Alliance Alert: The U.S. Government Accountability Office (GAO) has issued a report that underscores a reality our movement has known for years: Assisted Outpatient Treatment (AOT) lacks clear evidence of effectiveness.
Since 2016, the federal government has invested $146 million in AOT grants across 63 sites, yet evaluations from HHS and SAMHSA show no conclusive evidence that AOT improves outcomes for people with serious mental health challenges. The GAO cited key weaknesses in the federal assessments — including reliance on self-reported data, lack of uniformity across programs, and limited ability to distinguish AOT’s impact from that of voluntary services.
This latest report makes one thing clear: it’s time to halt any further expansion of forced treatment until we have a real understanding of what actually helps people heal and thrive.
As our CEO, Harvey Rosenthal, said: “In an environment that challenges the way federal funding has been used, this finding calls into serious question the value of the $146 million in assisted outpatient treatment grants that have been awarded to 63 grantees since the program’s inception in 2016.”
It’s not just this report. A 2009 Duke University study commissioned by the New York State Legislature also questioned whether court orders — rather than sustained engagement in voluntary support — were the real driver of positive outcomes. A follow-up study commissioned by the NYS Legislature now underway by the University of Pittsburgh and Human Services Research Institute may offer further insights.
We call on both federal and state governments to stop doubling down on coercion. Instead, they must prioritize expanding voluntary, community-based mental health and substance use services that support people in healing without threatening their rights or autonomy.
We need to invest in what works: housing, peer support, crisis response teams, culturally relevant services, and trauma-informed support — not court orders.
GAO: HHS Findings on Assisted Outpatient Treatment Reveal No Clear Results
By Valerie A. Canady | Mental Health Weekly | July 18, 2025
Despite nearly a decade of investment in and implementation of assisted outpatient treatment (AOT) programs across dozens of states, federal evaluations of the AOT programs have produced inconclusive results, according to a report released last week from the U.S. Government Accountability Office (GAO).
Directed by Congress, the GAO-25-107526 report evaluated how the HHS assessed the effectiveness of AOT programs funded through federal grants. AOT is a court-ordered, community-based treatment for individuals with serious mental illness who are deemed unable to live safely without supervision.
The report, “Serious Mental Illness: HHS Assessments of Assisted Outpatient Treatment Have Yielded Inconclusive Results,” describes HHS’s efforts to assess the effects of the grant program on participants’ health and social outcomes and summarizes what the assessments have revealed.
Under AOT, adults with a serious mental illness can be ordered by a judge in a civil court proceeding to adhere to community-based treatment in accordance with applicable state laws. It is generally intended for individuals who have been assessed as unlikely to be able to live safely in the community without supervision.
In 2014, federal law authorized HHS to award grants to organizations to implement assisted outpatient treatment programs. The Substance Abuse and Mental Health Services Administration (SAMHSA) has awarded approximately $146 million in AOT grants since the program’s inception in 2016.
The GAO received a mandate from the Consolidated Appropriations Act of 2023 included in the law, a provision to assess the HHS’s evaluation of AOT, Michelle B. Rosenberg, director with the GAO health care team, told MHW. This report represents GAO’s first evaluation of the treatment program, she said.
Federal agency assessments
Two HHS agencies — the Office of the Assistant Secretary for Planning and Evaluation (ASPE) and SAMHSA — have made efforts to assess the grant program since 2016: ASPE conducted a cross-site impact evaluation and SAMHSA generated two outcome reports based on routine collection of data from grantees. Topics studied included participant outcomes such as treatment adherence, psychiatric emergency room visits, hospitalizations and arrests.
GAO interviewed a nongeneralizable sample of six AOT grantees, which GAO selected to obtain var-iation in geographic location and levels of urbanization. GAO also interviewed representatives of six stakeholder organizations, including mental health professional associations and advocacy groups. The groups were selected to provide a range of views on AOT.
“This is pretty controversial program,” said Rosenberg. “We spoke to some grantees themselves from a number of different stakeholder organizations — organizations that are proponents of AOT and ones that are opponents.”
Rosenberg noted that as part of the GAO’s original background work, officials gained their thoughts from mental health professionals and advocates on the effectiveness and ethics of AOT. However, that was not the focus of their report, she said. “We did our due diligence,” she said. “We got the perspective from stakeholders, which varied significantly across stakeholder groups.”
Some advocates who support AOT believe the program is important for people with a serious mental illness who do not realize that they are ill and will not accept treatment voluntarily, Rosenberg indicated. Others who oppose AOT feel that people’s civil liberties are threatened.
GAO determined that the AOT assessments were inconclusive based on four primary reasons, said Rosenberg. The GAO found challenges with the methodology on the program as a whole, she said. The reasons include the following issues.
Variation among grantee programs. Programs included in the ASPE evaluation varied in whether they enrolled participants with pending criminal charges and whether participants could be enrolled voluntarily. Enrolling participants with pending criminal charges and enrolling participants voluntarily (or declining to order individuals to participate if they do not agree) pose challenges to evaluating AOT as defined as a civil, involuntary process.
Self-reported data. Most of the data ASPE used in its evaluation were self-reported by program participants through interviews with AOT grantee program staff. Self-reported data are prone to bias, as individuals may not remember all incidents of interest, and they may be hesitant to candidly answer questions on sensitive topics such as substance use.
“Participants may have felt pressure to respond more positively [than] they could have and [that] depends on them being honest,” said Rosenberg. They were asked sensitive questions, such as have they used substances, been arrested or are they taking their medications, she said.
Participant nonresponse and risk of bias. ASPE did not conduct a nonresponse bias analysis to determine whether AOT participants who did not agree to be interviewed differed systematically from those who did. Participation in data collection was voluntary, and not all participants provided data (referred to as participant nonresponse).
“The assessment by HHS provided no information on how many participants chose not to respond as part of data information,” said Rosenberg.
Reduced ability to detect differences between AOT and voluntary treatment. ASPE’s analysis comparing AOT participants to individuals enrolled in voluntary treatment programs of similar intensity was limited to data from one grantee. The outcome report states that this analysis had a reduced ability to detect differences between the two groups. Factors such as small sample size, or a large amount of variation, can hamper the ability to detect true differences, particularly when differences are small.
The GAO did not make recommendations to HHS or to SAMHSA on how to improve data collection and program evaluation, Rosenberg said. “Some of the challenges they experience are inherent in the program,” she stated. For example, AOT is governed by state laws, which can create variability in the program. “Some challenges they face are not within their control,” she added. Also, AOT involves the courts, which can also vary in the way they operate their program and its level of involvement. “There’s a lot of variability,” she added.
Advocates weigh in
The Alliance for Rights and Recovery (formerly NYAPRS) says the new report only reinforces what the advocates have been saying for years. “People improve when they are connected to quality support — not because they are forced into treatment by a court order,” the Alliance stated.
“The GAO’s findings are yet another finding that the evidence for the use of involuntary outpatient treatment orders (AOT) is flawed, imbalanced and inconclusive,” Harvey Rosenthal, CEO of the Alliance for Rights and Recovery, told MHW. “In an environment that challenges the way federal funding has been used, this finding calls into serious question the value of the $146 million in assisted outpatient treatment grants that have been awarded to 63 grantees since the program’s inception in 2016.”
Rosenthal pointed to the 2009 Duke AOT study called for by the New York Legislature, which found that “it is difficult to assess whether the court order was a key ingredient in promoting engagement or whether comparable gains in engagement would have occurred over time with voluntary treatment alone.”
“That’s why the legislature authorized a new study of the AOT program that is currently being conducted by the University of Pittsburgh and Human Services Research Institute,” he added.
“The GAO’s evaluation highlights many important concerns about the SAMHSA’s AOT grant program,” Shanti Silver, senior AOT research advisor, told MHW. “We hope this publication prompts SAMHSA leadership to continue ensuring all programs who receive AOT grant funds utilize those funds appropriately and encourages SAMHSA to require more rigorous data collection efforts, beyond the limited, self-reported data required by the National Outcome Measures (NOMs).”
Silver added, “Improved data collection efforts are especially important given that, even while AOT expands nationally, there are still many unanswered questions about best practices for its implementation. Accordingly, we hope that SAMHSA leadership uses this report as an opportunity to improve the AOT grant program rather than abandoning a promising intervention for a critically underserved population.”
To review the full report, visit GAO-25-107526, SERIOUS MENTAL ILLNESS: HHS Assessments of Assisted Outpatient Treatment Have Yielded Inconclusive Results.