A clinical evaluation of the "roll-over test" for pregnancy-induced hypertension.

A study of the capability of the roll-over test to predict pregnancy-induced hypertension was undertaken using our private patients. Sixty primigravid and 60 multigravid patients were studied between the twenty-eighth and thirty-second week of gestations. These patients were chosen at random by one of our two nurses who conducted all these studies. All results were recorded but were not available to the physicians. Eighteen months later, after all study patients had delivered, the hospital charts and the patients' office records were evaluated to determine if preganancy-induced hypertension had occurred. In primigravid patients a positive test accurately predicted the later development of pregnancy-induced hypertension only 50 per cent of the time while a negative test accurately predicted that it would not develop 93 per cent of the time. In multigravid patients, only 25 per cent of the patients who had positive tests later developed hypertension. The negative test in multigravid patients was accurate 94 per cent of the time.