Since the early 1930s, it has been recommended that a TBN-ST be read at no earlier than 48 h. However, this was not always the case. In one of his initial reports, Mantoux stated that a positive tuberculin reaction is apparent by 24 h. Despite this, it became clear in subsequent years that although most reactions are present by 24 hi7 a very small number of subjects will have false-positive reactions at 24 h, and a somewhat larger number of subjects will not show significant reactions until 48 h or even later. These results were from studies using either old tuberculin or only 1 U of PPD.
Despite the changes in purity and amount of antigen since these studies, there have been few contemporary studies which have examined the yield of reading a TBN-ST before 48 h. However, two studies have reported that 24-h readings yield a sensitivity of approximately 83 percent when compared with 48-h readings.
The present study does not allow an exact determination of the sensitivity of an early reading of a TBN-ST. However, extrapolations from two separate subgroups provided estimated sensitivities which were in close agreement. These results suggest that for patients whose tests were read between 24 and 47 h, a sensitivity of 91 percent is achieved. The sensitivity drops to 39 percent for patients whose tests were read before 24 h (Table 2).