Background: Ascertaining agreement between DSM-IV and DSM-5 is important to determine the applicability of treatments for DSM-IV conditions to persons diagnosed according to the proposed DSM-5.
Methods: Data from a nationally representative sample of US adults were used to compare concordance of past-year DSM-IV opioid, cannabis, cocaine and alcohol dependence with past-year DSM-5 disorders at thresholds of 3+, 4+, 5+ and 6+ positive DSM-5 criteria among past-year users of opioids (n=264), cannabis (n=1622), cocaine (n=271) and alcohol (n=23,013). Substance-specific 2 × 2 tables yielded overall concordance (kappa), sensitivity, specificity, positive predictive values (PPV) and negative predictive values (NPV).
Results: For DSM-IV alcohol, cocaine and opioid dependence, optimal concordance occurred when 4+ DSM-5 criteria were endorsed, corresponding to the threshold for moderate DSM-5 alcohol, cocaine and opioid use disorders. Maximal concordance of DSM-IV cannabis dependence and DSM-5 cannabis use disorder occurred when 6+ criteria were endorsed, corresponding to the threshold for severe DSM-5 cannabis use disorder. At these optimal thresholds, sensitivity, specificity, PPV and NPV generally exceeded 85% (>75% for cannabis).
Conclusions: Overall, excellent correspondence of DSM-IV dependence with DSM-5 substance use disorders was documented in this general population sample of alcohol, cannabis, cocaine and opioid users. Applicability of treatments tested for DSM-IV dependence is supported by these results for those with a DSM-5 alcohol, cocaine or opioid use disorder of at least moderate severity or severe cannabis use disorder. Further research is needed to provide evidence for applicability of treatments for persons with milder substance use disorders.
Keywords: Concordance; DSM-5; DSM-IV; Diagnosis; Kappa; Substance use disorder.
Published by Elsevier Ireland Ltd.