METHODS: Between 1993 and 1998 data of 27,000 patients with hip fractures and 89,000 patients undergoing cholecystectomy were collected by the department of external quality assurance at the chamber of physicians of Westphalia-Lippe. The data were analyzed for the relationship between volume and outcome, specified as mortality and morbidity. Logistic regression was used to adjust the results for demographic and clinical risk factors. RESULTS: The risk-adjusted probability of death from cholecystectomy was 89% higher in the group of small volume clinics (less than 30 procedures per year) than in the group of large volume clinics (more than 120 procedures per year) (odds ratio 1.89; 95% CI 1.19-3.00). Mortality from hip fractures was 33% higher (odds ratio 1.33, 95% CI 1.09-1.63) for the group of small volume clinics (less than 15 procedures per year) than the group of large volume clinics (more than 45 procedures per year). Among the group of small volume hospitals were some with excellent results, but they are more likely to be found in the group of large volume hospitals. On the other hand there were some departments with poorer results in the group of large hospitals. CONCLUSIONS: There is no threshold for good results. In general there is an inverse relationship between volume and outcome. High volume, though, is not a substitute for good results. It is not possible to infer good quality from high volume alone. Small-volume departments should document quality of care and results thoroughly.
METHODS: Between 1993 and 1998 data of 27,000 patients with hip fractures and 89,000 patients undergoing cholecystectomy were collected by the department of external quality assurance at the chamber of physicians of Westphalia-Lippe. The data were analyzed for the relationship between volume and outcome, specified as mortality and morbidity. Logistic regression was used to adjust the results for demographic and clinical risk factors. RESULTS: The risk-adjusted probability of death from cholecystectomy was 89% higher in the group of small volume clinics (less than 30 procedures per year) than in the group of large volume clinics (more than 120 procedures per year) (odds ratio 1.89; 95% CI 1.19-3.00). Mortality from hip fractures was 33% higher (odds ratio 1.33, 95% CI 1.09-1.63) for the group of small volume clinics (less than 15 procedures per year) than the group of large volume clinics (more than 45 procedures per year). Among the group of small volume hospitals were some with excellent results, but they are more likely to be found in the group of large volume hospitals. On the other hand there were some departments with poorer results in the group of large hospitals. CONCLUSIONS: There is no threshold for good results. In general there is an inverse relationship between volume and outcome. High volume, though, is not a substitute for good results. It is not possible to infer good quality from high volume alone. Small-volume departments should document quality of care and results thoroughly.