The efficacy of the tuberculin test: an analysis based on results from 33 countries.

Is it always justified to take a tuberculin reaction over a certain size limit as indicative of infection with tubercle bacilli and reactions below the limit as indicative of absence of such infection? Information on this point is here derived from a quantitative analysis of test reactions found in general population groups. The data were collected by specially trained international teams operating simultaneously with or as a preliminary to large-scale tuberculosis control programmes. Because of the international character of the work it has been possible to compile a picture which shows the tuberculin sensitivity in a large part of the world, yet which is based on highly uniform techniques. It is hereby demonstrated that the pattern of tuberculin sensitivity varies widely between different populations, but follows a definite geographical trend. In temperate and subtropical countries almost all test reactions are either clearly "positive" or clearly "negative", indicating that the test is highly efficient. In tropical regions, on the other hand, a large proportion of the reactions are intermediate in size, and distinction between two kinds of reaction is therefore difficult. The data strongly suggest that the cause of the intermediate reactions is that the population is being massively exposed to certain unidentified agents producing cross-reactions to tuberculin. In tropical regions a clear-cut distinction between tuberculosis infected and uninfected evidently cannot be made by means of the present tuberculin test.