Tuberculosis isolation rooms were tested with smoke approximately every 3 months. The failure rate ranged from 6.1 percent to 21.7 percent (mean, 16.5 percent). One room tested with sulfur hexafluoride had 4.9 ACH. The author suggest that their data imply that the improvements in PPD conversion rates were primarily the result of improved administrative controls since changes mirrored improved isolation as a result of the new policies. They argue that since room negative pressure was demonstrated to be frequently suboptimal, engineering controls were not the major factor in the improvements. Likewise submicron masks appeared to be adequate. The new policies resulted in only one of eight patients placed on tuberculosis isolation having culture-confirmed tuberculosis.

Columbia-Presbyterian Medical Center in New York City had control measures detailed by Bangsberg and colleagues (17). They revised their tuberculosis control guidelines to be consistent with the CDC guidelines. Prior to June, 1992, medical house staff were PPD tested at baseline and were then instructed to be tested annually by their primary physicians. Starting in June 1992, PPD testing was done every 6 months on medical house staff. The overall rate of participation was 92 percent.

Revised tuberculosis control measures included stricter isolation policy (implemented in May 1992) so that patients with HIV infection or HIV in fection risk factors or who were homeless and presented with pneumonia or evidence of tuberculosis were placed in tuberculosis isolation until three sputum samples were AFB negative and the patient was judged noninfectious by pulmonary and infectious disease consultants. Tuberculosis isolation rooms were installed in the emergency department (ED) in July 1992. A tuberculosis service was implemented at the end of June 1993. In July 1993 3M respirators (type not stated) were instituted.