Variation among physicians in use of laboratory tests. II. Relation to clinical productivity and outcomes of care.

Previous studies from this institution have shown extreme variation in laboratory and x-ray use among comparably trained physicians caring for similar patients. In addition, essentially no correlation (r = --.13) existed between a physician's lab use profile and subjective estimates of clinical competence. This study compares variations in lab use with both clinical productivity and outcome of care. Costs of lab tests of 149 long-term ambulatory hypertensive patients cared for by 13 faculty internists during one year were computed. Variation in mean annual lab costs per patient among the internists was great (range, $8-$161; standard deviation, $42). Outcomes of care were estimated using hypertension as an indicator condition. The physicians were scored according to percentage of hypertensive patients with systolic and diastolic pressures below specified levels. Correlation between lab use profiles and outcomes was negative (r = --.42) but not significant. Clinical productivity was estimated by two methods: adjusted panel size and subjective estimates of efficiency by the clinic administrator. Correlations between lab use behavior and each estimate of productivity were negligible (.13 and --.16 respectively). These data indicate that in this setting there is no positive association between a physician's frequency of lab use and either clinical productivity or outcomes of care.