The St. Louis investigation shows that restraint and seclusion become systemic civil-rights risks when districts rely on written policies and annual training without monitoring what happens inside classrooms. Enforcement is increasingly moving toward operational proof: incident-level data, repeat-use thresholds, parent notification, contractor oversight, post-incident review and evidence that behavioural interventions reduce restrictive practices.

This week’s deep dive covers:

The Special School District of St. Louis County secluded more than 300 students nearly 4,000 times and restrained almost 150 students 777 times over two school years. The frequency, concentration, and routine use of these interventions point to an operating failure. The case shows how weak behavioral capacity, inconsistent reporting, and absent central monitoring can allow emergency measures to become standard classroom practice.

The numbers from St. Louis County are difficult to explain as a collection of individual mistakes.

Across the 2022–23 and 2023–24 school years, employees of the Special School District secluded students an average of 11 times per school day. Nearly 40% of students enrolled in its high-needs schools experienced seclusion. At one school with fewer than 100 students, staff recorded 1,667 seclusions. At another, one student spent 101 hours, the equivalent of 17 school days, confined in a seclusion room.

These interventions were frequently used in situations that did not involve an imminent threat of physical harm. Federal investigators documented students being secluded for spilling coffee, refusing to attend music class, drawing on a chair, swearing and being disrespectful. Staff also placed students engaging in self-harm into seclusion, despite the risk that isolation could intensify their distress.

The district had policies limiting restraint and seclusion to emergencies. Missouri law imposed similar restrictions. Neither prevented these practices from becoming routine.

That distinction matters. A policy violation by one employee can occur inside an otherwise functioning system. Systemic failure begins when the same conduct appears across classrooms, schools, staff members and students without triggering meaningful intervention from central leadership.

The most consequential finding was how long the district continued operating without treating student seclusion numbers as an institutional warning.

Repeated use should have created multiple opportunities for intervention.

A student subjected to restraint or seclusion again and again should prompt an immediate review of the student’s Individualized Education Program, behavioral supports, and classroom environment. A classroom with unusually high incident rates should trigger observation and coaching. Repeated interventions by one employee should bring supervisory review. A school generating a disproportionate share of district incidents should require central-office involvement.

In St. Louis County, these feedback loops either failed or were absent. Restrictive interventions became a default response because the district did not consistently connect incident records to behavioral planning, staff supervision, or resource decisions. Annual training focused heavily on how to conduct physical interventions, while staff lacked sufficient preparation in prevention, de-escalation, and individualized behavioral support.

Staffing pressure also matters. Nationally, 34% of schools reported shortages in special education roles in late 2024. When districts lack behavioral specialists, trained paraprofessionals, and other high-need personnel, classroom employees are more likely to rely on reactive interventions during behavioral crises. Understaffing does not excuse improper restraint, but it helps explain why written policy can collapse under daily operating conditions.

The warning signs were therefore distributed across the organization: repeated incidents, insufficient behavioral capacity, weak staff preparation, missing escalation rules, and limited central monitoring. Viewed separately, each could be treated as a local problem. Viewed together, they described an operating model that was failing the students it was designed to serve.

Why is the St. Louis case part of a wider enforcement pattern?

St. Louis County is one of several districts where federal investigators have found that restrictive interventions were used as ordinary behavior management rather than emergency protection.

A 2024 Justice Department settlement required the Anchorage School District to eliminate seclusion after investigators found that students with disabilities had been secluded and restrained for noncompliant behavior, with some incidents contributing to self-harm and suicidal ideation.

In Michigan, the Montcalm Area Intermediate School District agreed to close its seclusion rooms, create classroom-wide behavior plans, and evaluate affected students for compensatory education and counseling. In Texas, Denton Independent School District entered an agreement addressing restraint by untrained employees and contracted school resource officers, alongside major documentation gaps.

Spokane Public Schools provides evidence that districtwide change is possible. After federal scrutiny, the district ended isolation, reduced class sizes, and expanded positive behavioral interventions and supports. Recorded restraint and isolation incidents fell from 6,087 in 2017–18 to 183 in 2023–24.

These cases share a common structure. Restrictive practices become embedded when districts fail to monitor patterns across students, personnel, programs, and external providers. Parents are notified inconsistently. Staff use alternative terms such as “quiet room” or “reset room,” obscuring whether an incident meets the district’s reporting definition. Contractors and school resource officers operate without sufficient alignment with special education and civil rights requirements. Central offices receive records but lack the thresholds, ownership or capacity to act on them.

The national data explains why federal agencies increasingly treat these failures as civil rights concerns. Students with disabilities represent approximately 14% of public-school enrolment but account for 76% of physical restraints and 68% of seclusions. Black students and boys are also disproportionately affected.

The legal framework remains fragmented. There is no comprehensive federal statute regulating restraint and seclusion in public schools, leaving states to establish their own definitions, restrictions and parent-notification requirements. Federal intervention instead relies largely on disability civil-rights protections under Section 504 of the Rehabilitation Act and Title II of the Americans with Disabilities Act.

That fragmentation makes district operating discipline more important. A board-approved policy may satisfy a procedural requirement, but it cannot show whether individual schools are following the same definitions, whether repeated incidents are reviewed, or whether behavioral supports are reducing reliance on force.

The St. Louis investigation exposed the end result of that gap. Emergency interventions had become part of routine school operations, while the district’s monitoring systems failed to recognize the scale of the problem. By the time federal investigators assembled the records into a districtwide pattern, the evidence of systemic failure had already accumulated for years.

A board-approved policy and annual training records do not show whether restraint and seclusion are being controlled in practice. Districts need evidence that repeated incidents trigger intervention, parents receive timely notice, staff and contractors follow common standards, and leaders can identify unusual patterns by student, classroom, employee, and school. Compliance now depends on continuous monitoring and corrective action.

Most districts can produce a restraint policy. Many can also show that employees completed the required training. The harder question is whether those controls change what happens when a student is in crisis.

Federal investigations increasingly reconstruct the full operating chain behind an incident: what happened beforehand, which de-escalation strategies were attempted, whether the intervention met the district’s emergency threshold, how long it lasted, who reviewed it, when the parent was notified, and whether the student’s support plan changed afterward.

A completed form provides only one part of that evidence.

Do repeated incidents automatically force a response?

The clearest test of an effective control system is what happens after the first incident, and what changes after the second or third.

Louisiana requires a formal review when a student is restrained or secluded more than three times during a school year. The threshold creates an automatic point at which the school must reconsider behavioral supports and determine why existing interventions are failing.

Without a defined trigger, repeated incidents can be documented without producing a decision.

Federal reviews have found students restrained dozens of times before districts meaningfully reconsidered their Individualized Education Programs or Behavior Intervention Plans. In one Florida district, a student was restrained 77 times in three and a half weeks without an adequate team review. In California, another student spent more than 38 hours in a “Refocus Room,” yet the district did not convene an IEP meeting until the end of the school year.

Districts, therefore, need escalation rules that operate independently of individual judgment. A defined number of incidents, an injury, an extended seclusion or repeated use by the same employee should automatically generate central review.

The purpose is not to impose one universal numerical threshold. It is to ensure that frequent use cannot continue indefinitely because every incident appears reasonable when viewed alone.

Can leaders see where incidents are concentrated?

Districtwide totals can make a severe problem look manageable.

A relatively low annual count may conceal one classroom, employee, or program responsible for a large share of incidents. Leaders need reporting that allows them to compare use across:

These comparisons should be based on rates as well as raw totals. A specialized program serving a small number of students may generate fewer incidents than a large comprehensive school, but expose a much larger proportion of its students to restrictive interventions.

The district also needs common definitions. Terms such as “quiet room,” “reset room,” “time away,” or “refocus room” can obscure whether a student was prevented from leaving and whether the event should have been reported as seclusion. When schools classify similar incidents differently, central-office data becomes unreliable before anyone begins analyzing it.

Boards should therefore ask whether every building, program, and contractor records the same behavior in the same way.

What must happen immediately after an incident?

Timely parent notification is one of the most visible indicators of whether a district takes restrictive interventions seriously.

Several states now require same-day notice. Districts operating under less specific rules should still establish a clear written deadline rather than leaving communication to staff discretion. Parents should receive enough information to understand what occurred, why staff believed intervention was necessary, how long it lasted, whether the student was injured and what review will follow.

The student should also receive an independent physical and emotional safety check. A nurse or other qualified professional should assess injuries, distress and any risk associated with the restraint position used.

The administrative review should then reconstruct the sequence of events:

Where a student lost substantial instructional time or experienced harm, the district may also need to consider counseling, compensatory education, or other remedial services.

The review has little value if it closes with a finding that the form was completed correctly. Its function is to reduce the probability of recurrence.

Does oversight extend beyond district employees?

District responsibility can continue when students receive services outside a traditional district classroom.

Non-public schools, specialized placements, behavioral-service providers, transportation contractors, and school resource officers may all become involved in managing student crises. Their employees may operate under different training systems, reporting tools, and organizational policies.

Contracts should resolve those differences before an incident occurs.

Districts need common definitions of permitted and prohibited interventions, minimum training requirements, parent-notification procedures, reporting deadlines, data-access rules and authority to investigate contractor conduct. School resource officers and law-enforcement partners also require clear guidance on disability protections under Section 504 and Title II of the Americans with Disabilities Act.

A district cannot reliably monitor patterns if contractor incidents remain in separate systems or reach central leaders only after a parent complaint.

What does effective training look like in practice?

Course completion is an input. It does not demonstrate that employees can prevent or safely manage a crisis.

Annual training often concentrates on the mechanics of restraint because those techniques appear to address the most immediate physical risk. Yet districts reduce restrictive interventions by building capability before the point of physical escalation.

That requires staff to practise recognising triggers, adjusting demands, communicating calmly, creating physical space and applying the student’s individual behavioural supports. It also requires observation and coaching after formal training, particularly in classrooms with high incident rates.

Turnover makes implementation harder. New paraprofessionals and temporary staff may begin working with high-need students before completing the same preparation as permanent employees. Districts need role-based training requirements, rapid onboarding and controls that prevent untrained staff from independently managing behavioural crises.

Successful reform depends on this implementation depth.

Maine School Administrative District 75 recorded 176 restraints and 152 seclusions involving just 15 students before adopting the Collaborative & Proactive Solutions model. Following intensive coaching, staff modelling and changes in how employees understood student behaviour, annual use fell to two restraints and seven seclusions by 2022–23.

Spokane Public Schools similarly reduced reported restraint and isolation incidents from 6,087 in 2017–18 to 183 in 2023–24 after ending isolation, reducing class sizes and expanding positive behavioural supports.

These examples show why districts need outcome measures alongside training records. Leaders should expect to see falling incident rates, fewer repeat cases, more timely plan reviews, improved documentation and consistent implementation across schools.

The standard is now whether the district can show that its controls identify risk early, compel a response, and change the conditions that caused the incident.

Districts need a single operating system for restraint and seclusion that connects incident reporting, special education planning, staff supervision, contractor oversight and board governance. The immediate priority is not another policy revision. It is assigning ownership, defining automatic escalation thresholds, and making repeated use visible before it becomes normalized. Leaders should be able to show what changed after every serious or recurring incident.

The St. Louis investigation demonstrates what happens when responsibility is distributed widely but accountability sits nowhere.

A classroom employee records the incident. A principal reviews the form. The special education team manages the student’s plan. HR tracks staff training. A contractor maintains its own records. The board receives an annual compliance update.

Each function may complete its assigned task while no one determines whether the same student, employee, classroom or placement is generating a recurring pattern.

Districts need one accountable executive or district-level intervention coordinator with authority to connect those records and compel action. That role should oversee data quality, incident escalation, school follow-up, behavioural support, contractor compliance and reporting to senior leadership. Montcalm Area Intermediate School District adopted this model as part of its federal settlement, alongside classroomwide behaviour plans and reviews of affected students for compensatory services.

Which incidents should trigger central review?

Districts should establish automatic escalation rules rather than relying on principals to decide when a pattern has become serious.

Triggers may include:

The trigger should initiate a defined response: central review, an IEP or Behaviour Intervention Plan meeting, observation of the classroom, staff coaching and a written corrective-action plan.

The purpose is to prevent incident reporting from becoming an archive of decisions that no one revisits. Louisiana’s three-incident threshold provides one model, but districts should also define triggers for severity, concentration and documentation failures.

Are district records describing the same conduct consistently?

Leaders should audit the language schools use when recording student isolation or physical intervention.

Terms such as “refocus room,” “quiet room,” “reset room” and “time away” can describe legitimate voluntary supports. They can also conceal situations in which a student was prevented from leaving and the event should have been recorded as seclusion.

A district cannot compare schools or identify unusual patterns when staff use different labels for the same practice. Leaders should define reportable interventions operationally, test those definitions against actual incidents and review whether low-reporting schools are genuinely using fewer interventions or recording them differently.

This audit should include contractor records. Students may be placed in specialised programmes, non-public schools or partner settings where reporting standards differ from those used by the district. The absence of incidents in the central system does not establish that no incidents occurred.

How should training accountability change?

Boards should stop treating completion rates as the primary measure of staff preparedness.

The relevant questions are whether employees can recognise escalation early, apply the student’s support plan, create safe space, call for appropriate assistance and avoid turning physical intervention into a routine response.

Districts should therefore combine training records with:

High turnover requires particular attention. Paraprofessionals and substitute staff may be placed with high-need students before they have received the same preparation as permanent employees. Leaders should identify roles that cannot work independently until required training and supervised practice are complete.

Can leaders prove that the district changed course?

The strongest evidence is a visible interruption in the pattern.

Maine School Administrative District 75 reduced annual use from 176 restraints and 152 seclusions to two restraints and seven seclusions after adopting the Collaborative & Proactive Solutions model and investing in intensive coaching and staff practice.

That result required more than replacing one training programme with another. The district changed how staff interpreted behaviour, how they responded before escalation and how leaders supported implementation.

Every district already holds evidence of whether its system is functioning: incident records, IEP reviews, training histories, staffing gaps, parent complaints and school-level patterns. The governance failure occurs when those signals remain separate until investigators assemble them into a civil-rights case.

K-12 Leadership Intelligence is for superintendents, district executives, and education leaders navigating board relations, state mandates, labor constraints, and political pressure.

This is one of our six education and learning-related publications spanning K-12, Higher Education, and Workforce. Our education newsletters reach tens of thousands of senior decision-makers across the U.S. and key international markets.

Ping us if you’d like to learn more, explore Enterprise Subscriptions, or would like to partner in other ways.

The Intelligence Council is a next-gen B2B media and business intelligence platform built for people who make strategy, allocate capital, and carry operating risk.