Although not as compelling as a directed study of the impact of control measures, studies correlating implementation of control measures with relevant outcomes can also provide insight into the efficacy of the CDC guidelines.
One such example is from a Society for Healthcare Epidemiology of America (SHEA)-CDC survey of 1989–1992 tuberculosis control practices reported by Fridkin and colleagues (23, 24). The survey was sent to all members of SHEA in March 1993. Members from 210 hospitals responded. Part II of the results (24) focused on the efficacy of control measures. It showed that “high-risk” employees (e.g., respiratory therapists and bronchoscopists) were more likely than other health care workers to have PPD conversion if ≥6 tuberculosis patients per year admitted, if the hospital was “large” (≥437 beds), or if MDR tuberculosis was present. The most significant impact on both high risk and other PPD conversions was whether the hospital admitted ≥6 tuberculosis patients (for non-high risk health care workers, PPD conversions of 1.2 percent in high-volume hospital versus 0.6 percent in low-volume hospitals; for high-risk health care workers, PPD conversion rates were 1.9 percent versus 0.2 percent).
The authors evaluated four criteria from the 1990 CDC guidelines: (a) placing known/suspected tuberculosis patients into single patient room (or cohorting), (b) negative-pressure ventilation, (c) air exhaust directly to outside, and (d) ≥6 ACH. Hospitals with ≥ tuberculosis patients meeting all four criteria had PPD conversion rates of 0.60 percent, whereas they were 1.89 percent for hospitals that did not (p = 0.02). Hospitals meeting at least criteria a to c had PPD conversion rates of 0.62 percent whereas the rate was 1.83 percent for those that did not (p = 0.03). The data suggested that having negative pressure or outside exhaust versus not having one or the other also reduced rates, but this did not reach statistical significance. The use of a submicron mask versus a surgical mask made no difference in conversion rates. For hospitals with less than six tuberculosis admissions per year, no difference in PPD conversion rates could be shown to be related to control measures.
A similar survey on tuberculosis control measures was sent to members of the Association for Professionals in Infection Control and Epidemiology (APIC) in March 1993, as reported by Sinkowitz and colleagues (25). It also covered practices from 1989 to 1992. Data were obtained from 1,494 hospitals. Compared with the SHEA-CDC survey, the hospitals in this APIC survey were more likely to be a community hospital and more likely to not have any tuberculosis admissions in 1992. Results of the survey are summarized in Table D-5.
Whether or not tuberculosis isolation rooms met CDC criteria was also reviewed, but the data are not summarized here.