Comments on Harvey Bialy: GIGO - or Why HIV Antibody Tests in South Africa Are WorthlessTypePad2007-02-26T09:27:01ZOtishttps://barnesworld.blogs.com/barnes_world/tag:typepad.com,2003:https://barnesworld.blogs.com/barnes_world/2007/02/harvey_bialy_gi_1/comments/atom.xml/Anthony I. Okoh commented on 'Harvey Bialy: GIGO - or Why HIV Antibody Tests in South Africa Are Worthless'tag:typepad.com,2003:6a00d83451609269e200e550635d0f88332007-02-27T15:13:08Z2008-02-22T08:52:21ZAnthony I. OkohLet me first thank you for making this compilation available as you have. It provoked an almost all day discussion...<p>Let me first thank you for making this compilation available as you have. It provoked an almost all day discussion here at the Univ. of Fort Hare among myself and a number of my colleagues and our students.</p>
<p>We all agree that the title, however, is a bit disingenuous, as the text, which is an insider, historical introduction to the larger question of "AIDS and Africa" and its political ramifications, belies the extravagant headline in so far as it presents data that call into the most serious question only one antibody test, which may not even be used any longer at MEDUNSA.</p>
<p>We do understand that a title that read, "GIGO - or Why HIV Antibody Tests in South Africa <i>May</i> Be Worthless" is not nearly as catchy, but it started us to thinking of a larger study that I will do my best to implement in the next months (not years).</p>
<p>It seems fairly certain that the reason the test that failed so miserably did so was because it was not calibrated or tested for performance against typical, impoverised African blood, which is biochemically quite different from the blood likely to be drawn from an average European.</p>
<p>It is disturbing to say the least that no one has ever systematically addressed this question, which is so obviously important.</p>
<p>What we therefore are proposing to do is gather all of the 7 or 10 antibody test kits that are currently in use in SA, and test them, as you did, for internal quality control against blood from (a) reasonably well nourished and healthy folks like us, and (b) blood drawn on different hospital wards, and in the villages where poverty, lack of clean water and crowded living conditions are perennial problems.</p>
<p>We are not interested in "HIV" status per se, but rather in seeing how reliably the tests perform on these different samples.</p>
<p>We will of course, since we are not racially biased, include as many of our colleagues and countrymen of European extraction as we can, realizing that not too many will be found in the infectious disease hospital wards and in the villages. </p>
<p>If we do find the kinds of horrendous problems that you did in the Pilot Study apply to most or all of the tests, we will then proceed in a systematic way, as you described, to see if we can improve any of them to the point of biochemical reproducibility.</p>
<p>But until such time, it is perfectly clear from your own results, that any African given a positive antibody diagnosis should demand that the test used to give him or her a death sentence is as internally reliable as the one used to tell them their blood sugar levels.</p>
<p>Professor Anthony I. Okoh<br />
Microbiology Unit<br />
University of Fort Hare <br />
Alice, South Africa</p>