Physician attrition from clinical practice increased from 4% in 2013 to 5% in 2019, with female physicians and those practicing in rural settings experiencing disproportionately higher departure rates, according to a nationwide longitudinal study.

In the analysis of 712,395 physicians treating Medicare beneficiaries between 2013 and 2022, investigators found that attrition rates rose across all specialties, geographic regions, and age groups over 35 years, wrote lead study author Lisa S. Rotenstein, MD, MBA, MSc, of the University of California at San Francisco, and colleagues. The study represented one of the first comprehensive examinations of actual clinical practice departures, as opposed to stated intentions to leave practice.

Physician Characteristics and Attrition Patterns

Female physicians showed a 44% higher risk of leaving clinical practice compared with male physicians in multivariable Cox models adjusting for physician factors and calendar year. Physicians practicing in rural areas showed a 19% increased risk vs their counterparts practicing in urban areas.

Unadjusted attrition rates increased more steeply in rural compared with urban settings during the study period—from 4% to 6% in rural locations compared with 3% to 5% in urban areas.

Specialty-Specific Departures

Psychiatrists and obstetrician-gynecologists showed the highest baseline attrition rates in models adjusting only for physician characteristics. Psychiatry attrition ranged from 7% in 2013 to 10% in 2019, whereas ObGyn showed attrition rates of 6% in 2013 and 11% in 2019.

However, after adjusting for Medicare beneficiary characteristics, hospital-based specialists exhibited the highest attrition risk, with psychiatry and ObGyn showing lower risks compared with hospital-based specialties when patient panel composition was considered.

Patient Panel Characteristics

Patient complexity measures showed strong associations with physician attrition. Caring for Medicare beneficiaries in the highest quartile of risk scores showed a 15% increased attrition risk compared with the lowest quartile. The highest quartile of average beneficiary age carried a 30% higher risk vs the lowest quartile.

Physicians caring for patient panels in the fourth quartile of dual-eligible beneficiaries—those qualifying for both Medicare and Medicaid—faced a 57% higher attrition risk compared with the first quartile. Each additional group of 100 Medicare beneficiaries who received care was associated with a 13% lower attrition risk.

Patient panels with the highest proportion of Black patients, Hispanic patients, and patients of other races and ethnicities were each associated with reduced attrition compared with lower proportions.

Geographic Variation

Puerto Rico and other territories showed the highest geographic risk compared with the Northeast reference region. The West showed a 29% higher risk, while the South showed a 7% higher risk.

Age-Related Patterns

All age groups, except physicians younger than 35 years, experienced increased attrition rates during the study period. Physicians aged 65 years or older showed the largest absolute increase, from 11% in 2013 to 14% in 2019.

Study Methodology

The investigators used 2013 to 2022 Centers for Medicare & Medicaid Services Medicare Physician & Other Practitioners and 2013 to 2021 Medicare Data on Provider Practice and Specialty data sets. Clinical practice attrition was defined as permanently leaving the practice, evidenced by not meeting a threshold of at least 50 evaluation and management services for at least 3 consecutive years.

The sample of 712,395 physicians comprised 71% male physicians and 91% practicing in urban settings, distributed across hospital-based specialties (28%), medical specialties (20%), primary care (28%), surgical specialties (17%), psychiatry (4%), and ObGyn (4%).

Physicians in the sample cared for a median of 280 Medicare beneficiaries. The median age of these beneficiaries was 72 years. Their median risk score was 1.5. A median of 57% of physicians' Medicare beneficiaries were female, 91% were White, 0% were Black, and 0% were Hispanic. A median of 25% of beneficiaries were dually eligible for Medicare and Medicaid.

Study Limitations

The analysis focused exclusively on Medicare fee-for-service claims, potentially not reflecting physicians' complete practices or those not treating Medicare patients. The requirement of at least 50 evaluation and management services yearly may have classified some active physicians as having left practice. The study couldn't capture physicians transitioning to nonclinical medical roles.

The study received funding from the Physicians Foundation, which had no role in study design, conduct, analysis, or publication decisions.

Source: Annals of Internal Medicine