IMPORTANCEAlthough physician concerns about medical malpractice are substantial, national data are lacking on the rate of claims paid on behalf of US physicians by specialty.OBJECTIVE To characterize paid malpractice claims by specialty.

Over a recent 10-year period, a small number of physicians with distinctive characteristics accounted for a disproportionately large number of paid malpractice claims.

BACKGROUNDPhysicians with poor malpractice liability records may pose a risk to patient safety. There are long-standing concerns that such physicians tend to relocate for a fresh start, but little is known about whether, how, and where they continue to practice. METHODSWe linked an extract of the National Practitioner Data Bank to the Medicare Data on Provider Practice and Specialty data set to create a national cohort of physicians 35 to 65 years of age who practiced during the period from 2008 through 2015. We analyzed associations between the number of paid malpractice claims that physicians accrued and exits from medical practice, changes in clinical volume, geographic relocation, and change in practice-group size. RESULTSThe cohort consisted of 480,894 physicians who had 68,956 paid claims from 2003 through 2015. A total of 89.0% of the physicians had no claims, 8.8% had 1 claim, and the remaining 2.3% had 2 or more claims and accounted for 38.9% of all claims. The number of claims was positively associated with the odds of leaving the practice of medicine (odds ratio for 1 claim vs. no claims, 1.09; 95% confidence interval [CI], 1.06 to 1.11; odds ratio for ≥5 claims, 1.45; 95% CI, 1.20 to 1.74). The number of claims was not associated with geographic relocation but was positively associated with shifts into smaller practice settings. For example, physicians with 5 or more claims had more than twice the odds of moving into solo practice than physicians with no claims (odds ratio, 2.39; 95% CI, 1.79 to 3.20). CONCLUSIONSPhysicians with multiple malpractice claims were no more likely to relocate geographically than those with no claims, but they were more likely to stop practicing medicine or switch to smaller practice settings.