The review, led by Martha Canfield and Elena Dimova of Glasgow Caledonian University with colleagues from the University of Edinburgh, the University of York and the University of Worcester, searched Embase, Medline, Web of Science and Google Scholar, alongside grey literature databases and hand searches of reference lists. The team screened 688 titles and abstracts after deduplication, assessed 58 full-text manuscripts, and ultimately included 24 studies published between January 2010 and January 2025. Nine studies came from Canada, five from the United Kingdom, five from Australia, three from the United States, and one each from the Netherlands and New Zealand. Most were conducted in urban settings, and the study designs ranged from qualitative interview work and ethnography to retrospective administrative analyses and quasi-experimental comparisons with usual care.

From this body of work, the researchers identified four broad categories of transfer-of-care model. The most common, appearing in fifteen of the included studies, were co-response models, in which police officers are paired with mental health professionals under shared protocols to respond jointly to crisis calls. Examples span New Zealand’s Crisis Response Teams, Canada’s Mobile Crisis Rapid Response Teams and Mobile Crisis Intervention Teams, Australia’s Police Ambulance Clinician Early Response (PACER) programme and its variants, and the United Kingdom’s Mental Health Street Triage. The second category, liaison models, was represented by two studies: the Crisis Intervention Team approach in the United States and the Netherlands’ psychiatric ambulance, a dedicated vehicle staffed by clinicians who can administer sedatives and transport patients without police involvement. The third category comprised designated places of safety, such as crisis centres in the United States and Scotland’s legislative Places of Safety. The fourth consisted of structured screening tools, including the interRAI Brief Mental Health Screener and an electronic Mental Health Screener, both evaluated in Canada, designed to help officers assess mental state, risk and appropriate level of care at the point of contact.

The outcomes associated with these models paint a promising picture at the organisational level. Two quasi-experimental studies compared co-response approaches directly with usual care. New Zealand’s Crisis Response Team was associated with lower rates of coercive treatment and involuntary detention, reduced emergency department utilisation, less time spent in police custody, and a non-significant decrease in hospital admissions. Australia’s PACER model also reduced involuntary detention, although notably it was associated with an increase in hospital admission rates, a reminder that shifting people toward care can look different depending on which outcome is measured. Other co-response studies reported reductions in involuntary detention, emergency department use and the time police officers spent waiting at emergency departments or custody suites for handover to occur.

Some of the most striking quantitative findings came from the screening tools. In the Canadian evaluation of the interRAI Brief Mental Health Screener, police service calls rose by nearly 30 percent over the study period, likely reflecting increased officer awareness of mental health indicators, while hospital diversions and involuntary referrals each fell by more than 30 percent, suggesting more informed decision-making and reduced reliance on emergency departments and detention. The Dutch psychiatric ambulance study offered an equally dramatic result: the use of police car transport for people in mental health crisis dropped from 96 percent to just 1 percent after implementation. Hospital admissions remained stable overall, but involuntary hospitalisations declined, and clinicians noted that sedation administered by ambulance staff may be less traumatic for patients than physical restraint by officers.

Yet across all four model categories, one factor emerged again and again as decisive: communication. Evidence from Australia, Canada, the United Kingdom and the United States converged on the finding that when effective communication and inter-service collaboration are lacking, transfer of care breaks down. This applies at every stage of a person’s journey, from information sharing and joint decision-making between police officers and mental health professionals at the scene, through communication between triage teams and hospital services, to clear signposting and handover between hospital and community services. The Australian A-PACER evaluation illustrated what good communication looks like in practice. In that model, a police officer and a mental health clinician, stationed roughly a kilometre apart at a police station and a hospital psychiatric triage unit respectively, respond jointly to crisis calls, allowing assessments and care plans to be developed onsite. Service users reported that handovers to hospital staff were markedly better than previous experiences marked by communication breakdowns, and they valued the continuity of care as information was passed to case managers and community services after discharge.

Training emerged as a second critical lever. A Canadian study of Mobile Crisis Intervention Teams found that nurses and police officers often had limited understanding of each other’s professional cultures, prompting calls for cross-sector training: crisis de-escalation and mental health system knowledge for police, and safety awareness and police culture for nurses. In Australia, Mental Health Intervention Team training increased officers’ empathy, patience and confidence during mental health-related events and improved relations with health agencies. In the United States, a 40-hour Crisis Intervention Team training programme, covering psychiatric diagnoses, psychotropic medications and de-escalation skills, led officers to increase their use of a dedicated mental health crisis centre and decrease their reliance on emergency departments, and trained officers were willing to travel farther than untrained colleagues to reach appropriate care.

Despite these encouraging signals, the review’s central conclusion is sobering: the evidence base is not yet mature enough to demonstrate the broader value of any of these models. No randomised controlled trial was found among the 24 included studies, a gap that mirrors shortcomings identified in previous reviews of police responses to mental distress. Most studies examined organisational-level outcomes such as call volumes, detention rates and waiting times, while the experiences and outcomes of the people actually living through these crises received scant attention. Only five of the 13 studies collecting primary data included service users’ perspectives, and the social and demographic characteristics of study populations were rarely reported, suggesting the models largely operate on a one-size-fits-all assumption. Evidence on longer-term outcomes, including follow-up care, repeat crises and sustained engagement with community services, was particularly scarce, with mixed findings: the New Zealand co-response team reduced emergency department attendance within one month of follow-up, while the Australian PACER evaluation found higher post-detention hospitalisation rates than usual care.

The review also raises uncomfortable questions about equity and policy direction. Health equity was largely absent from the studies reviewed; the experiences of people from different ethnic and cultural groups, those whose primary language differs from the country they live in, and individuals with neurodevelopmental conditions were simply not addressed. The findings may also prove difficult to implement in jurisdictions where police are actively withdrawing from mental health-related work, such as England and Wales, where the Right Care, Right Person policy has been introduced to shift crisis response away from policing. The authors argue that interagency approaches must explicitly prioritise marginalised communities and that future models should embed trauma-informed understanding within mandatory police training, supported by improved mobile technologies offering real-time access to relevant health information for faster, safer and more coordinated decision-making.

The implications for policy and research are clear. The authors call for meaningfully involving people with lived experience in the design, delivery and evaluation of crisis services, for greater investment in rigorous and scalable studies to overcome the methodological limitations that have constrained the field, and for researchers, police, health services and commissioners to collaborate in building research capacity. The review followed established rapid review guidelines and did not undertake formal quality assessment of the included studies, a limitation the authors acknowledge, and the scope focused specifically on the police role in transfers, leaving the ambulance services’ contribution underexplored. Nevertheless, the review offers the first overview of this literature that has not previously been available, providing a foundation for future systematic reviews and research investment at a moment when demand for mental health-related police response continues to strain both policing and healthcare systems worldwide.