MEDICARE

Research Objective Lower extremity joint replacements (LEJR) are among the most common and expensive inpatient surgeries for Medicare beneficiaries. The Comprehensive Care for Joint Replacement (CJR) Model of 2016 is a bundled‐payment initiative for LEJR that was mandated for hospitals in 67 Metropolitan Statistical Areas (MSA) covering the episode of acute and postacute care (PAC). It provides financial incentives for participating hospitals to improve outcomes and reduce episode cost of care. Approximately 30% of LEJR beneficiaries receive institutional PAC (ie, care in skilled nursing facility [SNF], inpatient rehabilitation center, or long‐term hospital) after an inpatient stay. This study aims to determine the impact of CJR on (1) institutional PAC use, readmission rate, and SNF care outcomes for patients discharged to SNFs (the most common type of institutional PAC); and (2) existing gaps in these measures between Medicare‐only and dual‐eligible (Medicare beneficiaries who are also eligible for Medicaid) patients. Study Design We conducted difference‐in‐difference models with MSA sampling weights using Medicare claims and assessment files, and hospital and PAC facility datasets from 2013 to 2017. We compared patient outcomes in hospitals of 75 initially randomly assigned MSAs (participation in CJR mandated) and outcomes of hospitals in 120 control areas. In analyses on all LEJR patients, the outcomes included institutional PAC utilization rate, all‐cause 30‐/90‐day readmission rates, and total payments for readmissions. In analyses on patients discharged to SNF, outcomes included SNF five‐star ratings, rate of successful community discharge, rate of transition to long‐stayers, SNF length of stay (LOS), and SNF total spending. All models adjusted for important patient, facility, and market characteristics. Population Studied 457,472 patients from 763 CJR‐hospitals and 597,764 patients from 924 control hospitals during 2013‐2017. Principal Findings CJR was associated with a 2.35‐percentage‐point ( P  < .001) decrease in institutional PAC utilization, and a 1.74‐percentage‐point ( P  = .001) and a 1.97‐percentage‐point ( P  < .001) decline in 30‐day and 90‐day readmission rates, respectively. The corresponding reductions in total payments for readmissions were $629.89 ( P  = .003) and $676.74 ( P  = 0.006), respectively. Among patients discharged to SNF, the CJR policy was associated with a 4.58‐percentage point ( P  = .002) higher rate of discharge to 4‐ or 5‐star SNFs, 0.76 days of reduction ( P  = .001) in SNF LOS, and $481.82 ( P  = .019) less SNF care spending. Compared to Medicare‐only patients, dual‐eligible beneficiaries with partial benefits had $1911.53 ( P  = .017) lower in total payments for 30‐day readmission and had $2545.08 ( P  = .040) lower in aggregate spending for 90‐day readmission in the treatment hospitals, under the CJR. No significant relationships were found between CJR and other SNF outcomes, or between CJR and disparities in other outcomes examined in this study between Medicare‐only and dual‐eligible LEJR patients. Conclusions The mandatory CJR bundled payments reform led to reduced use of institutional PAC and readmission rates, but had no effect on important SNF outcomes, such as successful discharge to community, for LEJR patients initially discharged to SNF. Furthermore, the CJR had no effect on disparities in postdischarge outcomes associated with Medicaid insurance. Implications for Policy or Practice Future Medicare alternative payment models and revisions to the mandatory CJR and other voluntary initiatives could incorporate incentives for improved SNF outcomes and equality of care. Primary Funding Source National Institutes of Health.