Longitudinal Patterns and Economic Consequences of Emergency Department Visits among Medicaid Enrollees
Date of Graduation
School of Pharmacy
Thomas K Bias
Objective.;The objective of the dissertation was to examine the patient- and county-level factors associated with the Emergency Department (ED) visits and economic consequences associated with persistent ED use among adult fee-for-service (FFS) Medicaid beneficiaries. The first study examined the patient- and county-level factors associated with the number of ED visits and the second study examined the longitudinal patterns of ED visits among FFS Medicaid beneficiaries. Further, in both the studies ED visits due to primary care sensitive conditions were also examined. The third study examined the patient- and county- level factors associated with persistent ED use followed by an estimation of the excess healthcare expenditures associated with persistent ED use.;Methods.;Both cross-sectional and longitudinal study designs were implemented using a retrospective observational claims data of Medicaid beneficiaries residing in Maryland, Ohio, and West Virginia. Study population included adult, alive, FFS, not dually enrolled in Medicare, non-pregnant and continuously enrolled Medicaid beneficiaries. Data on patient-level factors were obtained from the Medicaid Analytic eXtract (MAX) files for the years 2006-2010. MAX files consisted of personal summary, other therapy, inpatient and prescription drugs claims. The personal summary file included demographics, Medicaid eligibility, county federal information processing standard (FIPS) codes, Medicaid managed care enrollment, and Medicare eligibility status. The inpatient claims file included information related to hospital stays, dates of service, Medicaid payment, and the International Classification of Disease, Ninth Revision, Clinical Modification codes (ICD-9-CM) and ICD-9-CM procedure codes. The other therapy claims file included information on dates of service, types of service, Medicaid payment, ICD-9-CM, and Current Procedural Terminology (CPT) codes. The prescription drugs claims file included information on the date of prescription filled, days supplied, Medicaid payment and national drug code (NDC). All these files were linked using encrypted identification numbers. Data on county-level factors such as socio economic status, healthcare resources, and obesity rates were obtained from the Area health resource and county health ranking files. Frequencies, means, inter-quartile range, and 90th percentile were used to examine the characteristics of the study population and distribution of ED visits. In the first study, unadjusted and adjusted negative binomial regressions (NBR) were conducted to examine the patient- and county-level factors associated with the number of ED visits. In the second study, multivariable hurdle models with logistic and NBRs were used to analyze ED visits over time, after adjusting for all other independent variables. In the third study, chi-square tests and logistic regression was conducted to examine the patient- and county-level factors associated with persistent ED use. Further, adjusted generalized linear models with log link function and gamma distribution were conducted to examine the excess expenditures. All analyses were conducted using STATA version 14.0.;Findings.;In the first study, it was observed that more than half of the study population had one or more ED visit. Patient-level factors such as complex chronic illness, fragmented primary care use, poly-pharmacy, and tobacco use were associated with higher number of ED visits. Residents in counties with higher number of urgent care centers had lower number of ED visits. Almost, half of the ED visits were preventable. In the second study, the likelihood of ED use did not change from year to year. However, among ED users, the estimated number of ED visits increased over time with a small magnitude. More than half of the ED visits were primary care sensitive in each panel year. In the third study, one in ten Medicaid beneficiary had persistent ED use i.e. they had 4 or more ED visits in both index and follow-up years. There were significant differences between persistent ED users and non-users in patient- and county-level characteristics. Individuals with complex chronic illnesses, fragmented primary care use, poly-pharmacy and tobacco use were more likely to be persistent ED users. In multivariable regression, persistent ED users had significantly higher total healthcare expenditures as compared to non-users.;Conclusions.;Adult FFS Medicaid beneficiaries with complex healthcare needs had higher number of ED visits. The number of ED visits increased over time with a small magnitude. Almost, half of the ED visits are preventable with timely care. Medicaid beneficiaries also had persistent ED use and had higher excess healthcare expenditures associated with persistent ED use. Taken together, these findings suggest that only access to primary care may not reduce ED visits. There is a need to have targeted interventions focused on this particular subgroup of the population who is consuming higher healthcare resources as compared to others. Cost containment may be achieved by providing comprehensive care management to individuals with complex healthcare needs. Access to county-level resources such as urgent care centers may contribute in reducing the number of ED visits and cost containment as care provided in these settings is less expensive as compared to ED.
Agarwal, Parul, "Longitudinal Patterns and Economic Consequences of Emergency Department Visits among Medicaid Enrollees" (2015). Graduate Theses, Dissertations, and Problem Reports. 5032.