[Basic study on criteria for the reactivity of non-stress tests by means of computer simulation]

The non-stress test has become one of the most popular tests for antenatal surveillance in modern obstetrics. However, the fact that there are so many criteria for 'reactivity' in this test has made it rather difficult to interpret the results of these tests and compare them with each other. Using computer simulation of the occurrence of acceleration in fetal heart beats, the sensitivity and specificity of eight criteria which are in general use were studied. Diagnosis by Brown's criteria produced the best results for both sensitivity and specificity, but the long time necessary for this test is a drawback. However, the predictive value positive (probability of poor fetal outcome with nonreactivity) is only 62% even according to Brown's criteria when the test is done for antenatal care screening of an unrestricted population. The results of the test are far more convincing when it is reactive because the predictive value negative is more than 96%, even when strictly interpreted.