ACO Participation Did Not Lower Care Access for Vulnerable Patients

Physician groups did not reduce their share of vulnerable patients after joining an accountable care organization despite claims of the opposite, a new study from the Perelman School of Medicine revealed.

Accountable care organizations (ACOs) are the largest payment reform experimentation, with over 1,000 of the organizations covering more than 32 million patients in 2018, researchers reported. But there are concerns that the incentive structure of the alternative care delivery model reinforces or possibly exacerbates disparities in healthcare quality.

Prior research has shown that ACOs typically form in geographic areas with fewer black residents and lower rates of poverty, fewer uninsured patients, and fewer patients without high school education. Evidence has also demonstrated that ACOs with more medically complex patients and higher-cost physicians and beneficiaries were more likely to drop out of the ACO.

However, the study published in JAMA Network Open last week found that physician groups in ACOs did not engage in what the industry calls “cream-skimming” by altering the share of socially vulnerable patients in their panels after joining an ACO.

Cream skimming occurs when a physician group stops treating vulnerable patients to prevent high-risk patients from jeopardizing the group’s ability to earn shared savings payments. The practice has been known to impact the treatment of black patients, patients dually eligible for Medicare and Medicaid, and patients in areas with high poverty and unemployment rates.

But cream-skimming is not practiced as frequently as previously thought, researchers found.

The study of 76,717 physician groups in Medicare’s largest ACO program from 2010 through 2016 found no statistically significant changes associated with ACO participation in the proportions of vulnerable patients attributed to participating physician groups compared with non-participating groups.

Specifically, the magnitude of change after joining an ACO was 0.0 percentage points for black patients, -0.1 percentage points for patients dually enrolled in Medicare and Medicaid, 0.2 percentage points in poverty rates of patient zip codes, and -0.4 percentage points in unemployment rates of patient zip codes.

The findings reassured researchers that ACO providers are generally not avoiding vulnerable patients who stand to benefit the most from the alternative care delivery model. But ongoing monitoring is recommended, they stated.

“While there were no changes in the proportion of vulnerable patients cared for in ACO-participating physician groups in our main analysis, there are other ways that ACOs can perpetuate or worsen disparities,” researchers wrote in the study.

“Our unadjusted descriptive data showed that, based on patient characteristics in 2010, patients who were later attributed to ACOs during the study period were less likely to be black or dually enrolled in Medicare and Medicaid, and lived in zip codes with lower poverty rates. This is consistent with prior research showing that ACOs more often form in areas with fewer black residents and lower rates of poverty.”

This could continue and potentially get worse as ACOs increasingly take on downside financial risk.

The Medicare Shared Savings Program bases shared savings payments to ACOs largely on financial benchmarks set using the attributed population’s prior expenditures. All ACOs in the flagship ACO program can earn the incentive payments by reducing their population’s spending, but only a small group has to repay Medicare if spending increases above the financial benchmark.

However, that is changing under new program rules. CMS overhauled the Medicare Shared Savings Program in December 2018 to force more participating ACOs to take financial accountability for their population’s expenditures.

The program now dubbed Pathways to Success currently has 37 percent of its organizations in these downside risk tracks versus fewer than 10 percent during the 2012 through 2017 performance periods.

This could spell trouble for vulnerable patients despite the belief that downside risk adoption will lead to greater quality improvements and lower spending.

“[A] growing minority of MSSP contracts have downside risk, meaning that ACOs that fail to meet the financial benchmark share losses. As that incentive structure becomes more common, there may be changes in physician group behavior related to vulnerable patients,” researchers stated.

Researchers advised industry leaders to continue monitoring for changes in behavior that could reduce access for vulnerable patient populations, as well as for quality improvements for these patients.