MOST psychiatrists would agree that the practicing psychiatrist has a paucity of general laws and proven concepts on which to base clinical decisions. Good rules are not available for predicting the natural course of most mental diseases,¹ the response to drugs,² hospitalization,³ or psychotherapy.⁴ The clinician is left to develop his own rules, and, where patterns and types can only be identified through the processing of large numbers of patients and vast amounts of data, he is unable to do so. In view of this lack of fundamental scientific knowledge, major commitments to the furtherance of psychiatric knowledge are indicated, and a possible source of data is the clinical record.^5,6

Since 1966 the University of Missouri Medical School at the Missouri Institute of Psychiatry and the Missouri Division of Mental Diseases have worked jointly on a program to make clinical record data maximally retrievable and