Supporting carers to improve patient safety and maintain their well‐being in transitions from mental health hospitals to the community: A prioritisation nominal group technique

Abstract Introduction Carers of people with mental illness may face distinct challenges, including navigating fragmented health and social services during discharge from mental health hospitals. Currently, limited examples of interventions that support carers of people with mental illness in improving patient safety during transitions of care exist. We aimed to identify problems and solutions to inform future carer‐led discharge interventions, which is imperative for ensuring patient safety and the well‐being of carers. Methods The nominal group technique was used which combines both qualitative and quantitative data collection methods in four distinct phases: (1) problem identification, (2) solution generation, (3) decision making and (4) prioritisation. The aim was to combine expertise from different stakeholder groups (patients, carers and academics with expertise in primary/secondary care, social care or public health) to identify problems and generate solutions. Results Twenty‐eight participants generated potential solutions that were grouped into four themes. The most acceptable solution for each was as follows: (1) ‘Carer Involvement and Improving Carer Experience’ a dedicated family liaison worker, (2) ‘Patient Wellness and Education’ adapting and implementing existing approaches to help implement the patient care plan, (3) ‘Carer Wellness and Education’ peer/social support interventions for carers and (4) ‘Policy and System Improvements’ understanding the co‐ordination of care. Conclusion The stakeholder group concurred that the transition from mental health hospitals to the community is a distressing period, where patients and carers are particularly vulnerable to safety and well‐being risks. We identified numerous feasible/acceptable solutions to enable carers to improve patient safety and maintain their own mental wellbeing. Patient and Public Contribution Patient and public contributors were represented in the workshop and the focus of the workshop was to identify the problems they faced and co‐design potential solutions. Patient and public contributors were involved in the funding application and study design.


| BACKGROUND
Informal carers provide unpaid help to a friend or family member needing support. According to the 2021 census, more than 5.7 million people are estimated to be informal carers in the United Kingdom. 1 Carers Week in their 2022 report have estimated that the number of informal carers in the United Kingdom could be as high as 10.6 million. 2 A total of 1.5 million people care for someone with a mental illness. 3 Patient safety policies increasingly encourage carer involvement in reducing patient harm. One recent study found that carers who intensively engaged during hospital care provided patients with greater protection, but typically experienced negative consequences for themselves. The authors concluded that carer involvement in patient safety needs to be better understood, especially from the carers' perspective and negative consequences for carers need to be mitigated by practice improvements that value their contributions. 4 This is especially true for carers of people with mental illness who may face distinct challenges because mental health problems are not seen. They might experience what is called 'hidden caring' where family carers may not recognise themselves as carers due to which they are less likely to access support. 5 However, they might be providing all sorts of help including emotional support, encouragement, practical help with daily tasks and advocacy. 6 Carers may have serious concerns about the safety of the people that they care for and experience stigma. They may feel responsible/guilty when the patient's health deteriorates or when patient safety incidents occur. 7 These feelings are especially heightened when they care for people with mental illness. 8 Thus, it is not surprising that caring is seen as a chronic stress experience. 9 Carers often experience psychological distress symptoms, including anxiety and depression, and may fail to meet their own health needs or make unhealthy lifestyle choices. 7 All these challenges and especially those related to patient and carer safety, might worsen during the transition from inpatient mental health services to the community. 10 Inpatient mental health settings pose unique challenges for patient safety which also influence the discharge process including interpersonal violence, coercive interventions, safety culture, harm to self and safety of the physical environment. 11 Discharge is often described as a chaotic time with multiple threats to patient safety. For example, the weeks after discharge have been associated with numerous adverse outcomes, including self-harm, medication safety incidents, suicide and violence. 8 Adverse social outcomes which include loneliness and homelessness have also been reported. 12 Systems feel fragmented to many carers whereby social and clinical services seem funded and operated separately and miscommunication is common. 13 Carers must coordinate and navigate fragmented health and social services when their loved one is discharged from mental health hospitals. 7 They are often the individuals that must advocate for the patient and act as a 'boundary spanner' between fragmented services. 14 Fragmented services can be defined as a lack of coordinated care between health and social care for patients and carers. 15 Fragmentation of services is linked to the quality of care provided and poor clinical and social care outcomes. 16 Carers transfer important information between services and are sometimes the only constant in the patients' health and social care network. 17 Having effective social support (often provided by carers) is thought to reduce the likelihood of adverse events, such as suicide postdischarge. 18 The important role of carers during transitions of care has been highlighted during the pandemic. 19 This was because large numbers of patients were discharged from inpatient mental health services while access to community services was limited. 20 A recent systematic review found only 12 carer involvement interventions to improve patient outcomes (e.g., readmissions) during mental health transitions. 21 From the 12 interventions, the interventions which supported carers across the full care pathway were the most promising. 21 A recent study involving patient engagement activities found that many carers had concerns about the safety of the person they cared for and their own mental health and safety during care transitions. 8 They described feeling unsupported, lonely and depressed in the community, and were unable to optimally support their loved ones. 8 Moreover, carers for those with mental illness felt they were not involved in discussions and decisions around discharge and had insufficient information about available support services in the community. These feelings were amplified during the Covid-19 crisis. 22 Another study found that competence and listening skills of staff members, concerns about waiting times, staffing levels MCMULLEN ET AL. | 2065 and overall resourcing of services were key safety issues for mental health service users, carers and professionals. However, that study did not identify solutions and did not focus specifically on care transitions. 23 Supporting carers of people with mental illness during transitions of care is imperative for ensuring patient safety and the well-being of carers themselves. 24 Patient safety and carer wellbeing challenges during discharge from mental health hospitals span social care, primary care, secondary care and public health with carers being expected to skillfully interact with the multiple professionals across diverse disciplines who are involved in the care of their loved ones.
Therefore, interdisciplinary lenses are needed to develop multiagency solutions. 8 To address this, in the present study, we conducted a workshop that brought together patients and carers as well as expertise academics (some with clinical/social care backgrounds) within primary care, mental health, social care and public health. The work aimed to identify problems and solutions to inform future carerled patient safety interventions after hospital discharge as well as interventions to support the carer's well-being during this challenging time.

| Study design
The widely used nominal group technique (NGT) was used which combines both qualitative and quantitative data collection methods. 25  Participants were asked first in smaller breakout rooms to broadly identify problems that affect safety of people discharged from mental health services and the well-being of carers. The results of each breakout room were fed back to the full group.

| Phase 2 solution generation
Participants within each breakout room were asked what services could do to solve the problems initially highlighted in phase 1. All solutions identified within each breakout room were then discussed with all participants as one group.

| Phase 3 decision making
Participants decided which of the potential solutions should be prioritised considering also their feasibility (the possibility that can be achieved is reasonable) and acceptability (how much they like the idea) leading to a reduction to the list generated in phase 2.

| Phase 4: Prioritisation, implementation and intervention development
The study team reviewed and merged solutions to create a list that is manageable for participants to understand and rank. A participants were asked to fill out their demographic data: age, ethnicity, gender, stakeholder group (patient, carer, practitioner, academic, other), professional role/job title and the number of breakout room (1, 2, 3, 4).

| Participants' recruitment and eligibility
Participants were recruited via already established contacts within the study team and through social media. Links with relevant universities and third-party groups were also established. To help recruit people from underserved communities and maximise the relevance of the findings to the community, the recruitment approach was carefully designed to be inclusive (format/type/ language used on adverts) of patients/carers with diverse backgrounds to capture the voices of research underserved groups.
The eligibility criteria were broad so that patients/carers were not excluded based on demographic factors, social and economic factors, and factors related to health status and health conditions as The eligibility criteria for the NGT were the following: The eligibility of interested participants was further confirmed by the study team using the following screening questions. To take part in the workshop participants would have to return a completed screening pro-forma with the following items: 1. Your primary stakeholder group (are you a patient, informal carer, professional-state job title).
2. Do you have direct experience of discharge from inpatient mental health services (Yes OR No).
3. Would you prefer to be in a patient-/carer-only group or a mixed group with other stakeholders (question presented only to carers).

Please provide 3-4 sentences about yourself and why you would
like to take part in the workshop.
Once eligibility was confirmed, participants were sent the information sheet, topic guide (outlining the itinerary for the workshop and discussions) and a consent form to sign and return to the study team. Once the consent form was signed and returned, participants were sent the Zoom link for the workshop.

| Data sources
Relevant data from phases 1-3 were collected by the host and breakout room facilitators in the form of hand-written notes. Further to this, the whole workshop including the breakout rooms was audio recorded. For phase 3, anonymous rankings were collected using Qualtrics and were analysed independently by the two researchers within the study team (S. M., N. T.). Handwritten notes collected during phases 2 and 3 (solution generation) were used to provide context towards the ranking exercise (phase 4).

| Analysis
For the analysis, we used the Qualtrics data generated in the survey.
We asked participants to rank the solutions within each theme by feasibility and acceptability (1 being the most feasible/acceptable and the last as the least). We reverse-coded the data which gave us the cumulative ranking from the most acceptable/feasible solution (highest number) to the least (lowest number).

| Demographics
Twenty-eight participants took part in the workshop and 17 of them (61.8%) also completed the ranking survey on Qualtrics sent by email after the workshop. During phases 1 and 2 of the workshop (Zoom event), breakout rooms one to three were an equally distributed mix of stakeholder groups. Breakout room four consisted only of carers because some carers had stated that they would prefer to be in a separate group during the screening stage. Table 1 describes the demographics of the 17 participants who took part in the workshop and completed the online survey. We did not collect demographic data from participants who did not complete the survey. Participants were a mixed group of patients (29.4%, n = 5), carers (35.2%, n = 6) and academics (35.2%, n = 6), many academics also had a clinical/social mental health professional background. Two social care professionals and one academic also had lived experience as a carer. The mean age was 43 years (26-55) exact age was reported; however, we presented it categorically to increase anonymity. Eleven participants were female (64.7%), 10 participants identified as White British (White or British) ethnicity (58.8%), 2 (11.8%) Mixed and 1 person (5.9%) identified as Black Caribbean, Greek, Asian British and British Pakistani, 1 (5.9%) did not disclose.

| Phase 1: Problem identification
Several distinct patient safety and carer well-being concerns were identified by the four breakout rooms. One commonly identified concern across the breakout rooms was a lack of support for carers with regard to their mental health and difficulties navigating the discharge process (such as a lack of awareness on how to best support the patient). Table 2 lists the highlighted safety concerns for patients discharged from mental health services and carer well-being concerns by all four groups.

| Phase 2: Solution generation
There were many potential ideas generated across the four breakout rooms. These ranged from specialist teams to support patients and carers when they are transitioning back into their home, to family therapy and training techniques for patients and carers. Techniques suggested included behavioural activation and problem-solving skills to manage everyday stressors.

| Phase 3: Decision making
After a group discussion with regard to the potential solutions highlighted in phase 2, the list was combined (solutions that were very similar grouped together) into 34 potential solutions generated across the four breakout rooms. Potential solutions ranged from improving access to services to family therapy and carer wellness interventions. Table 3 lists the 34 potential solutions generated.

| Phase 4: Prioritization, implementation and intervention development
The list of 34 potential solutions was reduced to 20 solutions by the research team, many of the potential solutions were similar so were combined. Workshop participants were then sent a Qualtrics survey which comprised four main themes and were asked to rank each of the solutions within each theme by feasibility and acceptability. The four main themes (carer involvement and improving carer experience, patient wellness and education, carer wellness and education, policy and systems improvements) and the solutions within these themes were derived from the discussions in phase 3. Participants felt it too difficult to rank such a large list of solutions, we collectively so decided to create four smaller themes, based on the primary problem the solutions aim to address. Table 4 lists out the four themes and the solutions within each theme including the cumulative ranking (by reversing the ranks and adding them together across participants) for each solution within the themes, Figure 1 outlines the most feasible (the possibility that can be achieved is reasonable) and acceptable (how much they like the idea) and acceptable solutions within each theme.   (1)

| DISCUSSION
This study involving a diverse stakeholder group identified many patient safety concerns and carer well-being risks during care transitions from inpatient mental health hospitals to the community.
These include lack of information and support, distress and isolation.
Multiple solutions were generated based on the collective knowledge of the diverse stakeholder group; the most highly ranked feasible and acceptable solutions were naming a dedicated family liaison worker and adapting better implementing existing approaches. Four broad themes of solutions were uncovered based on their nature and purpose: carer involvement in improving patient experience, patient wellness and education, carer wellness and education, policy and system improvements.
Our findings are consistent with previous research which highlighted that carers are concerned for their own mental wellbeing and patient safety during this precarious time. 13 The outcomes of this The evidence base on which interventions can effectively improve patient safety and carer well-being is relatively limited. This is despite carer-led suggestions and the gradually growing awareness of the need to consistently include carers in care quality and safety improvement programmes for people with mental illness and in T A B L E 3 Potential solutions were generated during phase 3 of the workshop.
(1) Improving access to services

2) Quick access to therapies
(3) Dedicated family liaison worker: To act as a bridge between hospital and home, including checking the safety of home to return to, etc.
(4) Working with carers to develop a full pathway of mental health support, from emotional support and prevention to everyone stepping up to more intensive support for those who need it. Perhaps delivered by carers champions within services like IAPT.
(5) Having collaborative discharge planning from the hospital with the patient, family worker (as suggested above) and carer. Perhaps adapting existing approaches such as activity scheduling/BA type approach to help the patient plan what they need to do and what they need help with and plan where that help will come from, e.g., cooking, bills, cleaning, medication, etc.
(6) More education for families on conditions (7) Tailored peer support intervention for carers, in a format that is convenient to the people involved, that may be face to face, remotely, telephone, etc. Person-centred care is something that is well recognised and understood but remains extremely difficult to actually deliver, an intervention that provides a practical way of delivering this would be useful

| Strengths and limitations
One strength of the study was the promotion of open dialogue between diverse stakeholders including carers, academics with different perspectives and health professionals. We also provided the option of a carers-only group to reduce any potential negative power dynamics and promote inclusivity. As a result, we generated a series of potential intervention ideas that were agreed upon by this diverse group of stakeholders.
However, this study had also some key limitations. The most important limitation is that the attrition rate between the online workshop (phases 1-3) and the final online questionnaire to rank solutions (phase 4) was considerable. We originally planned to complete phase 4 during the workshop, but as there were so many solutions generated the task became too complicated to manage in one session.
We, therefore, arranged a follow-up exercise, which resulted in a moderate attrition rate and incomplete collection of the demographic information of the participants in the workshop. Furthermore, the online nature of the workshop could have resulted in digital exclusion for some carers (especially those who are lacking e-literacy).
The sampling decision to use established contacts within the study team and social media enabled us to access academic expertise across the three academic fields of knowledge we hoped to combine expertise (Primary Care, Social Care and Public Health). However, in the future, including a greater number of patients and carers and using techniques to access more diverse groups would be beneficial.

| Implications and conclusions
The members of our stakeholder group concurred that the transition from mental health hospitals to the community is a particularly distressing period of the care pathway, where patients and carers are particularly vulnerable to safety and well-being risks. This study identified several feasible and acceptable solutions to enable carers of people transitioning from mental health hospitals into the community to improve patient safety and maintain their own mental wellbeing. Clear policies and financial investments are required to convert these feasible and acceptable solutions into intervention components using a comprehensive co-production approach. Once co-produced, these intervention components could be evaluated preferably as a care bundle using a factorial trial design to better understand which components work best for which patient and carer outcomes. This coproduction and evaluation approach would generate crucial knowledge to ensure the longevity and cost-effectiveness of care-inclusive transitional care interventions.