Objective: A rise in the incidence of purified protein derivative (PPD) skin test conversions among employees at our medical center between 1991 and 1993 prompted an examination of factors associated with PPD skin test conversion. We focused on the effect of bacille Calmette-Guérin (BCG) vaccination on PPD skin test conversion because of changes in employee health service policies in 1990 regarding testing of persons who had received BCG vaccination.
Methods: The study took place in a university teaching hospital employee health service. Charts of employees who had PPD skin test conversion (> 10 mm increase in induration of the PPD response within 2 years if younger than 35 years of age or > 15 mm if older than 35 years of age) between 1988 and 1993 were reviewed for factors that could have influenced PPD skin test conversion and compared with data from 271 randomly selected charts of employees who underwent annual employee assessments in 1993 but did not have PPD skin test conversion.
Results: PPD skin test conversions rose from 0.06% (1/1604) to 1.3% (22/1760; p = 0.000001) in employees tested between 1988 and 1993. Of 41 persons with PPD skin test conversion between 1991 and 1993, 29 (71%) had received BCG vaccination. Only 21% of control subjects (56/271) had received BCG vaccination (p < 0.000001 for comparison of BCG vaccination among those with PPD skin test conversion with that among control subjects). When BCG recipients were not included as having PPD skin test conversion, there was no significant increase in PPD skin test conversions. Twenty-three BCG recipients had PPD skin test conversion on their second PPD skin tests.
Conclusion: A large proportion of PPD skin test conversions at hospitals that employ large numbers of health care workers who have received BCG vaccination may not represent recently acquired tuberculosis. Rather, these conversions may be effects of previous BCG vaccination. Two-step initial PPD skin testing may help to eliminate nearly 80% of such false-positive conversions.