Isoniazid Prophylaxis-Reply

In Reply.—  The study Dr Moulding refers to, from which we derived the risks of isoniazid hepatitis, was the Public Health Service cooperative surveillance study.¹ The investigators found 0.578 deaths per 1000 isoniazid recipients. We calculated the 95% confidence interval to range from 0.2 to 1.0 deaths per 1000 isoniazid recipients. Among other well-documented studies that were sufficiently large to uncover this low risk, the rates vary from 0.1 to 0.9 deaths per 1000 isoniazid recipients.^2-4 Consequently, we used 0.578 deaths per 1000 isoniazid recipients as our best estimate, and used 0.1 and 1.0 deaths per 1000 isoniazid recipients as the low and high estimates in our sensitivity analysis. Even at the high estimate, the decision favors taking isoniazid.Dr Moulding suggests that the tuberculosis case fatality rate may be lower in people compliant enough to have taken isoniazid chemoprophylaxis. To our knowledge, this hypothesis has not been