Abstract

Experience with a new serological method for the diagnosis of tuberculosis is reported in a predominantly black population. We have found that in only 69% of 167 patients was there agreement between serology and the presence or absence of tuberculosis. Both false positive and false negative results were common. Of 47 healthy controls, 80% were positive. These results are less satisfactory than previous studies but differences in the reading of the results seems an unlikely explanation. Differences in the populations studied may be an important factor.