How Communities Must Use 988 to Improve Care and Correct Crisis System Disparities

Most people cannot remember a time before 911, when people in an emergency dialed 0 for an operator or directly called or ran to the nearest police or fire station. For years, behavioral health¹ advocates and experts have pushed for a three-digit number for mental health, substance use, and suicide crises—a number that would be on par with and, hopefully, as ubiquitous as 911 is for medical emergencies. And, like 911, the innovation wouldn’t be just a number. Instead, experts envisioned a well-funded, interconnected emergency system for behavioral health crises, with a continuum of services and supports that would help foster parity while reducing stigma and criminalization.

The United States took a vital step toward this goal in July 2020, when Federal Communications Commission (FCC) Chairperson Ajit Pai announced a nationwide three-digit number that would “effectively establish ‘988’ as the ‘911’ for mental health emergencies.” Under Pai’s guidance, the FCC adopted rules requiring all phone service carriers to direct 988 calls by July 16, 2022, to the existing National Suicide Prevention Lifeline (2), which the Substance Abuse and Mental Health Services Administration (SAMHSA) funds. The Lifeline, run by Vibrant Emotional Health, is a network of about 200² accredited call centers. All 988 calls—and, potentially, texts—will go to the centers, filtered by area code unless callers press 1 to reach the Veterans Crisis Line, the national crisis call system operated by the Veterans Administration. Three months after Pai’s announcement, Congress passed the National Suicide Hotline Designation Act of 2020, supporting 988 implementation and allowing states to apply a monthly telecom customer service fee to pay for 988-related services. Many communities across the U.S., increasingly aware of gaps in the overall crisis response continuum, had already started or planned to re-engineer and build out their systems, including mobile crisis services and crisis facilities. Expansion plans have also included the evolution of crisis call centers into hubs that can dispatch mobile crisis teams and provide real-time data on and access to crisis facilities. As states move toward 988 readiness, communities must capitalize on this momentum to meet local needs.

This paper describes examples and new resources to improve the care of people experiencing behavioral health crises. We focus on imminent opportunities and provide links to resources helpful in regional or local crisis redesign efforts.

¹ The term behavioral health has evolved to include both mental and substance use disorders.
² As of November 2021, there were 194 Lifeline accredited call centers, with seven more in the final onboarding phase.

In recent years, crisis care reform has expanded to include not only addressing long-standing problems in providing appropriate care to people in mental health crises but also ensuring that people experiencing substance use crises receive such care. This expanded framework for reform has been accompanied by greater recognition of inequities associated with reliance on law enforcement as the de facto responder for behavioral health and quality-of-life crises. Tools such as the Sequential Intercept Model, developed by Munetz and Griffin (4), and resources promoted by SAMHSA’s GAINS Center, as well as Stepping Up (5), a national initiative to reduce the number of people with mental illness in jails, have helped advance this cause. However, a confluence of factors has magnified the need for change, including the COVID pandemic and increased awareness of racial and ethnic disparities in law enforcement interactions brought into focus by the murder of George Floyd by four Minneapolis police officers in May 2020.

As communities across the nation debate the role of police and explore alternative response programs, Ron Bruno, a retired Utah police officer and executive director at Crisis Intervention Team (CIT) International (10), has pointed out that while police have a critical role in mental health crisis response, it is in the position of support and only when necessary. He notes, “We have to challenge the belief that mental health crisis services must come in a police car” (11). According to Bruno, most calls to 911 don’t require a police response, and call takers can transfer calls to a crisis line where “the majority of calls, 80% and upward, are resolved at that level, and there’s no need for police involvement.”

The CIT program, developed by now-retired Memphis Police Maj. Sam Cochran, is a first-responder, police-based crisis intervention designed to reduce the role of law enforcement in behavioral health crisis response. The program in Memphis was one of the earliest and most critical innovations to address the criminalization of mental health, and it was sparked by a lethal police interaction that took place more than three decades ago. In 1987, police officers in Memphis responded to a 911 call from a mother worried her son would kill himself (12). Joseph DeWayne Robinson had cut himself 120 times with a butcher’s knife. He had schizophrenia and was experiencing paranoia. When police officers arrived at LeMoyne Gardens public housing project, the perimeter was tight, and they asked Robinson to drop the knife. He didn’t. The officers shot and killed him. They said Robinson lunged at them, but witnesses said he did not. The killing resulted in community outrage and prompted racial bias charges against the city’s administration. Robinson was Black, and the two officers were White.

Following the police killing of Robinson, Maj. Cochran led a diverse stakeholder effort and law enforcement commitment to develop teams specially trained to work with people in a mental health or substance use crisis. However, although many communities have emphasized the police training approach, they have often focused less on another core component of the model: community partnership between law enforcement, behavioral health, people with lived expertise, and other advocates. Maj. Cochran has said that the platform is designed, through robust collaboration, to remove system barriers that prevent people with mental illness from getting the help they need (13). That includes ensuring that there are crisis facilities where law enforcement can, obstacle-free, take people in distress for care—instead of booking and custody. These facilities are an essential component of a comprehensive behavioral health crisis system that continues to be absent in many communities.

The Council of State Governments’ Justice and Mental Health Collaboration Program also fosters partnership between criminal justice and behavioral health systems (14). Integrated into the program are 14 peer-to-peer learning sites that have demonstrated best practices when responding to behavioral health issues and houselessness. Linking crisis systems, starting with 911, should be a primary target for communities as they plan 988 implementation. Less examined, but also vital, is how 988 will connect to social service numbers for people experiencing quality-of-life challenges—sometimes 311 or 211, depending on the community—to ensure linkage beyond “a warm handoff.”

Historically, states and counties have developed behavioral health crisis systems without input from the community and essential partners and then expected them to fall in line. To ensure that system reform does not repeat the same mistakes, partnership and cooperation must be part of the 988 system’s development at every step of planning and implementation. That doesn’t mean considering only the populations a system serves, but instead convening a work group that includes these populations along with other stakeholders—law enforcement, hospitals and emergency departments, state and local behavioral health officials, Lifeline-988 call centers, judges, prosecutors, public defenders, social workers, peers, advocates, and representatives from city, county, and state agencies.

Using a community strategic planning tool to inform its recommendations, the work group should track how people with behavioral health and quality-of-life challenges come into contact and move through the criminal justice system. For example, the Sequential Intercept Model does just that at critical intercept points: community services, law enforcement, initial court hearings and detention, jails and courts, reentry, and community corrections (15). Communities, such as Los Angeles County, have used a modified version of the model (16), which adds intercept 00 (17), focusing on a holistic, community-based, system-of-care infrastructure and recommendations to prevent justice system involvement and for community reentry after incarceration. In addition, many communities already have cross-system and consumer-family representation work groups and collaborations through participation in the previously mentioned Stepping Up initiative, a national effort sponsored by the Council of State Governments (5). The initiative has four key jail diversion measures for people with mental illness: reduce the number of people booked into jail, shorten their average length of stay, increase their connections to treatment, and lower recidivism rates. Other communities are turning to their CIT steering committees, which seek to grow the program’s efforts beyond police training and toward a more systematic approach. Some communities are doing all three. These efforts provide a strong launching pad for implementing 988 to strengthen crisis response.

What has led to the high involvement of police in psychiatric and quality-of-life crises is multifold and based on precedent, expectation, and the vast underfunding and insufficient reimbursement of behavioral health and community-based services. Police also become involved because 911 is an easy-to-remember number that people know to call in an emergency, and public-safety answering points (PSAPs)—such as 911 call centers—are GPS-enabled, allowing first responders to respond quickly. Although 911 is the number people dial to reach police, fire, and emergency medical services, it has long been the number people call expecting a law enforcement response, even when the call is nonemergency and there has been no crime. In fact, when 911 launched in New York City on July 1, 1968, calls for a law enforcement response spiked by over 40% (going from 12,000 to 17,000 calls a day [18])—and more than half were nonemergency (19). Years later, a study analyzed 1977 data from over 26,000 911 calls to police in Rochester, St. Louis, and Tampa-St. Petersburg (20). The researchers found that most calls were for nonemergencies. Less than 20% were for what callers perceived to be criminal incidents of any kind. Yet callers continue to depend on an in-person police response.

Many 911 callers are not people in a behavioral health crisis themselves, but instead they are third-party callers reporting “strange behavior.” This dynamic creates a system ripe for blind spots because of its dependence on an observer, a viewpoint easily influenced by a person’s biases, to identify what another person might be experiencing. Dr. Rebecca Neusteter, executive director at University of Chicago Health Lab and former policing program director at the Vera Institute of Justice, has said that the nature of the 911 call taker role is “to hear one person’s perspective of a situation and treat it as though it’s true” (19). An analysis of more than 18 months of 911 call data in Austin, Texas—roughly one million calls—revealed that 64% of mental health–related calls originated from disturbance and trespassing incidents and requests for welfare checks on people (21). A similar study in Atlanta found that of 3.4 million 911 calls in the metropolitan area over 3.8 years, 600,000 calls (18.4%) may have been suitable for diversion from 911 (22). The researchers labeled these calls “calls of focus.” Most were coded as “suspicious person,” “directed patrol,” “criminal trespass,” “street/sidewalk hazard,” “person injured/down,” “demented person,” or “public indecency.” None were coded as mental or behavioral health. Also, a multicity 911 analysis conducted by Dr. Neusteter and her former colleagues at the Vera Institute of Justice revealed that call takers were unlikely to classify calls the same way (23). The lack of uniformity and transparency of 911 protocols and the local nature of 911 PSAPs make it challenging to divert people experiencing mental health, substance use, or quality-of-life crises.

It is likely, at least in part, that because of the obscurity of 911 call coding, criminal justice diversion has primarily focused on pre-arrest diversion as the starting point, with communities turning to co-responder models (teams that comprise a police officer or emergency medical technician partnered with a behavioral health professional) and alternative street responder teams. However, diverting people experiencing behavioral health or quality-of-life crises from a law enforcement response requires understanding what leads to police involvement in the first place, which often happens at the 911 call center, making it dire that communities crack open their 911 “black box” (23). With a 911-988 partnership, calls to 911 can be diverted to 988, and crisis resources—such as mobile crisis teams—can be dispatched instead of police officers. Los Angeles, Harris County, Austin, and Tucson are already working to improve linkages between their local 911 PSAPs and behavioral health crisis call centers. Sometimes that includes physical or virtual co-location of call counselors at 911 call centers, who are integrated into the PSAP computer-aided dispatch system, commonly called CAD.

For example, in Austin, Integral Care’s Expanded Mobile Crisis Outreach Team (EMCOT) clinicians are physically co-located at the Combined Transportation, Emergency, and Communications Center/Austin Police Department 911 Call Center and work on the same platform (24). When a person dials 911, the call taker asks, “Are you calling for police, fire, EMS, or mental health services?” The integration allows call counselors to see incoming calls and identify those with a behavioral health nexus that they can divert. Before integration with the 911 center’s communication system, Marisa Aguilar, the EMCOT’s program manager, said call takers were transferring calls over to the service by phone, and “most calls were getting dropped” (24). Since co-location, the Austin program has diverted roughly 85.4% of behavioral health calls from law enforcement and transferred them to their EMCOT clinicians (25).

Physical co-location of call counselors at a 911 PSAP, such as in Harris County, Texas, helps bridge the gap between two very different disciplines as they develop trust through partnership, continued communication, and a streamlined feedback loop. The Harris County program also includes dedicated mobile teams to respond to diverted 911 calls (26). However, physical co-location is not an option everywhere, especially in rural communities and those with a high number of PSAPs. Los Angeles County, for instance, has 78 primary 911 PSAPs. Although county leaders plan to co-locate their 988 call center with the local PSAPs (27), they will do so virtually (28).

In Tucson, the COVID pandemic forced call counselors to shift to virtual co-location temporarily. However, it created an opportunity for more call takers. Crisis line leads in Tucson are now trained to pick up calls from the Public Safety Communications Department if the two crisis professionals trained on the CAD dispatch system are already on 911 calls. The diversion program plans to increase the number of call takers on the crisis line who are CAD-trained so it can take on even more 911 calls and is actively looking to expand its virtual co-location to other 911 PSAPs in Arizona. During the pandemic, the stabilization rate after a dispatched Community Bridges or Community Health Associates mobile crisis team has interfaced with a person in crisis has hovered around 85%, regardless of whether the person called 911 or the crisis line (29). However, more 911 calls—roughly 60% of the 911 calls that the team diverts—result in a mobile crisis team dispatch. When people dial 911, they expect an in-person response.

Designing a robust 988-911 partnership can be maximized only if there is a reliable corresponding behavioral health crisis system to which first responders can divert people. However, this is not the case in many communities. When law enforcement does reach out to crisis care providers, they often experience inconsistent mobile crisis unit response or get tangled in a web of provider exclusion, such as when the provider deems the person too aggressive, acute, or intoxicated. Or providers require police officers to first take the person to a hospital emergency department, commonly known as the emergency room, to get “medical clearance”—a medical evaluation of a patient whose symptoms appear to be psychiatric. Communities should address this barrier in their crisis system redesign by building capacity for brief medical assessment. Unfortunately, these hurdles deter first responders from reaching out and continue to thrust them into the role of de facto behavioral health responder while taking them away from their primary responsibilities.

Although law enforcement has been the default behavioral health crisis responder, the emergency department has been the default provider. Because most communities do not have a robust behavioral health crisis continuum, the emergency department is the go-to place for people with a broad range of mental health needs, including those who have run out of medication or do not have anywhere else to turn. But when they get there, they wait. People wait for urgent psychiatric care and then wait for continued care, sometimes placed in an observation room or strapped to a gurney in the hallway. Wait times in the emergency department can be hours or days—sometimes weeks or months.

In February 2020, SAMHSA published national guidelines for mental health and substance use crisis care (34). This best-practices toolkit identifies four core elements that an integrated crisis system needs: regional or statewide crisis call centers that coordinate in real time, centrally deployed 24/7 mobile crisis services, 23-hour crisis receiving and stabilization programs, and essential crisis care principles and practices. Threaded throughout the guidelines is the recommendation for partnership with first responders and a “no wrong door” approach for rapid diversion, in which crisis service providers have a no-refusal policy, accepting all referrals, walk-ins, and first responder and mobile crisis team drop-offs.

What 911 is for medical crises, the regional or statewide 988 call center will become for behavioral health emergencies. The envisioned capacity is sometimes called a call hub or “care traffic control.” According to a 2018 survey of Lifeline centers, nearly 98% of crisis calls to accredited centers are de-escalated, diverting people from costly and highly restrictive law enforcement and emergency medical service responses (35). To divert calls from law enforcement, other first responders, and the emergency department, call hubs must be well connected with 911 call centers, first responders, and all other parts of the behavioral health crisis system, such as mobile crisis services and 23-hour crisis receiving and stabilization programs.

As states plan to build their own regional or statewide call hub, many seek to emulate features of the Georgia Crisis and Access Line call center, commonly called GCAL. The hub is data driven and uses technology to coordinate care, giving people a single entry point into the behavioral health and developmental disabilities crisis systems. The center provides phone, text, and chat intervention services, and call counselors can make urgent and emergency appointments for callers. In addition, call counselors have real-time access to available crisis and detox beds throughout the state and 24-hour mobile response coverage within 100 miles of every Georgia community, using GPS, if needed, to dispatch a mobile crisis team to the person in crisis.

Determining potential 988 call volume is still a bit of a guessing game. It is clear, though, that maximizing 988 will depend on marketing—so that people call 988 instead of 911—and how well communities use 988 to redesign their crisis system, with improved interconnections within the system and between 988 and the 911 emergency response system and first responders.

Another vital diversion target is to ensure that there is an alternative to law enforcement handling crisis call-outs. Mobile crisis teams, often including a behavioral health professional and a peer specialist, should be the first line of defense when people in crisis need more than telephonic support. Coordination and the development of shared protocols can clarify the circumstances in which the mobile crisis team—instead of 911 first responders—can handle a call and those situations (e.g., presence of a weapon at the crisis) where co-responding is appropriate. In Tucson, when integrated call counselors at the 911 call center need to dispatch a mobile crisis team, they tap into the crisis line’s dispatch optimization tool, an app connected to its electronic health record system. The app is GPS-enabled mobile phone software that allows call takers to see the location and status of each mobile crisis team and deploy one that is available and closest to the person in crisis (6). In addition, teams are color coded: green means a team is available, yellow means a team is activated and headed to a call, and gray is when a team is offline. However, across the United States, mobile crisis services are insufficient and unevenly developed, and—unlike the examples described in Austin, Harris County, and Tucson—they are often not connected to 911.

Although a crisis call or mobile visit can resolve most crises, some people require facility-based care. A person’s need might be as simple as a medication refill or counseling visit, assessment, brief stabilization and observation (sometimes called 23-hour care), or a short stay in intensive but out-of-hospital respite or residential treatment. If existing clinics maintain “open access” or walk-in capacity for people in crisis, the crisis system does not need to duplicate this. For example, as part of its crisis system redesign, the Greater Baltimore Regional Integrated Crisis System—GBRICS—is helping re-engineer existing community walk-in clinics so that they can reliably provide emergency access, rather than creating a new crisis clinic. Most communities have access to some residential options for more intensive crisis care, but capacity is usually inadequate. Also, available crisis facilities are often limited; they typically provide only short-term respite beds, exclude actively intoxicated people, and do not have intensive assessment/observation capacity. Also, many require that the person first visit an emergency department for medical clearance.

The “Arizona model” includes state-of-the-art crisis facilities. Statewide, this is in part because Medicaid in Arizona operates under a broad Section 1115 waiver, allowing for a managed care model, and provides a per member, per month payment for crisis services, which funds regional authorities to establish contracts for and oversight of crisis services. Also, a 2014 lawsuit settlement on behalf of people with serious mental illness in Maricopa County highly influenced crisis care in Phoenix (38). The settlement required community-based services and programs for people with serious mental illness and resulted in an ongoing partnership between law enforcement and behavioral health service providers that includes rapid mobile crisis team response times and a “no wrong door” policy, by which mobile crisis and local crisis facilities accept all first responder drop-offs and referrals (39). Crisis facility providers that pioneered the Arizona model are Connections Health Solutions (40) in Tucson and RI International (41) in Phoenix. Crisis facilities that offer a superior, comprehensive approach should include:

• Three levels of crisis care: clinic, 23-hour assessment/observation, and residential crisis stabilization;
• Medical and nursing staffing comparable to inpatient units, enabling comprehensive assessments and eliminating the need for outside medical clearance;
• Integration of peer support specialists;
• Special arrangements, such as sally ports, to allow expedited direct drop-off by police of people needing care. A sally port has two doors or two sets of doors set apart so that the first can be closed and locked before the second is unlocked.

Regional crisis systems should also offer a peer-run and peer-staffed behavioral health respite program that is an alternative to traditional professional services for people experiencing a mental health crisis.

Some communities have begun their 988 crisis system redesign by just building out some of the core elements in SAMHSA’s national guidelines for mental health and substance use crisis care, which will result in continued system challenges. A call center hub provides rapid de-escalation for most callers and coordination with mobile crisis services for those who need an in-person response. However, a 988 call center without ongoing 911-988 collaboration will continue to be siloed. Without mobile crisis services, law enforcement will continue to be the default behavioral health responder for people who need an in-person response. However, if communities build out the above but not crisis facilities, the emergency department and jail will remain the default behavioral health providers because law enforcement will have nowhere else to take people.

A long-awaited moment to increase access to behavioral health services and quality-of-life support for people experiencing behavioral health crises while reducing reliance on law enforcement, emergency departments, and jail has arrived. The creation of 988 and recognition of the need to enhance crisis response systems have resulted in national legislation, funding, and planning in state governments—often led by mental health agencies. However, comprehensive reform will not be “rolled out” of Washington or state capitols. It requires community involvement and leadership to capitalize on the opportunity and explore local or regional collaboration and re-engineering. Because of the vast variability in how law enforcement, mental health, and health care systems are organized, a local approach is necessary to ensure that crisis services and crisis responder models work in the way the community wants to be served.

The 988 initiative has arrived at a time of concern about mental health and addiction problems and new awareness about better solutions, giving states and communities an opportunity to correct crisis response disparities and redesign systems of care. By building systems of care and support for a community’s most marginalized, disenfranchised populations, these systems work better for everyone.