Forced Hospitalizations Can Increase Violence, Suicides, and Homelessness

A groundbreaking study finds that, at least for some groups, involuntary commitment backfires

A study was released this week that examines a number of key effectiveness outcomes of civil commitment. It produces bleak findings that are leaving some psychiatrists “incredulous”—but will not be surprising to many people who’ve personally experienced forced psychiatric hospitalization.

The paper is absolutely ground-breaking, and easily the most scientifically-sound study of the practical “effectiveness” of involuntary commitment on clinical outcomes ever done (as distinct from, say, studies surveying patients’ feelings about coercive experiences). It’s not difficult for me to say this somewhat definitively because, as I’ve written about often, even though coercion is a central feature of our mental health systems, there are very few studies of effectiveness outcomes globally, and there’s no evidence of the effectiveness of civil psychiatric commitment for helping people or for protecting public safety. As one (pro-force) psychiatrist wrote, involuntary commitment is “based on tradition rather than evidence.”

All of which makes this study still more interesting and important, as the researchers sought to understand the causal impacts of involuntary commitment on people’s subsequent health, mental health, safety, employment, housing status, and more.

I’ve divided my review into sections (and 15 mins total reading time):

• What They Measured and Why

• “Who” They Assessed

• How They Measured

• The Findings

• The Study’s Broader Significance

And trigger warning for anyone reading this who works in the business of civil commitment: The findings are not pretty.

What They Measured and Why

The paper is titled, “A Danger to Self and Others: Health and Criminal Consequences of Involuntary Hospitalization.” It’s authored by Natalia Emanuel of the Federal Reserve Bank of New York, Pim Welle of the Allegheny County Department of Human Services, and Valentin Bolotnyy of the Hoover Institution, Stanford University. They also wrote an associated Frequently Asked Questions (FAQ) document which summarizes some of their key findings.

While most researchers in this field have stakes in implementing civil commitment, these researchers came with unconflicted interests in the related policy and economic questions. Namely, the most common alleged or claimed intentions of civil commitment statutes are to reduce danger to self and danger to others, and to help prevent negative life consequences such as ending up homeless as a result of a mental disorder. These researchers decided to investigate: Does forcibly hospitalizing people help achieve these goals?

As the researchers note, at more than 300 detentions per 100,000 people, the low-end estimate of rates of psychiatric incarceration in the U.S. are nearly as high as the rates of imprisonment in this most carceral nation on Earth. So, when many states are enacting laws to allow more and longer involuntary commitments, should we not have at least some evidence this will actually help more than harm and backfire?

In Allegheny County, Pennsylvania, where the city of Pittsburgh is located, the researchers gathered and reviewed evidence from 16,630 evaluations for involuntary hospitalizations done by 424 physicians at 14 different hospitals over ten years up to 2023. (The rate of forced psychiatric detentions in the county is 357 per 100,000 people.)

They then sought to determine “how involuntary hospitalization changes the probability that a person is a danger to themselves… or a danger to others.” Their main measures were deaths by suicide or overdose, and violent crime charges from the FBI’s Uniform Crime Reporting Program during the ensuing three to six months after hospitalization. They also looked at a number of other related potential impacts that could be influencing deaths and violence, such as impacts on employment, mental health service use, and housing status—generally, they looked for concrete measurables linked to stated policy goals.

“Who” They Assessed

In Pennsylvania, laws allow police, medical professionals, and social workers—but also family members, friends, colleagues, neighbors, or virtually anyone—to call a public hotline “to request that the individual be evaluated for involuntary hospitalization.” A staggering 84% of these calls lead to people getting taken forcibly by police to hospitals.

After that, an emergency room physician or psychiatrist will evaluate the person and either release them immediately or transfer them to a psychiatric ward/hospital where they’ll be detained for up to 5 days. Through a hearing process, they can then be detained longer.

Out of 16,630 evaluations, 78% resulted in a decision to forcibly hospitalize an individual for up to 5 days. After that, about 37% of those got extended up to 20 days or longer.

One incidental finding is worthy of note: In Pittsburgh, a large urban center with a significant homelessness crisis, only 4.7% of detained psychiatric patients were incarcerated in the previous year and only 3% had spent any time in homeless shelters. On the contrary, the researchers wrote, “those who are evaluated for involuntary hospitalization include many individuals who are actively engaged in society, including through formal employment.”

In Your Consent Is Not Required, I argued that the hyper-focus by many politicians and news media on unhoused populations and civil commitment too often serves only to distract from the separate policy questions to do with both of these distinct issues—these numbers from Pittsburgh are further evidence to substantiate that.

Another of the researchers’ key findings was that the rates at which different physicians decide to hospitalize people varies widely—even when these doctors are in the same hospital emergency rooms and/or dealing with similar patients. The researchers identified 46 physicians who forcibly hospitalized literally all of the people they evaluated, while other physicians incarcerated as few as one in ten people they evaluated.

Again, as I and others have long argued, psychiatric diagnosing and determinations of the “need” to forcibly hospitalize are wildly subjective—and on that score, this confirmatory evidence alone hits home heavily.

This also points towards a logistical problem for an outcomes study.

How They Measured

If one finds negative, “dangerous” consequences of forced hospitalization, it’s easy for others to suggest this is just correlation rather than causation e.g. “all of those people were already dangerous to begin with.” Certainly, doctors who incarcerate 100% of the people they evaluate are going to argue that!

But these researchers rose to this challenge in a number of innovative ways. They openly explain their methodologies at length and in detail, and transparently test various assumptions and hypotheses on their data—which is another major strength in the paper, and I strongly encourage anyone with questions to read those sections.

I will generally summarize it in this way: The researchers cut off the extremes and focused on the mid-range of physicians who incarcerated either more or less frequently—and then they focused on patients who were first-timers being randomly assigned to these physicians. The researchers managed to identify key traits of the patients who were most often “judgment call” cases—that is, the types of people who would be evaluated and then would only sometimes get incarcerated for five days or longer by physicians in this middle range. These people then became the study’s comparison groups—very similar people, separated into those who got forcibly hospitalized for five days or more versus those who instead got immediately released.

As the researchers themselves acknowledge, their sophisticated statistical maneuvering to “find” this group of “judgment call” patients leaves room for possible invisible confounding errors or improper extrapolations to different types of patients or scenarios. On the other hand, a real strength of their approach is that these particular people comprise the most “policy-relevant” population. That is, when we talk about passing legislation or establishing policies aimed to dial up or dial down how many people get involuntarily committed, it’s these kinds of “judgment call” cases that will be most affected.

So, what did the researchers find?

The Findings

“[T]he aim of any policy is to achieve its explicitly stated goals,” the researchers comment. Yet they found that, for their study population, involuntary commitments quickly backfired, and created higher risks of danger to self and danger to others.

For these “judgment call” cases, they found, “involuntary hospitalization nearly doubles the risk of being charged with a violent crime in the months following evaluation.”

Involuntary hospitalization, they found, also “nearly doubles the risk of dying by suicide or overdose.”

In more detail, they state:

“involuntary hospitalization increases the probability of being charged with a violent crime by 2.6 percentage points (off of a base of 3.3%) and increases the probability of death by suicide or overdose by 1.0 percentage points (off of a base of 1.1%) in the three months following an evaluation.“

The researchers also found that involuntary hospitalization “causes destabilization” in a myriad of ways.

Among these “judgment call” cases where some physicians might have hospitalized the person and other physicians wouldn’t have, the people who were hospitalized for up to five days or longer suffered an ensuing overall 19% decrease in employment earnings over the following few months.

These people also ended up using homeless shelters more than people who hadn’t been hospitalized, when they’d not used a shelter at all in the previous year.

Forcible hospitalization also did not make these people more likely to continue to take medications or engage with outpatient mental health services than people who had not been forced.

The researchers could even track negative impacts to specific physician tendencies e.g. “a 10 percentage point increase in a physician’s tendency to hospitalize is associated with a 2.6 percentage point increase in being charged with a violent crime within the following 3 months.”

The policy implications, of course, are huge — moving ahead, there’s every reason that this study should be regularly cited and discussed in public discussions about involuntary commitment. It’s evidence is basically saying that, if you expand involuntary commitment statutes or change policies to try to capture more of the people who are currently ‘sometimes slipping through the cracks,’ doing so will make those people more likely to die, more likely to become violent, and more likely to become homeless.

The Study’s Broader Significance

It’s important to re-emphasize that their findings only applied to the “judgment call” cases they studied—for example, someone who met all of the criteria to very likely be hospitalized by all of the psychiatrists in the study might have different outcomes.

Nevertheless, as Awais Aftab, an increasingly high-profile psychiatrist who writes a Substack column (and interviewed me two years ago), notes, this study “upends” conventional thinking in psychiatry. Aftab says:

“This is the opposite of what most psychiatric clinicians are actively taught… Clinicians are trained to be conservative, to be risk-averse, to be “better safe than sorry.” When it comes to psychiatric commitments, the clinical culture in the United States is to involuntarily admit a person if one is in doubt. The findings by Emanuel and colleagues indicate that we should be doing the reverse.”

Aftab says he was “incredulous” when he saw these findings and “felt sure that there must be a confounder at play… some factor other than hospitalization.” However, he ultimately concluded, “the analysis is quite rigorous and robust and… the results are genuine.”

However, Aftab then concedes that the root problem may be that helping people and reducing harms are not actually the goals of many involuntary commitments, anyhow—instead, the goals are often self-protection of the clinician and institution. He comments:

“In many cases, the true purpose of involuntary hospitalization is risk management and liability control rather than meaningful improvement in clinical outcomes. You may think I am being cynical, but I am being frank, as a psychiatrist who is intimately familiar with what happens behind the scenes.”

It’s a sad admission—and it’s also something I’ve heard often and, so far, none of Aftab’s many clinician followers seem to have disputed it in his post’s comments forum, either.

Nevertheless, it actually reinforces the value and importance of this paper: The study demonstrates that psychiatric incarceration worsens key risks—and therefore, in future, medical staff could conceivably be held liable for worsening those tangible, risky impacts by detaining someone. Such a case would be all the more possible to make if the physician in question knew they were increasing those risks through hospitalization.

So, tell every psychiatrist you know about this study—if nothing else, it should make them more thoughtful, careful, and balanced in their decisions.

Direct link to this post.