Tuberculosis tracking: Determining the frequency of the booster effect in patients and staff

Background: Two-step tuberculin skin testing, which is recommended to exclu de the booster effect as a cause of converting nonreactive skin test respon ses to reactive responses, can be expensive and logistically challenging. W e studied the booster effect in our patients and staff to determine its fre quency and to identify factors that might predict its occurrence. Methods: Hospital staff members and long-term care patients were given 2-st ep Mantoux tests and evaluated prospectively. Results: Of 619 staff members tested, the initial tuberculin response was r eactive in 39 (6.3%). Of the 97 nonreactive staff members who presented for retesting 6 to 30 days later, 6 were now reactive (6.2%; [95% CI, 2.3%-11. 8%]). Twelve (13.6%) of eighty-eight patients were tuberculin reactive on t heir initial skin test. Of the 37 nonreactive patients appropriately admini stered their second test, 2 (5.4%, [95% CI, 0.7-18.2%]) were tuberculin-rea ctive. Thirteen percent of tuberculin-reactive staff members and 16% of tub erculin-reactive patients were identified only after their second skin test . Foreign birth (P = .02) was associated with purified protein derivative r esponse boosting in staff members; anemia was associated with boosting in p atients (P = .05). Conclusions: Our results support 2-step skin testing of all new employees a nd patients who are likely to receive periodic retesting. In our population , age alone is an inadequate criterion for selecting candidates for retesti ng.