The problems produced by using 95% confidence limits in evaluating a test result as normal for a patient or satisfactory when proficiency testing a laboratory are presented. The authors provide statistical tables for probability estimates to be used in performing multiple tests on a patient or specimen. Suggestions for reducing the problems of the “abnormal” normal in the screening of patients and the “unsatisfactory” satisfactory in proficiency testing of laboratories are offered.
No takes yet. Share an insight, caveat, or question.
Schoen et al. (1970) studied this question.