Study design

The research design was an exploratory sequential mixed methods study, based on the approach outlined by the James Lind Alliance (JLA) for identifying research priorities [17], re-purposed to identify priorities for improving services. Phase 1 involved a qualitative approach for initial priority gathering, based on an open-ended (qualitative) online survey and interviews. Framework analysis [18] was used to identify a long-list of specific priority improvements to be used in a second phase of data collection. Phase 2 involved a quantitative survey, where participants were asked to rank their five priority improvements from the initial long-list. The phase 2 survey results were also used to inform a final stakeholder meeting, where a small group of stakeholders (n = 12) decided on a final set of priority improvements, to be evaluated in further detail using epidemiological and economic modelling.

Participants

Three groups of stakeholders were invited to participate across phases 1 and 2 of the study– 1) stroke survivors, 2) family or main carers of stroke survivors, and 3) professionals involved in stroke delivery in Ireland, including healthcare professionals (HCPs) and people working in research, policy and advocacy, and senior clinical and policy leaders.

Phase 1 recruitment

Purposive sampling was used. Stroke survivors and family member/main carers were recruited for interviews through multiple sources. This included social media platforms and a range of voluntary organisations representing and providing services to stroke survivors in different parts of the country, including those catering to specific groups such as survivors with aphasia. This helped to ensure a heterogeneous group that was as reflective as possible of the stroke population. Participants were not known to the interviewer prior to the study. Detailed information on participant recruitment and inclusion criteria is provided in the protocol [15]. Briefly, inclusion criteria for stroke survivor/main carer participants were: 1) had (or cared for someone who had) a diagnosis of stroke more than 6 months prior; 2) aged 18 years and over and resident in the Republic of Ireland; 3) able to converse in and understand English; and 4) capable of providing informed consent. Sample consent forms and information leaflets are available at the OSF link above, including aphasia-friendly materials and accessible materials for individual with cognitive impairment. The recruitment target was 12–15 stroke survivors and 12–15 carers. We also aimed to carry out a small number of interviews (8–10) with individuals identified as key clinical, policy and research leaders in stroke.

Recruitment for the interviews began in April 2021. As recruitment progressed, we identified gaps (e.g., geographical area, severity, specific patient/service experiences) and re-targeted recruitment strategies accordingly. For example, we expanded the criteria to include people who had, or were caring for someone with, a diagnosis of vascular dementia (as a result of stroke or other vascular pathology, such as multiple transient ischemic attacks). Recruitment continued until March 2022, with data collection concluding in April 2022. The decision to stop recruitment was made in consultation with the study steering group, on the basis of data saturation along with feasibility and time constraints. Data saturation was considered at two levels–code saturation (no additional issues were identified) but also meaning saturation (no further dimensions, nuances or insights of issues were identified) [19].

HCPs and policy/research/advocacy stakeholders were recruited for the online qualitative survey through multiple approaches (see protocol for more detail [15]). Information about the study was disseminated through social media, personal networks of the research team and collaborators, a range of professional organisations and networks, and through voluntary bodies representing and providing services to stroke survivors in Ireland. Professionals had to be involved in stroke care delivery in the Republic of Ireland, either directly or in a policy/research/advocacy role. Survivors and carers were given the option of completing the online survey if they preferred that to an interview. Recruitment for the survey began in July 2021 and was open until November 2021. The target for recruitment was a minimum of 80–100 responses.

The interviews were primarily targeted at survivors and carers, although they were also given the option of completing the online survey if they wished. It was anticipated that the interviews would be easier for some individuals in this group, due to varying levels of computer literacy and access to technology. More importantly, the phase 1 survey questions were open-ended, requiring potentially lengthy written responses, and more effort than a verbal interview. Consistent with this, few survivors and carers availed of the survey option (n = 8). The interview format allowed us to generate highly relevant rich narrative data from survivors and caregivers on their experiences.

Phase 2 recruitment

Phase 1 participants who consented to be contacted again were invited to participate in a Phase 2 survey, either by email or text message. We also recruited further participants through the same channels as Phase 1 –social media, voluntary organisations, and professional bodies. The inclusion criteria were the same as for Phase 1. Recruitment for the survey took place from November 2022 to January 2023. As participation in the Phase 2 survey was through an anonymous link, it was not possible to determine how many of these participants also took part in Phase 1.

We decided that a target sample size of 96 would be feasible, with a 10% margin of error in estimates of the proportion of participants choosing each improvement, when considering the total sample [20]. A minimum of 43 participants in the professional group, and 43 in the combined survivor and carer group, would allow estimates specific to these groups with a 15% margin of error. We considered that this was a reasonable level of precision, as the survey results were not the final stage of the process, but were discussed and finalised in a stakeholder meeting, as per the JLA process.

Stakeholder meeting

In February-March 2023, a small group of stakeholders (target n = 12, balanced across stakeholder groups) were invited to take part in a meeting, based on the nominal group technique [19], to decide on a set of priorities to be examined further using epidemiological and economic modelling, using an already established population-based model [5]. Participants were randomly selected from the list of Phase 2 survey participants who indicated an interest in attending a meeting.

Informed consent

All participants at each stage provided written informed consent. For the interviews, consent was assessed on a continual basis, following the principle of process consent–at the initial consent form stage, immediately prior and during the interview, and after the interview has concluded. It was assumed that individuals had the capacity to consent, unless otherwise demonstrated, as per the Assisted Decision-Making Capacity Act (2015) and the National Consent Policy [21]. The consent process and all aspects of data management are described in detail in the study protocol [15].

Data collection and analysis

Phase 1 –Initial priority gathering—Participants

In-depth interviews were conducted with 18 survivors (8 women, 10 men, age range 39–81) and 13 family/main carers (12 women, 1 men, age range 34–76). There was a total of 29 survivor/carer interviews—two were carried out with survivor/carer dyads, and 27 were individual interviews. Interviews were conducted with 8 senior leaders (6 women, 2 men). Of the 37 total interviews, 23 were phone interviews, 13 Microsoft Teams calls, and one face-to-face. Interviews lasted between 30 and 130 minutes, with an average length of 69 minutes. One interview was split over two occasions to accommodate interviewee fatigue. There were 88 responses to the online survey– 80 professionals working in stroke care, 5 family/main carers and 3 survivors.

One survivor and one carer who initially provided contact details did not respond to follow up phone calls, and one survivor and one carer declined to participate in an interview following review of the participation information leaflet. None of the interview participants who signed and returned a consent form withdrew from the study. In addition to the 8 senior leaders who completed interviews, a further 10 were invited, with no response received from 8, and 2 declining to take part. For the online survey, 47 participants consented to participate but did not respond to any of the survey questions. Of the 88 eligible responses who answered at least one question, 33 answered 9/9, 39 did not answer 1 or 2 questions, and 16 did not answer more than 2 questions.

S2–S5 Tables display the participant characteristics in detail. All geographical areas of Ireland were represented within each stakeholder group. Among the professionals responding to the survey, 49/80 (61%) classified themselves as an allied health professional, and 21% as another health professional (e.g., doctor, nurse). The remainder included researchers, managers and voluntary sector staff. The most common service area was acute services (45/80, 56%), followed by rehabilitation services (24/80, 30%).

Phase 1 –Initial priority gathering–Results.

The final analytic framework consisted of 13 higher level categories. Six related to care domains–pre-hospital care, acute care, rehabilitation, primary prevention, secondary prevention and life after stroke. Four categories related to navigation–the healthcare environment, navigation behaviours, individual characteristics, and provider factors. There were three cross-cutting categories–national policy, quality of care, and social support. Within these higher-level categories, there were 37 sub-categories. These were mapped to themes related to 15 priority care areas, grouped within seven broader areas. These seven broad priority care areas are described briefly below, and in Table 1. Detailed descriptions of the analytic framework, including the 13 higher-level categories, 27 sub-categories and corresponding codes, and sample quotes are available at the OSF link above.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Phase 1 priority areas of care based on qualitative responses from survivors (n = 20), carers (n = 18) and professionals (n = 80).

https://doi.org/10.1371/journal.pone.0297072.t001

Primary and secondary prevention

All stakeholders identified awareness and understanding of stroke risk factors as a key priority for prevention. Stroke survivors and carers described missed opportunities to prevent a stroke, including failures to detect high blood pressure, intermittent atrial fibrillation (AF), or recognise the symptoms of or adequately manage a transient ischemic attack (TIA). Less well known risk factors such as sleep apnoea and stress were mentioned frequently.

Professionals working in stroke care emphasised the need for structured, evidence-based secondary prevention programmes, involving education, advice and support for behaviour change. In general, survivors found support for lifestyle change fairly limited, with advice tending to be general, with “nothing concrete” as one survivor [S_03] put it. Gym access, adapted exercise equipment and exercise classes were identified as helpful for staying active. An adapted bike provided by local community services was highlighted by three participants (C_09, S_17 and C_02) as very helpful for staying active, however this bike was provided only for a limited time due to a shortage. One survivor paid for a similar bike out of pocket following this, and saw it as essential to staying active—“I was never going to improve if I didn’t use that you know.” [S_17].

Support for carers and carer involvement

Carers described supporting survivors in multiple ways—acting as advocates, supporting navigation, meeting basic care needs, supporting recovery and prevention, and social and emotional support. Many carers felt isolated and unsupported as they cared for their loved one, while also adjusting to the changes to their lives as a result of the stroke. Despite this, some carers found it difficult to access support, with many carers saying they did not feel “ready” for counselling or felt they could not prioritise their own mental health–“I didn’t have time for support really” [C_11].

Family carers often did not feel involved enough in the survivor’s care–they were not kept sufficiently informed, did not have opportunities to be involved in therapy, and did not feel that they had sufficient training. Carer involvement was particularly poor during Covid-19 restrictions.

Awareness, research and policy

Survivors and family carers often encountered health-care staff with limited understanding of post-stroke cognitive and communication difficulties, and how to best support individuals with these problems. Friends, extended family, and the general public also often had limited understanding of stroke, particularly hidden disability such as problems with fatigue, cognition or communication.

Professionals working in stroke care highlighted the need for implementation of existing policies and strategies, and improved routine data collection, audit, and research. The need to recruit and retain more specialist stroke staff was a frequent concern among professionals, and seen as a necessary pre-condition to improving services.

Stroke survivors and family carers identified a need to be involved in service development—“it’s about bringing resources, and it’s about bringing knowledge, and it’s about bringing I suppose the capabilities to the table” [S_02]

An over-arching theme of under-resourcing

A common thread or over-arching theme running across all seven priority areas was a general scarcity of staff and resources to match the needs of the stroke population. This was reflected in waiting lists for needed services and pressure on staff leading to poor communication and information provision. It was notable that communication was often experienced as better in rehabilitation settings, which tended to be less busy and pressured relative to acute settings. Navigation challenges also appeared to be linked with this underlying scarcity of resources. For example, one key barrier to effective navigation was the use of unclear or arbitrary eligibility criteria for services, such as age limits—one senior leader described this as a “crude” way of “rationing” services that were limited in supply [SL_08]. The variation in service provision around the country can also be linked with scarcity, as gaps in services have been filled by small-scale, local-specific initiatives developed by local champions and voluntary organisations, resulting in what one senior leader (SL_04) described as a “patchwork” of services.

Long-list of improvements

The seven priority areas of care were mapped to a long-list of 45 potential improvements. There were 6 improvements identified for acute care, 7 for communication and navigation, 8 for rehabilitation, 10 for life after stroke, 7 for prevention, 2 for carer support, and 5 for awareness, research and policy. This complete long-list is available at the OSF link above.

Phase 2 –Priority setting survey—Participants

The survey was completed by 95 participants: 42 professionals (all online), 34 survivors (28 online) and 19 family carers (17 online). Two family carers and three survivors completed the survey by post. Three survivors completed the survey via facilitated interview, two by phone and one in-person. S5 Table displays the characteristics of the Phase 2 survey participants, with 2/17 family carers and 8/33 survivors aged 65+, and 5 survivors aged 70+. Across groups, the highest number of participants were Dublin-based, but all geographical areas of Ireland were represented. Among professionals, there was a mix of profession and service area.

Phase 2 –Priority setting survey—Results

There were five priority improvements that were selected by at least 20% of the participants as a whole (see Table 2). These were 1) improved access to community neuro-rehabilitation, 2) a comprehensive community stroke support service, 3) investment in staff recruitment and retention, 4) a service to provide information and support with navigation, and 5) improved access to stroke specialist acute care.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Phase 2 top priority improvements selected by professionals, survivors/caregivers and overall.

https://doi.org/10.1371/journal.pone.0297072.t002

Within the stroke survivor/carer group, 25% prioritised exploring ways to improve access for strokes with atypical presentation. Among professionals, 24% prioritised specialist inpatient rehabilitation, and 24% prioritised early supported discharge. No improvement within the prevention area of care was selected by at least 20% of any group. The full results of the survey are available on OSF.

Stakeholder meeting

20 participants were invited to take part in the stakeholder meeting, with 10 agreeing to participate and attending on the day (4 survivors, 2 family carers, and 4 professionals working in stroke care). The professionals included a clinical nurse specialist, a speech and language therapist, a clinical neuropsychologist and an occupational therapist. A further two study steering group members participated in the meeting and the voting–the study PPI representative, and an academic researcher (not the lead researcher).

Prior to the meeting, a candidate list of priority improvements was compiled, corresponding broadly to the top priorities listed in Table 2. Additional priorities were included that were selected by either at least 20% of survivors or at least 20% of carers. This included, for example, support for mental health difficulties (>20% of survivors) and support for carer involvement (>20% of carers). A number of these additional priorities related to information and navigating the system, and these were combined into a single improvement of services for providing support with information and system navigation.

“Exploring ways of improving access for atypical presentations” was excluded from consideration for evaluation with modelling, due to insufficient evidence for effectiveness of potential interventions, and an absence of any relevant evidence-based guidelines. “Investment in recruitment and retention of stroke specialist staff” was excluded from consideration due to difficulties quantifying a direct benefit for stroke survivors or their families. These exclusions were discussed and agreed in the meeting.

In order to maintain the focus on stakeholder priorities, it was decided to disregard the pre-specified criteria that required a prevention improvement would be included, as none met the criteria for the candidate list. All of the candidate improvements had RCT evidence of effectiveness, and almost all had systematic review evidence. This included, for example, Cochrane reviews providing evidence for the effectiveness of community rehabilitation [34, 35] and information provision [36]. The list of candidate improvements, and the advance information provided to meeting participants, including information on effectiveness, is available at the OSF link above.

The general meeting discussion echoed the Phase 1 findings. Meeting participants emphasised that services need to be well-organised and co-ordinated, stroke-specific where possible, and national and consistent, with the same services provided everywhere. There was broad consensus in the discussion around the final top 5 improvements, and this was also reflected in the voting. These were 1) Specialist community neuro-rehabilitation (12/12, 100%), 2) Services to provide information and support access to services (11/12, 92%), 3) Community support service with long-term access to meaningful, social and beneficial activities and programmes (9/12, 75%), 4) Specialist acute care (9/12, 75%) and 5) Support for mental health difficulties (9/12, 75%).

These aligned broadly with the top priority improvements identified in Phase 2, with the exception of support for mental health difficulties, which had been slightly below the threshold for a top priority, but which meeting participants felt was important to include (see Table 2). Family carer involvement was not selected as a final priority, but participants emphasised that this should be a key consideration across services. S1 Fig shows how priority improvements were identified at each stage of the research.

Implications for policy, practice and future research

There is an urgent need to implement specialised community neuro-rehabilitation for all areas of the country, accessible to all age groups, at a sufficient level of intensity, and based on a standardised assessment of need. This is a key element of the managed clinical rehabilitation networks that are envisaged for Ireland [7]. This finding supports the increasing international emphasis on the need for greater hours and intensity of rehabilitation post-stroke, evidenced by recent UK guidelines [37]. The guidelines also emphasise that rehabilitation must be tailored to individual needs, consistent with our findings. Recent research suggests that variability of individual needs (along with resource constraints) can complicate guidelines-based delivery of rehabilitation, indicating that a “one-size-fits-all” approach is not feasible [38].

Findings from this study also support an emerging recognition internationally that time-limited neuro-rehabilitation is not sufficient, and that stroke survivors need longer-term support, for this key “Life After Stroke” phase [22]. There is no current consensus on an optimal model of care for this phase of care, although there is work ongoing [39]. Important elements likely include evidence-based therapies such as self-management [40], and that address key areas of unmet need such as fatigue and emotional support [14]. These types of services are currently offered in Ireland by voluntary organisations, but in a fragmented, piecemeal fashion, and need to be offered and publicised in a consistent way around the country.

Issues related to service navigation and communication are mentioned in key stroke national and international policy documents [22, 32], but are arguably not addressed at a level of detail that reflects their importance to stakeholders. The findings from this study indicate that a comprehensive plan is needed for information provision for stroke, optimally co-designed with stroke survivors and families. This would involve active information provision [36] at multiple time-points, in multiple formats, tailored to the individual, and with family members and carers included.

Delivery of stroke services is complex due to the range and individual nature of post-stroke deficits, and as such requires both specialisation of tasks and co-ordination across tasks. Healthcare systems often prioritise specialisation at the expense of co-ordination [31], and this may explain the frequent challenges encountered internationally with care continuity and patient navigation. The solution to this likely lies in both services to support patient navigation (e.g. individual key workers, group-based supports) and clearer, more standard and transparent care pathways that are easier to navigate [31].

There has been considerable focus on reducing pre-hospital delay in stroke policy, with the emphasis on detection of FAST-positive cases (a mnemonic for the public for facial drooping, arm weakness, speech difficulties; time to call emergency services). However, the need to detect non-FAST cases has been relatively neglected in research. Tools such as BE-FAST broaden the criteria to include balance and visual symptoms [41], but do not include non-traditional symptoms such as mental status change which can be frequent particularly in women [42]. Many people with milder strokes, including those with “non-disabling” symptoms, can experience barriers and delays in treatment, despite the fact that these strokes can be associated with longer-term neurological deficits [43]. Further research is needed in this area to help the general public and healthcare professionals identify strokes associated with non-traditional stroke symptoms quickly and earlier, without the risk of casting the net too wide and overwhelming services with stroke mimics. Machine-learning is one approach that may show promise in this area [44].

The study findings indicate a general pattern of under-resourcing along the stroke care pathway, with stakeholders recognising particular deficits in staffing. There is a need for greater understanding of the level of demand for stroke services in Ireland, and the resource required to meet that demand. Policy-makers in Ireland have emphasised the need for population-health-based planning to inform service decisions [45]. A key next step of the current project will be to use population-based modelling to estimate current and future need for stroke services in Ireland [5, 15], and evaluate alternative policy and service options in terms of cost and cost-effectiveness.

Strengths and limitations

A key strength of this study has been the combination of rich qualitative data with more specific quantitative data. The scope of this study was more open and less pre-defined than previous qualitative studies, allowing participants to highlight issues that may be less well known to researchers and clinicians. We used an inclusive approach to recruitment and data collection at all stages to facilitate inclusion of participants with cognitive and communication difficulties. Limitations include the small sample size, particularly for caregivers in the Phase 2 survey. Caregivers were a particularly difficult group to recruit, perhaps because they are uniquely busy, and may not be as engaged in relevant organisations. This likely meant that carer priorities were not adequately represented, and that the needs of people with more severe stroke were under-represented. Another key limitation was the absence of any prevention improvements in the final priority list. The focus on perspectives and experiences of stakeholders may have led to prevention strategies being neglected, as participants were pre-occupied by the more immediate needs of stroke survivors and their families themselves. Although these findings are specific to the Irish context, they are of international relevance as similar challenges and priorities are likely to be shared across different countries and healthcare systems. Identifying where there are similarities and differences in challenges across healthcare systems is also crucial for identifying potential solutions.

Conclusion

This study has generated broad and rich evidence on stakeholder views and priorities for improving stroke care in Ireland, as well as a set of specific top priority improvements. Evidence from the literature indicates that many of these concerns are shared across countries. These findings have the potential to inform policy and service development in stroke in Ireland, and also internationally. They also indicate a number of areas for future research. The next step in this project will be to use an already established epidemiological and economic model of stroke [5] to evaluate improvements identified in this research, to provide concrete evidence to policy makers in relation to population demand, cost and cost-effectiveness.