MEASURING SAFETY, QUALITY, AND VALUE

Research Objective More comprehensive primary care is associated with lower costs and better outcomes for patients. Several primary care interventions under the ACA target practice site level redesign. Thus, it is helpful to have measures to assess whether the comprehensiveness of the practice site is associated with beneficiary outcomes. Building on prior work developing and validating primary care physician‐level measures of comprehensiveness, we developed two practice‐level measures of primary care comprehensiveness and tested their associations with beneficiary outcomes. Study Design Lagged analysis of claims‐based outcomes for Medicare FFS beneficiaries during a year to practice‐level comprehensiveness measures in the previous year. Data are from 2013‐2014 from the evaluation of the Comprehensive Primary Care Initiative. The practice‐level measures assessed, across all primary care physicians in a practice site, the extent to which the physicians in the practice were involved in care for their patients’ conditions (IPC = involved in patient conditions) and managed their patients’ new problems (NPM = new problem management). Regression models controlled for beneficiary, practice, and market characteristics. Population Studied 1,343 primary care practices where 5,336 physicians cared for over 1 million Medicare fee‐for‐service beneficiaries. Principal Findings The two measures varied across primary care practices and captured different aspects of comprehensiveness, as intended. Comparing practices in the 75th versus 25th percentile for each of these measures of comprehensiveness, practices where physicians had greater involvement in patient conditions (IPC) had 2.3% lower Medicare expenditures ( P  = .03), and 1.7% lower ED visit rates per 1000 beneficiaries ( P  = .08). Practices where physicians demonstrated greater new problem management (NPM) had 1.6% lower Medicare expenditures ( P  = .04), 1.8%, lower hospitalization rates per 1000 beneficiaries ( P  = .03), 2.6% lower overall ED visit rates ( P  = .00), and 2.6% lower ED outpatient visit rates ( P  = .01). Conclusions These measures demonstrate strong content and predictive validity and reliability. Medicare beneficiaries in primary care practices where physicians deliver more comprehensive care had slightly lower Medicare expenditures, and lower rates of hospitalizations, ED visits, and ED outpatient visits. Implications for Policy or Practice These new measures may be useful indicators of 2 aspects of practice‐level comprehensiveness (involvement in patient conditions and new problem management) for studies that target the practice as the unit of change. Additional research should focus on the robustness of these findings for the full Medicare population as well as for other populations. Future research to test interventions to promote primary care comprehensiveness may also be useful. Primary Funding Source Centers for Medicare and Medicaid Services.