Research examining Medicaid data from 78,795 individuals who experienced cocaine and methamphetamine overdoses between 2016 and 2020 identifies specific patterns of vulnerability. The findings highlight both individual characteristics and community-level factors that predict overdose risk, offering critical information for prevention efforts.

The Scale and Scope of Stimulant Overdose Risk

The study documented four distinct categories of stimulant-involved overdoses requiring hospitalisation or emergency department treatment. The largest group comprised 34,498 cases of methamphetamine, ecstasy, or other psychostimulant overdoses without opioid involvement. Cocaine-involved overdoses without opioid involvement represented 29,206 cases, whilst 8,459 cases involved cocaine with opioids, and 6,632 involved other stimulants with opioids.

According to the research findings, drug overdose deaths involving methamphetamines increased five-fold from 2015 to 2022, whilst overdose deaths involving cocaine increased three-fold during the same period. In 2022, approximately 26% of drug overdose deaths involved cocaine and 33% involved other psychostimulants with misuse potential such as methamphetamine.

Individual Risk Factors: Previous Diagnoses as Key Predictors

The research demonstrated that previous substance use disorder diagnoses served as the strongest predictors of future overdose risk. For cocaine-involved overdoses with opioid involvement, previous opioid use disorder (OUD) diagnosis emerged as the most significant risk factor. The study’s predictive models showed excellent performance, with Harrell C statistics ranging from 0.868 to 0.923, indicating strong to excellent discriminative ability.

Mental health conditions also played a substantial role in predicting overdose risk. The research identified bipolar disorder, depression, and schizophrenia diagnoses as significant predictors across multiple overdose categories. Additionally, individuals with previous stimulant-related disorder diagnoses faced substantially elevated risk for methamphetamine and other psychostimulant overdoses.

The median time to overdose across all categories ranged from 173 to 175 days, highlighting the relatively short timeframe in which risk manifests. This finding underscores the urgency of early identification and intervention strategies.

Community-Level Factors That Amplify Stimulant Overdose Risk

The study revealed that area-level socioeconomic indicators significantly contributed to overdose risk prediction. For cocaine-involved overdoses without opioid involvement, the Gini index—a measure of income inequality—emerged as one of the top five predictive factors. Areas with greater income disparity showed elevated risk levels.

Housing instability indicators proved particularly significant. The research found that the proportion of renter households spending at least 35% of their income on rent was amongst the top predictors for cocaine overdoses. For methamphetamine overdoses without opioid involvement, the percentage of population living in crowded housing units ranked as the highest contributing factor to the predictive model.

Areas with higher percentages of residents living with disabilities showed increased risk across multiple overdose categories. This factor ranked as the second-highest predictor for methamphetamine overdoses both with and without opioid involvement, highlighting the intersection of disability and substance-related vulnerabilities.

Demographic Patterns and Risk Distribution

The study population had a mean age of 42.2 years, with 58% male and 42% female individuals. The research showed that over 75% of predicted overdose cases fell within the highest two risk deciles across all four overdose categories, demonstrating the models’ ability to identify those at greatest risk effectively.

Geographic distribution varied significantly, with 50.2% of methamphetamine overdose cases occurring in Western states during the development period (2016-2019). The research also noted that more than half of amphetamine-related hospitalisations in the United States occur amongst Medicaid-insured persons, highlighting this population’s particular vulnerability.

Implications for Prevention Strategies

The predictive models developed in this research achieved Brier scores ranging from 0.028 to 0.129, indicating good to excellent accuracy in risk prediction. These findings suggest that readily available Medicaid and community-level data can effectively identify individuals at highest risk for stimulant overdoses.

The strong predictive value of previous substance use disorder diagnoses indicates that individuals with documented histories require enhanced support and comprehensive prevention approaches. Early intervention for those showing initial signs of substance use challenges could prevent progression to overdose events.

The significance of structural factors—including income inequality, housing instability, and community resources—demonstrates that effective prevention must address broader social determinants of health. Prevention strategies focusing solely on individual behaviour whilst ignoring these structural factors may prove insufficient.

Moving Forward: Evidence-Based Prevention Approaches

The research findings support the development of multi-level prevention strategies that address both individual and community risk factors. Given that mental health conditions emerged as significant predictors, integrated approaches addressing both mental health and substance use concerns simultaneously appear essential.

Communities with higher structural vulnerabilities, particularly those with significant income inequality and housing instability, require targeted prevention resources. The research suggests that investment in community infrastructure, affordable housing, and economic stability could create protective environments that reduce overdose risk.

The study’s ability to predict risk with high accuracy—particularly the finding that over three-quarters of overdose cases fell within the highest-risk groups—offers valuable opportunities for targeted prevention. By identifying those at greatest risk before overdose events occur, prevention programmes can allocate resources more effectively and intervene at critical moments.

Source: JAMA Network