I've completed the investigation. Here is the full report.

---

Sedated by the State: How Foster Children Are Chemically Restrained With Federal Money, and Why No Agency Has Ever Lost a Dollar Over It

For fifteen years the federal government has known that children in foster care are medicated with antipsychotics and sedatives at rates that cannot be explained by illness, that the drugs are overwhelmingly prescribed for uses the FDA never approved in children, and that in many states no one can produce a document showing who authorized the prescription or whether anyone ever checked on the child afterward. Congress responded in 2011 by requiring every state, as a condition of federal child welfare funding, to have "protocols for the appropriate use and monitoring of psychotropic medications" under 42 U.S.C. § 622(b)(15). It required a plan. It did not define what the plan must contain, did not require states to report a single data point about medication, and did not attach a penalty. The result is a fifteen-year paper exercise: HHS's Office of Inspector General and the Government Accountability Office have documented the failure in report after report, state auditors have confirmed it, federal judges have called it appalling, and neither the Administration for Children and Families nor the Centers for Medicare & Medicaid Services has ever withheld a Title IV-E or Medicaid dollar from a state because a child in its custody was chemically restrained. The enforcement remedy, in audit after audit, is training.

The Setup: A Statute That Requires a Description, Not a Standard

The Child and Family Services Improvement and Innovation Act of 2011 (P.L. 112-34) amended the Title IV-B state plan at 42 U.S.C. § 622(b)(15) to require each state to develop — with its Medicaid agency, pediatricians and other health experts — "a plan for the ongoing oversight and coordination of health care services for any child in a foster care placement," including "the oversight of prescription medicines."

Every operative word is discretionary. ACF reviews these plans through five-year Child and Family Services Plans and Annual Progress and Services Reports, and samples cases during Child and Family Services Reviews. But ACF never set a floor. As OIG put it in 2018, ACF "allows States flexibility in implementation" and merely suggests states consider practice guidelines from the American Academy of Child and Adolescent Psychiatry.

The gap that flexibility produces is measurable. A 2025 study in the peer-reviewed literature, Foster care psychotropic prescribing regulations are complex and insufficient, found that only 61% of states have published a foster care psychotropic guideline at all; nine states (18%) have none, and the researchers could not locate one for another eleven. More than one-third of America's foster youth — 38% — live in states with no psychotropic monitoring requirement whatsoever. A parallel 50-state policy review by The Imprint in 2025 found at least ten states with no child welfare policy on psychotropics for caseworkers, caregivers or prescribers; agencies in Massachusetts, North Dakota, Arkansas, West Virginia, Kansas, Alabama and Kentucky either confirmed or did not dispute that they have none.

Those states are nonetheless in full compliance with § 622(b)(15), because the statute asks them to describe what they do, not to do anything in particular.

The Numbers: Off-Label Is the Rule, Not the Exception

The disparity is not subtle. GAO's foundational 2011 study, Foster Children: HHS Guidance Could Help States Improve Oversight of Psychotropic Prescriptions, analyzed 2008 Medicaid data in Florida, Massachusetts, Michigan, Oregon and Texas and found foster children prescribed psychotropics at 2.7 to 4.5 times the rate of other children on Medicaid. Hundreds of children were on five or more psychotropic drugs at once — a regimen GAO's experts said has no supporting evidence in adults or children. Thousands exceeded the maximum doses in Texas's own FDA-label-derived guidelines. And both foster and non-foster children under one year old were prescribed psychotropics, for which GAO's experts said there is no established use in infants.

A 2023 study in the Journal of Child and Adolescent Psychopharmacology by Keefe and colleagues found foster youth had 6.8 times higher odds of being dispensed a psychotropic than other Medicaid enrollees — after controlling for age, gender, and mental health and developmental diagnoses — and were more likely to be medicated in the absence of any such diagnosis. By the 2018 OIG report's accounting, nearly 30% of the 400,000 children then in U.S. foster care were taking at least one psychotropic. In Kansas, 35% of child-welfare-involved youth aged 2 to 17 received a psychotropic in 2019, more than three times the rate of their non-foster Medicaid peers; researchers have identified 24 states medicating at least 30% of child-welfare-involved youth.

The most damning federal number concerns what the drugs are actually for. OIG's Second-Generation Antipsychotic Drug Use Among Medicaid-Enrolled Children (OEI-07-12-00320, March 2015) pulled Medicaid claims from California, Florida, Illinois, New York and Texas — about 39% of national Medicaid spending on second-generation antipsychotics — and had physician reviewers read the underlying medical records. The findings:

• Quality-of-care concerns in 67% of claims for antipsychotics prescribed to children; two or more concerns in 49%.
• Lack of monitoring in 53% of claims — no routine height, weight, vital signs, blood pressure, abnormal-involuntary-movement screening, liver function, glucose or lipid testing, or EKGs, for drugs whose signature harms are metabolic and neurological.
• Only 8% of antipsychotic claims were for a medically accepted pediatric indication. Ninety-two percent were off-label — most commonly for bipolar disorder (20%), mood disorders (13%) and autism spectrum disorders (8%) in children outside the approved age range or without the specific approved symptom.

The FDA's pediatric antipsychotic indications are narrow by design: aripiprazole for schizophrenia at 13–17, bipolar mania at 10–17, irritability in autism at 6–17. There is no approved antipsychotic indication for a preschooler. There is no biomarker for any pediatric psychiatric diagnosis. And as Texas's own prescribing manual concedes, the FDA "does not limit the way a practitioner may prescribe an approved drug."

The best-documented state is the one that reformed earliest. Texas built statewide prescribing Parameters in 2005 and a Psychotropic Medication Utilization Review process, and rates fell. Even so, its SFY 2024 data report, released in January 2026, shows that of 52,773 children in care, 9,335 (17.7%) received a psychotropic — including 261 children aged 0 to 2, 128 three-year-olds and 804 children aged four to five. One thousand and one children were on four or more psychotropics; 301 were on five or more, two of them under age three. That is the reform model.

The Money: Sedation Is Cheaper Than Staff

Nobody sets out to drug a seven-year-old. The system does it because every incentive points that way.

Residential care is the expensive end of foster care and the place where medication substitutes for supervision. Minnesota's published rates put Qualified Residential Treatment Program per diems at $524.60 to $530.90 a day — roughly $190,000 a year per child — and Title IV-E is an uncapped entitlement that matches state spending. A QRTP is required to have nursing and clinical staff. Staff are the single largest cost. A sedated unit needs fewer of them.

The Senate Finance and HELP Committees' two-year investigation, published June 12, 2024 as Warehouses of Neglect, examined four operators serving foster and low-income children: Universal Health Services, Acadia Healthcare, Vivant Behavioral Healthcare, and the nonprofit Devereux Advanced Behavioral Health. Investigators documented staff isolating children while administering sedatives — a practice that violates federal Medicaid rules on restraint and seclusion. The committee valued Acadia at roughly $6.5 billion and UHS at $11.8 billion, and found that in 2023 Medicaid supplied more than half of Acadia's revenue and more than a quarter of UHS's. On October 9, 2024, Chairman Ron Wyden referred the findings to the Justice Department as potential Medicaid fraud and civil rights violations.

The National Disability Rights Network reached the same conclusion three years earlier in Desperation Without Dignity (2021): for-profit programs used psychotropic medication "as a chemical restraint to control behavior for staff convenience," not for the health and safety of the child.

The enforcement that followed is instructive in scale. In September 2024 Acadia Healthcare agreed to pay $19.85 million to resolve False Claims Act allegations that from 2014 to 2017 it billed Medicare, Medicaid and TRICARE for medically unnecessary inpatient behavioral health services, failed to develop adequate treatment plans, and "knowingly failed to properly staff its facilities" — resulting in assaults, elopements and suicides. Medicaid's share was $6.37 million. Against a $6.5 billion enterprise built substantially on Medicaid revenue, that is a rounding error, and it arrived only because two whistleblowers filed suit in 2017.

Devereux, the nonprofit in the group, is a case study in how federal money flows to these operators outside the child welfare line item entirely. Project Milk Carton's federal award tracking identifies $91.8 million in tracked federal awards to The Devereux Foundation, including $48.2 million from HHS in FY2022 under CFDA 93.566 and $30.4 million under 93.676 — the Unaccompanied Alien Children program. Devereux is simultaneously facing more than fifty consolidated sexual abuse suits in the Eastern District of Pennsylvania, and a Georgia jury awarded $60 million to a girl assaulted by a staff member at one of its facilities, a verdict litigated up to the Georgia Supreme Court as Taylor v. Devereux Foundation, 316 Ga. 44.

And the drugs themselves were sold into this population deliberately. In November 2013 Johnson & Johnson paid $2.2 billion, including $1.391 billion in civil False Claims Act liability, to resolve charges that it marketed Risperdal off-label to pediatric and geriatric populations. Risperdal became a fixture in Texas foster care and state institutions. No child received a dollar of that settlement.

The Conflict: The Company Paid for the Pills Reviews the Pills

The consent chain is where the system's design becomes indefensible.

Under Texas Family Code Chapter 266, a court designates up to four "medical consenters" — typically the child's caregiver plus a caseworker and another CPS staffer. The state's own mandatory consenter training says plainly that "the vast majority of children in DFPS conservatorship do not need psychotropic medications" and that traumatized children may display behaviors that are a normal reaction to what they have survived. Under Georgia policy, the youth and parents may offer feedback but the final decision rests with county child welfare directors. Washington bars caseworkers and caregivers from authorizing psychotropics — but accepts a court order that is a general authorization for whatever a physician later directs. California requires a juvenile court JV-220 order. Maryland's suit alleges that 72.1% of medicated foster children outside Baltimore City had no documented psychiatric diagnosis at all.

Then there is the reviewer. In Texas, nearly all foster children are enrolled in STAR Health, operated by Superior HealthPlan — a subsidiary of Centene Corporation. Centene acquired Magellan Health in a $2.2 billion deal that closed in January 2022, adding 16 million pharmacy members and 2 million pharmacy benefit management members. The managed care organization that administers the pharmacy benefit is also the entity that runs the Psychotropic Medication Utilization Review determining whether the prescribing was appropriate. GAO flagged exactly this structural blind spot in GAO-14-362 (2014), recommending additional federal guidance for states overseeing psychotropics administered through managed care. Twelve years on, no binding federal standard governs how an MCO must review prescribing for children in state custody.

The Cases: What the Record Actually Shows

Texas. In M.D. v. Abbott, 152 F. Supp. 3d 684 (S.D. Tex. 2015), Senior U.S. District Judge Janis Graham Jack found that "rape, abuse, psychotropic medication, and instability are the norm" and that children "almost uniformly leave State custody more damaged than when they entered." The court-appointed monitors' 2023 report, after fourteen multi-day site visits, found that of 161 children's files reviewed, 75 (47%) were on four or more psychotropics, the youngest aged eight; nearly three-quarters had never had a clinical review or had not had one within the past year. Four or more medications automatically triggers a required PMUR. A PMUR had been completed for only 21 of the 75. Of three reviews conducted after the monitors showed up, all three found the child "outside parameters" with "opportunities to reduce polypharmacy" — and afterward, one child's medications were reduced, one child's stayed the same, and one child had a new psychotropic added. Judge Jack has held Texas in contempt twice and weighed a third contempt finding and a federal receivership in December 2023.

The named plaintiff D.P. described being chemically restrained: injected with a drug "cocktail" prescribed by a physician who, in her account, "left it up to the staff to administer however many shots they wanted to."

C.G. spent seven years in Texas foster care. After three psychiatric hospitalizations for suicidality in 2019–2020, she was discharged in March 2020 on at least three psychotropics. She saw a psychiatrist once, virtually; at that single visit the psychiatrist made three medication changes. Her safety plan required staff observation at all times, and the monitors found that the observation form auto-populated the 15-minute checks — no staff entry required. C.G. hanged herself in an emergency shelter bathroom and lay dead on the floor for thirty minutes before anyone checked.

Florida. Gabriel Myers, age seven, hanged himself in a Margate foster home in 2009. He was on Seroquel and other psychotropics. Florida's own investigation found that neither his parents nor a judge had approved them, and the medications were not reflected in his case file. Twelve-year-old Denis Maltez died in 2007 at the Rainbow Ranch group home in Hialeah from a combination of restraint and overmedication.

Litigation. Four class actions covering more than 18,000 foster youth have been filed on this single issue. M.B. v. Tidball (W.D. Mo.), filed 2017 by Children's Rights, the National Center for Youth Law, Saint Louis University's legal clinics and Morgan Lewis, was the first federal class action focused solely on psychotropic drugging in foster care; Judge Nanette Laughrey certified a class of all children prescribed psychotropics in Missouri's custody and approved a settlement covering more than 13,000 children in 2023. Bryan C. v. Lambrew (D. Me.), filed January 2021, settled with Judge Nancy Torresen's approval in 2024, committing Maine to timely medical records that travel with the child, a real informed consent process, and post-administration clinical review. Y.A. v. Padilla, filed in January 2023 by Disability Rights Maryland, the ACLU of Maryland, Children's Rights and Morgan Lewis, alleges Maryland drugs foster children as "a form of chemical restraint" when their behavior is hard to manage; it remains live.

Every one of these reforms was imposed by a federal judge on a state that was, throughout, in full compliance with its § 622(b)(15) state plan.

The Accountability Gap: Audits Without Consequences

In 2018 OIG published Treatment Planning and Medication Monitoring Were Lacking for Children in Foster Care Receiving Psychotropic Medication (OEI-07-15-00380). It selected the five states with the highest share of foster children on psychotropics — Iowa (35.7%), North Dakota (37.3%), Virginia (37.2%), New Hampshire (36.1%) and Maine (32.7%) — and reviewed 589 children's case files against each state's own rules. The findings:

• 34% received neither the treatment planning nor the medication monitoring their own state required. Eight percent got neither.
• 20% had no treatment plan; 23% received no medication monitoring. In Iowa, 48.3% had no medication monitoring by a prescribing professional.
• Where states had specific treatment plan criteria, 52% of the plans that existed failed to meet them.
• None of the five states required that medication dosages or potential adverse effects be documented in the child's case file at all.

ACF's response is the heart of this story. The agency "concurred with some of our recommendations but not others" and, OIG noted pointedly, "did not specify which." It replied that it "already has a well-established approach to program implementation that includes a regulated mechanism to identify and correct compliance issues." OIG answered that one in three children were going unmonitored, "which suggests the current approach to identifying and correcting compliance issues is insufficient." ACF disagreed outright with the recommendation that states report treatment planning and monitoring data, calling it outside what can be reliably captured administratively.

The state-level audits that followed found the same thing and produced the same non-consequence. In Florida (A-05-22-00009), of 85 sampled children on psychotropics, 36 had medications not recorded in the state system, 56 had no medication logs, and 33 had no prescribing authorization on file; 57 of 60 children's opioid prescriptions were unrecorded. Florida elected not to comment on the draft report. In California (A-05-22-00007), psychotropics for 22 children were unrecorded, and for 28 children the court authorizations were not in the state system and California could not produce them from the children's case files at all — meaning that for those children there is no evidence a judge ever approved anything. Ohio, Indiana and Michigan drew parallel findings. The recommendations are listed as open. The recommended fix, in nearly every case, is training and a data-sharing agreement.

The stakes are stated on the cover of each of those audits: "To receive Federal funding for child welfare services, States are required to have a plan for the oversight of prescription medications." States that cannot document who authorized an antipsychotic for a child in their legal custody continue to receive that funding in full.

In May 2026 — fifteen years after GAO's first report — ACF Assistant Secretary Alex Adams told a gathering of state officials that the agency would sharpen its focus on overmedicated foster youth, announcing a summer convening with CMS aimed at finally obtaining state-by-state prescribing data so ACF could "more objectively intervene." The admission buried in that pledge is that the federal government still does not know, nationally, which children in its care are on what.

Meanwhile Congress is moving the other way. The Ensuring Medicaid Continuity for Children in Foster Care Act of 2026, introduced in March by Reps. Julia Brownley and Gus Bilirakis, would exempt Qualified Residential Treatment Programs from Medicaid's institutions-for-mental-disease exclusion — expanding federal payment for congregate care without attaching a single new prescribing safeguard.

Why It Matters, and What Would Actually Work

Up to 80% of children enter foster care with significant mental health needs. Almost none of them need an antipsychotic. What they need — psychotherapy, placement stability, a consistent adult who knows their history — is labor-intensive, poorly reimbursed, and in chronically short supply. A pill is reimbursable the day it is dispensed. That asymmetry, not malice, is the engine.

The costs are permanent. Second-generation antipsychotics in children cause rapid weight gain, metabolic syndrome, type 2 diabetes, prolactin elevation and movement disorders that can become irreversible. OIG found monitoring for those harms absent in more than half of claims. Children are being placed at lifelong medical risk by drugs prescribed for indications the FDA never approved, authorized by adults who in many cases never met them, and monitored by no one.

Five changes would close the gap, and none require a new entitlement:

1. Put a floor under § 622(b)(15). Convert "describe your protocols" into minimum federal standards: a documented diagnosis, a documented trial of non-pharmacological intervention, informed consent by someone who has met the child, and mandatory independent second opinion for any antipsychotic under age six and any regimen of four or more concurrent psychotropics.
2. Make the data reportable. ACF added 62 elements to AFCARS in its December 2024 final rule. Psychotropic prescribing is still not among the enforceable set. It should be — with a Medicaid-claims-to-child-welfare-records match, the exact fix OIG recommended to California and Florida and that neither has completed.
3. Attach consequences. Condition a defined slice of Title IV-E administrative match on verified compliance. A penalty that has never once been imposed is not a deterrent; it is a description.
4. Break the review conflict. The entity that owns the pharmacy benefit cannot be the entity that certifies the prescribing was appropriate. Utilization review for children in state custody should be independently contracted and publicly reported.
5. Fund the alternative. Every settlement that has worked — Missouri, Maine, New Jersey, Michigan, Tennessee, and the New York community mental health agreement approved in January 2026 — required the same thing: real informed consent, medical records that travel with the child, clinical review after administration, and enough community mental health capacity that sedation stops being the cheapest available intervention.

The federal government has commissioned the studies, published the findings, and read the audits. The one thing it has never done is make any of it cost a state anything.

---

Sources: - GAO-12-201, Foster Children: HHS Guidance Could Help States Improve Oversight of Psychotropic Prescriptions - GAO-17-129, Foster Care: HHS Has Taken Steps to Support States' Oversight of Psychotropic Medications - GAO-14-362, Additional Federal Guidance Could Help States Better Plan for Oversight of Psychotropic Medications Administered by Managed-Care Organizations - HHS-OIG, OEI-07-15-00380, Treatment Planning and Medication Monitoring Were Lacking for Children in Foster Care - HHS-OIG, OEI-07-12-00320, Second-Generation Antipsychotic Drug Use Among Medicaid-Enrolled Children - HHS-OIG, Florida Did Not Comply With Requirements for Documenting Psychotropic and Opioid Medications (A-05-22-00009) - HHS-OIG, California Did Not Comply With Requirements for Documenting Psychotropic and Opioid Medications - 42 U.S.C. § 622, State plans for child welfare services - Texas HHS, Use of Psychotropic Medications for Children in Texas Foster Care, FY2002–2024 - Hiba Loya, Prescribed Injustice: Examining State-Sanctioned Over-Medication of Foster Youth in Texas, 24(2) Hous. J. Health L. & Pol'y 301 (2025) - Foster care psychotropic prescribing regulations are complex and insufficient (2025) - MindSite News / The Imprint, 'All I Did Was Sleep': Despite Years of Damning Reports, States Fail to Rein in Psych Meds for Foster Youth - The Imprint, A Detailed Window Into State Policies on Psychotropic Prescribing - The Imprint, Top Federal Child Welfare Official Focuses on Psychotropic Use - Senate Finance Committee, Warehouses of Neglect (June 12, 2024) - Wyden DOJ referral on youth residential treatment facilities (Oct. 9, 2024) - DOJ, Acadia Healthcare to Pay $19.85M to Settle Allegations Relating to Medically Unnecessary Inpatient Behavioral Health Services - DOJ, Johnson & Johnson to Pay More Than $2.2 Billion to Resolve Criminal and Civil Investigations - NDRN, Desperation Without Dignity - Children's Rights, M.B. v. Tidball (Missouri) - Children's Rights, Bryan C. v. Lambrew (Maine) - Children's Rights, Y.A. v. Padilla (Maryland) - California State Auditor, Report 2015-131 - Texas Tribune, Judge considers holding state in contempt a third time over foster care conditions - Rep. Julia Brownley, Ensuring Medicaid Continuity for Children in Foster Care Act of 2026 - Fierce Healthcare, Centene completes acquisition of Magellan Health - Project Milk Carton CivicOps database — federal award records for The Devereux Foundation (USAspending/IRS Form 990 linkage)