VA Records Raise Red Flags on Cannabis Use Disorder Diagnoses

September 4, 2026

Photo by Xperia6 via Pixabay.

A diagnosis code doesn't disappear after the appointment ends. Once "Cannabis Use Disorder" lands in a veteran's VA medical chart, it travels with them — into disability evaluations, into future mental health visits, into the assumptions a new provider makes before they've even asked a follow-up question. That's the part of this story that tends to get lost in debates about legalization and scheduling: the label itself has weight, independent of whether it was ever earned through a careful clinical workup.

A batch of records obtained through public records requests and reviewed by the Veterans Action Council is now raising uncomfortable questions about how often that weight gets assigned carelessly. The documents suggest a pattern in which mentioning cannabis — sometimes even just CBD — during an unrelated appointment can result in a diagnostic label with no real assessment behind it. For veterans navigating VA disability ratings, pain management, and PTSD treatment, that's not a paperwork technicality. It's the difference between being treated as a whole person and being treated as a chart flag.

How Cannabis Use Disorder Became a Catch-All Diagnosis

Photo by https://kaboompics.com/ via Pexels.

Before 2013, the DSM-4 kept cannabis diagnoses split into two categories: Cannabis Abuse, a milder classification, and Cannabis Dependence, reserved for more serious, compulsive patterns of use. When the DSM-5 arrived, the American Psychiatric Association folded both into a single diagnosis — Cannabis Use Disorder — scored along a severity spectrum from mild to severe based on how many of eleven criteria a patient meets. The stated goal was modernization: aligning cannabis with the way psychiatry already diagnosed alcohol and other substance disorders, on a continuum rather than a binary.

In a private practice setting, that severity scale can work as intended, with a clinician sitting down and walking through the criteria one by one. Inside a system the size of the Veterans Health Administration, which serves millions of patients across hundreds of facilities, the consolidation created a different problem. A single diagnostic code now has to cover an enormous range of behavior — someone using cannabis a few times a month with zero functional impairment, and someone whose use is genuinely compulsive and interfering with daily life — under the same three words in a medical record.

That ambiguity doesn't resolve itself. It gets resolved, or doesn't, at the point of documentation, by whichever provider happens to be typing notes that day. The FOIA records reviewed by the Veterans Action Council point to exactly this gap: a diagnostic category broad enough to swallow casual disclosure and clinical severity alike, applied inside a system processing an enormous volume of patient encounters, with no guarantee that anyone actually checked which end of the spectrum a given veteran falls on before the label got typed in.

What the FOIA Records Actually Show

Among veterans surveyed in NESARC-III, 7.3% reported using cannabis, but only 1.8% actually met the diagnostic criteria for Cannabis Use Disorder—suggesting most cannabis-using veterans do not have a clinically diagnosable disorder.

The numbers in the released records are the clearest evidence that something is off. Citing data from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC-III), the records note that 7.3 percent of U.S. veterans reported using cannabis in the past year — but only 1.8 percent met full DSM criteria for Cannabis Use Disorder. Do the arithmetic on that and it means roughly three out of every four veterans who use cannabis show no diagnostic evidence of a disorder at all. Use, on its own, is simply not a reliable stand-in for pathology, and the data the VA's own researchers are working from says so plainly.

That gap matters because it undercuts the reflexive assumption — still common in some corners of clinical practice — that disclosed cannabis use is itself a red flag worth documenting as disordered. Most veterans who use cannabis are doing so without meeting any recognized threshold for a substance use disorder, full stop.

What's striking is that this isn't a critique coming only from outside advocates. Bruce I. Friedland, Media Branch Chief for the VA's Office of Research and Development, acknowledged in the records that officials are trying to thread a genuinely difficult needle: treating veterans who have real, diagnosable substance use disorders, while also supporting a growing body of research into cannabis as a potential therapeutic tool for conditions like chronic pain and PTSD. That's a candid admission from inside the agency, and it implies VA officials themselves see the tension between current diagnostic habits and where the evidence and the research agenda are actually heading.

A Pattern Across the Series: Documentation Without Assessment

Photo by RDNE Stock project via Pexels.

The reporting series this op-ed is responding to lays out a pattern that goes well beyond statistics. A July 2026 installment described a veteran who mentioned cannabis and CBD use during a routine dental visit — a disclosure that was overwhelmingly about non-psychoactive CBD, not recreational or heavy cannabis use. That detail didn't survive contact with the record. In a later, unrelated ADHD evaluation, the same veteran was described in chart notes as a regular cannabis user, with no documented inquiry into frequency, dosage, purpose, or any sign of functional impairment. A passing mention at a dental appointment had, months later, hardened into a clinical descriptor.

A June 2026 installment examined VA's own screening policy and found the standard is actually fairly rigorous on paper: providers are supposed to assess frequency, quantity, method of use, signs consistent with CUD, and how cannabis use might interact with medications, mental health conditions, and overall functioning. The records suggest that standard isn't consistently applied in practice, which is a gap between policy and bedside behavior that's hard to catch except through exactly this kind of records request.

A May 2026 installment added important context on the regulatory ceiling providers are working under: VA clinicians currently cannot recommend cannabis to patients or certify them for state medical marijuana programs, a restriction tied to cannabis remaining a Schedule I substance federally — one that could shift if rescheduling to Schedule III moves forward.

Peer-reviewed research backs up the broader pattern. A 2022 study in the American Journal of Psychiatry, led by Livne and colleagues using VHA records from 2005 to 2019, found CUD diagnoses rose disproportionately among veterans who already had other psychiatric diagnoses. A 2025 follow-up in Lancet Regional Health Americas examined how state-level legalization tracked against CUD diagnosis trends within the VA system, adding further texture to a documentation pattern that looks less like careful clinical assessment and more like a habit.

Why the Label Matters Beyond the Chart Note

Photo by Age Cymru via Unsplash.

None of this stays confined to a single visit. A CUD diagnosis becomes part of the permanent record, and future providers — often working from limited time and a thick chart — tend to read prior diagnoses as established fact rather than as a data point that might warrant a second look. A label entered casually during a dental appointment can end up shaping how a completely different provider, months or years later, interprets a veteran's request for pain medication, sleep aids, or PTSD treatment.

That has a chilling effect that works against everyone's stated interest. Veterans have described hesitating to talk openly with VA providers about their cannabis use, worried it could affect disability benefit determinations or get flagged in a way that colors how their broader care gets managed. That's precisely backwards from what good medicine requires, since providers need accurate information about cannabis use to catch real drug interactions and avoid genuine safety problems.

Fontan, writing as a member of the Veterans Action Council, frames the core argument plainly: overdiagnosis doesn't protect veterans, it discourages the honesty that VA providers actually need to do their jobs well. And the tension VA officials themselves acknowledge — balancing legitimate addiction treatment against a research agenda increasingly interested in cannabis as therapy — isn't going away on its own. It remains unresolved while cannabis's federal scheduling status sits in limbo, still pending final action as of this writing. Veterans considering how openly to discuss cannabis use with a VA provider should also keep in mind that state law on cannabis varies widely and doesn't govern federal VA policy, so confirming both the local legal landscape and current VA guidance is worth doing before assuming either protects or exposes them.

Rescheduling cannabis to Schedule III would change real things for VA providers — potentially opening the door to research partnerships, altering pharmacy protocols, maybe even eventually touching what clinicians are allowed to discuss or recommend. But it wouldn't touch the documentation habit these records expose. That habit predates the scheduling debate and would keep operating under Schedule III exactly as it does now, because it isn't really a scheduling problem. It's a training and accountability problem, rooted in how a single, broad diagnostic category gets applied by providers moving fast through an enormous caseload.

Fixing it doesn't require Congress or the DEA to do anything. It requires VA facilities to actually enforce the screening standard that's already written into their own policy — asking about frequency, dosage, purpose, and impairment before a diagnosis goes into the chart, instead of after a veteran happens to mention it in passing. Until that happens consistently, veterans would do well to ask their own providers directly what's actually documented in their record and why, rather than assuming a diagnosis reflects a conversation that never took place.