Methods for Estimating Costs for Stays at Inpatient Rehabilitation Facilities and Long-Term Care Hospitals

Abstract

ObjectiveTo describe and compare 3 methods for estimating stay-level Medicare facility (Part A) costs using claims and cost report data for inpatient rehabilitation facilities (IRFs) and long-term care hospitals (LTCHs), the 2 hospital-based postacute care providers. DesignWe calculated stay-level facility costs using different methods. Method 1 used routine costs per day and ancillary cost-to-charge ratios. Method 2 used routine and ancillary cost-to-charge ratios (freestanding IRFs and LTCHs only). Method 3 used facility-specific operating cost-to-charge ratios from the Provider Specific File. For each method, we compared the costs with payments and charges at the claim and facility levels and examined facility margins. SettingData are from 1619 providers, including 266 freestanding IRFs, 909 IRF units, and 444 LTCHs. ParticipantsThe analyses included 239,284 claims from 2014, of which 86,118 claims were from freestanding IRFs, 92,799 claims were from IRF units, and 60,367 claims were from LTCHs. InterventionsNot applicable. Main Outcome MeasuresCosts and payments in 2014 United States Dollars. ResultsFor freestanding IRFs, the mean facility stay-level costs were calculated to be $13,610 (method 1), $13,575 (method 2), and $13,783 (method 3). For IRF units, the mean facility stay-level costs were $17,385 (method 1) and $19,093 (method 3). For LTCHs, the mean facility stay-level costs were $36,362 (method 1), $36,407 (method 2), and $37,056 (method 3). ConclusionsThe 3 methods resulted in small differences in facility mean stay-level costs. Using the facility-level cost-to-charge ratio (method 3) is the least resource-intensive method. Although more resource-intensive, using routine cost per day and ancillary cost-to-charge ratios (method 1) for cost calculations allows for differentiation in costs across patients based on differences in the mix of services used. As policymakers consider postacute care payment reforms, cost, rather than charge or payment data, needs to be calculated and the results of the methods compared.