Philip Musgrove rebuts Attaran

Philip Musgrove, one of the authors of the paper that Amir Attaran misunderstood writes:

The one--and only--point on which Attaran and his co-authors are right
is that in the Appendix to the Booster Program document, someone at the
World Bank who had not participated in the analysis of the Brazilian
program failed to notice that the data for 1996 referred to only half a
year. This is because the paper by Akhavan et al., of which I am a
co-author, developed a cost-effectiveness analysis and the cost data for
the Bank project ran only through half of that year. It was a mistake
to interpret the 1996 data as if they referred to a whole year, and the
Bank has since corrected that error. It was not, as Attaran et al.
assert, a deliberate attempt to mislead anyone by claiming that malaria
cases fell by 60% in one year. It is perfectly clear in Akhavan et al.
that no such claim was ever made; it is inexcusable to cite our article
so as to give the impression that it was made. Authors who were intent
on learning what actually happened, rather than looking for an excuse to
attack the World Bank, would have pointed out the mistake to the Bank
and waited for the necessary correction, rather than falsely accusing
the Bank--and by implication, Akhavan and his co-authors--of chicanery
and bad faith.

The more important point on which Attaran and co-authors are completely
wrong, and on which they would have been wrong even if the analysis had
stopped at the end of 1995, is their insistence that "[a] declining
trend in malaria cases during and after the Bank's program...ia
necessary to adjudge it a success." That would be appropriate, if
malaria incidence had been stable in the years before the program, but
it wasn't; as Figure 1 in Akhavan et al. shows, the number of cases had
been increasing every year from 1975 to 1988. The program is rightly
judged a success because it reversed that trend, overall, and more
dramatically in the case of falciparum malaria. The estimates of 1.73
million cases prevented and almost 231,000 deaths averted are the
difference between what actually happened and what we take as the
counterfactual, what would have happened in the absence of the program
and particularly without the change of strategy in 1992-93. Our
assumptions about that counterfactual--about what would have happened to
incidence, severity and lethality--are carefully spelled out in our
paper, and had the support of Brazilian and other malaria experts. I
can only conclude that Attaran et al. do not understand the concept of a
counter-factual and have failed to notice what was happening to malaria
cases in the whole decade preceding the program. People who do not
understand such basic matters, and who do not trouble to read carefully
before launching into an attack, should refrain from criticizing those
who do understand them.

Philip Musgrove
Ex-World Bank (retired)
Deputy Editor, Health Affairs

Mr. Lambert, please feel free to add this note to your weblog. This same
explanation has been provided, at somewhat greater length, for
publication in the Lancet. I and my co-authors hope that will put an
end to this discussion. It remains only for Attaran and his co-authors
to apologize for accusing the Bank of deliberate falsification.