Design

National interdisciplinary one-day prioritisation workshop held face-to-face with a range of stakeholders: patient and carer representatives, clinicians, researchers, and policymakers. Reporting according to REPRISE guidance (see supporting information) [41].

Patient and public involvement

A Patient and public involvement group supported development of the study aims and methods; three public contributors participated in the workshop, contributing to focus groups and the ranking exercise, and commented on the findings.

Participants and participant identification

Participants were clinicians from any profession or specialty with experience in providing care to people with serious illnesses, researchers and policymakers with an interest in this area, people with lived experience of serious illness and their informal carers. We sought a group with relevant knowledge and experience of uncertainty, and who had an interest in the area [42]. Based on previous research prioritisation work in this field, we aimed for an a priori sample size of at least 30 participants [43]. To reach this group we first disseminated the workshop invitation to a national interdisciplinary group of clinicians and researchers with a special interest in serious illness uncertainty. We then invited clinicians and researchers via the Applied Research Collaborative (ARC) East of England, and through local, regional and national clinical networks, focusing particularly on palliative care, geriatrics, intensive care and acute medicine (based on the likelihood of these groups having experience of caring for those with serious illness), and also including a national nursing forum. People with lived experience of serious illness and their carers were identified through local and regional patient and public involvement fora. Invitations were sent from December 2022, and registration was open until 27^th February 2023. The workshop was held on 28^th February 2023. Workshop attendees were informed of the research component in advance.

Workshop process

The workshop drew on existing approaches to prioritisation, incorporating idea generation, consolidation, and ranking stages [44]. Written informed consent and self-reported demographic data were collected from workshop participants at the start of the day, including age, gender, ethnicity, participants’ status as a researcher, clinician, policymaker, or patient representative, and details of their field where relevant. To encourage debate, the workshop began with presentations on the “state of the science” concerning uncertainty in serious illness. Models of uncertainty and evidence gaps from relevant literature reviews were outlined [11, 32, 34, 45, 46]. The focus was primarily on those uncertainties that are irreducible i.e. cannot be eliminated.

We then held focus groups with participants to explore their views on uncertainty and identify key areas for future research. This approach mirrored the idea generation stage of the nominal group technique and the first round of a Delphi process [44, 47]. The topic guide was developed by the research team and was informed by literature review: it focused on experiences of uncertainty, views on desired outcomes when addressing uncertainty, and ideas for research questions concerning uncertainty in serious illness. Focus groups were led by clinicians and researchers with expertise in facilitation. Conversations were audio recorded and a scribe within each group recorded the research questions identified by participants.

Research areas from the focus groups were collated into a summary list during the day. In the final session, this summary list was presented to participants who were invited to anonymously rank the top three areas in order of priority using the online ranking tool “Slido” (© 2023 Cisco Systems, Inc.). Items were presented in random order. See supporting information for full workshop programme and topic guide.

Data analysis

Participant details

Thirty-three participants took part in the focus groups, and 34 in the ranking survey, of whom 30 provided demographic information. The average age was 47 years (range 33–67), and 67% were female. 80% were of white ethnicity, 10% asian, and 10% from mixed or multiple ethnic groups. 70% were clinicians, 43% researchers, 10% patient or carer representatives, and 7% policymakers (participants could state multiple roles). Of the clinician participants, twelve had a background in palliative care, three geriatrics, two nursing and one each of intensive care, general practice, psychology, and physiotherapy.

Item generation and ranking

Five focus groups with six to seven participants in each were of 53 to 64 minutes’ duration. The groups generated over 60 research questions, which we condensed to produce the 10 priority areas that were then ranked by participants (Table 1). Communication of uncertainty was the highest-ranked item, scored first by 15 participants. Participants ranked the next four priorities almost equally: coping with uncertainty; training health professionals; optimising clinical approaches to uncertainty; understanding in-depth experiences of uncertainty. The other areas received lower priority scores.

Download:

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

Table 1. 10 priority areas for future uncertainty research in serious illness (n = 34).

https://doi.org/10.1371/journal.pone.0289522.t001

Detailed research priorities

We explored the 10 priority areas raised by workshop participants during qualitative analysis, and identified detailed sub-questions within each area (Table 2).

Download:

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

Table 2. Detailed research priorities from framework analysis.

https://doi.org/10.1371/journal.pone.0289522.t002

Priority areas are listed in this table in the order they were ranked

The highest priority research areas were related to uncertainty management, e.g. through improving communication, supporting individual coping mechanisms, or providing training:

“So it’s I think about that management of uncertainty … just having the confidence to say something, however vague it might be that will still provide some comfort.” Patient representative, focus group 2

Participants particularly focused on how and when to have conversations about uncertainty, and how to individualise communication to promote psychological wellbeing:

“Then there’s a really important research need around understanding more about how we communicate uncertainty and the consequences of how we communicate uncertainty and how we can talk about uncertainty in a way which is more patient-centred and supportive and considers sort of psychological wellbeing.” Social scientist, focus group 3.

Training of health professionals at varying stages was a recognised priority. This included training professionals on toleration of their uncertainty as well as how to address the uncertainty experienced by others. One participant highlighted that inadequate training of doctors to manage uncertainty has been a longstanding issue and queried the best timing of such training:

“We’ve been inadequate at teaching them [medical students] how to actually manage this complexity for a long time, although you could argue that it’s quite difficult to teach until they really get into the nitty-gritty of practising.” Geriatrician, focus group 1

Participants recognised it is unlikely a single intervention can address all the nuanced multilevel aspects of uncertainty, but nevertheless, some felt interventions could play a role in addressing uncertainty, as long as they were situated in a broader societal context:

“There are interventions that are being developed out there but there are no interventions that deal with all the different types and layers of uncertainty and they can’t by nature. And I think there is that kind of relational uncertainty, the kind of organisational uncertainties, there are uncertainties on macro, meso, micro levels and you can deal with one component, but you can’t deal with all the different components, and it’s how those interventions work within the broader societal context of uncertainty.” Occupational therapist, focus group 4

Older frail patients were seen as a priority for future research into how uncertainty can be managed. Participants noted a lack of knowledge about approaching uncertainty in the context of frailty:

“For old and frail elderly there’s no support you know, for that anxiety, fear, so it’s how do we look at that, you know, and how do we maybe look at how we can support, supporting that frailty and older.” Professor of end-of-life care, focus group 5

Participants identified research questions concerning experiences of uncertainty, though these were generally given lower priority than questions relating to managing uncertainty. Participants identified illness contexts where uncertainty experiences could be explored further, including critical illness and frailty. They also acknowledged it is important to explore how the experience of uncertainty varies depending on perspective (patient, carer or health professional), or individual characteristics:

“We’d want to look at different groups such as sort of learning disability, neurodiversity, sort of hard-to-reach areas whether deprivation, LGBTQ+, sort of that kind of differences you might get.” Palliative care consultant, focus group 1.

Participants identified potential positive impacts of uncertainty and its utility in certain situations. They suggested it was important to understand why some health professionals thrive when required to manage uncertainty, whereas others exhibit a lower tolerance. Participants noted the relationship between uncertainty and hope and reasoned that positive aspects of uncertainty should be explored further:

“I wonder….whether there’s something about learning to cope with uncertainty or to tolerate uncertainty and whether people, patients and families can see that as a positive as well.” Bereavement practitioner, focus group 4

Additional areas for future research

Some questions identified from focus group transcripts were not identified as research priority areas by participants. Additional areas included consideration of the legal and regulatory implications of uncertainty and its management, the link between uncertainty and patient safety, how to prepare the public for serious illness uncertainty, and the resource impacts of different levels of tolerance to uncertainty:

“Have we studied the use of resources around uncertainty? Because I’m sure loads of tests are done completely unnecessarily because people just want to be sure.” Palliative care consultant, group 1

Summary of findings

This study co-produced clinically-focused research priorities to address known evidence gaps concerning uncertainty in serious illness using a stakeholder workshop approach. Optimising communication of uncertainty was the top research priority. Research questions relating to managing uncertainty were considered higher priority than research investigating experiences of uncertainty.

Discussion of main findings

Communicating uncertainty was the top priority for participants. In some senses this is not surprising as it reflects key evidence gaps and recommendations in this field [34, 49], however this is, to our knowledge, the first time that communication of uncertainty has been demonstrated as the highest research priority by consensus methodology. In their narrative review of uncertainty communication, Simpkin and Armstrong identified a number of evidence gaps, including identifying individuals’ communication preferences and tailoring communication to those preferences [34], and a key finding here was the number of detailed questions participants identified in relation to communicating and managing uncertainty (Table 2). For example, the question of how to maintain hope whilst communicating uncertainty was noted as a priority; this has been explored in cancer care [50], but remains a key question in other serious illnesses. Additionally, participants raised several sub-questions in terms of how to discuss uncertainty, reflecting the need for implementation-focused communication research.

After communication, participants prioritised other aspects of managing uncertainty: identifying how individuals can be supported to cope with their own uncertainty; investigating how we can optimise clinical approaches to uncertainty; understanding more about how to equip health professionals to deal with uncertainty through training. Clinicians in particular focused on the importance of improving health professional training. Though management of uncertainty has been recognised as a core component of medical training for decades, curricula still make limited reference to uncertainty, and filling this gap should be a priority.(2) Whilst we have an improving understanding of how physicians manage uncertainty [31], the evidence base for other professional groups is very limited, yet nurses and allied healthcare professionals often lead the clinical care and support of older people and their families. To date, uncertainty management and communication interventions have had variable impact in serious illness [32], and often prove challenging to evaluate [33, 51]. There is scope for further intervention development [52], especially work that focuses on evaluation of uncertainty-focused interventions.

Whilst many of the questions identified by participants related to investigating experiences of uncertainty, these were usually considered lower priority, perhaps because much is already known about uncertainty experiences [7, 53]. For example, conceptual taxonomies of uncertainty are well developed [6] and there have been evidence syntheses of experience in some specific areas e.g. multimorbidity [11]. However, there are still evidence gaps concerning how uncertainty affects individuals in other clinical contexts, in particular older patients living with frailty. Given the rapidly increasing complexity of the healthcare system and unpredictability of the frailty trajectory, this is an area that warrants urgent exploration [34]. Understanding more about the impacts of uncertainty on experiences of illness, care, and bereavement would enable us to develop interventions focused on the real-world problems uncertainty can cause.

Strengths and limitations

The rapid prioritisation approach we used enabled a diverse group of interested individuals to generate and rank research priorities in a single day. Those ranking the priorities had spent the entire day considering uncertainty in serious illness and were well placed to express considered views when ranking the list presented to them. By identifying evidence gaps before the workshop and communicating these to participants during initial presentations, we were able to focus on areas where more research is needed and incorporate the key stages of a traditional prioritisation exercise. By additionally incorporating formal qualitative analysis we increased rigour and developed a robust priority list. This approach was a more feasible and pragmatic alternative to lengthier methodologies such as the Delphi process[42, 54]. However, the rapid nature of the prioritisation process meant there was limited time to condense the findings of the focus group discussions, which risked a loss of accuracy and ranking was limited to broad research areas. We ameliorated this by subsequent analysis of focus group transcripts, which enabled us to identify detailed research questions identified by participants, including areas that participants mentioned even when they weren’t identified as priorities at the time.

We incorporated patient and carer experiences, but most participants were healthcare professionals or researchers, thus these groups were not equally represented in the ranking process. t. We recruited a UK-wide sample, but this was not an international study, and future work should explore if these findings hold internationally, though literature from 17 countries reported consistent findings on a similar topic [11]. Whilst not specifically excluded, social care professionals did not attend this workshop, which limits the findings to healthcare. The anonymous nature of the ranking means we could not account for the background of participants when analysing ranking data, which means we were unable to look for systematic differences in response between stakeholder groups, however in the qualitative analysis we did not identify any major differences between professional groups, or between patients, clinicians or researchers. We sought to include a wide range of views of people with expertise, experience, and interest in this topic, and achieved a varied sample comparable in size to other similar studies [43], but our sampling method meant there was no way to calculate a response rate, and it is possible that a different group may have identified different priority areas. The largest group of clinical participants were from palliative care backgrounds which may have shaped their views; however, a range of health professionals and researchers were represented, and the degree of agreement, particularly regarding the top priority of communication suggests the findings represent true consensus.