What do Brief Substance Use Screening Instruments Actually Measure? A Stratified Meta-Analysis in Correctional Populations

Abstract

Substance use disorders are highly prevalent in adult correctional and forensic populations. However, brief screening instruments are often interpreted without clear differentiation between diagnostic, validation, and predictive purposes. In this study, we synthesize evidence on commonly used substance use screening tools using a stratified meta-analytic framework designed to clarify their legitimate inferential roles in custodial settings. Evidence was organized into three analytic tiers: Tier 1 (CORE: Diagnostic Accuracy) included studies permitting formal estimation of diagnostic accuracy against explicit clinical reference standards. Tier 2 (Extended Forensic Validation) comprised extended forensic validation studies employing context-specific or severity-based frameworks. Tier 3 (Predictive Validity) addressed predictive validity for substance-relevant post-release outcomes. Quantitative synthesis was restricted to Tier 1 studies and indicated high sensitivity with moderate specificity for brief screening instruments when evaluated against structured diagnostic assessments. Given the limited number of eligible studies, these pooled estimates should be interpreted as preliminary indicators rather than stable population parameters. Tier 2 studies demonstrated broadly consistent performance across diverse forensic contexts but substantial heterogeneity in reference standards, precluding pooled diagnostic inference. Limited Tier 3 evidence suggested that screening-derived severity classifications may be associated with substance-relevant post-release outcomes. Overall, the findings indicate that brief screening instruments support distinct, tier-specific functions. Evidence for one inferential purpose should not be generalized to others.