Cardiologist Participation in Accountable Care Organizations and Changes in Spending and Quality for Medicare Patients With Cardiovascular Disease

Abstract

Background: Despite widespread adoption of Medicare accountable care organizations (ACOs), healthcare spending reductions have been modest. This may relate to variable participation in ACOs by specialist physicians, who disproportionately drive spending. To examine whether specialist participation in Medicare ACOs was associated with changes in healthcare spending and clinical quality, we analyzed national Medicare data.

Methods and results: Working with a 20% random sample of Medicare beneficiaries (2008 to 2015), we identified those with cardiovascular disease. We estimated linear regression models at the beneficiary-quarter level to evaluate changes in healthcare spending and clinical quality after the start of the Shared Savings Program in 2012. We then examined whether changes in spending and quality across ACOs were conditional on cardiologist participation. Our study included ≈1.6 million beneficiaries per year. Although the number of ACOs increased over the study period (from 114 in 2012 to 392 in 2015), the proportion with any cardiologist participation remained stable (from 80% in 2012 to 83% in 2015). Compared with unaligned beneficiaries, those cared for by ACOs without cardiologist participation were associated with a spending reduction (per quarter) of -$75 (95% CI, -$105 to -$46; P<0.001). Care receipt in an ACO with cardiologist participation was associated with an additional difference in spending of -$56 (95% CI, -$87 to -$25; P<0.001), driven by lower spending for skilled nursing facilities, evaluation and management services, procedural care, and testing. While heart failure admission rates were similar among aligned and unaligned beneficiaries, ACO care was associated with fewer all-cause readmissions (P<0.001) and emergency department visits (P<0.001). Rates of these outcomes did not vary by cardiologist participation.

Conclusions: Annual spending for beneficiaries with cardiovascular disease was ≈$200 lower when cared for by ACOs with cardiologist participation (compared with those without). These spending reductions did not come at the expense of clinical quality.

Keywords: Medicare; accountable care organizations; cardiologist; cardiovascular disease.

Figure 1:. The density of cardiologist participation within Accountable Care Organizations by year.

This distribution of cardiologist participation in ACOs as demonstrated by the number of cardiologists per 1000 ACO beneficiaries (x-axis) and the percentage of ACOs (y-axis). The total number of ACOs per year is noted at the top of each graph. Note: 11 ACOs were missing provider information in 2014. Those ACOs and their assigned beneficiaries were dropped from the analysis for the year 2014.

Figure 2:. Adjusted total spending per beneficiary-quarter by sensitivity analysis.

Adjusted total spending per beneficiary-quarter: (A) when cardiologist participation is stratified by the density of cardiologists in an ACO per total beneficiaries; and (B) among ACOs with cardiologist participation stratified by ACO organizational structure. To account for inflation, all payments were adjusted to 2015 dollars using the Consumer Price Index. Results are presented after controlling for patient demographic variables, clinical severity, quarter, year, hospital referral region, ACO participation, and the number of cardiovascular specialists in a hospital referral region per 100,000 Medicare beneficiaries. Abbreviation: ACO = accountable care organization * Among ACOs with cardiologist participation.