June 16, 1993

Gauging the Accuracy of Laboratory Testing From Medicare Data-Reply

Author Affiliations

Brigham and Women's Hospital Boston, Mass
University of Alabama at Birmingham

JAMA. 1993;269(23):2985-2986. doi:10.1001/jama.1993.03500230066024

In Reply.  —In our article we made one fundamental point: health insurance claims data can be fruitfully used to track patient outcomes and to screen for laboratories that are having a higher than expected frequency of bad outcomes. We illustrated this use of claims data by examining Medicare patient outcomes following outpatient PT testing. We found that risks for stroke and myocardial infarction were higher when patients had been tested in low-volume physician office laboratories, when the site of testing switched between successive tests, and when the patient had no history of testing in the prior 6 months. We also found that patients in states with stronger regulatory climates had lower risk for stroke; similar results for myocardial infarction were not statistically significant.According to our approach with a large database of laboratories and clinicians, all factors that lead to adverse outcomes that are attributable to physicians' activities once the