Slightly different than in the SHEA-CDC study, bronchoscopists at hospitals with one to five tuberculosis patients per year were more likely than other health care workers to convert their PPD test results. This was not true for hospitals with ≥6 tuberculosis patients per year. Like the CDC-SHEA survey, the type of respiratory protection in use did not correlate with PPD conversion rates. However, unlike the CDC-SHEA survey, PPD conversion rates at hospitals were not related to control measures.

A result similar to that of the APIC result was found in a review of tuberculosis control measures in the 13 hospitals of a midwestern health system, as reviewed by Woeltje and colleagues (26). This survey was performed in 1994–1995. All hospitals had a tuberculosis plan, and all had annual testing of at least selected employees as recommended by the guidelines. Six of 13 (46 percent) of the hospitals were considered very low risk, 6 (46 percent) were considered low risk, and 1 (6 percent) was considered intermediate risk.

Tuberculosis isolation rooms were available at 10/13 (77%) of hospitals; however, only 44 to 100 percent of rooms (median, 88 percent) actually had negative pressure. Dust-mist-fume respirators were used most commonly. PPD conversion rates in 1994 ranged from 0 to 1.0 percent (median, 0.3 percent). The hospital location (urban/rural), type of respiratory protection, tuberculosis risk category, number of tuberculosis isolation rooms, percentage of tuberculosis isolation rooms that were actually at negative pressure, and number of tuberculosis cases were not correlated with PPD conversion. Only the tuberculosis case rate approached significance (p = 0.06, but this may have been spurious, as noted in the discussion section of this paper). In the discussion the authors note that actual compliance with CDC guidelines fell short of the hospitals’ written policies.

Although not as compelling as studies showing the before-and-after effects of implementing control measures, the experiences of hospitals