Value-Based Care News

Advanced Bundled Payment Model Reduced Payments for High-Risk Patients

Participation in the BPCI-A helped reduce Medicare payments for high-risk patients and was associated with a $1,412 decrease in costs per episode for frail patients and a $1,045 reduction for patients with multimorbidity.

Medicare payments, high-risk patients, bundled payment model

Source: Getty Images

By Victoria Bailey

- Hospital participation in Medicare’s Bundled Payment for Care Improvement Advanced Model (BPCI-A) decreased Medicare payments and did not negatively impact health outcomes for high-risk patients, a Health Affairs study found.

CMS launched the BPCI-A in October 2018 as a voluntary Alternative Payment Model (APM) in which hospitals are held accountable for the costs and outcomes of 90-day episodes of care. The model has been associated with reductions in Medicare payments per episode and improved clinical outcomes.

However, hospitals must either decrease utilization or attract a less sick patient population to meet spending targets, raising concern about the model’s impact on patients with substantial healthcare needs.

For example, some stakeholders worry that hospitals and physician groups may eliminate necessary rehabilitative care or discourage enrollment of patients with medical vulnerability.

Researchers used Medicare claims data from January 2017 to September 2019 to determine how BPCI-A participation impacted Medicare spending and clinical outcomes for patients with frailty, multimorbidity, and dual enrollment in Medicare and Medicaid.

January 2017 to September 2018 was considered the pre-period, while October 2018 to September 2019 was the program period.

The final sample included 811 hospitals participating in BPCI-A and 811 matched comparison hospitals, representing 6.8 million high-risk patients.

Patients with frailty, multimorbidity, and dual enrollment had higher Medicare payments per episode in the pre-period compared to low-risk patients. But payments decreased during the program period.

From the pre-period to the program period, Medicare payments at BPCI-A hospitals decreased by $1,412 per episode for frail patients. At comparison hospitals, payments fell by $1,209 per episode, indicating a change of -$204.

Medicare payments for non-frail patients declined by $631 at BPCI-A hospitals and by $550 at comparison hospitals, for a differential change of -$81.

Medicare payments fell by $1,045 for patients with multimorbidity at BPCI-A hospitals and decreased by $899 at comparison hospitals. Among patients without multimorbidity, payment decreased by $787 and $663 at BPCI-A and comparison hospitals, respectively.

The differential change per episode was -$147 for patients with multimorbidity and -$124 for patients without

The differential change for dually enrolled patients was -$367 per episode compared to -$66 per episode for non-dually enrolled patients. The difference in per episode payment at BPCI-A hospitals relative to comparison hospitals was -$301 for dually enrolled compared to non-dually enrolled patients.

In addition to reducing Medicare payments, the BPCI-A did not negatively impact health outcomes for high-risk beneficiaries. Readmission rates differed among high-risk and low-risk patients, but there were no differential changes between patients at BPCI-A hospitals and those at comparison hospitals.

Similarly, there were no differential changes in mortality related to BPCI-A participation for frail, multimorbidity, and dually enrolled patients.

Meanwhile, BPCI-A participation was associated with an increase in healthy days at home for frail and non-frail patients, with a differential change of 0.24 days per episode for frail patients and 0.15 days per episode for non-frail patients relative to comparison hospitals.

BPCI-A was also associated with an increase in health days at home for patients with (0.15 days) and without multimorbidity (0.22 days) and for dually (0.38 days) and non-dually enrolled patients (0.15 days).

Despite concerns about the BPCI-A and its impact on patients with medical or social vulnerability, the study findings indicate that the model reduced Medicare payments per episode and did not adversely affect health outcomes for this population.

“These findings suggest that programs that explicitly incentivize clinicians to develop interventions for high-risk populations may have the potential for improving outcomes and reducing costs of care,” researchers concluded.