Statewide physician group incentive program focused on population health reduces total spending in Michigan

Efforts are being made to contain costs, enhance quality, and improve population health. Many of these efforts have been unsuccessful and most evaluations are limited in scope and time. We present results from an economic evaluation of a statewide physician group incentive program (PGIP) including 15,000 physicians and 1.8 million members over four years.

PGIP includes multiple initiatives developed collaboratively by physicians, physician organizations and Blue Cross/Blue Shield of Michigan (BCBSM). Program activities reward primary care and population health, such as patient-centered medical home, health information technology (e.g., registries), LEAN redesign, and efforts to improve care transitions and coordination.

The study population included over 3.2 million adult enrollees from 2008 to 2011. 634,035 individuals were excluded because they were not enrolled for at least 12 continuous months. 14,518 physicians were included; 1,664 physicians were excluded because they did not meet eligibility criteria. Patients were attributed to physicians using a standard attribution process.

A difference-in-difference (DD) approach was used to isolate the effect of PGIP on total spending. The pre-intervention period was 2008; the post-intervention period was 2009-2011. The control group included all patients treated by BCBSM physicians who did not join PGIP during 2008-2011.

The dependent variable was average per member per month (PMPM) spending, including BCBSM spending and enrollee cost sharing. Prescription drug spending was excluded because a majority of enrollees did not have that coverage through BCBSM.

We controlled for age categories, gender, risk score, and secular trends to correct for differences in individual traits across intervention and control groups. Risk scores were calculated using Episode Risk Groups (ERG 7.0) provided by Ingenix.

Average spending during the study period was $289 PMPM. PGIP expenditures were $4.17 PMPM lower than non-PGIP (p<.001) indicating that lower cost practices were more likely to join PGIP. Including secular trends and other control variables, the DD analyses shows that the policy effect of PGIP was -$3.07 PMPM (p <.05). We also found that the first year a physician practice joined PGIP, it experienced, on average, $2.72 PMPM higher costs, roughly offsetting the savings in that year.

This study suggests that a statewide program developed collaboratively with physicians and emphasizing primary care and population health achieved significant spending reductions over time, though these reductions were not immediately evident. This may be due to increased spending in the initial program period, as providers initiate prevention and other activities aimed at improving population health.