Synopsis

Introduction

Rationale for research and background

Delirium is a prevalent neurocognitive disorder that poses significant challenges in the healthcare setting, particularly for older adults. This acute condition is primarily characterised by disturbances in attention and awareness, often accompanied by cognitive impairments and behavioural changes. Delirium frequently arises as a direct physiological consequence of other medical conditions, marked by an acute onset and a fluctuating course.1 The prevalence of delirium in hospitalised older patients is high, with studies indicating rates as high as 20% among those admitted acutely overall and up to 45% among individuals with pre-existing dementia.2,3

The consequences of delirium are profound, affecting not only immediate health outcomes but also the long-term quality of life. Patients experiencing delirium face extended hospital stays, a heightened risk of complications, emotional distress, poor functional recovery and increased mortality.4–9 Importantly, cognitive and functional deficits can persist long after the acute episode, with research reporting that 21% of those affected may exhibit persistent symptoms 6 months post discharge.10 Such prolonged challenges not only hinder patient recovery but also impose substantial burdens on caregivers and healthcare resources, with inpatient costs related to delirium varying between £12,575 and £49,689 per episode.11

While previous research has primarily focused on delirium prevention during hospital admissions and short-term management guidelines,12,13 a critical gap exists in the literature regarding post delirium recovery, particularly concerning the development of effective intervention strategies. Recent National Institute for Health and Care Excellence (NICE) guidelines for dementia underscored the pressing need for further investigation into clinically and cost-effective non-pharmacological interventions tailored for long-term recovery in individuals experiencing delirium superimposed on dementia.14

The urgency of this research was amplified by evidence indicating that many patients with delirium do not fully regain their cognitive and functional abilities, leading to an increased likelihood of requiring additional care or institutionalisation.5,7,10,15–19 Furthermore, carers of individuals with delirium report experiencing distress and uncertainty regarding the support needs of their loved ones.20 Existing studies note a decrease in the performance of activities of daily living (ADL) post delirium, reinforcing the need for effective rehabilitation strategies to address ongoing care requirements.4,21 Despite the recognition that rehabilitation could be beneficial for older adults with cognitive impairments, there is a pervasive misconception that such individuals lack rehabilitation potential.22 This highlights the need for evidence-based interventions specifically designed to support recovery after delirium.

The RecoverED study was grounded in the principles outlined by the Medical Research Council (MRC) guidance for the development and evaluation of complex interventions.23 The study involved a systematic exploration of existing theory and evidence through realist review and comprehensive qualitative research, aimed at refining the understanding of the recovery domains needed for patients post delirium. Engaging patients, carers and healthcare professionals (HCPs) in the co-design process allowed the development of an intervention tailored to meet the complex needs of those recovering from delirium.

The RecoverED study sought to improve patient outcomes by establishing a rigorous evidence base for rehabilitation strategies that promote recovery. By addressing the gaps in current knowledge and practice, the study aimed to address research questions related to the therapeutic components of the intervention, its feasibility, acceptability among diverse populations, and ultimately, its effectiveness and cost-effectiveness compared to usual care. Specifically, it set out to develop and test an intervention designed to improve recovery after an episode of delirium, ensuring that it would be suitable and accessible to all affected individuals, including those with previously diagnosed dementia and vulnerable groups.

A summary of any alterations to the programme’s original aims/design

The RecoverED study underwent several substantive amendments to work package 2 (WP2). Notably, the following key changes were made:

• Midway intervention review: A substantial amendment (dated 26 January 2023) introduced an individual participant midway intervention review, whereby a therapist accompanied the rehabilitation support worker (RSW) during 1 of the 10 intervention sessions. This allowed for real-time evaluation and adjustment of the intervention, ensuring participants received tailored support.
• Eligibility criteria adjustments: Eligibility criteria were refined (dated 25 October 2023) to specify a ‘clinical diagnosis of delirium while admitted to the hospital’, replacing the previous condition requiring delirium to last for > 48 hours.
• Geographical exclusion: The eligibility criteria also excluded participants living outside the geographical reach of the intervention teams, focusing the study on those who could realistically receive the intervention.
• Recruitment process modifications: The design was adapted to allow rapid consent and baseline procedures for patients discharged from the hospital within 1 week, using a face-to-face visit from a clinical researcher, thus streamlining the recruitment process.
• Compensation for HCPs: A financial compensation of £35 was introduced for HCPs [physiotherapists (PTs), occupational therapists (OTs) and RSWs], participating in the process evaluation, to help address recruitment challenges (dated 6 February 2024).

The recruitment period for the RecoverED study was extended twice to address unforeseen challenges at multiple sites. Delays occurred in opening two of the planned six sites due to difficulties in securing therapists and RSWs, alongside slow setup progress. Additionally, two other sites faced issues with poor screening and documentation practices for delirium, compounded by limited staff capacity, resulting in recruitment rates falling below target. Following discussions with the Programme Steering Committee (PSC), sponsor and funder, the recruitment period was extended from 31 October 2023, first to 31 December 2023, and then to 31 March 2024. However, recruitment at two sites could not continue beyond January 2024 due to persistent staff capacity issues.

All amendments were submitted and approved by the London South East Research Ethics Committee (REC), ensuring compliance with ethical standards. The scientific justification for these changes centred on the need to enhance participant support, improve recruitment and adapt the study design to logistical realities, thereby strengthening the overall feasibility and impact of the intervention. These modifications were documented and endorsed by the PSC, sponsor and the funder prior to submission, ensuring alignment with the study’s objectives and ethical considerations.

Objectives

• To convene an expert panel to design an intervention to improve recovery after delirium.
• To conduct a feasibility study assessing the acceptability of the intervention as well as the feasibility of conducting a future full-scale randomised controlled trial (RCT) and cost-effectiveness analysis (CEA).
• To examine the acceptability of the intervention for people from underserved populations.
• To conduct a definitive RCT and CEA of the intervention in people who have had delirium compared with usual care, with parallel process evaluation.
• To conduct an implementation study of the intervention.

Research pathway diagram

Figure 1 outlines the WPs for the RecoverED study. WP1 focused on designing the intervention using insights from realist reviews, qualitative studies and expert panel guidance while incorporating feedback from patient and public involvement and engagement (PPIE). WP2 conducted a multicentre, single-arm feasibility study to refine the intervention through process evaluations and data analysis.

FIGURE 1 Diagram of RecoverED study pathway. (Figure not reproduced here; see the original article.)

Following scheduled check point reviewer, the programme was discontinued after WP2 due to concerns about the feasibility of progressing definitive RCT and implementation study.

Work package 1: intervention design

Aims

Work package 1 aimed to develop a robust programme theory to understand what improves recovery after delirium, for whom and in what context. This served as the foundation for designing a complex rehabilitation intervention for older adults (aged ≥ 65 years) recovering from delirium following acute hospital admission.

Methods

Program theory and logic model

The RecoverED project adhered to the MRC guidance on developing and evaluating complex interventions, which emphasises the importance of taking a theory-driven approach to understanding how complex interventions can effect change and how they interact with the diverse contextual factors and wider dynamic systems within which they are implemented.24 Using a realist approach, a programme theory of what works to improve recovery from delirium, for whom and in what context was developed using a multistage process involving:25

1. Rapid realist review: The first stage involved undertaking a rapid realist review of literature to synthesise existing literature on mechanisms and strategies to improve recovery from delirium.26
2. Qualitative study with stakeholders: In the next stage, in-depth realist interviews were undertaken with older people with delirium (N = 8), their carers (N = 14), health and social professionals (N = 24). Semistructured realist topic guides were developed based on the preliminary programme theory, which was developed into a deductive codebook consisting of context–mechanism–outcome configurations (CMOCs). The aim was to understand the perceived rehabilitation and support needs of older people after experiencing an episode of delirium to identify what rehabilitative strategies might improve recovery. Realist analysis was performed on the data, deductively testing the hypothesised CMOCs against evidence from the data while inductively identifying novel areas of interest.
3. Expert panel meeting: A panel of 11 independent experts and 19 research team members was engaged in a full-day workshop in January 2022 to discuss, critique and offer recommendations on hypothesised CMOCs, as well as helping synthesise the findings into a coherent underlying theory of delirium rehabilitation. The panel comprised professors and consultants of geriatric medicine, professors of clinical psychology, ageing, dementia and rehabilitation sciences, professors in health and social care, patient and public involvement representatives, applied research and implementation science experts, health economists, statisticians and qualitative researchers.
4. Programme theory development meetings: A series of 3-weekly meetings were held between March 2022 and June 2022, with participation from the core team to revise the programme theory and develop a logic model of what works to improve recovery from delirium, for whom and under what circumstances.
5. Theory-based evaluation: As part of the embedded process evaluation assessing implementation fidelity and acceptability of the intervention, an evaluation of the underlying programme theory was undertaken using realist approaches. In-depth interviews were undertaken with older people (five) who received the intervention, their carers (nine) and HCPs who planned and delivered the intervention (eight were interviewed, while seven participated in focus groups). Quantitative and qualitative data were also collected through case report forms (CRFs), training and supervision log, and audio recordings of sessions. Qualitative data were analysed with aim of refining the previously developed CMOCs about potential mechanisms of change and how these might vary by context of implementation. Data were also analysed using the modified Conceptual Model for Implementation Fidelity.27 Quantitative data were analysed using descriptive statistics, and mixed-methods analysis was conducted using a triangulation table.

Multidisciplinary approach

Informed by the International Classification of Functioning, Disability and Health, the intervention design adopted a holistic, multidisciplinary approach.28 A PT, OT and clinical psychologist drafted the intervention materials, integrating inputs from a senior geriatrician and other specialists. Each component was crafted to work synergistically, addressing the complex needs of patients with delirium.

Patient and public involvement and engagement

Active involvement of stakeholders ensured that the intervention remained practical and patient-centred. Collaborating with Innovations in Dementia, and meeting regularly, PPIE contributions were crucial in tailoring educational resources and developing culturally relevant materials. Stakeholder feedback was integral to ensuring the intervention’s applicability and effectiveness, particularly in the design of the physical recovery component.

Key findings

Some text in this section has been reproduced with permission from O’Rourke et al.26 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.

A rapid review and 2-stage realist synthesis of 52 studies contributed to the development of a preliminary programme theory, identifying 3 interdependent recovery domains and 4 key recovery facilitators. The recovery domains were: (1) support for physical recovery through structured exercise programmes; (2) support for cognitive recovery through reality orientation and cognitive stimulation and (3) support for emotional recovery through talking with skilled helpers. The recovery facilitators included: (1) involvement and support of carers; (2) tailoring intervention to individual needs, preferences and abilities; (3) interpersonal connectivity and continuity in relationships and (4) facilitating positive expressions of self.

Building on this theory, qualitative analysis of stakeholder interviews provided deeper insights into the programme and its implementation. It reinforced the need for physical and cognitive rehabilitation and highlighted the importance of social connections, emotional support, delirium education and signposting to resources for individuals and carers. Guidance was also emphasised on optimising nutrition, hydration, sleep–wake cycles and addressing underlying health conditions. A person-centred approach was important, with rehabilitation goals and delivery tailored to individual needs and preferences. The intervention was integrated into daily activities and aligned with usual care to support continuity and relevance.29,30

Expert panel discussion rooted in lived experience and professional expertise further refined the theory of community-based delirium rehabilitation. The panel helped consolidate the findings from the qualitative study with the preliminary programme theory, ensuring a robust theoretical foundation aligned with current research and practice.

Using realist principles, the core team synthesised findings into CMOCs, which informed a structured refinement of the programme theory and logic model. This process focused on identifying the mechanisms by which the intervention works and the contexts that enable its effectiveness. The resulting RecoverED intervention model hypothesises that effectiveness depends on delivering core components in a person-centred, tailored, manner-aligned with individuals’ needs, preferences and abilities. Additionally, the approach should actively engage carers, integrate intervention goals into daily activities and ensure continuity in the relationship with the professional carer. Figure 2 illustrates the intervention’s core components and delivery approach.

FIGURE 2 RecoverED intervention components and delivery approach. (Figure not reproduced here; see the original article.)

The revised logic model can be found in Appendix 1, Figure 4.

Limitations

The process of developing a theory of delirium rehabilitation engaged older people with delirium, their carers and HCPs from a single geographical area. As a result, the insights and perspectives may not fully capture or reflect the broader diversity of experiences, needs and contexts of stakeholders in other regions of the UK. A wide range of HCPs were involved in the development of the intervention, but this did not include RSWs who may have been able to comment on their educational needs for the intervention. In addition to geographical variation, other dimensions of diversity – including ethnicity, culture, language and gender identity – were not adequately represented. These factors, alongside local government policies, service availability, socioeconomic factors may affect the transferability of the recommendations, as well as the implementation and evaluation of the intervention and its outcomes.

Work package 2: feasibility study and intervention refinement

Aims

Work package 2 aimed to assess the feasibility and acceptability of a home-based rehabilitation intervention designed for older adults who have experienced delirium during acute hospitalisation. Specifically, it aimed to explore the feasibility of delivering the intervention across individuals with diverse characteristics through a process evaluation, test the feasibility of collecting data to inform the selection of primary and secondary outcome measures for the definitive RCT and within-trial economic evaluation, and iteratively refine the intervention before full-scale testing.

Methods

Work package 2 employed a multicentre, single-arm feasibility study design with an embedded process evaluation, following the protocol outlined by Allan et al.31 Participants were recruited from six NHS hospitals in locations across the UK (North Devon, London, Birmingham, Nottingham, Newcastle upon Tyne, Edinburgh), with eligibility criteria focused on individuals aged over 65 years, who were diagnosed with delirium and expected to return home. Initiated within 2 weeks’ post discharge, it included assessments by PTs and OTs, with a RSW delivering up to 10 sessions over 12 weeks. A RSW is an unregistered healthcare practitioner who works under the supervision of registered professionals (such as PTs and OTs) to provide rehabilitation to patients. Detailed descriptions of the intervention can be found in the protocol paper.31

Medical history and demographic information (e.g. age, sex and ethnicity) were collected. A detailed delirium assessment was completed using standardised Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition criteria,1 with additional assessments including the Observational Scale of Level of Arousal32 and Memorial Delirium Assessment Scale.33

Follow-up assessments were scheduled at 3 and 6 months post discharge, with a ± 2-week time frame to accommodate logistical considerations while ensuring data integrity. These included measures of cognitive function, ADL, quality of life and other relevant outcomes, administered through pseudonymised CRFs and questionnaires. Data collection was facilitated through the Research Electronic Data Capture platform, ensuring efficient and secure handling of study data throughout the assessment period.

Feasibility outcomes were assessed through recruitment and retention rates, with an embedded, mixed-methods process evaluation to assess implementation fidelity and intervention acceptability, as well as refining the underlying programme theory. Descriptive statistics were used to report feasibility parameters (see more information on statistical analysis in Report Supplementary Material 1).

The resources required to deliver the intervention were identified through discussion with the research team, intervention developers and providers, and these included staff time, travel, training, supervision and materials. These resources were measured within-trial using participant-level CRFs, with additional information provided from intervention developers and providers, and were valued using nationally recognised UK unit costs, including the National Schedule of NHS Costs,34 Unit Costs for Health and Social Care35 and the Adult Social Care Activity and Finance Report36 for 2022–3. Where published sources were not available, costs were identified in consultation with the intervention developers and providers.

To capture other relevant resources used during the study period, a bespoke Resource Use Questionnaire (RUQ) was developed by the research team and the Patient Advisory Group for the study, informed by recent guidance on items for inclusion in resource use measures.37 The RUQ was designed for proxy completion by carers and included NHS and Personal Social Services (PSS) resources, that is primary, secondary and social care resource use, as well as broader societal resources, including informal care/support, Voluntary, Community and Social Enterprise Sector services and support, and additional participant purchases and costs. The RUQ was administered at baseline and at 3- and 6-month follow-ups. NHS and PSS resource use was valued using nationally recognised UK unit costs for health and social care services.34–36 Informal care was costed using median hourly earnings sourced from the Office for National Statistics (Office for National Statistics, Newport, UK) Annual Survey of Hours and Earnings 2023,38 and costs of additional purchases were reported directly by carers completing the RUQ. Use of Voluntary, Community and Social Enterprise Sector services and support was rarely reported, and these were therefore not costed for the purposes of the feasibility study.

The intended primary economic outcome measure for a definitive trial was quality-adjusted life-years (QALYs), derived from health-state values obtained using the EuroQol-5 Dimensions, five-level version (EQ-5D-5L) health-related quality of life (HRQoL) instrument.39 Further information on the HRQoL of patient participants was obtained from a dementia-specific measure, the dementia quality of life (DEMQOL).40 Due to the potential for some patient participants to lack capacity to complete questionnaires at one or more time points, both self-report and proxy-report versions of the EQ-5D-5L (EQ-5D-5L Proxy Version 2) and DEMQOL (DEMQOL-Proxy) were administered to all patient participants at all time points, with the carer providing the proxy reports. In addition, information on patient and carer well-being was collected using the relevant version of the ICEpop CAPability (ICECAP) measures and was used to calculate well-being-adjusted life-years (WALYs). Data on HRQoL and well-being for patient and carer participants were collected within the trial at baseline, and at 3- and 6-month follow-ups, as set out in Table 1.

Health-state values were derived from responses to the EQ-5D-5L and EQ-5D-5L Proxy Version 2, by mapping between EQ-5D-5L responses and the published UK health-state value set for the EuroQol-5 Dimensions, three-level version using an approved algorithm,41 as recommended by the NICE.42 Health-state values were derived from responses to the DEMQOL and the DEMQOL-Proxy, using published tariffs.43 Well-being values were derived from responses to the ICEpop CAPability measure for older people (ICECAP-O) and ICEpop CAPability measure for adults (ICECAP-A), using the published sources of values for these instruments.44 QALYs and WALYs were estimated from the derived health-state values and well-being values, respectively, through application of standard area under the curve methods, using baseline, month 3 and month 6 assessments. More information on the health economic analysis can be found in Report Supplementary Material 2.

Progression criteria

Progression criteria for the study were agreed with the PSC and the funder (Table 2).

Key findings

Recruitment and retention

Out of 435 patients across 6 hospitals identified with delirium, only 36 (8.2%) met the eligibility criteria. Among 399 patients deemed ineligible, the most common reasons included: discharge not being to a private home, no carer, no capacity or consultee, carer not being a family member or friend, patients being at the end of life or communication challenges (e.g. patient or carer unable to communicate, or language barriers). Some patients were ineligible for more than one reason. Nineteen (N = 19; 53%) of those who were eligible consented to participate among patient and carer pairs, with 13 participants (68%) starting the intervention and 10 (53%) completing the final follow-up (Table 3 and Figure 3).

FIGURE 3 RecoverED flow diagram of participants. a, Some participants ineligible for more than one reason; b, approached are defined as patient/carer pairs who have been confirmed eligible and have been given a participant information sheet (see Report Supplementary Material 3–5) to consider taking part in the study; c, no decision given is defined as a patient/carer pair who were eligible and given a participant information sheet but did not specify either way their decision about taking part in the study; d, withdrawn represents only the participant-pairs for whom primary outcome data at 6 months will not be provided. (Figure not reproduced here; see the original article.)

All participants were White, the mean age was 83.8 years (range 73–97) and 12 were men and 7 were women. Mean clinical frailty score of the participants was 5.3 [standard deviation (SD) 1.8]; 5 participants (26%) had dementia as a comorbidity (Table 4). The carer participants were mostly women (15), and all carer participants were White. Carers had a mean age of 66.6 years (range 43–86), with 11 of the carers being a spouse or partner of the participant; 7 were children and 1 was a grandchild.

Recruitment was higher in men (60%) than women (44%), and 26% of participants had dementia as a comorbidity. Fifty-three per cent of participants attended six or more sessions. The disability assessment for dementia score was 43.7 (SD: 27.1) at baseline and 64.5 (SD: 27.9) at 6 months (%, higher score is better). At baseline, 79% showed delirium markers, but none at follow-ups.

Health economic evaluation

Data were successfully collected for most resources required for delivery of the intervention (direct staff contact time, travel time, mileage, training and materials). Data collection was unsuccessful for staff time spent on non-contact activities (session preparation, updating records and supervision); therefore estimates for this component were provided by the research team. The mean cost per participant of the intervention was £1249. The resources required for delivery of the intervention and their associated costs are detailed in Table 5.

While missing data rates for individual RUQ items were largely acceptable – ranging from 0 to 2, missing observations per item for the 10 carers at 6-month follow-up, and 3–4 for out-of-pocket expenses – the missing items varied among participants. Because only cases with complete data at both follow-up points can be included in the complete case economic analysis, total NHS/PSS costs were based on only three participants, dropping to two when societal costs were added. Responses to the RUQ at the 6-month follow-up and disaggregated costs of resource use across the combined follow-up period are reported in Tables 6 and 7.

Table 8 presents patient QALYs based on self-reported and proxy-reported EQ-5D-5L and DEMQOL, carer QALYs based on self-reported EQ-5D-5L, patient WALYs based on the ICECAP-O and carer WALYs based on the ICECAP-A. These figures indicate the number of participants providing complete data across all time points (baseline, 3- and 6-month follow-up). Results show QALY and WALY gains across the 6-month period for patients and carers, reflecting positive impacts on HRQoL and well-being in both groups. However, as this was not a RCT, these findings should be interpreted with caution. Greater improvements in QALYs were observed when the EQ-5D-5L was self-reported, rather than proxy-reported, and when the condition-specific DEMQOL measures were used compared to the generic EQ-5D-5L.

Process evaluation

The process evaluation was embedded within the feasibility study and used a mixed-methods design to examine fidelity, acceptability and contextual factors influencing delivery of the intervention. The modified conceptual model of implementation fidelity45,46 was used as a framework for data collection and analysis. Participant–carer pairs enrolled in the trial were simultaneously invited to take part in the process evaluation.

Of the 19 pairs recruited to the feasibility study, 8 patients and 11 carers consented, with 5 people with delirium and 9 carers completing interviews. Eleven HCPs involved in delivery and management of the intervention also participated, with 8 completing interviews and 7 taking part in online focus groups.

Semistructured interviews were carried out by trained qualitative researchers using topic guides (see Report Supplementary Material 6 and 7) that had been codeveloped with the PPIE group and informed by the programme theory. Interviews were conducted by telephone after participants had completed the intervention and explored experiences of delivery, engagement and perceived outcomes. All interviews were audio-recorded, anonymised and transcribed verbatim for analysis. The focus groups, held online, drew on a participatory action research approach to capture professional reflections on implementation, adaptation and collaborative delivery.

Quantitative data, including CRFs, training and supervision logs, and demographic information, were analysed descriptively. Qualitative data from interviews, focus groups and free-text responses were analysed deductively using the modified implementation fidelity framework. Findings from both strands were triangulated to provide a comprehensive account of implementation. These analyses directly informed refinement of the intervention by identifying which components were delivered as intended, where adaptations occurred and how participants perceived their relevance and value.

In total, 31 HCPs were trained across 6 sites, ensuring that, at each site, at least 1 OT, 1 PT and 1 RSW received training. The intervention was widely regarded as acceptable and valuable by older people with delirium, their carers and HCPs. The intervention was delivered with good fidelity to both content and approach, with components and activities flexibly tailored to meet participants’ needs and abilities. A person-centred approach was central, reflecting the high value that participants placed on setting their own goals, thereby fostering autonomy and agency. The same HCP team delivered the intervention throughout for each participant, ensuring excellent continuity of care. Psychosocial support was frequently provided, even when unplanned, highlighting its importance for both participants and carers. Mobility-related tasks were sometimes underdelivered due to external factors such as poor weather or participant ill health.

Adherence data were limited for those with complex needs, as health deterioration or care home placement often led to withdrawals. Importantly, withdrawals were due to health-related issues, not dissatisfaction with the intervention. HCPs required ad hoc support, particularly for delivering psychosocial elements. While they found the training and manual thorough, HCPs expressed a preference for more interactive, practical training rather than purely manual-based learning. Strategies for improving delivery included forming dedicated, colocated RecoverED teams, assigning participants closer to HCPs’ workplaces and establishing structured supervision to enhance efficiency and collaboration.

While the feasibility study demonstrated the potential to recruit individuals with delirium, significant challenges remained, particularly in achieving sufficient participation rates. Factors contributing to low screening numbers included the limited capacity of research nurses and the predominance of clinical teams diagnosing delirium rather than research staff. This situation revealed weaknesses in clinical screening processes in the UK, which have also been identified in relevant international studies.

The NICE guidelines recommend the routine use of the 4 As test (4AT) for delirium screening upon admission, yet this was not consistently applied at several sites, leading to fewer screens than the expected 1200 patients over 6 months; ultimately, only 435 were screened. The reasons for non-eligibility included transitioning to care homes and the absence of a carer to participate alongside the patient. Considering including individuals transitioning to care homes in future trials could be beneficial, though it may necessitate a redesign of the intervention, which is currently focused on activities in participants’ homes.

Additionally, it was noted that some patients were not approached for participation, potentially due to staff workload, discharge timing or concerns regarding the patients’ health status. Despite these challenges, the promising consent rates and reasonable level of data collection from participants who remained in the study suggest that a future trial could be feasible with improved recruitment strategies.

While retention was lower than anticipated, with withdrawals due to death and health deterioration being commonplace in this frail population, these findings underscored the implications for determining the sample size and retention strategies in future trials. Recruitment challenges and participant withdrawals highlighted the need to better define the target population for whom the intervention would be most suitable.

Further detailed findings from the process evaluation and feasibility outcomes are available in manuscripts in press (see titles listed in Publications).

Intervention refinement

In response to the process evaluation findings, several refinements to the intervention have been identified. These include updates to the training content, such as the incorporation of case studies that illustrate the flexibility of the intervention in practice. Additionally, consideration is being given to transitioning to a more interactive training platform to enhance engagement and learning outcomes.

Revisions to the intervention manual are also required. These will focus on clarifying guidance around goal setting and health and well-being checks. Furthermore, the manual will be updated to explicitly require the RSW to attend the initial assessment. Corresponding changes to the initial assessment paperwork will be made to further clarify goal-setting expectations.

Finally, a review of the supervision guidance for RSWs delivering the intervention is recommended. This will include the development of supervision templates and an assessment of the resource implications, including potential costs associated with increased levels of supervision.

Ethics

Ethical approval was obtained for the feasibility study from the London-South East REC (IRAS ID: 302675, 24 October 2022) and Scotland A REC (IRAS ID 319555, 28 October 2022).

Limitations

The limitations of the feasibility study include the lack of diversity in participant demographics, as all individuals were White, restricting the findings’ applicability to a broader population of older adults with delirium. The reliance on clinical teams for diagnosis rather than standardised research-based screening may have led to missed recruitment opportunities, and potential selection bias exists due to not approaching all eligible patients. External, non-participant assessments of fidelity through session audio recordings could not be conducted due to practical and consent-related challenges.

The process evaluation did not include long-term follow-up, although fidelity and acceptability may change over time. Implementation support was provided through ad hoc and peer sessions, which may have improved fidelity but limits generalisability to routine practice. The evaluation was also limited by the small number of participants and the withdrawal of more impaired individuals, raising the possibility that the intervention was better suited to less impaired participants and may require further refinement for those with greater impairment. All participants were White British, so the study could not assess acceptability in minority ethnic or linguistic groups, further limiting generalisability.

The insights gained highlighted the importance of developing a deeper understanding of the diverse needs, challenges and circumstances faced by delirium patients and their carers. Future studies should consider the complexities of this population, including the possibility that only a small subgroup of patients may benefit from active intervention, while others may require a more supportive or palliative approach. To improve recruitment and intervention strategies, future research should focus on tailoring approaches to the specific needs of different patient subgroups, ensuring that these vulnerable populations are better represented and supported in delirium recovery research.

Patient and public involvement and engagement

Aim

The aim of PPIE in the RecoverED study was to actively engage individuals with lived experience of delirium, either personally or as carers, to ensure that the study design, implementation and outcomes were aligned with the needs and preferences of this population. The PPIE group provided input throughout the study, helping to shape materials, process evaluation methods and intervention approach to improve the relevance and effectiveness of the project. They also contributed to discussions on the project findings.

Approach and methods

Patient and public involvement and engagement was integrated into the study from the outset, facilitated by Rachael Litherland from Innovations in Dementia and with support from the trial manager. The PPIE group initially consisted of five members with lived experience of delirium and was later expanded to include two individuals from diverse backgrounds. PPIE members participated in regular meetings, mainly virtually, but also in-person in some cases, to offer feedback on study materials, recruitment strategies and intervention plans. Key activities included reviewing participant information sheets, selecting a programme logo, providing input on questionnaires, participating in video recordings, contributing to the design and execution of the process evaluation, contributing to inclusion initiatives and advising on recruitment challenges. Feedback was solicited through both meetings and e-mail correspondence to gather broader insights across multiple study aspects.

Key outcomes

The outcomes of PPIE involvement in the RecoverED study have been both valuable and diverse, significantly contributing to various aspects of the research. A PPIE representative took part in the expert panel to design the intervention and provided valuable input to the design. One of the most substantial impacts of the PPIE group was in shaping the materials designed for participants. For instance, during the initial stages of the study, the PPIE members provided extensive feedback on the participant information sheets. These documents were crucial in ensuring that potential participants fully understood the study’s aims, procedures and expectations. PPIE members pointed out areas where the language was either too technical or unclear and helped reword key sections to make the information more accessible to individuals with varying levels of health literacy. This early feedback was vital, as it directly influenced how participants would perceive and understand their involvement.

Another contribution was during the development of the intervention materials. When the physical recovery interventions were being designed, PPIE members played an essential role in reviewing the content to ensure that it was user-friendly. One PPIE member even volunteered to be photographed for the activity instruction sheets, which demonstrated their willingness to be actively involved. This contribution was particularly beneficial as it added a personal touch to the materials, making them more relatable and practical for participants. The feedback helped to ensure that the instructions were not only clear but also representative of real-life home environments where participants would be carrying out the exercises. This level of detail and personalisation might not have been possible without the involvement of individuals who understood the lived experience of recovering from delirium.

The PPIE group was also involved in discussions about the study’s branding and communication strategies. For example, they were asked to choose a programme logo from several design options, and their input led to the selection of a logo that resonated well with both participants and the research team. This seemingly small decision had broader implications, as it contributed to a sense of ownership and engagement among the PPIE members, reinforcing their integral role in the study.

As the study progressed, PPIE members continued to provide input on more complex aspects of the research, such as reviewing questionnaires that would be used to collect data from participants. One of the key questionnaires discussed was the (proxy) RUQ. PPIE members provided detailed feedback on how this could be adjusted to better capture the realities of caregiving and post delirium recovery. Their feedback was essential in refining the questions to ensure that they were sensitive to the emotional and logistical challenges faced by both patients and carers.

The PPIE group was particularly instrumental in shaping the content and form of the Health and Care RUQ. The study Health Economist met online with the group to find out what health, social care and broader care services people with delirium tended to use and not use and to explore with them how service use might be affected by receiving the RecoverED intervention. The RUQ was then designed to include service use particularly relevant to this population. Members of the group then provided input on the draft RUQ.

As an additional piece of work, we took the opportunity to work with the PPIE group on exploring what would be considered as a ‘minimal important difference’ (MID) on the ICECAP-O. The purpose here was to use this information to inform future trials wishing to use the ICECAP-O and particularly for sample size calculations. The Health Economists and PPIE lead undertook a task with the PPIE to identify this MID. As outside the main scope of this grant, the findings will be reported elsewhere.

One particularly innovative aspect of PPIE involvement was their participation in the creation of educational videos for the study. In these videos, PPIE members shared their personal experiences with delirium, both as patients and carers, offering authentic perspectives on the condition. The videos added a critical dimension to the intervention, allowing participants to see and hear from people who had been through similar experiences. This not only enhanced the credibility of the study but also provided a level of empathy and understanding that would be difficult to convey through written materials alone. The PPIE members’ willingness to be filmed and share their stories demonstrated their deep commitment to the study and the value they placed on helping others navigate the complexities of delirium recovery.

The PPIE group worked closely with the study teams to shape the process evaluation. They advised on the development of topic guides, ensuring that questions were clear and relevant, and guided us on how to approach interviews in a sensitive and respectful way. They also contributed to discussions on data interpretation and presentation, helping us to centre the voice of the person with delirium and highlight issues of most importance to service users. Another key area of involvement was during the grant application process for the Programme Development Grant (PDG), which provided an opportunity to seek additional funding alongside the main Programme Grant for Applied Research that already supported our programme. The aim of the PDG application was to explore strategies to improve inclusivity, particularly by addressing language barriers through the translation of participant materials and the provision of interpretation during intervention delivery and outcome assessments. As part of this effort, we held a PPIE meeting with individuals from diverse backgrounds to ensure that the project was inclusive and addressed the needs of underserved populations. During this meeting, the PPIE members offered critical insights into issues related to language barriers, the challenges of communicating with HCPs through interpreters and their experiences with translated materials. This feedback was essential for shaping the grant application and ensuring that the proposed intervention would be accessible to all participants regardless of their linguistic or cultural background. The PPIE group highlighted the importance of face-to-face communication, cultural sensitivity and the use of interpreters who are not only linguistically compatible but also culturally aligned with the participants. This input played a significant role in influencing the design of the development work, as it ensured that the study would address these barriers and create a more inclusive environment for participants from minority ethnic backgrounds.

Despite strong and ongoing engagement from the PPIE group, reaching consensus on the most appropriate outcome measures in early 2024 proved challenging – highlighting the inherent complexity of these decisions. While initial attempts to gather views via e-mail provided a useful starting point, the diversity of perspectives reflected the nuanced nature of selecting a primary outcome for a complex intervention like delirium rehabilitation. In response, the research team convened a dedicated meeting to facilitate more detailed discussion. During this session, each outcome measure was introduced and discussed in relation to the intervention’s aims, allowing PPIE members to ask questions and share their insights. Although the meeting enhanced shared understanding, it also reinforced that outcome selection is a multifaceted issue requiring ongoing dialogue rather than a single definitive decision point. These discussions underscored the importance of continued coproduction in shaping future trials, particularly when it comes to balancing clinical relevance, feasibility and lived experience.

Discussion and conclusions

Overall, the PPIE input had a substantial influence on the RecoverED study. Their involvement ensured that study materials and interventions were more accessible, inclusive and patient-centred. Key study elements, such as the participant information sheets, programme logo and activity instruction sheets, benefited directly from PPIE feedback. Furthermore, the PPIE group contributed to refining recruitment strategies and provided ongoing reflections on the intervention’s delivery and evaluation processes. However, there were also notable challenges in integrating PPIE effectively throughout the study. One of the main challenges was the difficulty in gathering conclusive feedback on more complex or technical issues, such as the selection of outcome measures. The reliance on e-mail correspondence to collect feedback proved to be insufficient, as the responses were mixed and did not lead to clear decisions. Another challenge was that one of the PPIE representatives was not able to attend the expert panel meeting for personal reasons, meaning that contribution to the initial design of the intervention was limited to one person. However, the PPIE group did contribute the design of the intervention during the development following the expert panel meeting.

Reflection on what was and was not successful in the programme

The RecoverED programme developed a structured, theory-based intervention aimed at supporting recovery following delirium. Using an iterative co-design process, the team integrated expertise from geriatricians, PTs, OTs and a clinical psychologist. This ensured that the intervention was theory-based and clinically relevant.

Another achievement of this work was the development of a manual and structured training programme for RSWs. The training was designed to improve knowledge and confidence in delivering rehabilitation support, and post-training evaluations demonstrated improvements in the self-reported preparedness and ability of healthcare staff to implement the intervention. Feedback from trained staff highlighted that the framework was practical and well-structured.

Our programme also prioritised PPIE to ensure that the intervention was practical, patient-centred and culturally relevant. At the design stage, we engaged patients and their carers to shape key aspects of the intervention, including the education resource and delirium information sheet. PPIE members actively contributed to the development of materials for the physical recovery component and providing valuable insights to refine the design. This collaboration led to important refinements, particularly in supporting cognitive and physical recovery, enhancing the intervention’s relevance.

In WP2, we assessed the feasibility of delivering the RecoverED intervention, but recruitment challenges significantly affected the study. Only 9% of screened patients met the inclusion criteria. To improve recruitment, we broadened eligibility criteria to include all hospital-admitted delirium patients, introduced geographical restrictions to ensure feasible intervention delivery and streamlined consent procedures for recently discharged patients. Despite these efforts, delays in opening study sites, staff shortages and poor delirium screening practices kept recruitment below target. The recruitment period was extended twice (from 31 October 2023 to 31 December 2023, and then to 31 March 2024), but two sites had to cease participation by January 2024 due to staff capacity issues. In hindsight, the progression criteria should have included the total number of people recruited, but this was identified and informed the decision not to proceed with WPs 3 and 4.

Retention was also a challenge, as some participants withdrew due to health deterioration, hospital readmission or relocation. Despite these challenges, we were able to collect outcome data at baseline and follow-ups. While consent rates of 53% among eligible participants was moderate, strong engagement from those who remained in the study suggest the intervention’s potential benefits. The problems with recruitment and retention had a knock-on effect on the process evaluation, meaning that insights were limited to a smaller sample than planned.

Equality, diversity and inclusion

Language and terminology

In our study, fostering an inclusive and respectful environment was central to our approach, beginning with the language and terminology used throughout the project. Recognising that clear communication is important for participant understanding and engagement, we prioritised the use of plain, jargon-free language in all interactions and materials. This approach aimed to make the study accessible and comprehensible to participants with varying levels of education and familiarity with clinical research processes.

To achieve this, we engaged with PPIE members from diverse backgrounds to review all participant-facing materials, including invitation letters and consent forms. The PPIE contributors provided insights into potential areas of confusion or ambiguity, enabling us to refine language choice and presentation to better meet the needs of our audience.

Furthermore, we adopted terminology that was inclusive and respectful, avoiding medical jargon that could serve as a barrier to understanding. For example, instead of terms like ‘cognitive impairment’ or ‘functional decline’, we used more accessible phrases such as ‘changes in memory or thinking’ and ‘difficulties with everyday activities’. By involving PPIE members, who are representative of the communities we sought to engage, we ensured that our communication strategies resonated well with participants and addressed their concerns effectively. This participatory approach underscored our commitment to transparency and inclusivity, fostering trust and encouraging informed participation.

This strategy not only facilitated a better understanding of the study’s objectives and processes but also empowered participants by making them feel valued and respected. For example, in qualitative interviews, several participants expressed appreciation for the clear, respectful communication and noted that this helped them feel more involved in their care. These reflections – reported in the process evaluation paper – illustrate how inclusive language and collaborative engagement contributed to participants’ sense of trust and involvement. Such efforts are important in research involving vulnerable populations, as they help to assure participants that their perspectives and experiences are integral to the research process.

Methods

The study targeted older adults with delirium and their carers, making a concerted effort to include underserved populations, particularly individuals who lack capacity. By incorporating these participants through either direct consent or proxy consent, we achieved a representation of those affected by delirium. However, it is important to note that participation required having a carer willing to take part. This criterion excluded individuals who live alone, a group that is increasingly prevalent, especially among older adults. For instance, approximately 40% of people with dementia living in the community in England live alone, which highlights a significant barrier to inclusion. We acknowledge that people living alone may respond differently to interventions, and their inclusion could provide important insights into the effectiveness and cost-effectiveness of such interventions. While this was necessary for ensuring robust data collection, it also represents a neglected dimension of inequality that requires further attention in future research. Future studies should explore how to include these individuals while maintaining research rigor and inclusive practices

Our recruited population completely consisted of White participants. Despite efforts to recruit a more diverse sample, we were unable to include non-White participants in the study. This highlights a significant limitation in ethnic diversity and underscores the need for more inclusive recruitment strategies in future studies. While achieving ethnic diversity remains a challenge, we successfully included individuals across a diverse cognitive spectrum, including those without decision-making capacity, ensuring representation of groups typically under-represented in clinical research. Additionally, participants came from varied educational backgrounds, which contributed to an understanding of the intervention’s applicability.

Generalisability and transferability of evidence

The study’s geographical and epidemiological scope provides valuable insights that may be relevant to regions with similar healthcare systems and demographic compositions, although the limitations of the evaluation suggest that broader generalisations should be made with caution. The intervention, which included comprehensive cognitive, physical, physiological and psychosocial components, demonstrated feasibility and acceptability in an older cohort, reflecting its potential effectiveness. The inclusion of various settings – homes or NHS sites for follow-up – enhanced the study’s applicability by accommodating participants’ preferences and abilities, thereby increasing its external validity.

Efforts were made to ensure that the intervention materials were accessible, considering literacy levels and language preferences. However, the study only included White participants, highlighting a need for more extensive recruitment from diverse ethnic backgrounds to improve generalisability. The intervention was particularly accessible to those with a support network in place, such as family carers, indicating the necessity for strategies to enhance inclusivity for those who might be socially isolated or from minority ethnic groups.

Participant representation

The study prioritised recruiting older adults with delirium. Recognising the essential characteristics of our target population, such as age, available caregiving support and geographic location, we designed recruitment and retention strategies to optimise participant engagement and minimise attrition. Key to our approach was the establishment of partnerships with local NHS site and intervention teams.

Our inclusion and exclusion criteria were carefully considered to mirror the demographic and care dynamics of older adults with delirium. Enrolment was limited to individuals over 65 years old, diagnosed with delirium during hospitalisation, and those expected to return to a private dwelling post-discharge. Crucially, eligible participants were required to have a carer who could participate in the study. While this requirement was designed to ensure comprehensive data collection on participants’ experiences, it also introduced barriers, excluding potential subjects without consistent caregiving support. Furthermore, geographic constraints and language requirement further limited eligibility, focusing on individuals within the operational reach of our intervention team who could communicate effectively for the study.

Throughout the study, gaps in representation, particularly regarding ethnic diversity and participants without designated carers, were identified. To address these issues in future research will require more than simply broadening geographic reach – it will demand dedicated strategies, resources and partnerships tailored to engage under-represented and underserved groups. This includes developing culturally appropriate materials, building trust within diverse communities and exploring alternative models of support for participants without traditional carer arrangements.

Enrol and retain diverse participants

In our study, concerted efforts were made to minimise participant burden and enhance accessibility. One of the key strategies involved offering home assessments to accommodate participants’ preferences and logistical needs, thereby reducing the stress and inconvenience typically associated with traveling to clinical sites. Follow-up assessments were conducted flexibly, allowing participants the choice of attending in person or completing them via telephone. This approach aimed to cater to varying levels of mobility and personal circumstances, making participation feasible for a broader range of individuals.

Despite these efforts, the study encountered challenges in achieving demographic diversity, particularly concerning ethnic representation. Recognising this shortfall, we sought additional funding specifically to expand ethnic minority participation in our trial. This funding would have enabled us to provide interpreters and produce translated study materials, addressing linguistic barriers that might deter non-English-speaking participants.

For future studies, it is important to address these challenges head on by securing funding and resources that cater to diverse linguistic and cultural needs. Implementing enhanced communication strategies and adaptive study designs are vital steps towards inclusivity. Tailoring materials to meet the linguistic and cultural preferences of potential participants will not only improve their understanding and engagement but also foster greater trust and willingness to participate. Additionally, outreach programmes specifically targeting under-represented ethnic groups will be essential in ensuring more balanced and diverse participant enrolment.

Participant data

In our study, we prioritised the disaggregation of participant data by age, gender and ethnicity. Comprehensive data were collected from participants, including anonymised screening information from those over 65 years old, who screened positive for delirium during hospital admissions. For the consented participants, detailed demographic and intervention-related data were recorded through medical records, paper CRFs and participant questionnaires.

Research team

Our research team comprised individuals with a breadth of expertise across various fields relevant to the intervention, including members from diverse ethnic backgrounds. To foster growth and inclusivity, junior team members were actively encouraged to engage in professional development opportunities aimed at broadening their skills and experiences. This included participation in various training programmes, presenting at conferences and local events, and even leading a grant application, all of which contribute to their professional growth and the team’s overall capacity.

Limitations of the programme

The RecoverED programme faced several limitations that affected the robustness of its findings. A key limitation was the context in which the intervention was delivered. As a feasibility study, it relied on ad hoc and peer support from clinical teams, which could have influenced implementation fidelity. This reliance on the research teams’ support may limit the generalisability of the findings, particularly regarding the wider roll-out of the intervention. Furthermore, the feasibility study did not include participants from minority ethnic, racial or linguistic backgrounds.

Additionally, the study’s recruitment process had notable limitations. Although NICE guidelines recommend that all patients aged ≥ 65 years admitted to hospital should be screened for delirium, this standard was not consistently met across all study sites. As a result, it was often unclear whether eligible patients had experienced delirium. To compensate, research staff had to visit wards and rely on clinical teams to identify potential participants, a process that proved to be both time-consuming and inefficient – particularly given the availability of electronic health records. This reliance on clinical diagnosis, rather than a consistent, standardised screening protocol, limited timely identification and contributed to missed recruitment opportunities. The recruitment challenges led to a small sample for the process evaluation. This may have led to an overoptimistic evaluation of the suitability of the intervention for all as more impaired participants often withdrew and we could not assess the suitability of the intervention for this group.

Furthermore, selection bias may have affected the sample, as not all eligible patients were approached, often due to staff capacity constraints or concerns about participants’ frailty. These factors highlight the need for future studies to adopt more inclusive recruitment strategies, standardised screening protocols and ensure greater representation of vulnerable populations to enhance the external validity and inclusivity of delirium recovery research.

Finally, it was very difficult to identify therapy teams to deliver the intervention. It is possible that even if the interventions were shown to be effective, implementation in the NHS would be very difficult with current staffing levels. Alternative approaches to finding staff might be needed, such as using privately funded teams or training social care staff.

Conclusions for whole programme

The RecoverED programme has made significant strides in addressing a crucial gap in improving recovery after delirium. This complex rehabilitation intervention was designed to support recovery following an episode of delirium, an area with limited research and understanding. A single-arm feasibility study with an embedded qualitative process evaluation has provided some insights into the viability of a definitive trial to evaluate the intervention’s effectiveness. Although the feasibility study demonstrated some challenges, it also highlighted the positive reception of the intervention by participants we were able to recruit, who reported perceived value and benefits from their participation. A strength of the intervention was the inclusion of both physical and cognitive rehabilitation together with psychosocial support. Current rehabilitation approaches do not include such a comprehensive approach. We have designed a new intervention which shows some promise, but unfortunately, the stop-go criteria were not met in our feasibility trial and significant changes in recruitment will need to be achieved for a successful trial. The process evaluation revealed important considerations for future implementation, particularly in terms of staffing and facilitation strategies to address the complexities of delivering such a multifaceted intervention. It raised pertinent questions about the target population, specifically regarding individuals with deteriorating health or those requiring care home placements, suggesting that future interventions should focus on adapting to these needs. It is possible that more impaired participants and those whose delirium had not resolved would need a more refined intervention.

Despite challenges in recruitment and retention, the study achieved a 53% consent rate among eligible participants and observed positive engagement from those who remained in the study. These findings indicate the potential benefits of the intervention, particularly when considering the vulnerable population of older adults recovering from delirium. However, only a small number of potential participants were eligible. Other approaches may be needed for more complex patients and those without a carer or living in care homes.

While the feasibility study highlighted significant challenges in terms of recruitment and retention, it has provided a solid foundation for refining the intervention and improving future recruitment strategies. The intervention training for RSWs was successfully developed and implemented, empowering staff to deliver sessions in alignment with the intervention manual. Data on primary and secondary outcomes were successfully collected, though issues with resource use data completeness were noted. Moving forward, addressing these challenges will be critical to improving recruitment effectiveness, retention and the overall implementation of the intervention in future studies. Despite this progress, however, the challenges of identifying people with delirium have not been resolved. It is possible that the NHS will not be ready for new interventions in delirium care until the problem of identifying delirium can be solved.

Recommendations for future research

We are actively pursuing further funding opportunities to address the key feasibility barriers identified in this study – such as recruitment, screening and inclusivity – with the aim of refining the intervention and laying the groundwork for a larger-scale evaluation in a future trial. Our planned participation in upcoming conferences – including the Alzheimer’s Association International Conference, the European Geriatric Medicine Society Congress, European Delirium Association Conference, World Geriatrics and Aging Congress and the British Geriatrics Society Conference – will allow us to share findings with national and international stakeholders and explore potential collaborations.

To strengthen the evidence base for the intervention and improve delirium assessment, future research should consider the following areas:

• Employing dedicated staff for delirium assessment: Future studies should explore the feasibility and impact of employing dedicated staff, such as specialist nurses or research nurses, to undertake direct assessments of delirium using validated tools. This would enhance diagnostic accuracy, facilitate early identification and ensure consistency in data collection across study sites. However, one problem of this approach is that research nurses would not be available when the intervention is implemented into the NHS, meaning that without improvements in screening for delirium by clinical teams, the intervention would not be used. Alternatively, if an effective intervention was known to be available for delirium, clinical teams might be incentivised to improve screening for delirium.
• Recruiting participants in nursing homes: Given the high prevalence of delirium among nursing home residents and the unique challenges of post hospital rehabilitation in this setting, future research should extend recruitment to nursing homes. However, this would require careful consideration of appropriate outcome measures, as traditional ADL-based outcomes may have limited sensitivity in this population. Instead, outcomes focused on engagement in meaningful activity may offer a more suitable and person-centred way to evaluate intervention impact, particularly in terms of quality of life and psychosocial well-being.
• Inclusion of individuals without a participating carer: Current models often depend on the involvement of a family carer, which may unintentionally exclude a significant portion of the population affected by delirium. Future research must explore alternative approaches to include those without an available or willing carer – such as through the use of community support workers, professional advocates or tailored intervention delivery models. Addressing this gap is essential for making interventions more inclusive and for informing funding bodies like National Institute for Health and Care Research (NIHR) about the structural changes needed to support broader participation.

Implications for practice and lessons learnt

The findings from RecoverED have the potential to inform future research. While the study highlighted important recruitment challenges – particularly around recruitment and screening practices – it also helped to identify key areas requiring further exploration. These learnings provide a foundation for refining study design, enhancing participant identification strategies and ensuring future interventions are better aligned with the diverse and complex needs of people recovering from delirium.

Key lessons from this study include the importance of:

• Improving delirium screening and documentation in clinical settings to facilitate participant identification.
• Enhancing recruitment strategies through closer collaboration between research staff and clinical teams.
• Addressing workforce challenges, including research nurse capacity and site-specific limitations, to ensure successful trial delivery.
• Providing flexible approaches to participant engagement to accommodate the health vulnerabilities of this patient group.
• Gaining a deeper understanding of the diverse problems, trajectories and needs of people who experience delirium – potentially through a multisite cohort study or case series – to better inform future intervention development.
• Recognising that relying solely on NHS community teams to deliver research interventions may not be feasible; employing privately contracted therapists could offer a more sustainable and reliable approach to intervention delivery.