States Quietly Tighten the Rules on Restraining Children in Community Mental Health Programs

When a child in the grip of a psychiatric crisis becomes a danger to themselves or the people around them, the adults responsible for their care sometimes resort to one of the most controversial tools in behavioral health: seclusion, restraint, or other restrictive interventions. These practices, known in the research literature as SR/RI, range from a supervised time-out or a therapeutic hold to mechanical devices and sedating medication. Federal regulators describe them as a last resort, to be used only when de-escalation has failed, because the evidence that they actually calm an agitated child is thin, and the evidence that they can cause harm is not. Physical injuries are common, some restraints have ended in death, and survivors frequently describe the experience as humiliating and trauma-retriggering. Now a new study has done something no previous investigation managed: it systematically measured, across nearly two decades, how US states authorize and safeguard these interventions for children with serious emotional disturbance who are being treated not in hospitals, but in their own homes and communities.

The research, published in Community Mental Health Journal by Genevieve Graaf of The University of Texas at Arlington and colleagues, examined Medicaid’s 1915(c) Home and Community-Based Services waivers, the funding mechanism that allows states to pay for community-based psychiatric care as an alternative to institutional placement. Eleven states currently operate such waivers for youth with serious emotional disturbance, a federal designation covering the roughly five to six percent of children whose psychiatric conditions substantially impair daily functioning. These children often experience acute symptoms, including agitation expressed through hostile, destructive, or self-harming behavior such as head banging or elopement. Because the programs serve youth at elevated risk of restraint and seclusion, and because many of these youth carry trauma histories that make restrictive interventions especially dangerous, the question of what states permit, and what protections they demand, carries real clinical weight.

To quantify policy, the team collected 138 waiver applications from nineteen state programs approved between 2006 and 2022, obtained through the federal CMS database and, for older documents, through Freedom of Information requests. Each application includes a dedicated section on participant safeguards concerning restraint and restrictive interventions. Using directed content analysis with line-by-line coding, the researchers catalogued every type of intervention a state authorized and every safeguard it required, such as monitoring, incident reporting, staff training, and procedures for detecting unauthorized use. From these counts they built a novel metric, the Restrictiveness and Safety Index, which subtracts the number of authorized intervention types from the number of safeguards. A higher score means a state demands more protections relative to the coercive practices it permits; a negative score means the opposite.

The headline finding is a slow but broad national shift toward safety. Across the nineteen programs, the mean index score was 3.94, and the score rose in seventeen of the nineteen states over the study period, declining in only two. The national average for community-based waivers climbed from 2.18 in 2009 to 5.02 for waiver periods approved through 2027, a change the authors attribute to states adding safeguards, reducing the number of authorized intervention types, or both. The spread across states was striking: West Virginia topped the table with a score of 12.00, while Wisconsin began the study at negative 4.50, and Ohio ended it at negative 3.00. In other words, a child’s legal protection from restraint in a community program depends heavily on which side of a state line they live on.

The study also compared these community waivers with a separate, temporary class of programs. From 2007 to 2012, Congress authorized a demonstration project allowing states to use waiver authority specifically to divert youth from Psychiatric Residential Treatment Facilities, the locked institutions that remain the primary Medicaid-financed out-of-home setting for children with serious emotional disturbance. Nine states participated. The researchers analyzed these demonstration waivers separately and found their average index score was lower, at 2.62 compared with 3.94, though it rose from 1.63 in 2007 to 3.42 by 2017. The gap may reflect the fact that youth eligible for residential placement are judged more likely to require restrictive interventions, or simply that the demonstrations ended in 2017 and missed the later national upward trend. Either way, community programs consistently specified more safeguards per authorized intervention category than the institutional-diversion programs did.

Individual state trajectories reveal how policy evolves in practice. Louisiana’s score jumped from zero in 2015 to 4.33 the following year. Maryland’s demonstration program began at negative 3.00, the lowest of any PRTF waiver, and ended at 1.00. Kansas and South Carolina led the demonstrations throughout, finishing at 6.50. Texas and Michigan moved in the opposite direction, with Texas falling from 3.50 to 0.33 and Michigan from 3.00 to 1.17. Most intriguingly, when New York and Wisconsin each consolidated their mental-health-specific waivers into broader programs serving all children with complex healthcare needs, their scores rose sharply. New York’s index leapt to 11.00 after consolidation, and Wisconsin’s climbed from negative 1.00 to 4.33 by 2026. The authors speculate that adding physically or cognitively fragile children to the rolls may push administrators to tighten restrictions on coercive practices.

Perhaps the most sobering result concerns the federal Home and Community-Based Services Settings Rule, introduced in 2014, one of the largest regulatory changes in the program’s history. The rule requires that settings receiving waiver funding ensure participants’ freedom from coercion and restraint. Yet the data show no visible shift in state policies after the rule’s introduction or around its final compliance deadline in 2023. The rule does not outright prohibit restraint and seclusion, and the authors cite prior interviews with state officials suggesting the final version was perceived as having been watered down and lacking enforcement teeth. For a rule premised on protecting people’s most basic rights in publicly funded care, its apparent failure to move the policy needle in children’s mental health waivers is a finding that regulators will likely have to reckon with.

The technical achievement of the study lies in making an opaque policy landscape measurable. Waiver applications are dense administrative documents, and the researchers had to handle a structural change in 2014, when seclusion authorizations began being reported in a separate section from restraint, adjusting their scoring formula accordingly. They converted policy changes into quarterly observations aligned with effective dates, allowing year-by-year comparison within and across states. About twenty percent of possible amendment documents were missing, a gap the team mitigated by carrying forward prior values, and the authors acknowledge that states with very different policy provisions can produce identical scores, a limitation inherent to any composite policy index.

Crucially, the researchers are careful about what their numbers do and do not prove. The index measures written policy, not practice. Whether the safeguards states enumerate are actually enforced by provider organizations, whether they reduce the use of restraint in homes, schools, and parks, and whether they improve outcomes for children and staff remain open questions. The authors call for linking the index to administrative and clinical data to validate it, and for qualitative research into why state administrators change these policies. They also note that proven restraint-reduction strategies from inpatient psychiatry, including structured de-escalation training and trauma-informed care frameworks such as the Six Core Strategies, have never been rigorously tested in the less controlled environments where community-based care actually happens.

The deeper tension the study exposes is philosophical. These waiver programs exist precisely to move children out of restrictive institutions and into ordinary community life, yet restrictive interventions remain widely authorized within them, sometimes in settings such as family homes or public parks where oversight is weakest. The authors point to the Americans with Disabilities Act and the Olmstead decision, which established the right to care in the most integrated setting, as a legal lens through which community restraint practices deserve scrutiny. What the new index provides, for the first time, is a yardstick: a way to track, state by state and year by year, whether the promise of least-restrictive care for the country’s most vulnerable children is being written into the rules, or quietly left off the page.

Subject of Research: State Medicaid waiver policies authorizing seclusion, restraint, and safeguards for children with serious emotional disturbance in home and community-based programs

Article Title: Authorized Seclusion, Restraint, and Safeguards in Home and Community Based Medicaid Programs for Children With Serious Emotional Disturbance

Article References: Graaf, G., Fejer, A., Cañizares, M., Unnerstall, A., Friedman, C., & Purtle, J. (2026). Authorized Seclusion, Restraint, and Safeguards in Home and Community Based Medicaid Programs for Children With Serious Emotional Disturbance. Community Mental Health Journal. https://doi.org/10.1007/s10597-026-01725-3

Image Credits: AI Generated

DOI: 10.1007/s10597-026-01725-3

Keywords: serious emotional disturbance, Medicaid waivers, seclusion and restraint, home and community-based services, children's mental health, health policy, Restrictiveness and Safety Index, HCBS Settings Rule, psychiatric residential treatment, policy surveillance, de-escalation, trauma-informed care