Respiratory isolation of patients with suspected pulmonary tuberculosis in an inner-city hospital.

OBJECTIVE

To identify clinical factors that predict which patients presenting to the ED with pneumonia will require respiratory isolation for suspected tuberculosis and to evaluate a protocol for rapid identification of patients at risk for pulmonary tuberculosis (PTB).

METHODS

To identify potential clinical indicators of PTB, a case-control study was performed using patients admitted to an urban teaching hospital with the ED diagnosis of pneumonia (derivation sample). These predictors were then evaluated in a separate prospective observational study of 103 patients admitted to the same institution from July 1994 to February 1995. Adult patients with the admitting diagnosis of pneumonia were admitted to a respiratory isolation bed if they met 1 of the following criteria: 1) HIV-positive or unknown HIV status with a history of injection drug use; 2) chest x-ray consistent with PTB; or 3) pneumonia with 1 of the following: PPD conversion within 2 years, recent exposure to PTB, previous PTB, or hemoptysis. Patients who did not meet isolation criteria were admitted to the medical ward and had a PPD and anergy panel placed. Those who were anergic or PPD-positive were transferred to respiratory isolation.

RESULTS

Predictor variables identified during the first study phase were incorporated into the isolation guidelines noted above. Only 36 of 50 (72%) PTB patients were admitted to an isolation bed during this phase. During the second phase, 103 patients were admitted with the ED diagnosis of pneumonia-rule out PTB; 22 patients (22%) were culture-confirmed positive for PTB. The guidelines predicted PTB as follows: sensitivity, 0.96 (95% CI, 0.88-1.0); specificity, 0.14 (95% CI, 0.08-0.24); positive predictive value, 0.23 (95% CI, 0.17-0.35); and negative predictive value, 0.92 (95% CI, 0.77-1.0). The 1 patient who was not isolated was found to be anergic after 48 hours and subsequently isolated.

CONCLUSION

Respiratory isolation guidelines for patients admitted from the ED with pneumonia were developed and validated. These guidelines provide satisfactory guidance for isolation of patients at risk for PTB in a high-PTB-prevalence population.