Association for Public Policy Analysis & Management

In 2011, the Centers for Medicare and Medicaid Services (CMS) mandated that Medicare Advantage plans charge no more than $1,000 for the first 20 days of care in a skilled nursing facility (SNF). This cost-sharing cap was intended to ensure access to post-acute nursing care and prevent high out-of-pocket costs from prompting disenrollment among sicker beneficiaries. We evaluated how Medicare Advantage beneficiaries’ use of SNF services and patterns of disenrollment changed following the introduction of cost-sharing caps.

Because no national database details health care utilization for all Medicare Advantage plans, we linked the Medicare Beneficiary Summary File, Medicare Advantage enrollment records and benefit data, and Minimum Data Set assessments that captured SNF use across all payers. We identified 27 Medicare Advantage plans that reduced cost-sharing from 2010 to 2011 to meet the CMS caps. We compared trends in these plans’ monthly SNF utilization for 2008-2012 to concurrent trends for 21 plans that already had cost-sharing levels below the new cap. Using time-series regression analysis, we tested whether the level and slope of SNF utilization trends changed between the three-year pre-cap period and two-year post-cap period. We also examined whether disenrollment from plans changed following the implementation of the new cap. Our study population included 107,665 beneficiaries in plans with mandated cost-sharing reductions and 112,863 beneficiaries in plans that already met cost-sharing requirements.

In plans that were required to reduce cost-sharing, cost-sharing for the first 20 days of SNF care decreased from an average of $2,039 in 2010 to $991 in the first year of the cost-sharing cap. In adjusted analyses, plans with mandated cost-sharing reductions averaged 156.8 days (95% CI: 152.0, 161.7) per 1,000 members in the three years leading up to the cost-sharing cap. The monthly number of SNF days per 1,000 members increased by 14.3 days (95% CI: 3.8, 24.8, p<0.01), or about 10% in relative terms, following the implementation of the cap. We found no indications that the rate of change of SNF utilization (i.e. the slope of utilization trends) differed when cost-sharing caps were in place.

However, increases in SNF utilization did not significant differ between plans with and without mandated cost-sharing reductions. Results were suggestive, but not statistically significant, for beneficiaries age 80 and above. In plans with mandated cost-sharing reductions, the monthly number of SNF days per 1,000 members age 80 and over increased by 32.5 days (95% CI: 10.5, 54.5). This increase was 23.9 days greater (95% CI: -4.5, 52.4, p=0.10) than the increase seen among members age 80 and above in plans without mandated cost-sharing reductions. Disenrollment patterns did not change after the cap was implemented.

We did not find evidence that capping SNF cost-sharing amounts significantly increased SNF utilization among MA members, with the potential exception of MA members age 80 and above. These findings suggest that the CMS policy significantly reduced the burden of out-of-pocket costs for post-acute SNF care among MA beneficiaries without evidence that these mandated cost-sharing reductions increased the use of SNF care.