Behavioral Health: The Blind Spot and the Expired Excuse

Only 17% of hospitals send summary of care records to most or all behavioral health providers, the weakest link in hospital data exchange. The consequence is that an organization accountable for a patient’s total cost of care often cannot see the behavioral health treatment driving it.

That gap has an unusual cause. It is not simply that nobody built the connection. For decades, federal privacy law said not to, and that law was rewritten less than two years ago.

The barrier was real, and it was justified

Federal regulations at 42 CFR Part 2 have governed the confidentiality of substance use disorder records since the 1970s, and were deliberately stricter than the general health privacy rules that came later. Under Part 2, disclosures that HIPAA would permit routinely required patient consent, and records had to be segregated so they could not travel with the rest of the chart.

The reason matters. Records of addiction treatment have been used to fire people, remove their children, and prosecute them. A rule that made those records hard to move was protecting patients from documented harm, not indulging paperwork. Describing Part 2 purely as an obstacle to coordination tells half the story.

But the effect on coordination was real. A patient’s addiction treatment could not follow them to the emergency department that might prescribe them opioids, and the organization accountable for their total cost of care could not see it.

The barrier came down

In February 2024, the Department of Health and Human Services published a final rule (89 FR 12472) implementing section 3221 of the CARES Act, rewriting Part 2 to align more closely with the HIPAA Privacy Rule.

Three changes matter. The rule replaces consent for each disclosure with permission to use and disclose records “with a single consent given once for all such future uses and disclosures” for treatment, payment, and health care operations. It lets recipients redisclose those records under HIPAA’s ordinary rules. And it states that an entity receiving records under that single consent “is not required to segregate or segment such records.”

That last provision removes the technical mechanism that kept behavioral health data walled off inside electronic records. The rule took effect in April 2024, and compliance became mandatory on February 16, 2026.

That date has passed. The legal justification for behavioral health’s isolation, cited for fifty years, no longer applies in the form it once did. If the data still does not move, then the law was never the real constraint.

Where accountability reached behavioral health, outcomes went unmeasured

The case for pulling behavioral health inside the accountability tent is weaker than its advocates usually admit.

A systematic review of alternative payment models for mental health and substance use disorders, covering 27 studies, found these models were associated with “reductions in MH/SUD utilization, and decreases in spending.” But clinical outcomes were measured in only 5 of the 17 payment models evaluated (29.4%). Of the eight models assessed for gaming and adverse selection, one showed evidence of gaming and three showed evidence of adverse selection, that is, organizations avoiding expensive patients.

Put plainly: under models that made someone accountable for behavioral health costs, utilization fell, spending fell, and in most cases nobody checked whether patients were better. Where investigators looked for organizations avoiding costly patients, they often found it.

The pattern repeats in the most-studied commercial accountable care contract in the country. An evaluation of Blue Cross Blue Shield of Massachusetts’s Alternative Quality Contract found enrollees in participating organizations were “slightly less likely to use mental health services,” a decline concentrated in the organizations that had accepted financial risk for behavioral health. The contract’s own participants said the arrangements “did not meaningfully affect mental health care delivery in the program’s initial years.”

A fall in mental health utilization inside a risk contract is not self-evidently a win. It may mean care was delivered more efficiently. It may mean people did not get care. Those two possibilities look identical in a spending report and are distinguishable only by measuring outcomes, which mostly was not done.

What it means for value-based care

The lesson is not that behavioral health should stay outside the tent. Leaving it out produced exactly the fragmentation above, and the interoperability data shows the cost.

The lesson is that behavioral health is where the standard value-based playbook is most likely to be confidently wrong. In most of medicine, lower utilization with stable quality is a reasonable proxy for value. In behavioral health, where undertreatment is the historical norm, falling utilization is at least as likely to be a symptom as a solution. A model that rewards reduced spending on a chronically underserved population may just be paying for the status quo.

Three things follow.

• Outcome measurement is not optional here. An arrangement that puts an organization at risk for behavioral health spending without measuring whether patients improved is a budget cap, not a value-based model.
• Adverse selection needs active surveillance. The evidence that organizations avoid expensive behavioral health patients was found wherever people looked for it. Attribution and risk adjustment must be tested against that specific failure, or they will penalize whoever takes these patients on.
• The data question is now answerable. The largest excuse for not integrating behavioral health information has been removed and the compliance deadline has passed. Organizations managing total cost of care can now ask exchange partners a question that used to get a legal answer instead of a technical one.

What we do not know

The most important number does not exist yet. There is no national data on whether behavioral health information began moving after February 2026, because not enough time has passed to measure it. This brief asserts that the legal barrier changed, not that behavior did.

Other limits are worth naming. Part 2 governs substance use disorder records specifically; general mental health records have long been under HIPAA, so the change does not sweep as broadly as some suggest. The Alternative Quality Contract evidence is one commercial contract in one state, in its early years. The systematic review spans 1997 through 2019 and predates the rule change, so it describes the old regime, not the new one.

Those limits point one way. The next few years will produce the evidence that settles this, and the field should be measuring outcomes now rather than discovering in 2030 that it recorded a decade of savings and no outcomes.

Related

• Interoperability Reached 70% of Hospitals by 2023, but the Data Often Goes Unused
• Post-Acute Care: The Unclaimed Prize

Sources

• Confidentiality of Substance Use Disorder (SUD) Patient Records. Final rule. 89 FR 12472. Department of Health and Human Services; February 16, 2024. federalregister.gov (opens in a new tab)
• Carlo AD, Benson NM, Chu F, et al. Association of Alternative Payment and Delivery Models With Outcomes for Mental Health and Substance Use Disorders: A Systematic Review. JAMA Network Open. 2020;3(7):e207401. doi:10.1001/jamanetworkopen.2020.7401 (opens in a new tab)
• Barry CL, Stuart EA, Donohue JM, et al. The Early Impact Of The ‘Alternative Quality Contract’ On Mental Health Service Use And Spending In Massachusetts. Health Affairs. 2015;34(12):2077-2085. doi:10.1377/hlthaff.2015.0685 (opens in a new tab)
• Gabriel MH, Richwine C, Strawley C, et al. Interoperable Exchange of Patient Health Information Among U.S. Hospitals: 2023. ONC Data Brief No. 71. Office of the National Coordinator for Health Information Technology; May 2024. healthit.gov (opens in a new tab)