Little Impact of Medicare ACOs on Low- and High-Value Revascularization Decisions

Reassuringly, cardiologists don’t appear to be altering their usual practice, but new models may be needed to “bend the cost curve.”

Little Impact of Medicare ACOs on Low- and High-Value Revascularization Decisions

Despite concerns that Accountable Care Organizations (ACOs) could negatively impact clinical care, new research suggests that patients for whom coronary revascularization is considered to be of either low or high value are equally unaffected by their providers’ participation in a network designed to reduce healthcare spending.

“It doesn’t look as if ACOs as they are currently designed are having much effect on upfront treatment decisions,” said the study’s lead author, John M. Hollingsworth, MD (University of Michigan Medical School, Ann Arbor), in an interview with TCTMD. “We found that both low- and high-value groups were continuing on the same secular trend as they had been prior to implementation.”

For the study, Hollingsworth and colleagues looked at 298 provider groups that participated in a Medicare ACO contract at some point during the study period (2008-2014) and 1,329 provider groups who did not participate in an ACO. Medicare beneficiaries were divided into two cohorts. In one, patients had evidence of stable, asymptomatic ischemic heart disease without concomitant angina, congestive heart failure, or recent acute MI hospitalization. The value of coronary revascularization was considered to be lower for this cohort. The other cohort consisted of patients admitted in a given year with a primary diagnosis of acute MI. For this cohort, the value of coronary revascularization was considered to be high.

“There’s a huge concern about healthcare spending in this country, and ACOs are one of many approaches that are being tried to change incentives and transform the way we give care,” said Michael Esman Chernew, PhD (Harvard Medical School, Boston MA), who was not involved in the study. He told TCTMD the bottom line regarding the findings is that despite changing payment structures, cardiologists are still doing their job in pretty much the same way they always have.

“People want to know how the system is changing, for obvious reasons, and this study tells us a little bit about that,” Chernew said, adding that it is reassuring in the sense that such things as reduced access to care and incentive not to provide needed care do not go hand-in-hand with participating in an ACO.

The study was published online last week in Circulation: Cardiovascular Quality and Outcomes.

Slight Increase Seen in Higher-Value Procedures

Among the participating provider groups in the study, 53.7% included hospital partners and 62% were concurrently in a commercial ACO, but only 7.1% of those also participated in a Medicaid ACO. ACO participants were more likely than those who did not participate to practice at larger, not-for-profit, and teaching hospitals, and they were concentrated disproportionately in the Northeast and Midwest regions.

When the researchers examined rates of revascularization between nonparticipating and participating provider groups prior to joining ACOs, there was no difference between lower-value and higher-value cohorts. Once ACO contracts began, rates of lower-value revascularization remained unchanged between participating and nonparticipating providers (from 1.0 per 100,000 beneficiaries per year in 2008 to 1.0 in 2014). Rates of higher-value revascularization, on the other hand, rose by an average of 13.5% (from 51.5 per 100,000 beneficiaries in 2008 to 58.5 in 2014), but the difference was not statistically significant.

Multivariable analysis confirmed that the degree of change for low- and high-value coronary revascularization was not affected by participation in a ACOs (lower value: difference, −0.04 per year; 95% CI −0.11 to 0.03; higher value: difference, 0.96 per year; 95% CI −0.46 to 2.4). The findings remained similar when rates of PCI and CABG were examined individually.

Sensitivity analysis found no association between when providers joined an ACO during the study period (early or late) and revascularization rates. Other factors, including group organizational structure, degree of financial risk, and commercial or Medicaid ACO participation, also did not change the effect of ACOs on lower- or higher-value revascularization.

To TCTMD, Hollingsworth said that although there was a slight indication of a learning curve associated with joining an ACO, as evidenced by a reduction in lower-value revascularization and PCI in year 2 after joining, the trend did not continue beyond that point.

In exploratory analysis, there was a suggestion that provider groups with high specialist participation had greater decrease in lower-value revascularization and increased rates of higher-value revascularization, but neither of these were statistically significant.

Looking for Explanations

Hollingsworth said the initial hypothesis of the study was that there would be change in lower-value revascularizations after joining an ACO. He and colleagues suggest that one explanation for the lack of effect may be that these procedures are already relatively infrequent.

“As such, ACO leaders may have chosen instead to focus on higher-prevalence, lower-value activities where more room for improvement exists. An alternative explanation has to do with the fact that participating provider groups, and particularly specialists, have too little skin in the game,” they write in their paper.

Chernew said another way to look at it could be that the structure of ACOs haven’t changed incentives enough to make an impact on providers. Yet another explanation could be that coronary revascularization is too narrow a clinical area to evaluate when gauging the impact of ACOs.

Overall, Hollingsworth said the findings are reassuring that the high-value revascularizations are continuing to be performed, but also show that there is room for improvement.

“Given the role that specialty care plays in healthcare spending, if you really want to bend the cost curve, you need to target specialty care in some way. Our data would suggest there probably needs to be more involvement of specialty care in new iterations of these models, otherwise we may struggle to control what specialists physicians are doing,” Hollingsworth said.

Sources
Disclosures
  • Hollingsworth and Chernew report no relevant conflicts of interest.

Comments