Rasch Analysis of Cannabis Use Disorder in an Adult Inpatient Sample

Area of Science:

• Psychiatry
• Psychometrics
• Addiction Medicine

Background:

• The Diagnostic and Statistical Manual of Mental Disorders version 5 (DSM-5) defines cannabis use disorder (CUD) as a polythetic unidimensional diagnosis.
• Empirical evaluation of the latent structure of CUD criteria is limited.
• Rasch analysis is a validated psychometric technique for unidimensional scales.

Purpose of the Study:

• To empirically evaluate the latent structure of DSM-5 cannabis use disorder (CUD) criteria using Rasch analysis.
• To assess the unidimensionality and item fit of CUD symptoms in a clinical sample.
• To provide psychometric evidence supporting or refining the DSM-5 CUD diagnostic framework.

Main Methods:

• Rasch model analysis was applied to DSM-5 CUD criteria in a clinical sample (n=249) of adults undergoing inpatient substance use disorder treatment.
• Unidimensionality was assessed using the Martin-Löf test and Ponocny's nonparametric -T2 test.
• Item fit was evaluated using the chi-squared goodness of fit test.

Main Results:

• The Rasch analysis supported the unidimensional structure of the DSM-5 CUD criteria.
• Symptom #3 (hazardous use) was the least endorsed and highest severity item.
• Symptom #9 (craving) was the most endorsed and lowest severity item, with overall good model fit, except for symptom #8.

Conclusions:

• Rasch analysis of CUD symptoms in an inpatient sample broadly supports the DSM-5 CUD syndrome.
• The hazardous use criterion (symptom #3) warrants further examination for potential revision or removal in future DSM editions.
• Refining diagnostic measurement for CUD may improve clinical assessment and treatment.