Substance Use Referrals Among Inpatients With Serious Mental Illness

To evaluate the performance of the New Hampshire Hospital Screening and Referral Algorithm (NHHSRA) and describe the characteristics of inpatients aged 18-65 years, with serious mental illness (SMI) referred for substance use disorder (SUD) interventions. Two questions were evaluated: (1) the accuracy and utility of the NHHSRA in identifying patients appropriate for addiction-focused interventions and (2) associations between referral status and patient characteristics. Receiver operating characteristic (ROC) curve analysis evaluated diagnostic performance of the NHHSRA, with sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), precision, and F1 score assessed. Logistic regressions assessed associations between patient characteristics and referral outcomes. The reference standard ("need for SUD intervention") was defined as the addiction psychiatrist's clinical assessment following direct patient evaluation. The patients assessed in this study were admitted to New Hampshire Hospital between January 2023 and February 2025. The cohort (n = 927) was predominantly male (66.7%), with a mean age of 36.5 years. Opioid use disorder (OUD) was the most prevalent primary SUD (37.9%), followed by alcohol (23.9%), cannabis (15.1%), and methamphetamine (13.7%). The NHHSRA demonstrated excellent performance-sensitivity: 96.6%, specificity: 93.6%, PPV: 98.5%, NPV: 86%, and F1 score: 97.5%. ROC analysis yielded an AUC of 0.82, indicating strong discriminative ability. Logistic regressions identified higher odds of a positive NHHSRA screen among patients with OUD and lower odds among those with cannabis use disorder, after adjusting for age, sex, and psychiatric diagnosis. The NHHSRA is an accurate, objective tool that enhances identification of SUD intervention needs among inpatients with SMI. By addressing limitations of subjective clinical judgment and patient self-report, its implementation may improve access to addiction services and optimize treatment delivery. Patient characteristics associated with referrals inform targeted strategies for integrated care in this population. .