A study of referral failures for potentially suicidal patients: a method of medical care evaluation.

To learn more about the methodological problems inherent in medical care evaluation studies, a quality assurance committee conducted a study of rates of referral failure for potentially suicidal patients seen in the psychiatry division of a university hospital's emergency room. A simple 5-point scale was used to identify for study a similarly ill patient group: 296 patients with moderately to extremely severe suicide potential. The rates of referral failure for the emergency room's psychiatry division were 24 per cent for referrals within the university hospital system and 33 per cent outside the system, which compared favorably with other facilities. Outlier analysis-a statistical procedure that holds promise for medical care evaluation studies-was used to identify clinicians whose rates of referral failure differed significantly from their peer's rates of. As a result, an instance of substandard performance was identified and corrected through the transfer of information about exceptionally good performance.