In his editorial on tuberculosis, Sbarbaro [1] states that directly administered therapy has become the standard of care in the treatment of tuberculosis. He goes on to say that "Physicians deviating from this standard put their patients and society at risk." We report the case of a 61-year-old male farmer who was diagnosed with cavitary tuberculosis in November 1994. He received isoniazid, rifampin, and ethambutol daily for 2 months, followed by isoniazid (900 mg) and rifampin (600 mg twice weekly) under the direct observation of public health authorities. Two months into his intermittent therapy, he began to develop flu-like symptoms during the hours following the ingestion of his medications. The severity of these reactions increased after each dose. He suspected that his drugs were the cause but was persuaded to continue taking them. Ultimately, he presented to our hospital with fever, acute renal failure, hemolysis, thrombocytopenia, ataxia, and paresthesias. Eight days after his last dose of rifampin, these abnormalities resolved.

This patient had complications of discontinuous rifampin therapy described by Poole and colleagues [2] in 1971. When rifampin was administered in doses of 1200 mg twice weekly, almost one fourth of patients developed pyrexia, thrombocytopenia, or renal failure. In a study from Hong Kong [3], 30% of patients had side effects from 1200 mg of rifampin given weekly. Reducing the intermittent dose of rifampin to 600 mg reduces the frequency of but does not eliminate this reaction [4]. Although Dr. Sbarbaro counsels us to "review the literature of yesterday for insights," the persons observing our patient were obviously unaware of the danger of their therapy. The literature of yesterday also tells us that cure of this man's tuberculosis was almost certain if he received daily isoniazid with one companion drug. Let us be cautious in declaring "standards of care."