Stimulant Overdose in the Medicaid Population: Who is at Risk, and When Are They at Risk

The US drug overdose epidemic has grown dramatically in the past twenty years, with more than 70,000 fatal drug overdoses in 2019 alone. This growth in overdose deaths is a major contributor to the multiyear decline in US life expectancy that was seen even before the COVID-19 pandemic began. The drugs responsible for the largest increases in overdose deaths are synthetic opioids, cocaine, and other stimulants (primarily methamphetamine). In recognition of the increasing role that cocaine and other stimulants are playing in the drug overdose epidemic, the US Centers for Disease Control and Prevention (CDC) has called for increased surveillance and evidence-based prevention and response strategies to address overdoses involving these agents. While rates of stimulant use and overdose have been reported to vary by demographics, physical and mental health conditions, disability, and other factors, no prior research has linked individual-level data on demographics, disability, and social determinants of health together with granular measures derived from healthcare utilization records, with comprehensive, area-level data on social deprivation to develop knowledge about risk and protective factors for stimulant overdose. Further, no prior research has focused on stimulant overdose in Medicaid enrollees, a large, vulnerable, underserved population in whom half of all amphetamine-related hospitalizations occur. This project addresses Objective 2 of RFA-CE-21-002: to assess risk and protective factors for illicit stimulant use, use disorder, or overdose that can contribute to the development or adaptation of intervention strategies. The study will 1) develop and validate a model using both person-level characteristics (including demographic characteristics, household income, diagnoses, prescriptions, and healthcare utilization) and area-level characteristics (including a wide range of measures of socioeconomic deprivation) to identify, among Medicaid enrollees age 15 and older, who is at highest risk of an inpatient hospitalization or emergency department (ED) encounter for overdose from cocaine or other stimulants; 2) develop and validate a model to identify, among those Medicaid enrollees age 15 and older at highest risk of an inpatient hospitalization or ED encounter for stimulant overdose, when they are at highest risk; and 3) among those Medicaid enrollees age 15 and above with a prior inpatient hospitalization or ED encounter for stimulant overdose, to measure the rate of and identify risk and protective factors for a subsequent inpatient hospitalization or ED encounter for overdose from stimulants and/or opioids. The results will be useful in at least two ways. First, they will provide generalizable knowledge about the individual-level and social factors that predispose to or protect against stimulant overdose. Such etiologic factors can then be the targets of intervention at the national, state, county, and local levels to ameliorate the effects of these causes, as well as the basis of future research to better understand the underlying causal mechanisms. Second, the results can be used pragmatically to identify high-risk individuals for the purpose of targeting scarce resources for evidence-based approaches to overdose prevention.

Inclusion criteria

• Cases: Medicaid enrollees who (1) had an inpatient hospitalization or emergency department encounter indicating a stimulant overdose during 2016-2020, (2) had been continuously enrolled in Medicaid from January 1st of the year preceding the year in which the stimulant overdose occurred, and (3) were 15 years old or older as of January 1 of the year of the overdose.
• Random subcohort: Medicaid enrollees who (1) had been continuously enrolled in Medicaid for at least one calendar year during 2015-2019, and (2) were 15 years old or older as of January 1 of one year following the last year of continuous enrollment.

Exclusion criteria

• Medicaid enrollees whose data were in the format of Medicaid Analytic eXtract (MAX) in 2015
• Medicaid enrollees whose residence zip code is missing or in Puerto Rico or Virgin Islands
• Medicaid enrollees whose residence zip code does not correspond to any ZIP Code Tabulation Area (ZCTA) in the American Community Survey 5-year data