Patients' cooperation with a medical regimen. Difficulties in identifying the noncooperator

Abstract

Twenty-seven ward residents estimated their individual patients'' adherence to an antacid regimen (ATR). The physicians'' estimates were then compared with counts of the bottles of antacid actually consumed by the patients. The median patient took 46% of the prescribed medicine but the median physician''s error of estimate was 32%. Of the 27 physicians, 22 overestimated their patients'' ATR. Further, physicians were unable to distinguish those patients who adhered well from those who adhered poorly, even when the judgments were restricted only to those in which the physician indicated most confidence. The correlations between estimated and measured ATR clustered about zero. Accuracy did not vary with the amount of medicine prescribed, the number of patients judged, height of estimated or measured ATR, or the range or varability of the estimates.