Key Strategies for Improving Psychiatric Home‐Based Care in Spain: A Modified Delphi Study

ABSTRACT

Aims

This study aimed to identify and prioritize key strategies for improving a psychiatric home‐based care programme, the Crisis Resolution and Home Treatment (CRHT) intervention in Catalonia, Spain. The objective was to incorporate the perspectives of service users, family caregivers and healthcare professionals to guide quality improvement efforts.

Methods

A modified Delphi method was used to reach consensus among stakeholders previously involved in a qualitative evaluation of the CRHT programme. The expert panel included 130 participants: 31 CRHT service users/caregivers and 99 healthcare professionals. Over three rounds of online surveys, participants rated 18 pre‐identified improvement strategies across three dimensions: problem significance, change agency and feasibility of change. Service users and caregivers assessed only problem significance. Open‐ended responses were also collected and analysed.

Findings

Consensus was reached in the third round, with five strategies prioritized: (1) training on CRHT referral and eligibility criteria, (2) standardized discharge and continuity of care protocols, (3) feedback loop between CRHT and referring professionals, (4) scheduling and communication of visiting hours, and (5) available updates for CRHT slots. These areas were rated highest in terms of feasibility and change agency, though not necessarily in problem significance. Notably, the same individuals who participated in the initial qualitative phase contributed to the prioritization process, ensuring continuity and relevance in stakeholder input.

Conclusion

The study demonstrates the value of collaborative, consensus‐based approaches to quality improvement in mental health care. Prioritized strategies reflect feasible and actionable areas for enhancing CRHT services. Involving service users, caregivers and professionals in both evaluation and decision‐making processes strengthens the relevance, person‐centredness and effectiveness of care. Implementation of these strategies is underway, with further improvements planned based on stakeholder feedback and continuous assessment.

1. Introduction

Home‐based psychiatric care, also known as home‐based psychiatric hospitalization, offers an alternative to inpatient admission. Through daily home visits, a multidisciplinary team provides psychiatric interventions, facilitating person‐centred care and contextualized crisis management [1, 2, 3]. Different home care models have been implemented to promote community‐based care in psychiatric crises, and experiences with home‐based management in similar contexts have contributed to the deinstitutionalization of mental health crises [3, 4]. These approaches mitigate the stigma associated with crises, and users and their families highly rate the quality of care delivered through these home‐based models. Evidence also indicates that they effectively support crisis resolution, with benefits observed across clinical, functional and social domains [5]. Home care fosters closer and more reliable care by balancing the power relationship between professionals and users and their caregivers [6, 7, 8].

The deinstitutionalization of mental health care, implemented several decades ago, has promoted the integration of community services and the family and social environment in the recovery process. However, adjustments are still needed to adapt the model to the complex and sometimes unique needs of people with mental health problems. Factors such as coordination of care between different community resources and specialized training are a few examples of elements to be enhanced in home‐based psychiatric care [7, 8, 9].

Continuous quality improvement is an ongoing, cyclical process involving reflective, analytical and iterative steps. Some steps include identifying gaps, implementing action plans, evaluating results, providing feedback to stakeholders and proposing adjustments as needed [10]. The healthcare field has gradually adopted this philosophy as a key driver of system enhancement. Nevertheless, organizational culture and inadequate healthcare resources are often the most significant obstacles to improving care quality [10, 11]. Consequently, it is essential to evaluate and prioritize actions based on the specific context. This requires integrating evidence‐based best practices, clinical experience, and the preferences and experiences of service users, families and professionals [12, 13].

Processes of continuous improvement that involve patients, their families (or caregivers) and health professionals working together as an active partnership have demonstrated important contributions. The knowledge, experiential insights and practical suggestions contributed by health service users are recognized as essential for improving the design and delivery of services [14, 15]. These contributions, such as identifying service gaps, redefining priorities for improvement and co‐designing care pathways, have supported the development of patient‐centred initiatives tailored to patients' needs and preferences, enhancing care experiences and clinical outcomes [16]. This study forms part of a quality improvement project that assessed the impact of a home‐based psychiatric care programme and incorporated the perspectives of users, caregivers and professionals involved in these services [7, 17]. As part of this work, participants pointed to several areas where the delivery of home‐based psychiatric care could be strengthened, such as the need for more consistent continuity of care after discharge, challenges in coordinating home visit schedules and occasional difficulties in communication between services. These findings underscored the need for a structured and participatory process to determine which improvements should be prioritized within these models. Given these considerations, we adopted a consensus‐based, modified Delphi approach to support a structured and transparent prioritization process. This method enabled the inclusion of diverse stakeholder perspectives and reduced hierarchical influence through anonymous, iterative feedback, allowing convergence on improvements that were both relevant and feasible within these models [18, 19]. The study aimed to prioritize five key areas for improvement of a home‐based psychiatric care programme implemented in Catalonia, Spain, including the perspective of service users, families/caregivers and healthcare professionals.

2. Methods

2.1. Setting

Catalonia is an autonomous region located in the northeast of Spain, with a population of approximately 7.7 million inhabitants distributed across urban and semi‐urban areas. Mental health care in the region operates within Spain's decentralized National Health System (NHS), where services are organized through territorial networks that integrate hospital and community‐based care [20]. This structure makes geographical coordination central to service delivery and provides the context in which home‐based psychiatric care initiatives have emerged.

The Parc Sanitari Sant Joan de Déu (PSSJD) is a comprehensive health network that includes a regional referral hospital with inpatient/outpatient care for children and adults, an intermediate care centre for geriatric care and a mental health network. The adult mental health network includes acute, subacute and long‐term psychiatric hospitalization and community care, covering more than 700,000 people in the south of Barcelona (Catalonia, Spain).

At the end of 2019, the PSSJD Mental Health Network implemented a psychiatric home‐based hospitalization programme. The ‘Crisis Resolution and Home Treatment’ (CRHT) programme provides home‐based care through daily visits by a multidisciplinary team of nurses, psychiatrists, psychologists and social workers for a period no longer than a traditional hospital stay for a psychiatric crisis. Two CRHT teams were implemented in the catchment areas of three Community Mental Health Centers (CMHC) of the PSSJD's Mental Health Network in the Barcelona southern area.

As part of a quality improvement project led by the PSSJD Quality and Patient Safety Unit, the impact of the implementation of the CRHT and the experiences of all stakeholders were thoroughly evaluated between 2021 and 2022 [7, 17]. A qualitative approach was used to explore the experience of service users, family/caregivers and healthcare professionals with CRHT teams, and the results helped identify areas for improvement [7]. In collaboration with the PSSJD Mental Health Department, the research team identified 18 areas for improvement to be implemented in the short to medium term. These 18 improvement strategies were presented in a consensus exercise to prioritize the implementation of five of them.

2.2. Study Design

An online, modified Delphi study was conducted to prioritize improvement areas for the CRHT programme. The process consisted of three survey rounds in which participants rated 18 predefined strategies identified in the prior qualitative evaluation. This approach enabled the structured collection of perspectives from geographically dispersed stakeholders while ensuring anonymity throughout the process [18].

The Delphi methodology is widely used in healthcare to refine and prioritize improvement strategies through iterative feedback [21, 22, 23]. In this project, the number of rounds and the iterative structure were guided by the stability of the rankings across rounds. After each round, aggregated results were shared to support reconsideration of responses while avoiding group influence. This process facilitated the progressive convergence of perspectives among diverse stakeholders involved in the CRHT pathway.

In our study, we used a modified Delphi approach that differs in several ways from the traditional Delphi method. Classical Delphi studies typically begin with an open first round to generate items, followed by subsequent rounds aimed at refining and rating these statements [18, 24]. In contrast, our process began with a set of predefined improvement areas derived from a prior qualitative evaluation, eliminating the need for an exploratory first round. Moreover, instead of selecting a small panel of appointed experts, we included all stakeholders involved in the service pathway to ensure broad representation. The three survey rounds were conducted online, maintaining participant anonymity while providing structured, aggregated feedback between rounds to support iterative reflection [19].

2.3. Research Team

Consensus was led by the same research team that evaluated the impact of the CRHT implementation and the stakeholders' experiences, in collaboration with the Department of Quality and Patient Safety of the PSSJD Mental Health Network. The research team included M.R.‐V. (PhD, pharmacist), M.G.‐G. (PhD, pharmacist), A.M.B.‐C. (MSc, pharmacist), C.P.M. (MSc, psychiatrist) and L.B.H. (MSc, psychiatrist). M.G.‐G. and A.M.B.‐C. were full‐time researchers, while M.R.‐V. shared her research time with managing the Quality and Patient Safety Unit at the PSSJD. C.P.M. provided support as a quality technician in the Quality and Patient Safety Unit, and L.B.H. worked as deputy director in mental health and had considerable experience as a psychiatrist in both acute units and community care. With support from M.G.‐G. and M.R.‐V., A.M.B.‐C. designed and coordinated the online survey rounds. C.P.M. facilitated access for users and their families, and L.B.H. invited professionals to participate.

2.4. Improvement Areas to Prioritize

Table 1 presents the 18 improvement areas identified in the comprehensive qualitative evaluation, which were grouped into nine thematic categories. These include professional training and opportunities for CRHT teams to meet and share experiences. The referral and pre‐admission assessment process (covering both users and their family environments) was also included, along with the information provided by professionals, the organization and coordination of home visits during admission and care continuity upon discharge to community services. Additional categories cover family involvement and the availability of staffing and material resources.

TABLE 1.

Improvement strategies identified from experience evaluation.

Category	Improvement area	N°	Identified improvement strategies
Professional training	Training on CRHT referral and eligibility criteria	1	Provide regular training for professionals on CRHT eligibility criteria and referral procedures.
Referral process	Feedback loop between CRHT and referring professionals	2	Ensure that the CRHT team provides feedback on the user's health status to the referring professional.
	Optimization of referral registration process	3	Review and streamline the user referral registration process for CRHT.
	Available updates for CRHT slots	4	Ensure up‐to‐date information is available to professionals regarding CRHT capacity.
Pre‐admission assessment	Review and update of CRHT access protocol	5	Revise the CRHT access protocol to improve risk assessment at admission, minimize security incidents and clarify procedures for exceptional cases lacking a caregiver.
	Multidisciplinary pre‐admission assessment	6	Implement a multidisciplinary assessment of CRHT candidates before referral process (beyond psychiatrist‐only evaluation).
	Assessment of family/caregiver suitability	7	Assess the containment capacity, suitability and adequacy of the user's family environment before referral to CRHT.
Information for users, families and caregivers	User and family information at point of referral	8	Review the information given to users and families at the time of referral (e.g., referral process, initial visit, visit frequency and hours). Consider including visual materials (e.g., brochure or leaflet).
Home visit organization	Continuation of weekend in‐person visits	9	Continue conducting in‐person home visits on weekends for both CRHT teams.
	Scheduling and communication of visiting hours	10	Coordinate visiting hours with users, their families, and provide timely updates on any changes or delays.
Care coordination and follow‐up post‐discharge	Standardized discharge and continuity of care protocol	11	Standardize the discharge process and enhance continuity of care following CRHT admission, including coordination with the receiving service.
Family involvement	Structured family involvement in CRHT admission	12	Develop a protocol for involving families in CRHT admission, considering their role in supporting the user's adherence to pharmacological and psychosocial treatment.
	Family support plan development	13	Develop a support plan for family members, with special focus on families with children and caregivers of users with chronic mental health conditions.
CRHT team development and collaboration	Regular inter‐team meetings for CRHT teams	14	Establish regular meeting opportunities between the two CRHT teams to share procedures, address questions and discuss clinical perspectives.
	Training in systemic family therapy	15	Provide training in systemic family therapy for CRHT team professionals.
Staffing and material resources	Expansion of CRHT team staffing	16	Increase staffing for CRHT teams, including psychologists, administrative staff, social workers, psychiatrists and others as needed.
	Dedicated workspace and equipment for CRHT teams	17	Provide a dedicated workspace for the CRHT teams and equip them with necessary tools such as laptops, phones or mobile internet.
	Review CRHT vehicle identification for privacy	18	Use vehicle identification that avoids reference to the mental health network, displaying only PSSJD branding.

Abbreviations: CRHT, Crisis Resolution and Home Treatment; PSSJD, Parc Sanitari Sant Joan de Déu.

2.5. Expert Panel

The Delphi technique is based on multiple iterations to reach consensus on a subject [24]. Consensus is reached among a group of people with experience on the subject (experts), and iterations involve a feedback process, allowing each participant to re‐evaluate their initial responses using information from previous iterations. This gives everyone the opportunity to modify their answers based on their ability to evaluate the opinions of the other experts.

In this study, all participating stakeholders were considered ‘experts’ because they had direct, first‐hand experience with the CRHT programme, either through lived experience as service users or family caregivers, or through their involvement in referral, assessment or delivery of home‐based psychiatric care as healthcare professionals. Experiential knowledge is widely recognized as a valid and essential form of expertise in Delphi studies aimed at service improvement, where the goal is to prioritize actions grounded in real‐world practice [22].

In addition, the Delphi method allows participants' statements to remain anonymous throughout the process, preventing opinions from being influenced by group interaction, such as the influence of dominant or hierarchical individuals or group pressure to agree with the majority.

2.5.1. Panel Configuration

The expert panel included all stakeholders who participated in the qualitative evaluation phase of the CRHT quality improvement project. The aim was to ensure that each participant could express the level of importance they assigned to each area for improvement based on their own perspective and experience with the CRHT teams. The panel was composed of service users, family caregivers and healthcare professionals involved in the CRHT programme.

2.5.1.1. Inclusion and Exclusion Criteria

Participants were the same individuals from the prior qualitative study and were eligible if they: (a) were service users or family caregivers identified through the census provided by CRHT coordinators, or (b) healthcare professionals directly involved in CRHT delivery or in referring service users to CRHT (from CMHC, day hospital, acute inpatient unit, emergency department). Individuals who declined further participation after invitation and professionals not involved in CRHT care or referral pathways were excluded. It is worth highlighting that the same individuals who shared their experiences in the initial qualitative phase also took part in determining which improvements should be addressed first, ensuring that the actions selected truly reflect the needs and expectations of those with first‐hand experience.

• –Service users and families: Using the census of users and caregivers/family members provided by the CRHT team coordinators, C.P.M. invited them to participate in the project via phone calls and/or email. A total of 36 service users and caregivers were invited to participate; however, five declined further contact regarding the survey.
• –Professionals: Although traditional Delphi methods rely on a selected panel of experts, we adopted a modified approach tailored to our context. Rather than limiting participation to a small expert group, we invited all relevant professionals—including referring professionals (from CMHCs, day hospitals, acute psychiatric unit, emergency department and on‐call psychiatrist) as well as CRHT team members. This inclusive strategy was designed to capture a wide range of perspectives and promote collective ownership of the improvement process. To encourage open and honest responses, L.B.H. distributed the surveys by email, and responses were collected anonymously. Ninety‐nine healthcare professionals were invited to take part in the panel, and none declined further contact.

The expert panel consisted of 130 individuals, including 18 service users, 13 caregivers and 99 healthcare professionals, with a majority of female presence. Most caregivers were the users' mothers, and the professionals were primarily female psychiatrists working in the emergency department. All participants were informed about the purpose of the project, the voluntary nature of their participation, and the link between this phase and the prior qualitative evaluation of the CRHT programme. Tables 2 and 3 show the main characteristics of the expert panel.

TABLE 2.

Main characteristics of CRHT service users and caregivers.

		Service users (N = 18)	Caregivers (N = 13)
		N (%)	N (%)
Gender	Female	13 (72.2)	7 (53.8)
	Male	5 (27.8)	6 (46.2)
Relationship to service user	Mother	N/A	6 (46.1)
	Father	N/A	2 (15.4)
	Brother	N/A	1 (7.7)
	Partner	N/A	2 (15.4)
	Ex‐partner	N/A	2 (15.4)

Abbreviations: CRHT, Crisis Resolution and Home Treatment; N/A, not applicable.

TABLE 3.

Main characteristics of healthcare professionals.

		Healthcare professionals (N = 99)
		N (%)
Gender	Female	71 (71.7)
	Male	28 (28.3)
Credentials	Nurse	6 (6.1)
	Psychiatrist	48 (48.5)
	Psychologist	20 (20.2)
	Social worker	1 (1.0)
	Psychiatry resident	24 (24.2)
Mental health service	CMHC	25 (25.3)
	Day hospital	8 (8.0)
	Acute psychiatric unit	2 (2.0)
	Emergency department	49 (49.5)
	CRHT Teams	15 (15.2)

Abbreviations: CMHC, Community Mental Health Centers; CRHT, Crisis Resolution and Home Treatment.

2.6. Survey Design

The modified Delphi method involved three rounds of online surveys. A survey was developed listing the 18 specific improvement strategies previously identified (Table 1). Survey rounds were held between October and December 2023, and each round was conducted over a 3‐week period. Responses from each round were analysed and results calculated at the end of the 3‐week period. Each participant received a summary table of the previous round's results with an invitation to the next survey round. Participants and their responses remained anonymous throughout the process.

A pilot was conducted to test the appropriateness of the survey format and layout, wording of improvement areas and explanation of the dimensions to be assessed. The pilot test was applied to members of the research group who were aware of the results of the previous CRHT qualitative evaluation and who provided feedback on the wording of the statements to improve understanding.

Surveys were administered in Microsoft Forms and included two parts: (i) closed‐ended ratings of each pre‐identified improvement area on three dimensions (problem significance, change agency, feasibility; 5‐point Likert), and (ii) a single open‐ended question at the end of the instrument inviting participants to suggest additional improvements or comment on wording/clarifications. The same structure was kept across the three online rounds, with aggregated feedback from the prior round provided in tabular form to support iterative reconsideration while maintaining anonymity.

2.7. Data Collection

In Round 1, an email was sent to the expert panel referring to the previous phone call/email invitation to participate, explaining the objectives of the consensus methodology, the voluntary nature of participation and what it entailed and instructions for completing the survey. The email included the results report from the previous CRHT evaluation project as an attachment [25]. A link was also added to take participants directly to the online survey, which consisted of two phases. The first phase included an information sheet and informed consent form, which had to be accepted in order to continue. The survey was created and tested using Microsoft Forms.

After giving informed consent, participants were asked to rate each of the 18 improvement areas along three dimensions: the Problem significance, change agency, and feasibility of change on a 5‐point Likert scale from ‘very low’ (1) to ‘very high’ (5). These three dimensions reflect widely used prioritization criteria in public health and service‐improvement initiatives (e.g., significance and feasibility) and were adapted to the CRHT context to support a transparent, criteria‐based ranking process across rounds [26]. The dimensions were defined as follows:

1. Problem significance: Severity and prevalence of the issue. How much does it matter?

2. Change agency: Level of autonomy and control within the work team and/or institution, or at which professionals and institutions are independent and have sufficient resources to be agents of improvement or change. Can we do something about it ourselves?

3. Feasibility of change: Likelihood of success with reasonable effort/resources or extent to which it is considered possible to achieve an improvement or modification with a level of effort that can be assumed by professionals, users and/or institutions. Is it realistically achievable?

Users and family caregivers were only asked to assess problem significance, as evaluating change agency and feasibility required specific knowledge of the healthcare organization and its resources. After completing the closed‐ended ratings in each round, participants also answered a single open‐ended question inviting additional suggestions or comments on the predefined strategies (e.g., wording, missing aspects or implementation considerations). The same instrument structure was used across Rounds 1–3. Qualitative input from this field was collated by round and handled through rapid content grouping: the research team reviewed and grouped comments by theme, and mapped each suggestion to the closest predefined strategy (or noted it as out of scope when unrelated to the project timeframe). In Round 1, we received one comment from users/caregivers and five from professionals; these remarks led to minor wording refinements in the strategies related to scheduling/communication of visits and standardized discharge and continuity, without adding new items. In Rounds 2 and 3, comments were scarce and did not introduce new themes; therefore, no further modifications were required. Aggregated quantitative results and a brief synthesis of qualitative comments were shared prior to each subsequent round to support iterative reflection while maintaining anonymity.

2.8. Data Analysis

For each of the 18 improvement areas, the average score was calculated separately for each of the three dimensions. Since users and caregivers only rated problem significance (and were fewer in number than healthcare professionals), we applied a proportional weighting to ensure equal influence of stakeholder groups on that dimension. Specifically, if ƞ _pro is the number of professionals and ƞ _uc the number of users + caregivers, we multiplied the mean ‘problem significance’ score from users + caregivers by Weight = ƞ _pro/ƞ _uc and then averaged the group means so that users + caregivers and professionals contributed equally to the aggregate score. In other words, their responses on problem significance were given proportionally greater weight to balance representation across respondent groups [27, 28]. For each round, we calculated mean scores for each dimension and then combined them to generate item rankings, applying the same weighting approach for problem significance across stakeholder groups. Overall priority was obtained by averaging the three dimensions, with ties resolved first by higher feasibility and then by higher change agency. All analyses were based on available responses for each item and round, and means were calculated using the number of valid ratings provided for each dimension. All descriptive analyses were performed using Microsoft Excel (Microsoft 365), which was used to calculate means, weighting factors and rankings across rounds. Responses to the open‐ended question were analysed separately and categorized according to common themes, being fully integrated into the area of improvement addressed.

The improvement areas were then ranked in descending order based on their total score within each dimension. After each survey round, the aggregated results were shared with participants, who were asked to re‐rate the improvement areas considering the results of the previous round. Consensus was reached in the third round, where participants agreed on the top five priority areas. Once consensus was reached, no further rounds were conducted. In this study, consensus was defined as the point at which the ranking of improvement areas demonstrated stability across rounds [29]. Specifically, consensus was considered reached when no new items entered the top positions, the relative order of the five highest‐rated strategies remained stable between rounds, and these strategies showed consistent differentiation from the remaining items. As this pattern was observed between rounds 2 and 3, the process was concluded after the third round. The research team acted solely as facilitator and did not influence or participate in the panel's decisions. No funding entity participated in the process or had any influence over the decisions reached.

2.9. Ethical Considerations

Since this study was conducted as part of a quality improvement initiative by the PSSJD Quality and Patient Safety Unit, it did not require review by the research ethics committee. However, in accordance with internal institutional protocols, key ethical considerations were addressed throughout all phases of the project using a checklist. This included the project's scope; involvement of individuals, families/caregivers and healthcare professionals; informed consent; access to personal information; and potential risks.

All participants provided informed consent prior to completing the surveys. The consent form included detailed information about the study, its objectives and what participation entailed, and confirmed that their involvement was entirely voluntary. Furthermore, the surveys were fully anonymous, and no identifying information was collected, ensuring that participants' identities remained unrecognizable at any stage of the study.

The authors confirm that all procedures involved in this study were conducted in accordance with the ethical guidelines established by the appropriate national and institutional committees for research involving human participants and are consistent with the principles outlined in the Helsinki Declaration (1975), as last amended in 2013. This study was not prospectively registered.

3. Results

3.1. Participation

Online surveys were sent to the 130 individuals comprising the expert panel (31 CRHT users/caregivers and 99 healthcare professionals). The highest response rate occurred on the first round, with participation decreasing in subsequent rounds. Response rates for users and caregivers were 32% in the first round (n = 10), 19% in the second (n = 6) and 16% in the third (n = 5). For healthcare professionals, response rates were 27% (n = 27), 19% (n = 19) and 10% (n = 10). Figure 1 provides a detailed overview of the consensus exercise process using the online modified Delphi method.

FIGURE 1.

Prior qualitative phase and online modified Delphi process: participant flow and responses across rounds. The left panel summarizes the qualitative study conducted in 2021–2022, in which service users, family caregivers and healthcare professionals identified 18 improvement areas grouped into nine thematic categories. These areas were used as input for the subsequent online modified Delphi process. The right panel depicts the three‐round Delphi process, including the number of respondents per round (Round 1: Users/caregivers = 10, professionals = 27; Round 2: Users/caregivers = 6, professionals = 19; Round 3: Users/caregivers = 5, professionals = 10) and the attrition between rounds (Δ–4 and Δ–8 from Round 1 to Round 2; Δ–1 and Δ–9 from Round 2 to Round 3). Consensus on the top five improvement priorities was achieved in Round 3.

3.2. Online Consensus

Consensus was reached in the third round, considering the agreement of all participants and that no new areas for improvement emerged within the top 5 by comparing the results of the three completed rounds. No items or topics were modified or removed during the consensus process.

The scores for the five key areas for improvement ranged from 3.85 for first place to 3.52 for both fourth and fifth place, on a 5‐point scale. (1) Training on CRHT referral and eligibility criteria; (2) standardized discharge and continuity of care protocol; (3) feedback loop between CRHT and referring professionals; (4) scheduling and communication of visiting hours; and (5) available updates for CRHT slots were the areas with the highest overall score. The five key areas will guide the next steps in implementing specific improvements to CRHT teams. For transparency, the overall priority score was computed as the arithmetic mean of problem significance, change agency and feasibility of change. Ties were broken by higher feasibility, then by higher change agency. Round‐to‐round changes were tracked using the same scoring and ranking procedure; the stability of the top five between Rounds 2 and 3 is summarized in Table 4, and the final round ranking for all 18 items is provided in Table S1. Figure 2 shows the overall and dimension‐specific scores for the five improvement areas prioritized by the expert panel.

TABLE 4.

Top five prioritized improvement areas.

Rank	Improvement area	Problem significance	Change agency	Feasibility of change	Overall score
1	Training on CRHT referral and eligibility criteria	3.35	4.10	4.10	3.85
2	Standardized discharge and continuity of care protocol	3.70	3.90	3.90	3.83
3	Feedback loop between CRHT and referring professionals	3.35	4.20	3.60	3.72
4	Available updates for CRHT slots	2.95	4.00	3.60	3.52
5	Scheduling and communication of visiting hours	3.15	3.70	3.70	3.52

Abbreviation: CRHT, Crisis Resolution and Home Treatment.

FIGURE 2.

Five key areas for improvement by dimension. Y‐axis represents mean scores on a 5‐point Likert scale (1 = very low, 5 = very high). CRHT, Crisis Resolution and Home Treatment.

3.3. Importance and Feasibility

After applying the weighting factor, the highest score in the problem significance dimension was for Standardized discharge and continuity of care. On the contrary, the initiative related to the Expansion of CRHT team staffing (psychologists, administrative staff, social workers, psychiatrists, etc.), which was weighted second in terms of the problem significance, received the lowest scores from professionals in the dimensions of change agency and feasibility to change.

As shown in Figure 2, the five priority improvement areas did not receive high scores on the problem significance dimension. However, they received higher scores on the other two dimensions. In fact, of the 18 proposed areas for improvement, the five with the highest scores in terms of change agency and feasibility to change are the same prioritized areas considering the overall score.

Because users and caregivers contributed only to the problem significance dimension, the weighting factor was applied to ensure that their perspectives carried equal influence in this dimension despite the smaller size of their group. The weighting adjusted the aggregated score without altering the relative importance of their responses.

3.4. Open‐Ended Question

In the first round, one user/caregiver suggested increased home visit frequency and improved adherence to scheduled visits, as well as better follow‐up after discharge. These comments were incorporated into the strategies related to Scheduling and communication of visiting hours and Standardized discharge and continuity of care protocol.

Among professionals, five suggestions were made that included increasing CRHT staffing (particularly adding nursing roles), promoting user and family involvement in treatment decisions, improving time distribution among team members and raising awareness of CRHT services. While staffing increases were acknowledged as important, they were limited by institutional constraints, which was noted in the survey. Suggestions on user participation and CRHT visibility should be considered in future stages of the project.

4. Discussion

The consensus exercise highlighted clear patterns in the improvement priorities identified by stakeholders, particularly regarding referral training, continuity of care, communication with referring professionals and coordination of home visits. These priorities reflect shared expectations about what is needed to support effective crisis care and minimize fragmentation across service levels. Overall, the findings align with previous evidence underscoring the central role of communication and coordination mechanisms in sustaining coherent and responsive home‐based psychiatric care [30, 31].

The final ranking reflected the joint contribution of the three dimensions. Problem significance oriented attention to issues stakeholders considered most consequential, change agency indicated what could realistically be driven by local teams and institutions, and feasibility assessed the likelihood of implementation with available effort and resources. Items with high significance but low agency/feasibility (e.g., staffing expansion) were de‐prioritized for immediate action, whereas the selected top five combined moderate‐to‐high significance with consistently higher agency and feasibility, maximizing near‐term, context‐appropriate impact. Together, these dimensions functioned as complementary lenses: what matters, what is within scope to change and what can be achieved reliably in routine care.

Previously, the results of the qualitative assessment identified several improvements related to participation and support strategies for caregivers and family members, as well as initiatives related to the staff and physical resources available for CRHT teams. However, despite being relevant issues for the entire expert panel, these areas did not emerge as a priority for immediate implementation. In most cases, the score was high for the problem significance dimension but low for the dimensions of change agency and feasibility of change. Increasing staff in the CRHT teams was an example of this. It was the second priority area in the problem significance dimension; however, this initiative is subject to decisions made outside the institutional level. For this reason, despite it being a clear priority recognized by all stakeholders, considering it among the priority improvements to be implemented is limited by its low feasibility. These feasibility constraints also have implications beyond the immediate quality‐improvement cycle. While not actionable in the short term, areas such as staffing expansion remain relevant for longer‐term planning and institutional advocacy. In this sense, the Delphi results can contribute to strategic discussions by signalling where sustained investment, resource allocation or system‐level decisions may be needed to support more substantial organizational changes over time.

The referral and eligibility assessment, which considers the appropriateness of the clinical condition and social context of the service user to be admitted to CRHT, is a critical process often related to the eventual effectiveness of the intervention [9, 32]. It is important for professionals in charge of user referrals to maintain updated knowledge of the target population and the inclusion criteria for CRHT [1, 7, 33]. This presents a significant challenge, particularly given the high turnover rate among psychiatrists working in emergency mental health services.

With CRHT at the centre, continuity of care should span the initial process (assessment and referral) and follow‐up, including post‐discharge referrals to community services and, in some cases, hospital admission. It also encompasses all the ways in which the CRHT provides ongoing support (between visits and across resources), recognizing that consistent care is essential for quality mental health support [30, 34]. Service users recognize and appreciate CRHT efforts to promote collaborative work with other professionals and services, particularly community resources and families [1, 31]. However, discharge is often followed by a transition to community services where the intensity of follow‐up decreases. Although users usually have quick access to an initial contact, subsequent visit frequency tends to decline, which many experience as a challenging shift: the crisis may be contained but not fully resolved, and the drop in contact can foster feelings of being left alone, perceptions that services are not responding, or the sense that care is episodic rather than ongoing [7, 9].

At the professional level, clear and timely communication about users' health status, particularly between CRHT professionals and other services, is essential for the effectiveness of the intervention [30]. Effective communication and adequate, comprehensive information sharing between services, professionals and service users can help overcome the challenge of improving coordination and continuity between levels of care [1, 33, 35].

Visits scheduling remains an important issue to be addressed in accordance with the expert panel's priorities. Normally, CRHT teams coordinate visits according to user preferences whenever possible, which makes them feel respected and considered by the team and makes it easier for them to balance visits with their personal activities [7, 31, 36]. Professionals see coordinating the visit schedule with users as an opportunity to improve daily planning, not only to make the visit more efficient, but also to allow them time to seek assistance if needed [37]. However, some details still need to be adjusted in the coordination of visits. The consensus exercise presents as a priority opportunity the possibility of improving scheduling and communication in the event of any problems prior to the visits. The standardization and resolution of this concern has been identified as a feasible task by the panel.

4.1. Limitations

This study has several limitations. First, while the survey was designed to gather the expert panel's priorities for improving teams, no information on sociodemographic data on those who actually answered the survey was collected. Participation was low and declined over the survey rounds, particularly among health professionals. Before Delphi exercise, users and family members received personal phone calls explaining the consensus process and could ask questions. In contrast, professionals were invited via an email from hospital management outlining the process and survey instructions. Still, some professionals later reported difficulty understanding the evaluation dimensions and had no active space to ask questions. Additionally, the initial invitation came from someone in a higher position, prompting early participation but not sustained engagement in later rounds.

Participation attrition across rounds may have had some influence on the stability and representativeness of the consensus. Lower engagement in later rounds (together with differences in response patterns between users/caregivers and professionals) might have shaped the balance of perspectives, particularly given that users contributed only to the problem‐significance dimension. Although anonymity, structured feedback, predefined consensus criteria and proportional weighting were intended to mitigate these risks, a limited degree of attrition‐related bias cannot be entirely ruled out.

4.2. Strengths

To our knowledge, this is the first expert consensus prioritization resulting from a quality improvement project focused specifically on home‐based psychiatric crisis care. The Delphi exercise allowed us to explore the perspectives of service users, family members/caregivers and healthcare professionals on priority areas for improving the CRHT programme. The online approach allowed us to broaden the scope and try to contact as many people as possible within the processes among the resources related to CRHT.

4.3. Implications for Practice

Involving users, caregivers and professionals in prioritizing improvement areas helps ensure that home‐based mental health programmes remain responsive and person‐centred [38, 39]. The five prioritized strategies provide a clear roadmap for immediate action: strengthening referral and eligibility training, standardizing discharge and continuity protocols, improving feedback mechanisms with referring professionals, enhancing communication and scheduling of home visits, and ensuring timely operational updates on service availability. The findings further underscore the importance of focused training initiatives for both CRHT teams and referring professionals as a means of supporting more consistent implementation of CRHT models. In addition, the prioritization exercise offers a useful basis for examining how such improvement strategies are implemented and sustained over time in home‐based psychiatric care. These are concrete, feasible steps that CRHT teams can address through existing structures, and they guide the first phase of implementation now underway. Beyond these initial actions, the Delphi results also highlight additional areas (such as staffing needs and broader resource constraints) that warrant continued monitoring and longer‐term planning within the institution.

5. Conclusion

The consensus exercise, conducted using a modified Delphi approach, enabled the prioritization of the most relevant and feasible improvement areas from the perspective of all stakeholders involved in CRHT services. A key added value of this approach was the inclusion of service users and caregivers alongside professionals, ensuring that experiential knowledge contributed meaningfully to the decision‐making process. This inclusive and structured methodology strengthened the relevance, person‐centredness and practical applicability of the prioritized actions. In addition, the findings suggest directions for strengthening professional training and for further inquiry into how improvement priorities in psychiatric home‐based care are developed, implemented and sustained over time. By integrating these diverse perspectives, the study provides a robust foundation for implementing improvements that are both context‐appropriate and aligned with the real needs of those receiving and delivering care.

Author Contributions

Ana María Besoaín‐Cornejo: conceptualization, methodology, software, formal analysis, investigation, data curation, writing – original draft, writing – review and editing, visualization. Montserrat Gil‐Girbau: conceptualization, methodology, validation, formal analysis, resources, writing – review and editing, supervision. Luisa Baladón Higueras: validation, resources, writing – review and editing. Cristina Pou Matarranz: resources, writing – review and editing. Maria Rubio‐Valera: conceptualization, methodology, validation, resources, writing – review and editing, supervision, project administration.

Funding

Supported by the Instituto de Salud Carlos III (ISCIII) under the programme of Cooperative Research Projects Oriented to Health Outcomes (RICORS) 2021, with funds from the European Recovery, Transformation and Resilience Plan (PRTR). Programme funded by the European Union—NextGenerationEU, grant number RD21/0016/0018. This work was also supported by the National Agency for Research and Development (ANID)/Scholarship Program/DOCTORADO BECAS CHILE/2019–72200124.

Ethics Statement

Participants gave informed consent and completed fully anonymous surveys. All procedures followed national and institutional ethical guidelines and the Helsinki Declaration (1975, revised 2013).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Full ranking of all 18 improvement strategies.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.