Healthcare providers’ responses to healthcare delivery reform

Authors:

Abstract
Facing the escalating cost and unsatisfying quality of care, the Accountable Care Organization model was designed to as a remedy for the fragmented US healthcare system. Under the Patient Protection and Affordable Care Act of 2010, CMS established the Medicare Shared Savings Program (MSSP) as an alternative to the traditional Fee-for-Service payment model. However, little empirical evidence exists on the extent to which MSSP changes how participating providers deliver healthcare services. Using a mixture of models aimed at causal inference and national provider-level datasets, I examined MSSP effects on primary care physicians’ (PCP) provision of preventive care, hospital financial sustainability, and nursing homes’ (NHs)’ staffing level and care deficiencies. The first study estimated how MSSP participation affected PCP preventive service delivery. To extends the literature on preventive care and primary care delivery in the context of VBP transition, I utilized a Difference-in-Differences (DD) model comparing within-provider changes over time between participating and non-participating physicians and controlling for demographic and health characteristics of served patients. PCPs who participated in MSSP had an increase in the likelihood and the volume of services delivered for most preventive services included in MSSP quality performance metrics except cancer screening. The second study assessed the impact of hospitals’ participation in MSSP on their financial performance. I used a DD model and an event study design to analyze 2011-2018 hospital cost reports. I found that MSSP participation was associated with small increases in net patient revenue and Medicare revenue share. No significant effects were found on outpatient and inpatient revenue, operating margin, and inpatient revenue share. The third study examined the potential influence of MSSP-affiliation on NH staffing level and deficiency citations. I analyzed the NH Compare data in the national universe of Medicare NHs and relied on the DD model and Instrumental variable approach to provide the causal estimates of MSSP effect. The IV estimates suggest that after MSSP participation, NHs increased registered nursing (RN) staffing but reduced licensed practical nurse staffing, which led to the growth of 2.02 percentage-points of RN in total staffing. NH participation in MSSP was related to an increase of 3 counts in total deficiency citations. This dissertation offers novel evidence on how MSSP may affect healthcare providers and organizations by using a mixture of causal inference models; and it would assist policymakers in understanding healthcare providers’ responses to policy reforms and refining future program design.