Outcomes of Practice-Level Continuity in Primary Care

Problem

Continuity of care is a core attribute of primary care, traditionally defined by an individual patient-provider relationships but increasingly shifting towards teams of providers and even practices. While individual continuity is shown to reduce unnecessary care (as well as total costs and hospitalizations in our previously published work), less is known about the effect of continuous care at the practice level. Our objective was to develop a practice-level measure of continuity of care using U.S. Medicare claims data and assess its impact on total costs of care and hospitalizations.

Approach

We acquired 2011 U.S. Medicare (public insurance) service utilization data on all claims for 1,230,947 Medicare beneficiaries who received any care from a nationally representative sample of 6,837 primary care physicians (PCPs). After identify PCPs in the same practice, we calculated patient-level practice-level primary care continuity scores. Then, we obtained PCP-level scores based on those of their patients. We estimated impact on two principal outcomes, total expenditures and hospitalizations, using multilevel models, including controls for beneficiaries (age, gender, race and Charlson scores), physicians (graduation year, international training, and gender) and practice characteristics (size and rurality).

Findings

PCPs in larger practices and rural areas are more likely to provide practice-level continuity. Adjusted expenditures for beneficiaries cared for by PCPs in the highest quintile of practice continuity were 29% lower than for those in the lowest quintile (β=-.292, 95% CI: -.333 to -.251). Odds of hospitalization were 31% lower between the highest and lowest continuity quintiles (OR: .694, .647-.745). Practice-level continuity effects were higher than physician-level continuity. While patients obtaining care in larger practices are more likely to obtain continuous care from other PCPs in that practice, they are more likely to incur higher costs and hospitalizations.

Consequences

Primary care continuity has both individual and practice level dimensions that impact quality. This study comes from the United States, during a period of considerable experimentation and movement towards value-based purchasing. It demonstrates that practice-based continuity should be accounted for in quality measurement and value-based purchasing schemes, which continue to evolve in most OECD (wealthy nation) health systems.