Concordance between DSM-5 and DSM-IV nicotine, alcohol, and cannabis use disorder diagnoses among pediatric patients

Background: The recently published Diagnostic and Statistical Manual of Mental Disorders-5 (DSM-5) includes several major revisions to substance use diagnoses. Studies have evaluated the impact of these changes among adult samples but research with adolescent samples is lacking.

Methods: 525 adolescents (93% African American) awaiting primary care appointments in Baltimore, Maryland were recruited for a study evaluating a substance use screening instrument. Participants were assessed for DSM-5 nicotine, alcohol, and cannabis use disorder, DSM-IV alcohol and cannabis abuse, and DSM-IV dependence for all three substances during the past year using the modified Composite International Diagnostic Interview-2, Substance Abuse Module. Contingency tables examining DSM-5 vs. DSM-IV joint frequency distributions were examined for each substance.

Results: Diagnoses were more prevalent using DSM-5 criteria compared with DSM-IV for nicotine (4.0% vs. 2.7%), alcohol (4.6% vs. 3.8%), and cannabis (10.7% vs. 8.2%). Cohen's κ, Somers' d, and Cramer's V ranged from 0.70 to 0.99 for all three substances. Of the adolescents categorized as "diagnostic orphans" under DSM-IV, 7/16 (43.8%), 9/29 (31.0%), and 13/36 (36.1%) met criteria for DSM-5 disorder for nicotine, alcohol, and cannabis, respectively. Additionally, 5/17 (29.4%) and 1/21 (4.8%) adolescents who met criteria for DSM-IV abuse did not meet criteria for a DSM-5 diagnosis for alcohol and cannabis, respectively.

Conclusions: Categorizing adolescents using DSM-5 criteria may result in diagnostic net widening-particularly for cannabis use disorders-by capturing adolescents who were considered diagnostic orphans using DSM-IV criteria. Future research examining the validity of DSM-5 substance use disorders with larger and more diverse adolescent samples is needed.