables in some journals' online formats are the bane of citation reviewing. Thanks for pointing out what I missed.  WhatamIdoing (talk) 23:00, 26 March 2011 (UTC)

Would you agree that the following reference is good enough to support the use of "Hughes syndrome"? (Yes, D'Cruz and Cuadrado work at the same department as Hughes...) JFW &#124;  T@lk  19:13, 28 March 2011 (UTC)
 * I agree. That journal/series is highly reputable, and "Hughes syndrome" does not seem to be a controversial eponym.  -- Scray (talk) 21:32, 28 March 2011 (UTC)
 * Oh, sure. As I said above, I don't think this claim is either one of the  four things absolutely required to have an inline citation according to the content policies or one of the five things required for GA, so an inline citation is IMO optional, strictly speaking.  However, that one will do nicely, if you want to add it.
 * I expect to close this (as passed) tomorrow morning (~20 hours from now), assuming my last-minute checks don't turn up anything surprising. Scray, if you've got any objections, now's your chance to squawk.  WhatamIdoing (talk) 21:52, 28 March 2011 (UTC)
 * No objections. Nice work, all!  -- Scray (talk) 23:22, 28 March 2011 (UTC)

Kumar/Robbins
I'm comparing chapter 4 of the Robbins textbook (ref 1, "Kumar"), and seeing some differences from the article that seem important. I have already made some changes that were straightforward, but am getting into a bigger snag with the descriptions of HIT and PNH (described as rare, under acquired causes). Here is what Kumar says: "Among the acquired causes of thrombotic diathesis, the heparin-induced thrombocytopenia (HIT) syndrome and antiphospholipid antibody syndrome (previously called the lupus anticoagulant syndrome) deserve special mention.
 * Seen in as many as 5% of the population, the HIT syndrome occurs when administration of unfractionated heparin (for therapeutic anticoagulation) induces autoantibodies to complexes of heparin and a platelet membrane protein (platelet factor 4) (Chapter 12). This antibody binds to similar complexes present on platelet and endothelial surfaces, resulting in platelet activation and endothelial cell injury, and a net prothrombotic state. The occurrence of HIT syndrome can be reduced by using low-molecular-weight heparin preparations that retain anticoagulant activity but do not interact with platelets; these preparations have the additional advantage of a prolonged serum half-life.
 * Antiphospholipid antibody syndrome has protean manifestations, including recurrent thrombosis, repeated miscarriages, cardiac valve vegetations, and thrombocytopenia; it is associated with autoantibodies directed against anionic phospholipids (e.g., cardiolipin) or—more accurately—plasma protein antigens that are unveiled by binding to such phospholipids (e.g., prothrombin). In vivo these antibodies induce a hypercoagulable state, by inducing direct platelet activation or by interfering with endothelial cell production of PGI2. However, in vitro (in the absence of platelets and endothelium) the antibodies merely interfere with phospholipid complex assembly and thus inhibit coagulation (hence the designation lupus anticoagulant). Patients with antibodies to cardiolipins also have a false-positive serologic test for syphilis, because the antigen in the standard tests is embedded in cardiolipin."

PNH is certainly rare, whereas quite a large proportion of the population (5%) may be at risk for HIT. Does our article reflect this? Still working through the rest of our article's cites of this source. -- Scray (talk) 02:02, 27 March 2011 (UTC)
 * I have reviewed all of the cites of Robbins/Kumar, and the second paragraph of Thrombophilia is my only lingering concern. That paragraph is a mashup of very disparate conditions - from HIT (an iatrogenic complication of heparin therapy, fairly common and relatively easily recognized) to PNH (much less common and more difficult to diagnose, inciting cause not easily identified) and red cell dyscrasias.  I think it needs to be reworked - at least to provide proper context for HIT and PNH, and a conceptual framework for combining them in this way, but I'm a late arrival here and am loath to change too much at this stage unless there's a clear consensus.  -- Scray (talk) 02:16, 27 March 2011 (UTC)

This is the problem with using textbooks. I have no access to Kumar4 and personally have a preference for Harrison's. I'm not sure which editor added the content that was based on this source. Most of it was verifiable, and you will see that most assertions from Kumar4 are dually sourced to other sources.

I think our current mention of antiphospholipid syndrome is appropriately balanced and thoroughly sourced to the Lancet review. I don't think it needs much tweaking. With regards to HIT, I think that Kumar4 vastly overstates the risk. Keeling (and everything Theodore Warkentin has ever written) gives much lower figures, and of course it requires exposure to heparin.

I understand your concerns about the "acquired" paragraph. With regards to its inclusion criteria, I have tried to remain faithful to lists provided in Heit and Kumar (and Rosendaal2005, which contains a similar list). The only medical condition not traceable to that list is inflammatory bowel disease, which is not common but widely regarded as a prothrombotic state. If you have any idea on how to get this better, I am very much open to suggestions. JFW &#124; T@lk  08:43, 27 March 2011 (UTC)
 * I share your preference for Harrison's, and have not been a student of Robbins. Overall, I agree that the article (Thrombophilia) is in great shape, and most statements sourced to Kumar were fine.
 * In your second paragraph above, you state that, "I think that Kumar4 vastly overstates the risk" of HIT. For this article, we just need to settle on statements and sources.  It certainly varies by population and drug (sulfation of the UFH, in particular), but a recent review  published in a high-quality journal summarized the literature in this way, "This 10-year retrospective summarizes how the understanding of HIT has changed over the past decade. In contrast to previous opinions, restated above, it is now evident that the risk for HIT-associated thrombosis is substantial, even >50% in certain patient populations (2)."undefined The review goes on to say (in a section on "Frequency"):
 * "Warkentin & Kelton reviewed 13 prospective studies of HIT (12), accepting a case as a probable episode of HIT only if thrombocytopenia began on day 5 or later of heparin treatment. Based on this criterion, approximately 3% of 1336 patients who received UFH in these studies may have had HIT. Even more important, 14 of these 1336 patients (1%) developed thrombosis in association with thrombocytopenia, i.e. HIT-associated thrombosis occurred in approximately 1 in 100 patients who received heparin for 5 days or more. Such a high frequency of clinically significant adverse events for a medication as commonly prescribed as heparin indicates that HIT is one of the most important adverse drug reactions physicians will encounter. Indeed, in one prospective study, HIT-associated thrombosis occurred in about 1 in every 40 patients who received UFH (2)."


 * "Differences in patient populations may influence the frequency of HIT antibody formation. Visentin and coworkers (31) found that 52% of patients receiving UFH for heart surgery developed IgG anti-heparin/PF4 antibodies. This high frequency could reflect wide fluctuations in PF4 release into the circulation of these patients, thus leading to the optimal ratio of heparin/PF4 that can trigger an immune response. In contrast, medical cardiac patients receiving UFH had a much lower frequency (2.5%) of anti-heparin/PF4 antibody formation in one study (32)."


 * undefined
 * Clearly, HIT is a high-frequency complication of heparin treatment. All parties with an interest in thrombosis and anticoagulation should be cognizant.  -- Scray (talk) 18:08, 27 March 2011 (UTC)
 * I've refactored paragraph 2 under acquired into two separate paragraphs; I hope others agree that it's an improvement. I've avoided direct reference to frequency, pending consensus here.  -- Scray (talk) 18:20, 27 March 2011 (UTC)
 * At this point, I am satisfied that the article accurately reflects cited sources. -- Scray (talk) 18:29, 27 March 2011 (UTC)

Thanks for your comments and for your attention to the HIT/PNH paragraph. It does flow much better now and inter alia discusses the paradox that WhatamIdoing alluded to with regards to the nomenclature of HIT. Just to clarify: I do not dispute the importance of HIT and totally agree that it should be given the prominence we are giving it currently. At the same time, in absolute terms HIT-related thrombosis is rare compared to the other mechanisms. (COI I'm the main author of heparin-induced thrombocytopenia.) JFW &#124; T@lk  20:09, 27 March 2011 (UTC)

Passed
I have decided that the article in its current form meets the Good article criteria. Congratulations to JFW and all the other editors who helped the article along the way. WhatamIdoing (talk) 16:43, 29 March 2011 (UTC)