Riley and colleagues present an interesting question. Should a "standard of care" that has emerged from 30 years of heated professional debate, randomized treatment trials, retrospective case studies, and thoughtful consensus by experienced, knowledgeable physicians be sacrificed because of the lack of knowledge and possible poor judgment of a single provider? The obvious answer is no. If such decisions were made on the basis of such anecdotal incidents, no drug would be acceptable.

As noted, drug reactions related to rifampin were first reported in 1971. Familiarity with the older literature, however, is not a prerequisite for awareness of this risk. Warnings about potential drug reactions are included in each package insert for rifampin and are listed on page 1529 (rifadin/rifampin) of the 1996 Physicians' Desk Reference. Experienced providers are alert to these symptoms.

An advantage of directly observed therapy is that the provider sees the patient before administering each dose and is therefore in an ideal position to immediately identify such adverse reactions. Unfortunately, many patients continue to take a drug in the face of an adverse reaction while awaiting their next physician visit. I hope that Dr. Riley would not suggest that, because of such rare instances, no patient should be prescribed drugs.

Before returning to the days of long-term daily isoniazid therapy, Dr. Riley and colleagues might refresh themselves on the additional potential for isoniazid-related hepatitis, on the increased likelihood of patient noncompliance resulting in the emergence of drug-resistant organisms, and, finally, on the subsequent opportunity to address a seemingly unending assortment of adverse reactions that arise from treatment with second-line antituberculous drugs.

The problem of drug-resistant tuberculosis will continue to grow only if physicians fail to accept the direct administration of tuberculosis treatment as the "standard of care" as recommended by the Centers for Disease Control's Advisory Council on the Elimination of Tuberculosis [1] and the World Health Organization [2].

I hope that in "reviewing the literature" for insights, we won't reject the literature of the present. Directly administered tuberculosis treatment is one "standard of care" that should and must be met [3].