Dissertations & Theses (Open Access)

Date of Award

Spring 4-2020

Degree Name

Doctor of Philosophy (PhD)

Advisor(s)

James R. Langabeer Edd Phd

Second Advisor

Lee Revere Phd Ms

Third Advisor

Tiffany Champagne-Langabeer Phd Mba

Abstract

The opioid crisis in the United States was declared a public health emergency due to escalating and untoward human, financial, and systemic consequences and effects on the nation. Opioid use disorder (OUD) comprising opioid abuse and dependence is devastating because of its associated chronic relapsing nature, overutilization of healthcare services, rising morbidity and mortality rates, and high cost of care. Efforts to address this have not made significant positive impacts. It is thus imperative to reassess the influence of factors associated with OUD. This study answered the question, what patient-, hospital-, and state-level policy factors were associated with prevalence of diagnosing and treating OUD in U.S. emergency departments (ED), since the ED which were usually first point-of-contact with the healthcare system by patients with OUD witnessed significantly increased visits related to nonmedical use of opioids. A retrospective secondary data analysis of the cross-sectional Nationwide Emergency Department Sample of patients 12 years and older from January 1 to December 31, 2016; ASAM state reports; SAMHSA Office of Policy, Planning and Innovation State Medicaid coverage reports; and KFF report on opioid epidemic was performed. Outcome variable was prevalence of diagnosing and treating OUD in the ED. Primary predictor variable was OUD condition, and covariates included, patient characteristics – primary payer, annual median income, patient location, and ED event; hospital characteristics – control/ownership, region, and designation; and state-level policy characteristics – medication-assisted treatment (MAT) policy, MAT medication coverage, Medicaid expansion, and Medicaid section 1115 behavioral health waiver statuses. Descriptive statistics was reported for all variables. Pearson’s chi-squared was test used to determine statistically significant differences between opioid abuse and opioid dependence diagnosis. Hierarchical linear regression model (HLM) was used to estimate association between outcome and predictor variables. In total, 32,680,232 ED visits in 953 hospitals across 35 states and District of Columbia which when generalized to the entire United States amounted to 144,842,742 visits to the ED in 4,639 hospitals across the 50 states including the District of Columbia were analyzed. The total number of opioid-related incidents to the ED was 1,623,490. The overall prevalence of any opioid-related incident was 1.12% while overall prevalence of diagnosis and treatment of uncomplicated OUD in U.S. ED was 0.5%. Significant regional disparities existed in state-level opioid policies, prevalence of uncomplicated OUD and other characteristics influencing treatment of OUD in U.S. ED. Opioid dependence patients (55.6%) were preponderantly of upper-lower income class, micropolitan residents, covered by Medicare; admitted to same hospital they presented, attended to largely in privately owned not-for-profit ED, in micropolitan areas, and in Southern and Western U.S. Opioid abuse patients (44.4%) were predominantly of lower-lower income status, metropolitan dwellers, Medicaid covered; presented commonly to privately-owned not-for-profit ED, in metropolitan locations, and in Northeastern and Midwestern U.S. Combined, patient and hospital-level policy characteristics accounted for 25.4% (R2=0.254, Adj. R2=0.254, F change (3,734618)=31937.906, p<0.0001) of variance in prevalence of treating OUD in ED. Patient characteristics only accounted for 15.6% (R2=0.156, Adj. R2=0.156, F(5,734621)=27245.686, p<0.0001) and hospital characteristics only for 9.7% (R2 change=0.097, F(3,734618)=31937.906, p<0.0001) of the variance. Proportion of variance accounted for by each predictor variable was, control/ownership of hospital (9.67%), patient location (6.35%), annual median income (1.44%), hospital designation (1.21%), OUD diagnosis (0.20%), primary payer (0.04%), region of hospital (0.02%), and ED event (0.008%). Patient and hospital level characteristics significantly influenced prevalence of treating OUD in U.S. ED. Hospital-level characteristics contributed more that patient-level characteristics. A socioecological approach, which ensures an integrated and holistic method, is required to understand factors influencing OUD with the view to developing innovative policies and programs that can positively and significantly address the opioid crisis.

Download

Share

COinS