Triple combination antiretroviral prophylaxis for needlestick exposure to HIV.

rate of the algorithm of 39% (12 of 31) is somewhat misleading. This rate includes both patients for whom isolation was delayed for >24 hours (n=7) or not implemented (n=5) and does not differentiate by level of infectiousness. Of the 12 patients with pulmonary TB who were not placed in negative-pressure isolation rooms within 24 hours of admission to the MGH, 5 were AFB smear-positive, 6 were smear-negative, and 1 patient had no smear obtained (TB diagnosed on autopsy). If only high-risk patients (ie, smear-positive, more infectious) are prioritized for immediate isolation, then only 5 of 31 patients with pulmonary TB were not isolated appropriately (algorithm failure rate=16%). The timing and duration of isolation of smear-negative patients is more uncertain, given the knowledge that only a small minority of such patients will be found to be culture-positive. The use of more rapid and sensitive diagnostic tests that currently are undergoing evaluation at our hospital and others may assist in the assessment of these patients. Finally, we agree that clinical algorithms are subject to limitations and cannot substitute for careful clinical judgment. However, use of the TB algorithm at the MGH has improved the awareness of TB among clinicians and other healthcare workers and has assisted infection control personnel in the ongoing evaluation of TB control program needs and priorities throughout the hospital. We agree that new, as well as feasible, approaches to the management of this problem are needed.