Graded Response Item Response Theory in Scaling Suicidal Thoughts and Behaviors Among Trauma-Exposed Women With Substance Use Disorders

Abstract

Objective: The co-occurrence of substance use disorders (SUD) and trauma-exposure is a risk factor for suicidal thoughts and behaviors (STB). However, traditional methods of measurement for suicidal thoughts and behaviors are limited by an overreliance on dichotomous (i.e., yes or no) and averaged/summed scale score measurements. Further, among trauma-exposed individuals with SUD, it remains unclear which specific demographic factors, types of SUDs, and trauma sequelae (e.g., posttraumatic stress disorder [PTSD] symptom clusters) may be associated with elevated STB. The present study utilized item response theory to (a) generate empirically derived STB severity scores and, (b) examine which demographic factors, SUD diagnoses, and DSM-IV PTSD symptom clusters are associated with suicidality in a trauma-exposed sample with SUDs. Method: Female trauma-exposed participants with SUDs (N = 544) were recruited from community substance use treatment facilities in the National Drug Abuse Treatment Clinical Trials Network (CTN). Clinician-administered interviews assessed STB, SUDs, and PTSD symptoms. Results: Results indicated that the unidimensional item response theory (IRT) model used to estimate latent STB severity scores fit well, with strong local reliability at higher levels of latent STB severity. Regression predictors of elevated STB severity included younger age, opioid dependence, and higher PTSD reexperiencing symptoms. Conclusions: Clinicians are advised to screen for and target opioid use disorders and reexperiencing symptoms when addressing suicidal thoughts and behavior in trauma-exposed individuals with SUDs