Variation Among Physicians in Use of Laboratory Tests: Relation to Quality of Care

A previous study at this institution demonstrated extreme variation in use laboratory tests among similarly trained internists caring for similar groups of ambulatory patients. This report examines whether these variations can be related to differences in the quality of physician performance. The physician group studied consisted of 21 medical interns at the George Washington University Hospital. In order to insure that differences in laboratory use among the physicians were not attributable to clinical differences in their patient populations, measurement of laboratory use was restricted to a standardized subset of patients admitted to the coronary care unit with acute myocardial infarction or chest pain. Mean laboratory costs for the first three hospital days varied from 48 to 113 per patient among the interns. The relative clinical competence of the interns was independently assessed by five active faculty members. Concordance among the faculty raters was high and statistically significant (p <. 01). Rank order correlation between an intern's cost behavior and his assessed clinical competence was negligible (rs = —. 13). Strong correlation existed between the number of redundant (unnecessary) tests ordered per intern and his overall cost rank (p <. 01), but not between frequency of redundant tests and clinical competence. At least in this setting, a physician's laboratory utilization profile is not a good index of quality of care.