TABLE D-4. Results of Tuberculosis Control Measures at Parkland Memorial Hospital

Percent

Measure

1994

1995

p

Patient isolated on admission day

87

79

Patient isolated by 2nd hospital day

89

83

Patient isolated within 72 hours

91

86

Patient never isolated

4.1

7.5

Employee tuberculosis exposure rate

18

25

0.03

Health care worker compliance with PPD testing

49

74

PPD conversions

2.7

3.5

interpretation of apparent increased PPD conversion rates may be spurious because some employees not tested in 1994 may already have had positive PPD test results and would not have been counted as conversions in 1995.

This paper highlights the fact that despite implementation of a protocol and other measures, reductions in employee skin test conversions is not inevitable. Clearly, Parkland suffered from continued delayed isolation of patients and even an increase in employee exposures. This occurred despite an apparently energetic infection control program.

Very little is known about tuberculosis control in nonhospital settings. Nolan and colleagues (22) reported on the control of an outbreak in a shelter for homeless men in Seattle. During December 1986 and January 1987, seven cases of tuberculosis were diagnosed in shelter clients. This prompted mass PPD testing of all the residents of the shelter. Anyone with a positive PPD test result (≥5 millimeter) or symptoms suggestive of tuberculosis were offered chest radiographs. This resulted in the identification of six additional asymptomatic cases of tubreculosis. Persons with tuberculosis were excluded from the shelter, and isoniazid (INH) therapy for latent tuberculosis infection was offered to everyone with a positive PPD test result. The air-handling system (which provided minimal air changes—air was recirculated for economy of heating) was reengineered. Thirty-six UVGI lights were installed in the duct system. The intensity-time dosage was considered adequate to kill 95 percent of the Mycobacterium tuberculosis organisms exposed to it. These interventions led to an interruption of the outbreak. Only five residents were found to have active tuberculosis over the next 2 years. Although this shelter did not follow the CDC guidelines in the strictest sense, their control plan included implementation of administrative controls (identification of cases with subsequent isolation [i.e., removal from the shelter]) and engineering controls. Provision of therapy for latent tuberculosis infection (LTBI) was likely also an important aspect in preventing further cases of active disease in those already infected.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement