The tuberculin test. Studies of the dynamics of reactivity to tuberculin and Candida antigen in institutionalized patients.

To determine the importance of the recall phenomenon ("booster effect") in chronically hospitalized patients, we performed three sequential tuberculin and Candida antigen skin tests in patients at two hospitals. Twenty of 162 patients (12.3%) demonstrated a significant reaction on the initial tuberculin test, and 9 additional patients (5.5%) showed a significant reaction (booster response) to a second tuberculin test administered 3 wk later. Five patients (3.1%) demonstrated a significant reaction only when a third tuberculin test was administered 6 wk after the initial test. Only 1 patient (0.6%) with a significant tuberculin reaction on the first 2 tests was nonreactive on the third test. Nineteen patients (11.7%) demonstrated a significant response to Candida antigen on the initial test and a booster effect was noted in 11 (6.2%) and 6 (3.7%) patients, respectively, on subsequent tests. Four patients (2.5%) appeared to lose reactivity to Candida antigen on each of the 2 repeat tests. Only 25% of patients who demonstrated a tuberculin booster response had a significant reaction to the initial Candida skin test. Serial skin tests may be necessary to reliably determine the ability of a chronically hospitalized patient to demonstrate a response to tuberculin.