Conducting Economic Evaluations Alongside Substance Use Disorder Clinical Trials: Statistical Methods and Best Practices | AHSR 2020

On Monday, January 11, 2021, CHERISH Research Affiliate and Pilot Grantee Ali Jalali, PhD, alongside Methodology Core Director Kathryn E. McCollister, PhD and Consultation Service Director Sean M. Murphy, PhD, conducted a workshop entitled “Conducting Economic Evaluations Alongside Substance Use Disorder Clinical Trials: Statistical Methods and Best Practices” at the 2020 virtual Addiction Health Services Research (AHSR) Conference.

This CHERISH (Center for Health Economics of Treatment Interventions for Substance Use Disorder, HCV, and HIV) sponsored seminar provides an in-depth review of methods for conducting an economic evaluation alongside a clinical trial. Statistical methods frequently employed in applied health econometric studies (two-part, and generalized linear models, recycled predictions, mixed effects, longitudinal analysis, etc.), and best practices for data collection and analysis are discussed, with examples using Stata statistical software. This seminar is designed for investigators who have some familiarity with economic evaluations, but would benefit from a more detailed outline of the methods involved and examples of their application.

Slides are available here.

Criminal Justice Measures for Economic Data Harmonization in Substance Use Disorder Research

The intersection of criminal justice-involved populations and people who use substances makes criminal justice outcomes particularly significant for estimating the economic impact of substance use disorder interventions. New National Institute on Drug Abuse (NIDA) HEAL initiative funding opportunities in response to the opioid crisis include integrated studies that will develop, test, and validate evidence-based approaches to preventing and reducing opioid use disorders and reducing opioid overdose fatalities. One of these initiatives is the Justice Community Opioid Innovation Network (JCOIN), a national consortium of investigators developing and implementing effective approaches to addressing opioid use disorder in local and state justice systems. This initiative will fund clinical research centers, a Methodology and Advanced Analytics Research Center and a Coordination and Translation Center, collectively working to address gaps in the continuum of care for justice-involved populations in the United States.

To support integrative data analysis, multisite comparative effectiveness studies, and economic data harmonization opportunities in HIV, HCV, and SUD research, CHERISH Methodology Core Co-Director Kathryn McCollister, PhD led a study that examined baseline data from six NIDA-funded Seek, Test, Treat and Retain (STTR) studies, focusing on criminal justice outcomes. Data harmonization applies common measures in order to improve the quality and comparability of data across independent studies which can be synthesized in order to promote more rigorous and generalizable analyses of the impact of an intervention, program, or policy. NIDA considers data harmonization research in substance use disorders, HIV, HCV, and other related disorders/diseases to be a high priority research area.

The primary goal Dr. McCollister’s study was to understand the extent to which specific outcome measures were represented across all selected studies that are amenable to cost-effectiveness or cost-benefit analyses. She created a database of select crime/criminal justice measures which were represented in two or more of the six studies, sorting them into four categories used by the Bureau of Justice Statistics and FBI Uniform Crime Reporting System: property crimes, violent crimes, public order crimes, and enterprise crimes.

All six studies measured the number of times arrested and days of incarceration over different timeframes, and at least four also collected data on arrests and criminal offenses. Based on the available data in these STTR studies, Dr. McCollister and colleagues recommend four crime/criminal justice measures for data harmonization projects: number of arrests, number of convictions, days of incarceration, and times committing criminal offenses. These outcomes can be valued in dollars using monetary conversion factors (MCFs) found in published studies, government reports, and national data sets. Criminal justice administrative databases may serve as supplementary sources of economic data in SUD research.

Extreme variability in crime/criminal justice measures exists across jurisdictions and can be addressed in part by pooling individual data across studies and grouping by offense classification. Future data harmonization initiatives should address timeframe harmonization for more meaningful comparison across studies.

Dr. McCollister said of the implications of translating criminal activity measures into dollars, “estimating the economic impact of SUD and HCV/HIV interventions through reductions in crime or criminal justice contacts broadens the context for evaluating clinical effectiveness and provides policy makers and other stakeholders evidence to support efficient allocation of already-stretched budgets to public health programs.”

This article is available open access in Health & Justice. Co-authors include CHERISH investigators Sean M. Murphy, PhD and Bruce R. Schackman, PhD, CHERISH staff April Yang, MPH, MHS and Jared A. Leff, MS, and CHERISH Research Affiliate Daniel J. Feaster, PhD.