• March 1, 2023
• Public Safety

Who do you call for a medical emergency? Nowadays, we’re used to dialing 911 and receiving medical help in the form of an ambulance and trained emergency medical technicians (EMTs). However, before the 1970s, trained professional paramedics didn’t exist. According to the Journal of Emergency Medical Services the birth of emergency medical services (EMS) was in the late 1960s, thanks to a federal report that led to the standardization of emergency training and a national curriculum for “rescue squad personnel, police [officers], fire[fighters] and ambulance attendants.”

Yet who do you call for a mental health emergency? All too often, the answer is calling 911 and having the police show up — a response that can be deadly for people of color and those experiencing a mental health emergency or living with a mental illness. This question, and the history of EMS in the United States, is a crucial one for Jaymes Pyne, senior research associate at the Gardner Center at Stanford University and Thomas Dee, the Barnett Family professor and faculty director of the Gardner Center.

Roots of the Research

Over the past twenty years, cities and states have been experimenting with different responses to this question, by launching various types of mental health emergency teams. Although many are based on the Eugene, Oregon CAHOOTS model, they can take many different forms including sending unarmed mental health professionals, pairing trained peer counselors with EMS personnel, and sending out crisis trained police officers with a mental health expert. Dee and Pyne were interested in these programs being implemented in the Bay Area, and they were already in conversation with practitioners on the ground. But many of these programs are still at the pilot stage or are slowly expanding as funding and support grows; while there is evidence that these programs are working to divert those experiencing a mental health emergency into appropriate care, Pyne and Dee are the first to publish a rigorous research report on the impacts of the Denver Support Team Assisted Response (STAR) Program.

While Pyne and Dee research many different types of programs at the Gardner Center, they were initially drawn to STAR as they found it “radical” as one of the models that foregoes police altogether. There are many law enforcement agencies — including the Denver Police Department — that require police officers to undergo crisis intervention training (CIT), which aims to decrease stigmatizing attitudes among law enforcement officers and improve treatment of residents with mental illness or substance use disorder. However, according to Dee, “most of the studies on CIT police training show that it just doesn’t work that well alone.”

Once they began looking into STAR, interested to see how a non-police program measured, they were concerned that there wasn’t rigorous evidence on outcomes, good or bad, and discovered what Dee described as a “surprisingly large gap in the research.” In order to address this gap, Dee and Pyne decided to engage in a rigorous research project; in 2022 they pre-registered their work plan, attenuating the risk of “p-hacking” bias and building credibility.

Another aspect that made Denver an appealing study was the availability of transparent open data published by the city. While having access to the incoming 911 calls and how many were diverted from police and sent to clinicians would be ideal, Pyne and Dee could still analyze arrest data and show a reduction in crime with the publicly available data.

Lessons for Other Cities

The number one takeaway for other cities is the “robust evidence that [STAR] reduced reports of targeted, less serious crimes (e.g., trespassing, public disorder, and resisting arrest) by 34 percent.” Dee and Pyne do believe that this outcome can be replicated in other cities, and provided the following recommendations for local leaders hoping to implement similar programs.

1. Implement with intention. Every jurisdiction will be different due to local idiosyncrasies, but every municipality must take time laying the groundwork before starting even a pilot. One aspect is building relationships with the police, the community, and the treatment and health system to have everyone on the same page. Pre-implementation training is crucial, in particular training for emergency dispatchers as they will be the connection point for the community and the ones deciding which calls should be sent to police and which should be diverted. “If they get a call can the dispatcher understand the nuances in mental health, drug abuse, and behavioral calls?” asked Dee “And can the public?” The coordination, communication, and education required as a foundation will determine the success or failure of an alternative response program. 

2. Refine with data and evidence. Pyne and Dee recommend that cities use data formatively — and to consider partnering with university researchers if local leaders don’t have the capacity to measure outcomes themselves. Having strong impact assessments and evidence means that programs can be tweaked, leaders can pinpoint why something didn’t work, and what needs to be refined or amplified in each iteration. This does mean that a program like this will — and should — take time. “Promising, innovative policies can die a quick death when people expect their impact to be immediate,” cautioned Dee. 

3. Accomplish with collaborations. “There might be an urge to bring in a program at-scale to new cities,” said Pyne “but if you do that before getting local buy-in and making cultural changes, it could actually backfire.” Partnering with a broad and diverse group of stakeholders can help determine the best times of day to run a pilot, where teams should be operating, how many calls could be diverted based on previous 911 call data, and how to bring in clinicians and mental health professionals. Of course, residents are also a major stakeholder; running a successful alt- or co-response program “requires the hard work and good will of the community,” said Dee.

4. Analyze cost effectiveness. Another aspect of reimagining policing by including alternative or co-responses is understanding the cost effectiveness. According to Dee it is “four times as expensive to do the criminal justice route relative to the cost of STAR,” providing an economic justification for exploring similar programs. Over the six months they conducted research, not a single STAR call resulted in an arrest, representing significant cost savings. 

Changing Viewpoints

Despite the evidence from their research and Pyne and Dee’s recommendations above, implementing a program like this has unique challenges; the current national conversation about policing and police reform, funding, and racism means that many elected officials are treading lightly. However Dee “sees an opportunity for an uncommon political consensus around this work,” as alternative or co-response programs appeal to both sides of the current reckoning over police in America.

Many police officers support these types of programs since they know they aren’t well prepared for mental health crisis response. And this work is appealing to advocates who don’t want large police forces and operations, as these programs reduce police presence. “There is a fundamentally humane, common sense appeal to these reforms,” said Dee.

Part of the broader issue is in changing perspectives. “If someone trespasses — which is illegal — but community responders can work with them and get them help so they aren’t ‘punished’ but instead redirected to the hospital system…that’s changing how we view this,” said Dee. The researchers do point out that technically crime reduction isn’t necessarily the same thing as lack of crime; in the above example trespassing did occur. But “it definitely reduces crime to have a co-response,” said Dee “which is consistent with folks who need help actually getting it.

Pyne and Dee see their work as pioneering the conversation from the academic side, as there is “virtually no evidence base right now” — which could lead to haphazard implementation and actually damage the burgeoning field. “We’re at the beginning of an inquiry cycle,” said Pyne “but we’re building an evidence base to transform the way we do first response.”