October 1991

The Isoniazid Debate

Author Affiliations

Boston, Mass

Arch Intern Med. 1991;151(10):2105. doi:10.1001/archinte.1991.00400100153032

To the Editor. —  We read with interest Colice's article1 in which he repeats our analysis of isoniazid preventive therapy2 using somewhat different data. We were gratified to find that our model and assumptions were explicated in sufficient detail to allow its independent recreation. We disagree, however, with his analysis because of the rather selective biases in his data.Although Colice correctly used data from young adults3-5 in estimating the isoniazid-induced hepatitis case-fatality rate for a 30-year-old man, he then inappropriately used data primarily relevant to elderly persons in estimating the tuberculosis case-fatality rate. In the two studies that he cited,6-7 that rate was much lower (either zero or nearly zero) among young adults. This inconsistency is particularly troubling because reactivation of tuberculosis is most likely to occur in the years proximate to the discovery of the positive tuberculin reaction; the rate of reactivation declines over

First Page Preview View Large
First page PDF preview
First page PDF preview