The Data Is In, and It's Not Pretty: What ONC's 2024 Report Tells Us About Behavioral Health's Digital Divide

Every few years, the federal government takes a hard look at whether substance use and mental health treatment facilities have caught up to the rest of healthcare when it comes to electronic health records and data exchange. The Office of the National Coordinator for Health Information Technology recently released findings from the 2024 survey cycle, drawing on SAMHSA's National Substance Use and Mental Health Services Survey (N-SUMHSS), and the answer is: they haven't. Not by a long shot.

The gap is not a surprise to anyone who has spent time in this space. But the report puts numbers to a problem that too many health IT conversations treat as a footnote. For state health agencies, Medicaid programs, HIEs, and the treatment facilities themselves, this data should function as a call to action, not a shrug.

What the Data Actually Shows

The 2024 N-SUMHSS covers more than 21,000 behavioral health treatment facilities nationwide, asking about EHR adoption, clinical functionality use, and electronic exchange capabilities. The picture it paints is one of uneven progress, with pockets of genuine advancement surrounded by large stretches of facilities still managing care in ways that look more like 2005 than 2025.

Substance use treatment facilities continue to lag behind mental health facilities on virtually every technology measure, including basic EHR adoption, and both categories trail hospitals and physician practices by considerable margins. While 96 percent of non-federal acute care hospitals had adopted certified EHR systems by the early 2020s, behavioral health provider adoption rates have remained stubbornly lower. ONC's own earlier analysis of hospital survey data found that only 67 percent of psychiatric hospitals had adopted a 2015 Edition-certified EHR, compared to 86 percent of general acute care hospitals — and that figure covers only hospital-based psychiatric units, not the broader universe of freestanding community behavioral health programs..

Among freestanding substance use and mental health treatment facilities — the community-based programs that serve the vast majority of people in behavioral health treatment — adoption rates are lower still, and exclusive use of EHR systems for core clinical activities like progress notes, prescriptions, and lab monitoring remains elusive for most. Prior research found fewer than 25 percent of facilities in any state reporting fully paper-free operations for clinical documentation, and the 2024 data suggest that structural change has been slow.

The exchange side of the equation is where the gap becomes operationally consequential. Having an EHR matters. Having an EHR that can send and receive patient data with the rest of the care system is what actually moves the needle on care coordination, overdose response, crisis follow-up, and care transitions. On this measure, behavioral health facilities are not just behind — they are largely absent from the networks that hospitals and health systems have spent the last decade building.

Why This Happened, and Why It Has Stayed This Way

The disparity did not emerge from negligence. It emerged from a structural imbalance that HITECH and Meaningful Use created when they were designed. The Meaningful Use incentive program that turbocharged EHR adoption across hospitals and physician practices starting in 2009 was not available to most specialty behavioral health and substance use treatment facilities. The program rewarded Medicare and Medicaid eligible providers, and the majority of freestanding behavioral health organizations — particularly those funded through state grants, block grants, or serving uninsured populations — were simply not eligible.

The result was a federally funded modernization wave that swept through acute care settings and largely skipped behavioral health. While general hospitals went digital with financial assistance and regulatory pressure, substance use treatment programs and community mental health centers were left to self-fund whatever technology they could, or go without. ONC acknowledged this directly when launching the BHIT Initiative, noting that behavioral health providers' lagging adoption was "due in part to their ineligibility to participate in health IT incentive programs."

Compounding this is the regulatory complexity that surrounds behavioral health data. The 42 CFR Part 2 consent framework for substance use disorder records has historically created real and perceived barriers to data sharing that have made EHR vendors less likely to build robust behavioral health data exchange capabilities into their products, and made facilities more likely to see data sharing as legally risky. The 2024 final rule updating Part 2 aligned its consent structure more closely with HIPAA, representing meaningful progress — but implementation of that alignment is still working its way through provider and HIE operations. The rule went into effect April 16, 2024, with a compliance date of February 16, 2026.

The documentation model itself adds friction. Behavioral health treatment relies heavily on narrative clinical notes, complex symptom scale scores, individualized treatment plans, and encounter types that don't map neatly to the structured data fields that most EHR systems were built around. ONC and SAMHSA have both noted that this documentation complexity leads to difficulties in efficiently sharing behavioral health information with primary care providers and hospitals to support continuity of care. Data entry burden has been a persistent complaint from clinicians, and the EHR products available to smaller behavioral health organizations have historically offered less functionality at higher per-user cost relative to their ability to pay.

The Operational Consequences Are Not Abstract

When a patient in crisis leaves a hospital emergency department and steps into a community mental health center the next morning, what data travels with them? In most cases, precious little, if anything. The hospital is likely using a certified EHR that participates in ADT alerting and can query an HIE. The community mental health center is likely on a system that cannot receive or generate those notifications in any standard format. The care transition happens in a documentation vacuum.

When a person on medication-assisted treatment for opioid use disorder is admitted to a primary care clinic for an unrelated condition, does the prescribing provider know about their buprenorphine? In most cases, there is no reliable electronic pathway to find out. The substance use treatment program's records sit in a separate system, subject to separate consent requirements, with no bidirectional exchange with the HIE that serves the surrounding area.

When an overdose response team is dispatched and a patient is stabilized and transported, what information does the receiving emergency department have about prior treatment history, current medications, and care connections? The paramedics may have some information if their EMS system participates in health information exchange. The emergency department may have a prior visit on record. The substance use treatment program that has been working with this person for six months is probably invisible to everyone in that chain.

These are not edge cases. They are the routine experience of coordinated behavioral health care in the United States in 2025. The ONC data confirms that the technology infrastructure necessary to close these gaps exists in one part of the system and is absent in another.

Federal Movement Is Real, but Pilots Are Not Scale

The Behavioral Health Information Technology (BHIT) Initiative, launched jointly by ONC and SAMHSA with more than $20 million in SAMHSA funding, is the most significant federal investment in this problem in years. The initiative is piloting a new USCDI+ behavioral health dataset across nine states, testing whether standardized behavioral health-specific data elements can be incorporated into EHR systems in ways that reduce documentation burden and enable meaningful exchange. Nine pilot sites across Colorado, Connecticut, Delaware, Florida, Massachusetts, North Carolina, Oregon, Rhode Island, and Washington, D.C. are now live, with a goal of informing national standards and a Behavioral Health Information Resource targeted for potential nationwide adoption in 2027.

This is genuinely good work. Getting to a standardized set of behavioral health data elements that EHR systems can natively support is a necessary precondition for the kind of deep exchange that whole-person care requires. USCDI v4 additions of alcohol use assessments, SUD treatment interventions, and medication adherence data elements were meaningful steps in that direction.

But pilot programs test feasibility. They do not solve the broader problem of 20,000-plus facilities that lack the resources, the technical capacity, or in some cases the organizational incentive to modernize. Demonstrating that FHIR-based behavioral health data exchange is possible in nine well-resourced pilot settings is not the same as making it routine across the full landscape of substance use and mental health treatment in the United States.

The gap between what is technically possible and what is operationally deployed is where most of the work actually lives. And that work requires more than standards.

What Actually Moves the Needle

The facilities that have made meaningful progress on EHR adoption and exchange share common characteristics that are worth naming directly.

Hospital or health system affiliation is a significant predictor. Facilities that are part of a larger system have access to enterprise technology resources, legal and compliance infrastructure, and connectivity through their parent organization's HIE participation. This is one reason why the gap between freestanding community behavioral health programs and hospital-based psychiatric units is so persistent — the playing field is not level, and it was not designed to be.

Accreditation and payer mix matter. Research has shown that facilities holding Joint Commission accreditation, accepting Medicare, or carrying a higher share of insured clients demonstrate stronger EHR adoption rates. These are facilities with both the financial resources to invest in technology and external accountability structures that create pressure to modernize.

State-level policy levers are underused. A handful of states have implemented Medicaid incentives, technical assistance programs, or reporting requirements that have meaningfully accelerated adoption in behavioral health settings. A comprehensive HHS environmental scan found that states like Arizona have structured their Medicaid programs to incentivize HIE data submission from behavioral health providers, with the ability to process both Part 2 and non-Part 2 data. These models exist and they work. They are simply not widespread.

HIE outreach and onboarding investment changes outcomes. Behavioral health providers are not going to find their way onto statewide HIEs by accident. The technical complexity, the consent framework requirements, and the resource constraints of most community programs mean that passive participation models do not work. Proactive, facilitated onboarding with workflow support, consent management tools built into the process, and ongoing technical assistance is what actually closes participation gaps. This is operational work, not just policy work.

The Role of Health Information Exchange

For statewide HIEs, the ONC data is essentially a map of where the unfinished work is. The hospitals are connected. The large physician groups are connected. The payers are increasingly in the network. The last mile, to borrow a phrase that I think about every day, runs straight through behavioral health.

An HIE that does not have a serious, well-resourced strategy for behavioral health connectivity is not a whole-person care infrastructure. It is a medical records infrastructure with a gap where the mental health and substance use treatment system should be.

This is not a criticism of HIEs. The gap is real, the barriers are real, and the resources required to close it are significant. But the 2024 data makes clear that voluntary participation and standard onboarding processes are not producing behavioral health connectivity at the scale that coordinated care requires. Different approaches, with different levels of investment and intentionality, are necessary.

The strategies that work include dedicated behavioral health outreach lanes with staff who understand the clinical and operational context of substance use and mental health treatment. They include consent management infrastructure that makes 42 CFR Part 2-compliant exchange operationally manageable rather than theoretically possible. They include use case design that gives behavioral health providers a clear answer to the question of why participating in the HIE makes their clinical work better, not just why it satisfies a policy requirement. And they include honest engagement with the resource constraints of the providers being asked to connect, with technical assistance and workflow integration support that meets them where they are.

What This Means for States

For state health agencies and Medicaid programs, the data is both a benchmark and a brief for investment. The Rural Health Transformation Program, Medicaid value-based care transitions, and federal behavioral health initiatives all depend, to varying degrees, on data flowing between the treatment settings that serve the highest-need populations. If the data infrastructure is not there, the care model cannot function as designed.

State strategies that have worked combine several elements: Medicaid reimbursement or administrative match funding for HIE participation, technical assistance programs that address both technology adoption and exchange connectivity, grant programs that help smaller community behavioral health organizations afford and implement certified EHR systems, and reporting requirements that create accountability for progress over time. No single lever is sufficient. The combination of financial incentive, technical support, accountability structure, and ongoing operational assistance is what drives durable change.

The 2024 N-SUMHSS annual report is a useful baseline for states building behavioral health data infrastructure strategies. It documents where the system is, not where it needs to go. But knowing where you are is the first step toward knowing what it will take to get somewhere else.

The Bottom Line

The ONC 2024 data on EHR adoption and exchange capabilities among substance use and mental health treatment facilities is important precisely because it is not a success story. It is a documentation of a structural gap that has persisted for more than a decade and that has real consequences for the people who depend on the behavioral health system to coordinate their care.

Progress is happening. The BHIT initiative is investing in standards that could accelerate adoption. States are developing more sophisticated strategies. HIEs are building behavioral health engagement programs. And the 2024 42 CFR Part 2 final rule has removed one of the most persistent barriers to exchange.

But the pace of change is still slower than the scale of need. One in five Americans experience a mental health condition in any given year. Opioid overdose remains a leading cause of preventable death. The behavioral health system is not a secondary concern in whole-person care. It is a primary one.

The data infrastructure has to catch up to that reality. This report measures the distance between where we are and where we need to be. The work is in closing it.