Essays on the Impact of Financial Incentives on Medicaid Service Provision and Patient Outcome

Chapter 1: “The Effect of Physician-Hospital Integration on Medicaid Access to Care and Utilization: Findings from Louisiana’s Medicaid Market” The U.S. healthcare market has seen a rapid increase of physician-hospital integration with share of hospital-employed physicians increasing from 26% to 44% during 2012-2018. However, it is largely unknown how this trend affects the Medicaid population. This study intends to bridge that gap by relating county/parish-level measures of physician-hospital integration to patient-level measure of access to care and Medicaid service utilization. It focuses on Louisiana, where the entirety of the Medicaid population is under a Fee-for-service payment scheme during the study period of 2008-2010. Among 64 counties, county-level physician-hospital integration increased from 2008 to 2010 by a mean of 1.5 percentage points, with considerable variation in changes across counties (interquartile range, -4 to 29 percentage points). I find no statistically significant relationship between greater physician-hospital integration in a county on overall access to physician services for Medicaid beneficiaries. However, I find that where Medicaid patients see physicians changes following greater availability of integrated physicians in an area, namely from office to hospital outpatient department. This suggests that greater physician integration could imply a decrease in quality of care for Medicaid patients in terms of continuity of care and patient satisfaction. Greater integration also leads to a large increase (140%) in use of imaging procedures among children. On the other hand, I find minimal impacts on use of lab tests. These results on the use of ancillary services suggest that joining providers of complementary services may lead to increase in the use of lucrative procedures. I also estimate a precise zero impact on the use of inpatient services. Lastly, I find that greater physician-hospital integration results in reduced emergency department service usage (-85%) among adults, which may suggest that there may be some efficiency gain following integration. As the trend toward integration intensifies, it will be important for policy makers to consider the consequences of greater integration that may be unique to the Medicaid population. Chapter 2: “How Do Changes in Fee-for-Service Payment Impact the Utilization of Imaging Exams? Evidence from Medicaid” Diagnostic imaging services are expensive, and overutilization of the imaging exams contribute to increasing medical expenditure in the U.S. I test whether the Fee-for-service (FFS) payment scheme drives the utilization of these lucrative services using detailed Medicaid claims data in twelve states over a five-year period (2003-2004 and 2008-2010). Using a difference-in-differences design, I find that increasing Medicaid-to-Medicare fee ratio from 0.7 to 1 is generally associated with a large and positive effect on utilization of X-ray, diagnostic ultrasound, and CT scans and a negative effect on MRI use. The effect sizes tend to be larger in outpatient hospitals than in office. I also examine whether low-value imaging services, spine imaging among back pain patients and brain imaging among headache patients, are impacted by the fee change. I also find that the fee increase is associated with a positive probability of receiving the low-value imaging services, except for spine and brain MRIs. However, none of the estimates are statistically significant at the 5%-level or precisely measured. While this study does not provide conclusive evidence on the relationship between Medicaid FFS reimbursement rates and utilization response, it offers a perspective on how to think about physician’s decision to supply care to patients in terms of income and substitution effects in a unique Medicaid environment.