Thirteen records were identified for inclusion through the systematic search process (Fig. 1).

Flow diagram of data extraction

Thirteen publications were included [14,15,16,17,18,19,20,21,22,23,24,25,26], and eight services were identified. This included five peer-reviewed articles [14, 17, 23, 24, 26], one evaluation framework [25], and seven grey literature reports [15, 16, 18,19,20,21,22]. From the thirteen publications, nine provided information about service characteristics [14,15,16,17,18,19,20,21,22] and twelve provided information on service outcomes (two were specifically focused on visitors’ experiences) [14,15,16,17,18,19,20,21,22,23,24, 26].

All publications presenting data on each of the services included were subjected to quality appraisal [14,15,16,17,18,19,20,21,22,23,24, 26]. Complete appraisals and criteria explanation are provided in supplementary material (File 2). Two authors (DP and SP) completed the appraisals using the relevant JBI checklists (economic evaluations, qualitative research, and text and opinion) [27,28,29]. Disagreements were resolved via collaborative discussion.

The methodological quality must be considered when drawing conclusions about person-centred crisis support services from the publications provided. None of the publications were assessed as high-quality; three were fair-quality (two peer-reviewed qualitative studies and one grey literature report) [17, 19, 23], and, therefore, most of the publications included were rated as poor quality for the purposes of this review [14,15,16, 18, 20,21,22, 24,25,26]. Most common issues included limited description of methodology, rudimentary cost-effectiveness analyses, and inadequacy of methods such as survey methodology for interpretation of qualitative outcomes and trend-based assessments for quantitative and qualitative outcomes.

From the nine publications describing service characteristics [14,15,16,17,18,19,20,21,22], the main aspects identified in all services were that they aim to offer one-to-one support in a safer environment than EDs to people experiencing mental health crises where they may feel listened to and respected, avoid long waits and be connected to other community services (see Table 1). Key features included the leadership from peer workers with lived experience of mental health crises, the non-medical, calming, and relaxing environment, and the option of self-presenting to the services, which was considered an important aspect for avoiding EDs and long waiting periods. Only one service seems to differ in some of these core features, The Crisis Hubs [18], as they do not mention the lived experience workforce or a non-medical environment and have a team of medical staff. Three of the services are described as “Crisis Cafés” or “Safe Haven Cafés” [15, 16, 20, 21] but there are no specific differences from the other person-centred crisis intervention services mentioned, except that they are designed to resemble cafés. Two services are hospital-based [18, 21], while all others are community-based; no other differences were indicated between the two types of services.

In the services with peer workers, these staff are responsible for most of the support provided, including the one-to-one chats and working collaboratively with the visitors to develop their coping skills and safety planning. Clinicians provide support with clinical skills and external referrals when required. Although all services were presented with a general description, only one provided a comprehensive description of staffing roles and training procedures, as well as the physical space (where they discuss the choice of non-clinical furniture to compose a non-medical environment), and therefore, the only one to describe the layout of a non-medical environment [14]. A more thorough description of these aspects from the other services would be essential to better understand how they function.

Twelve publications displayed information on service outcomes [14,15,16,17,18,19,20,21,22,23,24, 26]. Findings from all twelve publications, alongside study design, methods and quality assessment are detailed in Table 2, and a summary of key findings is displayed in Table 3.

As illustrated in Tables 2 and 3, all services that reported on visitors’ experiences were considered safe and effective. Other key findings were that visitors reported that they were treated with dignity and respect, had a better experience than at EDs and avoided long wait times by visiting the crisis support services instead of the EDs [15,16,17,18,19, 21,22,23,24].

In the only two studies that focused specifically on visitors’ experiences (two fair-quality studies according to quality appraisal), participants reported that the services provided feelings of connectedness and being cared for without judgment and discrimination, where they can go for help and respite during a crisis. In contrast, their ED experiences were characterised by feelings of insecurity, loneliness, intimidation and stigma, which may exacerbate a crisis rather than resolve it [