A Novel Approach to Quantify Risk for SUD: Computerized Adaptive Testing
*Levent Kirisci, University of Pittsburgh 
Michael Vanyukov, University of Pittsburgh 

Keywords: Liability for SUD, IRT, CAT

Employed as a quantitative measure of Substance Use Disorder (SUD) risk, the Transmissible Liability Index (TLI) may be useful for detecting youths requiring prevention intervention. Adoption of an assessment tool for use in practical settings is, however, contingent on satisfying several important criteria such as the time required for administration and scoring. Accordingly, lengthy questionnaires not only detract from treatment delivery, but may also incur unacceptable cost. Computer adaptive testing (CAT) mitigates measurement error while maximizing efficiency since only the items pertinent to accurately measuring trait level are administered. This study points to the utility of the CAT version of the TLI to quantify risk at the individual level. Sample 1 consisted of 318 males and 107 females having a mean age of 18.8 years (SD = .49) who were evaluated two years later to determine transition to cannabis use disorder. Sample 2 consisted of 276 twin pairs. The sample was 19.94 (sd=4.50) years old. Instrumentation. Liability Index. The rationale and method of deriving the TLI has been previously described along with preliminary results documenting its construct and pediatric validity. This task was carried out by faculty at the NIDA-funded CEDAR. Statistical Analysis. The items were calibrated and IRT-based TLI scores were obtained using MULTILOG 7 for the entire set of items. Results. Sample 1. The TLI scores of the P&P and the CAT are almost identical (r=.95). The average number of items administered was 16.8 (SD=4.70). The TLI scores of the P&P (OR=2.94) and CAT (OR=2.23) versions predict cannabis use disorder diagnosis at age 22 with overall accuracy of 72% and 68%. Sample 2. The two versions is strongly correlated (r=.87). The CAT version required administering 18.6 (SD=3.07) items. The P&P and CAT versions predicted alcohol and drug use with sensitivity=75% (75%) and specificity = 64% (65%). Conclusion. Taking into account individual differences in SUD risk enables calibration of intervention intensity to risk severity. Although tailoring intervention intensity to severity of individual risk is established practice for prevention of medical disorders, it has not yet been adopted for prevention of psychiatric disorders, including SUD. Accordingly, the CAT version of the TLI advances this goal by objectively, efficiently and inexpensively quantifying an important component of risk for developing substance use disorder.