A congressional advisory panel tasked with developing a unified post-acute value incentive program is struggling with how to set fair benchmarks across healthcare settings, given differences in patient conditions as well as socio-economic needs that may vary by geography.
The Medicare Payment Advisory Commission on Friday met to discuss its upcoming mandated report on the value incentive program and reviewed various metrics. Staff has spent months reviewing data points that could be tied to outcomes to adjust reward incentives for providers who face steeper challenges in achieving performance goals.
The post-acute care value incentive program (PAC VIP) would apply to all skilled nursing, long-term care hospital, inpatient rehabilitation and home health settings in a manner “consistent with” the proposed SNF value incentive program.
Models presented to the advisory commission Friday rated performance based on hospitalizations, Medicare spending per beneficiary, and successful discharge to the community.
It also calculated ways to create peer groups by which performance rewards would be meted out. Providers could be judged on their share of dual-eligible residents or on socio-economic characteristics associated with the ZIP codes in which patients live.
“Peer grouping would help counter disadvantages some SNFs face,” said Carol Carter, the models’ co-author.
But one challenge of using either peer group suggested is that neither dual eligibility nor ZIP code association is tied to worse performance in long-term care hospitals, meaning it couldn’t be applied across the value incentive program.
One member pointed out that the ZIP code measure, the Area Deprivation Index based on census data and developed at the University of Wisconsin, wouldn’t necessarily capture challenges in rural areas. The measure tracks 17 social risk factors, among them car ownership (for access to appointment and prescription pick up) and phone use (for follow-up care and scheduling).
But Lynn Barr, executive chairwoman of Kansas City’s Caravan Health and formerly on the Government Affairs Committee of the National Rural Health Association, noted that “everyone” in rural settings has a car and phone. The measure might not reflect challenges in those areas, she said, asking instead for staff to consider federal Quality Improvement Organization interventions as a basis for peer grouping.
“The statistical issues here are enormous in a whole range of ways,” said commission Chairman Michael Chernew, Ph.D., healthcare policy professor at Harvard Medical School. “This is really hard. Issues of how to adjust for social determinants, what that means is challenging enough in one sector, let alone in a unified way.”
Staff members are continuing to look for additional ways to measure patient differences in advance of its final report, which is due to Congress in March 2022.