Find sources: Google (books · news · scholar · free images · WP refs· FENS · JSTOR · TWL

Inclusion of FibroTest equation

edit

Biopredictive, who has a self-acknowledged conflict of interest as an employee of the company that produces FibroTest, has asked me to remove the formula for the test, citing a variety of concerns. Quoted below is this exchange, in full. I copy it here because it's so relevant to the article. --Scray (talk) 11:43, 25 September 2009 (UTC)Reply

Hi,

I would like to thank you for your work on FibroTest article. I agree with all modifications you (and others) made on the article, but one.

You did add the FT formula you found from the patent in the article, but :

  • this is an issue for the owner of the patent (public APHP)
  • most importantly, usage of the formula is dangerous without its securities. This may lead to majors problems with up to 5% of users, with both false positives and false negatives. Diagnosis tests are not a matter of "just a formula".
  • the commercial strategy of the company licensed by Assistance Publique des Hopitaux de Paris (like any other) is the only one to build robust tests, validated with its initial population (HCV patients for FibroTest), but also for other populations that deserves a better diagnosis (HCV with all genotypes, HBV, HIV co-infected, NAFLD, Alcohol, children, 65+, all ethnies, etc.).

What I suggest is to keep the link to the US Patent website, explain the idea of the formula (a regression including all parameters) for the educational material (as suggested in your comment), but not show the formula itself.

I'm willing to talk about it if you want - I did not make these changes to the article myself before talking of it with you.

Thanx for your time and implication on Wikipedia.

Biopredictive (talk) 10:17, 25 September 2009 (UTC)Reply

Thank you for your thoughtful and civil note. I added the formula because it illustrates the concept of the test better and in a more concrete way than a text description would, and helps the curious see the relative weights that each part receives. I found it very interesting when I found it. I don't think the safety concerns are substantial, and they certainly do not outweigh the information value. Regarding the other concerns from the company's point of view regarding their patent, it's clear that no one should use this formula commercially - and I did specifically make it clear that it's the subject of a patent. The mandate here is to inform, as illustrated by the mantra, "Wikipedia is not censored". If you read the text behind that link, I think you'll see what I mean. I did not provide the formula saying that it should be used, and agree with retaining the text that describes the services, precautions, and safety measures that come with the official report. Above all, the formula is the subject of the article, and I feel it's relevant. I am going to copy this reply to the Talk:FibroTest page, so that others have a chance to weigh in. --Scray (talk) 11:43, 25 September 2009 (UTC)Reply

Comments welcome. --Scray (talk) 11:43, 25 September 2009 (UTC)Reply

I completely agree with Scray. We should include the equation as it is an important part of understand the function of FibroTest - it's not magic, after all. Wikipedia should never be used for medical advice or diagnoses. Fences&Windows 20:38, 25 September 2009 (UTC)Reply


Thank you for taking time to discuss about this. Would you please consider carefully these arguments :
  • We did read the article about censorship, and we don't feel it reflects our behavior,
  • You don't think the safety concerns are substantial but liver fibrosis is a widespread disease : we are talking about millions of people. Some of them may lead to cirrhosis or liver cancer. Having a possibility of false diagnosis (may it be false positive or false negative) is a major issue on such a large population.
  • You did ask us to read several articles related to Wikipedia rules. We did and we applied these rules. We would just like you to consider more carefully the article "Fibrosis biomarkers are not only a formula" (ref http://www.clinchem.org/cgi/eletters/51/10/1867 ). You will read that medical experts are having a consensus for best practice in this area : releasing such a formula is worse than not (think in terms of risk vs benefits - like any medical material).
  • Your motivation is to expose the principle of such an algorithm: let's explain it's a non-linear regression (linked to its wikipedia article), and let's explain weight of each parameter (most proeminent parameter is A2M : as A2M will grow, FibroTest will raise ; as bilirubin will down, FibroTest will grow, male have higher fibrosis than female, age raising the fibrotest will raise, etc. ). You will notice that some parameters have log, some have not : this likely will not allow a lot of readers to achieve the educational goal...
  • Thinking that people will read carefully the recommendations of use for such a test is theory. In real world, formula might be used without any precautions (commercially or not) and lead to false diagnostics. This is a major issue : we talk about this with experience. We can not afford such a responsibility.
  • Keeping the link to US Patent website will allow people willing to read the formula to find it easily (as you did), but will keep it from being spread insecurely as it is right now

We would like you to reconsider your position, and would prefer the article to be modified this way. I did not modify the article myself though.

Thanx for your time,

Biopredictive (talk) 11:24, 26 September 2009 (UTC)Reply

These arguments could be applied to much of the information on Wikipedia. Sophisticated information is included in Wikipedia, including complex mathematics and details of much more dangerous things like nuclear weapon design, production of arsenic, and suicide methods, to support the full range of interested readers. Let's see whether others weigh in, but I still think inclusion of the calculation that is at the heart of this test is relevant and informative. --Scray (talk) 18:59, 26 September 2009 (UTC)Reply
It's also worth noting that the "article" you've repeatedly cited is neither an article nor is it peer-reviewed. It is a letter, and the first author has an acknowledged conflict of interest: "Thierry Poynard is the inventor and has a capital interest in Biopredictive, the company marketing FibroTest, ActiTest, SteatoTest, NashTest and AshTest." --Scray (talk) 01:39, 27 September 2009 (UTC)Reply
This argument doesn't convince me. If you didn't want the formula known and potentially misused, you shouldn't have patented it. If you're worried about your intellectual property being infringed by others using the results of blood tests along with your algorithm, that's not Wikipedia's concern. Fences&Windows 00:04, 27 September 2009 (UTC)Reply
We heard your arguments of educational purposes. But our tests are medical devices developed in compliance with medical agencies requirements. Part of them is the heart of our motivation in this discussion : the security algorithms dedicated to detect false negative or positive profiles which are full part of the whole diagnosis service. Therefore and according to our initial discussions, we are adding a clear disclaimer to the article, so that responsabilities are clearly identified. Have a nice day and thanx for the time you give to Wikipedia Biopredictive (talk) 09:33, 2 October 2009 (UTC)Reply
I've toned down the "disclaimer"; the additional information is useful. Wikipedia informs, we do not instruct. Fences&Windows 01:24, 3 October 2009 (UTC)Reply
  • Came here from the note on WP:GI. I agree with what Scray and Fences and windows have listed here above. Specifically, I think that the equation listed and referenced to the patent is useful information for the article and should stay. The concerns regarding infringement of IP are not valid, as the equation is listed in the patent, and is referenced as such in the article -- Samir 05:54, 15 October 2009 (UTC)Reply

Ambiguity: 6 serum biomarkers stated, 5 appear in equation

edit

I can't tell from reading this article why there is a mismatch. Could someone please clarify this in the article. — Preceding unsigned comment added by 130.238.224.66 (talk) 11:34, 21 January 2021 (UTC)Reply