We agree with Dr. Stead's comments about the contagiousness of draining lesions, the changing face of tuberculosis in the era of the acquired immunodeficiency syndrome (AIDS) [1], and the risk posed by the likelihood of lymph node involvement that may lead to abscess and spontaneous fistula formation. Patients with AIDS often require prolonged hospitalizations, providing ample opportunity for nosocomial spread of tuberculosis from lesions harboring large numbers of organisms.

Our report [2] provides another important lesson. The diagnosis of tuberculosis was first considered at autopsy, when the skin lesion was examined histologically. Had the autopsy not been requested by the attending physician and agreed to by the family, the subsequent contact investigation would not have been initiated, and the infected hospital personnel would not have been provided with prophylaxis in a timely fashion. That two nurses who did not receive prophylaxis developed clinical tuberculosis suggests that additional cases might have occurred, with the potential for further secondary spread.

I offer the following recommendations to minimize the risk of future nosocomial outbreaks of tuberculosis from infected skin lesions:

1. Evaluation of draining skin lesions of unknown cause should include a smear and culture for acid-fast bacilli and consideration of biopsy.

2. If tuberculosis is a possible cause, a patient with an undiagnosed, draining skin lesion should be appropriately isolated until an alternate diagnosis has been established or until the patient is rendered noninfectious through treatment. Measures to prevent nosocomial spread of tuberculosis have been recently reviewed [3].

3. Wound care measures should be designed to minimize aerosolization of drainage or secretions.

Finally, physicians and the public, in general, should be re-educated regarding the value of postmortem examinations. In this age of resurgent tuberculosis, the autopsy is an important public health tool.