The Hidden Patient: How Gambling Disorder Slips Past Modern Health Systems

A patient presents to primary care with insomnia, low-grade anxiety, and a vague complaint of “stress at home.” The intake form is unremarkable. The PHQ-9 lands at a 7. The metabolic panel comes back clean. The visit ends with sleep hygiene advice and an SSRI starter sample, and the patient is discharged in twelve minutes.

The actual disease driving the visit — gambling disorder — never entered the conversation. It was not on the intake form, did not appear in the EHR’s problem list prompts, did not surface on any screener, and was not adjacent to any lab value the physician was trained to interpret. Of every chronic condition this clinic treats, this is the one with no biomarker, no claim code that fires routinely, and no SBIRT workflow. It is also increasingly likely to be silently driving the visit.

The First Behavioral Addiction Is Also the Most Under-Detected

The American Psychiatric Association reclassified gambling disorder in the DSM-5 as the first non-substance behavioral addiction, moved it out of impulse-control disorders, and placed it alongside alcohol and drug use disorders. That reclassification is now more than a decade old, but the clinical infrastructure built around it has not kept pace.

A recent systematic review of general practitioner management of gambling disorder found that none of the surveyed GPs routinely screened for it. Only 7–14% treated their gambling disorder themselves, and only 17–38% were confident in their knowledge of treatment pathways. Sixty-five percent acknowledged that managing gambling disorder fell within their scope, but in practice, the condition was overwhelmingly discovered by chance — typically when financial collapse forced the conversation into the room.

That gap matters more than it used to. Population prevalence of gambling disorder in U.S. adults sits at roughly 1%, but in primary care samples, the measured rate has been as high as 6.2% — six times higher than the general population, and meaningfully higher than several chronic conditions a clinic does screen for. The patients are there. The detection isn’t.

Why EHRs Are Blind to It

The structural problem is straightforward and largely a digital health one. Gambling disorder has no LOINC, no biomarker, no medication regimen to reconcile, and no routine ICD-10 trigger that fires from claims data. The most-used screening instruments — the Lie/Bet, the Brief Biosocial Gambling Screen, the South Oaks Gambling Screen — exist and are validated, but they are not built into the workflows that capture every adult patient walking through the door.

Substance use does have that infrastructure. AUDIT and DAST screeners are standard fields in many EHRs. SBIRT is reimbursable. Health systems run dashboards showing screening completion rates by clinic. Gambling has the screening tools but lacks the patient-engagement workflows that would make those tools routine practice. As a result, the only behavioral addiction the DSM recognizes as such is also the only one with no embedded place in adult intake.

The High-Functioning Profile Clinics Are Actually Seeing

The image of the problem gambler many clinicians still carry — the late-stage patient in crisis at a casino — is the wrong one. Outpatient programs treating this population see a different patient: high-functioning adults whose presentation matches what clinicians now describe as the clinical profile of digital gambling addiction. These are professionals, parents, students, and first responders using betting apps and crypto exchanges as a stress-management strategy, with financial damage compounding silently before medical or relational consequences force a visit.

The exposure is no longer marginal. Sports betting has expanded across roughly 38 U.S. states since Murphy v. NCAA in 2018, and helpline data from the National Council on Problem Gambling shows 43% growth in call volume in 2021 alone, with text volume up nearly 60% and chat volume up 84%. State-level data is starker: Pennsylvania’s helpline volume more than doubled between 2020 and 2023, and Ohio’s grew 55% in the first year of legal sports betting. The downstream demand is arriving in mental health caseloads, ED visits for anxiety with negative workups, and the financial-stress complaints that show up in primary care without their actual cause attached.

What Detection Would Actually Look Like

Detection infrastructure is the lowest-hanging intervention because the screening tools already exist and work. Three changes would move the needle.

First, embed a three-item gambling screener — the Lie/Bet or the Brief Biosocial Gambling Screen — into adult intake alongside the PHQ-9 and AUDIT that most primary care clinics already administer. The marginal cost is roughly one minute of patient time and one new column in the intake dataset, and the literature on brief screeners is clear that even one- and three-item instruments achieve adequate diagnostic accuracy for case-finding.

Second, build behavioral addiction prompts into telehealth and digital intake. A meaningful share of mental health intake now happens through screens, and adding gambling, gaming, and high-risk financial behavior questions to those flows is a configuration change, not a clinical innovation. Platforms operating in the digital mental health innovation space are well-positioned to lead here, and the data they generate could begin to populate the population-level dashboards that currently treat behavioral addictions as invisible.

Third, build proxy signals from existing data. Patients with gambling disorder leave traces in what clinics already collect — repeat ED visits for anxiety with normal workups, abrupt declines in adherence to medications they were previously stable on, and unexplained financial hardship flags captured in social determinants screening. None of these is diagnostic on its own. As pattern signals are routed to a behavioral health intake team, they would surface patients who would otherwise wait until the financial collapse to be seen.

Conclusion

The treatment models for gambling disorder are well-developed. Cognitive Behavioral Therapy and Dialectical Behavior Therapy, adapted for behavioral addictions, are evidence-based, dual-diagnosis programs that address the anxiety, depression, and trauma that frequently drive the betting behavior, and outpatient clinics specializing in digital-age presentations are running programming built around the actual patient profile arriving at their doors. The intervention pipeline works.

The bottleneck is detection. Behavioral addiction is the only DSM-recognized addiction with no claim code that fires, no biomarker, no SBIRT workflow, and no structured place in adult intake. It produces a six-fold higher prevalence in primary care than in the general population, and the system finds it almost exclusively by accident — after the financial collapse, the relationship breakdown, or the suicide attempt has already happened.

Closing that gap is a digital health problem more than a clinical one. The screeners exist, the treatment exists, and the patients are already in the system. They are arriving with insomnia, anxiety, financial stress, and clean labs. Whether the system finds them is a question of intake design, EHR configuration, and which behavioral addictions get a column in the dataset. The clinical case has already been made. What’s missing is the workflow.