to the case patient converted his PPD (and he was exposed while the patient was in the ICU on a ventilator, with negative AFB smears at the time). The report of Stead and colleagues implies that it was UVGI that was responsible for low PPD conversion rate, even though they had terrific ventilation. He cites one reference (42) regarding PPD conversion, despite 11 ACH in support of this. He also cites older data (43) on UVGI being effective even if no negative-pressure ventilation is available.
The paper by Jernigan et al. (28) cited previously included an interesting statement regarding National Jewish Hospital in Denver. A “personal communication” from L.J.Burton is cited stating that only two PPD conversions had occurred at National Jewish Hospital over a 10-year period, both associated with failure of an ultraviolet light system on a ward. I could not find a publication to confirm this.
Initial studies early in the 1990s suggested poor initial implementation of the 1990 CDC guidelines. Manangan and colleagues (47) reported on a 1992 survey of 180 Texas hospitals (of 475 in the state, of which 151 [83 percent] responded). In 1991, 122/151 (81 percent) had at least 1 tuberculosis admission (up from 98/151 [65 percent] in 1989). Overall, tuberculosis isolation rooms of any sort were not available at 25/140 (18 percent). Seventy-two percent of hospitals had at least one room meeting all CDC criteria. Of the hospitals that had tuberculosis isolation rooms, the rooms had negative pressure in 108/133 (81 percent), ≥6 ACH in 97/131 (74 percent), and air directly vented to outside in 109/131 (83 percent). Only 53/121 (44 percent) hospitals routinely checked the negative pressure in the tuberculosis isolation rooms. The rooms had a private bathroom in 125/134 (93 percent) hospitals. At 94/143 (66 percent) the door was kept closed at all times. Eighty-two percent of hospitals had only surgical masks available for health care workers. Ninety-seven percent performed baseline PPD testing, but only 91 percent performed PPD testing after an exposure.
Van Drunen (48) and colleagues presented data from a Minnesota survey of 17 hospitals carried out by APIC for 1989–1991. Overall there was a wide variety of practice. A total of 13/17 (76 percent) had tuberculosis isolation rooms available. Only three hospitals performed annual PPD tests; many hospitals let employees self-read the PPD test results. All