David Crowe Asks HIV Testing Expert Dr. Margaret Fearon to Supply a Citation for a Window Period Claim

In a 2005 paper:

Fearon M. The laboratory diagnosis of HIV infections. Can J Infect Dis Med Microbiol. 2005 Jan; 16(1): 26-30.

you stated that “HIV EIAs have become increasingly more sensitive and specific since HIV testing began in the early 1980s. This has shortened the ‘window period’, or the time from exposure to seroconversion, from up to 12 weeks or more in the early days of diagnostic testing to the current ‘window period’ of less than three weeks in most cases.”

Unfortunately you did not provide a citation for this statement. Could you do that now, provide a citation that supports this statement? I am interested in gathering scientific data on what the window period appears to be, and how that was determined.

Regards,

David Crowe

if you do a MedLine search you can find lots of references concerning serological window period for HIV as well as the detection time using an RNA assay, which is generally shorter – about 9-11 days. It is important to only review references from the last few years as the older assays were not as sensitive as the current ones. Seroconversion panels are used to determine an assays’ ability to detect early seroconversion. The Ontario Public Health lab has data on the performance of assays it has evaluated using such panels. These days most PHLs will also run PCR (looking for viral RNA) on an individual if the clinician has indicated recent risk and the antibody test is negative.

Dr. Margaret Fearon
Executive Medical Director
Medical Microbiology
Canadian Blood Services

as I no longer work at the Ontario Ministry of Health lab I think it is better if you direct your questions to the HIV lab there. They will be more up on the current literature in the HIV diagnostic field.

Dr. Margaret Fearon

Executive Medical Director, Medical Microbiology

Canadian Blood Services

Thank you very much for your response.

You did not answer my basic question though, what was the reference for your 2005 assertion about window periods? Note that 2005 is not a long time ago.

Also, although I could easily do a medline search I was assuming that you would be an expert in this area and would know offhand of at least one major paper on the window period issue.

My basic question is this: The moment of infection is unknown, all that is known is the first time that a person is antibody-reactive or PCR-reactive. How can you estimate or average a time period when the start of the time period is by definition unknowable?

Also, in some documentation that I saw a while ago, I saw a more detailed flowchart for HIV testing in Ontario. Unfortunately I was not in a position to get a copy of it. Do you know if this flowchart is available, and could I have a copy of it?

Regards,

David Crowe

Thank you for your email.

I understand that you are no longer with the Ontario Ministry of Health, perhaps you can let me know who your replacement was and I will talk to them.

However, you were still the first author of the 2005 paper, and I believe that you still have a responsibility to provide a citation for your claim about the HIV window period … or publicly retract that portion of your paper if it has no basis in fact.

Regards

David Crowe

I already told you that the statement can be backed up by the HIV lab data. The published references I used are listed at the end of the article. It is up to the editor of the journal to accept or reject those and publish an article accordingly. There are many additional references on HIV window period and I am sorry but I do not have time to do the MedLine search for you. If you phone the Ontario Ministry of Health, Laboratories Branch on 81 Resources Rd. they can put you in touch with the appropriate person.

Dr. Margaret Fearon

Executive Medical Director, Medical Microbiology

Canadian Blood Services

You wrote a specific statement in a 2005 paper about the window period. It is your responsibility to provide a reference for that statement, not mine. And if you can’t, then it is your responsibility to write to the journal and withdraw that portion of your paper.
Your claims that “the statement can be backed up by the HIV lab data” is mere hot air if you can’t provide the backup.

There are far too many things that are accepted today because all the experts know that it’s true, and when they’re challenged on it they assert authority and refuse to answer the question.

I sense that our communication is coming to a close, but I wanted to let you know that I will be publishing this exchange on a website. So, if you do have the references, it would certainly look better if you provided them.

I think it is important for the general public to know that the emperor has no clothes (metaphorically speaking).

Regards,

David Crowe

I have attempted to direct you to how to find the additional references you seek, short of actually doing the MedLine search for you. I do not understand your confrontational attitude to be honest. I have tried to be as helpful as I can.

Let me suggest you read any good textbook on infectious diseases such as Principles and Practice of Infectious Diseases, Mandell, Douglas and Bennett, 5th ed. (this would have been around at the time of the publication of the paper) pub. Churchill Livingstone Chapter 106 Detection of Human Immunodeficiency Virus Infection. If you read the section entitled Tests for the Detection of Human Immunodeficiency Virus Infection: Serologic Assays the first sentence reads ‘Antibodies to HIV do not develop until after the initial decline in HIV viremia and can first be detected within 2 to 8 weeks after infection.’

Our experience at the HIV lab was that the majority ( but not all) seroconversions were detectable within the the first few weeks. Because not all seroconversions can be detected this early, for various reasons, the lab always recommends follow up testing and nucleic acid testing when a physician suspects recent infection but the antibody test is negative. If the physician directly requests NAT because of suspicion of recent infection, the lab will do this.

I hope this is helpful.

Dr. Margaret Fearon
Executive Medical Director, Medical Microbiology
Canadian Blood Services

I fear we are going round in circles. My basic request is for a reference for a claim you made in a 2005 paper (”HIV EIAs have become increasingly more sensitive and specific since HIV testing began in the early 1980s. This has shortened the ‘window period’, or the time from exposure to seroconversion, from up to 12 weeks or more in the early days of diagnostic testing to the current ‘window period’ of less than three weeks in most cases.”). I think it is your responsibility to provide at least one citation or, if there is no citation, to publicly withdraw the claim as it is now in the published scientific literature.

In your latest email you tell me that your “Our experience at the HIV lab was that the majority ( but not all) seroconversions were detectable within the the first few weeks”. This raises the question of how you knew for sure that infection occurred on a certain date. For example, did you have recent negative HIV test results for all these people. And given the existence of a window period, even if you had a non-reactive antibody test a week before the high risk contact that is not proof that the person is HIV-uninfected (according to the theory that HIV antibodies only develop weeks after contact).

Let me illustrate why it is not satisfactory for you to tell me to do me my own research and to stop worrying about the lack of a citation. Let’s say I was to find a citation on the HIV window period and it was to contradict your claim about the length of the period. I could then claim that your paper was wrong, but I might not be talking about the same citation as you used, a problem that could only be resolved if I could compare my citation with your currently hidden citation.

There are other fundamental problems too. Let’s say that the citation you used was later found to be erroneous or, worse, fraudulent. Since you did not cite this paper there will be no way for the publishers of the withdrawn or modified paper to recognize that your paper also needs to be modified or withdrawn.

Let me quote from a relatively authoritative work in the area of scientific conduct, “On being a scientist: Responsible conduct in research” published by National Academy Press in 1995: “Citations serve many purposes in a scientific paper. They acknowledge the work of other scientists, direct the reader toward additional sources of information, acknowledge conflicts with other results, and provide support for the views expressed in the paper. More broadly, citations place a paper within its scientific context, relating it to the present state of scientific knowledge.”

You are also doing a disservice to readers who will hit a blind alley when they read your claim and cannot discover the data that is behind it. This means that they would not only not have the benefit of the research if it is still deemed reliable, but they would not be informed of any questions about the research such as through links to comments on the research listed in PubMed or elsewhere, or other important information on the status of the research.

I now formulate my own hypothesis: You do not actually have a reference to the scientific literature on the window period of HIV tests, and that your claim about the HIV window period in your 2005 paper (Fearon M. The laboratory diagnosis of HIV infections. Can J Infect Dis Med Microbiol. 2005 Jan; 16(1): 26–30.) is therefore not based on published scientific research.

It will be very easy for you to disprove my hypothesis, simply just produce the citation for your claim that “HIV EIAs have become increasingly more sensitive and specific since HIV testing began in the early 1980s. This has shortened the ‘window period’, or the time from exposure to seroconversion, from up to 12 weeks or more in the early days of diagnostic testing to the current ‘window period’ of less than three weeks in most cases.”

Alternatively, you can admit that my hypothesis is true and write to the journal indicating that this part of your publication is not supported by scientific evidence.

Regards,

David Crowe