Comparative Analysis of DSM-5-TR and ICD-11 for Substance Use Disorders
Paragraph Form
Both DSM-5-TR and ICD-11 represent valid but fundamentally different approaches to diagnosing substance use disorders, with DSM-5-TR offering superior dimensional assessment through its unified disorder construct, while ICD-11 maintains better diagnostic specificity through its traditional dependence-focused framework. 1, 2
DSM-5-TR Strengths
The American Psychiatric Association's DSM-5-TR eliminated the problematic abuse/dependence dichotomy that plagued DSM-IV, combining them into a single "substance use disorder" diagnosis with 11 criteria based on data from over 200,000 participants. 1 This unified approach overcame the poor reliability and validity of the former abuse diagnosis, which often captured patients with only one criterion (typically hazardous use) despite abuse supposedly representing a "syndrome." 1 The dimensional severity grading (mild: 2-3 criteria, moderate: 4-5 criteria, severe: 6+ criteria) provides clinically useful gradations of disorder severity. 2 The addition of craving as a criterion was supported by factor analyses demonstrating all criteria form a single unidimensional continuum. 1, 3 The system demonstrates excellent concordance with ICD-10 and DSM-IV (all κ ≥ 0.9). 4
DSM-5-TR Weaknesses
The two-criterion threshold casts a significantly broader diagnostic net than traditional dependence criteria, potentially capturing individuals with less severe problems and increasing false-positive diagnoses. 2 Concordance between DSM-5-TR and ICD-11 ranges from good for severe disorders to poor for mild disorders, primarily because DSM-5-TR's "mild substance use disorder" (2-3 criteria) does not align well with ICD-11's "harmful use" category. 4 The heterogeneous nature of "substance use disorder" combines what were previously distinct clinical entities—social consequences of use versus physiological dependence—into one diagnosis, potentially obscuring important clinical distinctions. 5 The lower diagnostic threshold may medicalize substance use patterns that do not represent true pathology, particularly in research contexts where prevalence estimates become inflated. 2
ICD-11 Strengths
The World Health Organization's ICD-11 maintains substance dependence as the "master diagnosis," preserving the well-validated dependence syndrome concept that demonstrated consistently high reliability and validity in DSM-IV. 1, 2, 5 This narrower diagnostic approach maintains better specificity and reduces false-positive diagnoses compared to DSM-5-TR's broader criteria. 6 ICD-11 underwent the largest participative revision process in classification history, with field studies demonstrating higher reliability and clinical utility compared to ICD-10. 2 The system retains the clinically meaningful distinction between harmful use (social/interpersonal consequences) and dependence (neurobiological syndrome), which has empirical support for alcohol, cannabis, and prescribed opioids. 5 ICD-11 expanded from 11 to 21 disorder groupings, enhancing coverage while maintaining diagnostic precision. 2
ICD-11 Weaknesses
The condensation of six ICD-10 dependence criteria into three pairs in ICD-11, where only one symptom within each pair must be present, may reduce diagnostic specificity compared to ICD-10. 6 This restructuring risks increasing false-positive dependence diagnoses because fewer distinct symptoms are required. 6 The criterion "substance use often continues despite the occurrence of problems" is more broadly worded than ICD-10's "persisting substance use despite clear evidence of overtly harmful consequences," potentially allowing diagnosis based solely on legal status of substances rather than true harm. 6 In the largest multinational study across 10 countries, ICD-11 diagnosed alcohol dependence approximately 10% more frequently than ICD-10, suggesting diagnostic inflation. 6 The new "harm to others" feature for harmful use showed very low endorsement rates and questionable diagnostic validity based on self-report. 4 ICD-11's categorical structure with optional dimensional assessments provides less systematic dimensional information than DSM-5-TR's integrated severity grading. 2
Suggested Improvements
For DSM-5-TR: Increase the diagnostic threshold from 2 to 3 criteria to improve specificity and reduce false positives, particularly for mild substance use disorder. 2, 4 Reintroduce subcategories that distinguish physiological dependence features (tolerance, withdrawal, impaired control) from social consequence features (role impairment, interpersonal problems) while maintaining the unified disorder structure, allowing clinicians to specify predominant syndrome type. 1, 5 Develop and validate biomarkers to supplement self-report criteria, improving diagnostic objectivity. 1 Conduct longitudinal studies examining outcomes (mortality, morbidity, quality of life) across the mild severity range to validate whether 2-3 criteria truly represent clinically significant disorder requiring intervention. 2
For ICD-11: Require at least two distinct symptoms within each of the three criterion pairs rather than allowing single symptoms to fulfill entire pairs, restoring diagnostic specificity. 6 Revise the "continues despite problems" criterion to require clear evidence of specific harm rather than vague "occurrence of problems," preventing diagnosis based solely on legal status. 6 Remove or substantially revise the "harm to others" feature for harmful use given its poor validity and low endorsement rates in population studies. 4 Develop standardized severity indicators within the dependence category similar to DSM-5-TR's approach, as the current system lacks systematic gradation of dependence severity. 2, 5 Conduct cross-cultural validation studies specifically examining the three-pair structure to ensure it maintains adequate specificity across diverse populations. 6
For Both Systems: Establish formal harmonization working groups to align diagnostic thresholds and improve concordance, particularly for mild/harmful use categories where agreement is poorest. 4 Develop shared biomarker standards and neurobiological assessment tools that can be incorporated into both classification systems. 1, 5 Create common data elements for research to allow direct comparison of prevalence, treatment outcomes, and long-term prognosis across diagnostic systems. 2 Prioritize longitudinal outcome studies examining mortality, morbidity, treatment response, and quality of life across diagnostic categories to validate clinical utility of both approaches. 1, 2
Chart Form
| Aspect | DSM-5-TR | ICD-11 |
|---|---|---|
| STRENGTHS | ||
| Diagnostic Structure | Single unified "substance use disorder" eliminates problematic abuse/dependence dichotomy based on 200,000+ participants [1] | Maintains well-validated dependence syndrome as "master diagnosis" with consistently high reliability [1,2,5] |
| Dimensional Assessment | Integrated severity grading (mild/moderate/severe) based on criterion count provides systematic gradation [2] | Underwent largest participative revision in classification history with demonstrated higher reliability than ICD-10 [2] |
| Criterion Validity | All 11 criteria form single unidimensional continuum; added craving criterion supported by factor analysis [1,3] | Preserves clinically meaningful distinction between harmful use and dependence with empirical support [5] |
| Concordance | Excellent agreement with ICD-10 and DSM-IV (all κ ≥ 0.9) [4] | Expanded from 11 to 21 disorder groupings enhancing coverage while maintaining precision [2] |
| Clinical Utility | Overcame poor reliability/validity of former abuse diagnosis that often captured single-criterion cases [1] | Narrower diagnostic approach maintains better specificity and reduces false positives [6] |
| WEAKNESSES | ||
| Diagnostic Threshold | Two-criterion threshold casts broader net, potentially increasing false positives and medicalizing less severe use [2] | Condensing six criteria into three pairs (one symptom per pair) may reduce specificity versus ICD-10 [6] |
| Concordance Issues | Poor agreement with ICD-11 for mild disorder category; heterogeneous construct obscures clinical distinctions [4,5] | "Continues despite problems" criterion too broad; may allow diagnosis based on legal status alone [6] |
| Diagnostic Inflation | Lower threshold may inflate prevalence estimates in research contexts [2] | Diagnosed alcohol dependence ~10% more frequently than ICD-10 in multinational study [6] |
| Specificity Concerns | Combines social consequences and physiological dependence into single entity, losing important distinctions [5] | "Harm to others" feature shows very low endorsement and questionable validity [4] |
| Dimensional Limitations | N/A | Categorical structure with only optional dimensional assessments provides less systematic severity information [2,5] |
| SUGGESTED IMPROVEMENTS | ||
| Threshold Modifications | Increase diagnostic threshold from 2 to 3 criteria to improve specificity [2,4] | Require two distinct symptoms within each criterion pair rather than single symptoms [6] |
| Criterion Refinement | Reintroduce subcategories distinguishing physiological vs. social features while maintaining unified structure [1,5] | Revise "continues despite problems" to require clear specific harm evidence [6] |
| Validity Enhancement | Develop and validate biomarkers to supplement self-report [1] | Remove or revise "harm to others" feature given poor validity [4] |
| Severity Grading | N/A | Develop standardized severity indicators within dependence category similar to DSM-5-TR [2,5] |
| Outcome Validation | Conduct longitudinal studies of mortality/morbidity/QOL across mild severity range [2] | Cross-cultural validation of three-pair structure to ensure adequate specificity [6] |
| SHARED IMPROVEMENTS NEEDED | ||
| Harmonization | Establish formal working groups to align thresholds and improve concordance for mild/harmful use categories [4] | |
| Biomarkers | Develop shared biomarker standards and neurobiological assessment tools for both systems [1,5] | |
| Research Standards | Create common data elements allowing direct comparison of prevalence and outcomes [2] | |
| Clinical Validation | Prioritize longitudinal outcome studies examining mortality, morbidity, treatment response, and quality of life [1,2] |