During period I, the likelihood of an HIV-infected patient getting tuberculosis decreased with distance from source patient room (but oddly, not related to the amount of time spent on the ward). Smear negative patients were not a source of nosocomial infection in period I. Crude rates of nosocomial tuberculosis were reduced from 8.8 percent during period I to 2.6 percent during period II and to 0 percent in period III. During period II, there was no association of nosocomial tuberculosis with distance from the source patient’s room.

The impact on health care worker PPD conversion rates could not be determined due to insufficient data. However, during period II plus period III, PPD conversion rates were higher on tuberculosis wards than on other wards (5/29 versus 0/15; p = 0.15).

The impact of implementing the CDC guidelines on employee PPD conversion rates at St. Clare’s Hospital in New York was reported by Fella and colleagues (19). Beginning in 1991, all health care workers with patient contact had PPD testing every 6 months; others were tested annually. Two-step testing of new employees was implemented in February 1993. Prior to 1991, no negative-pressure isolation rooms were available at St. Clare’s. The implementation of control measures and PPD conversion rates are shown in Table D-2.

In an abstract presented at the 1994 Annual Conference of the Society for Occupational and Environmental Health—Tuberculosis Control in the Workplace: Science, Implementation, and Prevention Policy, Koll and colleagues (20) summarized data from Beth Israel Medical Center (BIMC) in New York City. The hospital had large numbers of tuberculosis patients and admissions in the early 1990s. A comprehensive tuberculosis policy (based on the 1990 CDC guidelines) was implemented in mid-1992. tuberculosis isolation rooms with negative pressure, ≥6 ACH, and UVGI were