Date of Award


Degree Type


Degree Name

Master of Public Health (MPH)


Public Health

First Advisor

Lee Rivers Mobley

Second Advisor

Dora Il'yasova


Objectives: To assess how three channels of hospital care – Inpatient (IP), Emergency Department (ED), and Outpatient (OP) visit rate changes between 2005 and 2013 and which factors are associated with the changes.

Methods: Using nation-wide data from Area Health Resource File (AHRF) 2015-2016 Release, we compare the mean hospital utilization between 2005 and 2013 and perform paired t-test to identify significant changes over time. We perform Wilcoxon two-sample test to determine if utilization changes between Medicaid-expanding states and non-expanding states have significant difference. To explore which factors can affect the utilization rate, we perform a multiple regression with thirteen factors of interest based on available data and the Andersen conceptual framework model. We stacked our 2005 and 2013 data and add time interaction variable with Medicaid expanding status to determine if the interaction between expansion and time is associated with the utilization rate in the adjusted regression model.

Results: We find significant changes in hospital utilization rate between 2005 and 2013. There is a 23.3% decrease in inpatient (IP) days of care rate and an increase of 4.4% in emergency (ED) and 12.9% in outpatient (OP) visit rates. When we separate the counties by Medicaid-expanding status, we find a significant difference on IP, and ED utilization rate between states that did and did not expand Medicaid over time. However, after we test it again by inserting expanding status and time interaction in our adjusted regression model, we did not find any significant association with any of the channels of care.

Factors that influence the utilization rate also vary, we find that area mortality rate, female and non-English Speaker population proportion, residential segregation, poverty, income inequalities, urban areas, and hospital availability are significantly associated with increased hospital utilization.

Conclusions: There are significant hospital utilization changes in crude, unmodeled hospital utilization rates between 2005 and 2013. However, despite the concerns of hospital overcrowding in Medicaid-expanding states, we find that after adjusting for all other factors, expanding states variable alone is significantly associated with the outpatient care utilization rate (β=-413.09, p = 0.0240), but does not have any significant association with Inpatient and Emergency department utilization rate. Which implies that in the six Medicaid expansion states, the outpatient care utilization rate was significantly lower than in the non-expanding states. Furthermore, when Medicaid expansion is interacted with time, it does not show any significant association with any channels of hospital utilization rate. Thus, the fears that Medicaid expansion would crowd the system are unfounded for these six states.

Available for download on Wednesday, May 01, 2019