HIV and AIDS: Questions of Scientific and Journalistic Responsibility

36

HIV AND AIDS

suffered several fatal flaws, such as using a circular HIV-based
definition of AIDS, failing to quantify total drug use over time, and
ignoring drug-use-differences between HIV-positive and -negative
men."
The article "Debunking Doubts That H.I.V. Causes AIDS", by Gina
Kolata in the New York Times (11 March 1993, p. All) followed the
Nature press release in reporting incorrectly "a group of 1034 randomly
selected single men who lived in San Francisco and were 25 to 54 years
old in 1984, when the study began." The New York Times also reported
uncritically the misleading data from the Ascher et al. "Commentary".
Thus did the New York Times propagate the misinformation of the
press release and of the "Commentary".
I take no position here on the relative merits of the AIDS virus
hypothesis or the AIDS drug hypothesis (in whatever form they may
be formulated). I do take a position against the announcement of
purported scientific results via superficial and defective press releases,
and before scientists at large have had a chance to evaluate the
scientific merits of such results and the data on which such results are
purportedly based.
Some other scientists reacted negatively to Nature's publication. For
example, Richard Strohman wrote a letter to the editors of the San
Francisco Chronicle, which had rushed into print about the Nature
article. Strohman's letter was never printed, and I quote it in full:
Letter to the editors of the San Francisco Chronicle by Richard
Strohman (sent 11 March 1993, never printed). I am dismayed by
your treatment of the AIDS-drug hypothesis (4/11/93). As a piece
of reporting it is a masterpiece of scientific ignorance. First, in the
article in question all conclusions, dutifully reported by Mr.
Perlman, were drawn from hearsay. It is hearsay because the
article is not a scientific paper that survived any rigorous review
process; it was instead part of what is called "scientific
correspondence" that gets by with often cursory review by journal
editors. Second, as a result of lack of thorough review there is no
detail given on methods used to collect data. Third, without
details on methods we can not evaluate the data itself, never mind
conclusions drawn from that data. Thus, all standards of real
science are violated. What remains is only "scientific
correspondence", at best a mechanism for developing opinion or