rethinkTHC Search
Menu
Study breakdown

DSM-5 and ICD-10 Diagnose Cannabis Use Disorders Differently at the Moderate Level

Cross SectionalModerate evidence
The takeaway

While the DSM-5 and ICD-10 largely agreed on who has no cannabis disorder and who has severe dependence, about half of people with a "moderate" DSM-5 diagnosis received a "dependence" label under ICD-10.

Read this if you work in addiction treatment or want to understand how cannabis use disorders are classified in medical systems.

~50% of DSM-5 "moderate" cannabis diagnoses mapped to ICD-10 "dependence," potentially overstating severity.

What the researchers found

Researchers compared how the DSM-5 and ICD-10 diagnostic systems classify cannabis use disorders using data from 7,672 prison inmates. The two systems agreed well at the extremes: people with no DSM-5 diagnosis almost always had no ICD-10 diagnosis, and those with severe DSM-5 disorders almost always received an ICD-10 dependence diagnosis.

The disagreement emerged in the middle. Among those with a DSM-5 "moderate" cannabis use disorder, roughly half received an ICD-10 "dependence" diagnosis while the other half received the less severe "harmful use" label. This is clinically meaningful because the DSM-5 coding guidelines currently direct clinicians to assign ICD-10 dependence codes even for moderate diagnoses.

Why it matters

Since U.S. healthcare systems must use ICD-10 billing codes, the mapping between DSM-5 clinical diagnoses and ICD-10 codes has real-world consequences. When a moderate DSM-5 diagnosis gets coded as ICD-10 dependence, it may overstate the severity of the condition for billing, insurance, and clinical records purposes.

The numbers in context

7,672 inmates assessed. Overall prevalence rates were comparable across systems. Concordance was high for no-diagnosis and severe categories. Approximately 50% of DSM-5 moderate cases mapped to ICD-10 dependence, and 50% to harmful use.

How the study worked

Data came from routine clinical assessments of 6,871 male and 801 female inmates admitted to a state prison system between 2000 and 2003. DSM-5 and ICD-10 diagnostic determinations were made using algorithms corresponding to each system's criteria. Past 12-month prevalence rates were compared across systems.

What this study cannot tell us

The sample consisted entirely of prison inmates, who may have higher rates and different patterns of substance use than the general population. The data were from 2000-2003, before the DSM-5 was published, so diagnostic criteria were applied retrospectively using algorithms. This limits the ecological validity of the findings.

How to read the evidence

Moderate evidence from a large sample using validated diagnostic algorithms, though the incarcerated population limits generalizability.

When this study was published

Published in 2016. The ICD-11 has since been released with revised substance use disorder categories.

The bigger picture

Diagnostic labels shape treatment decisions, insurance coverage, and how people understand their own behavior. The finding that the same person can be labeled as "dependent" or merely engaging in "harmful use" depending on which system is applied highlights that cannabis use disorder severity exists on a spectrum that does not neatly fit either classification system.

Questions still open

  • Should the DSM-5 coding guidelines be revised to better align moderate diagnoses with ICD-10 categories? Do patients with DSM-5 moderate diagnoses who receive ICD-10 dependence codes get different treatment or outcomes than those coded as harmful use? Would these concordance issues apply equally in non-incarcerated populations?

Common questions

Do the DSM-5 and ICD-10 diagnose cannabis use disorders the same way?
They agree at the extremes (no disorder and severe/dependence) but disagree substantially in the middle. About half of people with moderate DSM-5 cannabis use disorder receive a dependence diagnosis under ICD-10, while the other half get the less severe "harmful use" label.
Why does this matter for patients?
Because U.S. healthcare uses ICD-10 codes for billing, a moderate cannabis use problem may be recorded as "dependence" in your medical records, which could affect insurance coverage, treatment decisions, and how future providers view your history.

Read the original research

Diagnostic Concordance between DSM-5 and ICD-10 Cannabis Use Disorders.

Addictive behaviors, 58, 117-22

Citation

Proctor, Steven L; Williams, Daniel C; Kopak, Albert M; Voluse, Andrew C; Connolly, Kevin M; Hoffmann, Norman G. (2016). Diagnostic Concordance between DSM-5 and ICD-10 Cannabis Use Disorders.. Addictive behaviors, 58, 117-22. https://doi.org/10.1016/j.addbeh.2016.02.034

Explore the wider topic