AQUEDUCT Main Trial
Research type
Research Study
Full title
Achieving Quality and Effectiveness in Dementia Using Crisis Teams (AQUEDUCT): A Randomised Controlled Trial of a Resource Kit for Teams Managing Crisis in Dementia
IRAS ID
289982
Contact name
Martin Orrell
Contact email
Sponsor organisation
Nottinghamshire Healthcare NHS Foundation Trust
ISRCTN Number
ISRCTN42855694
Duration of Study in the UK
1 years, 11 months, 30 days
Research summary
Research Summary
Currently there is evidence that crisis teams working with older people with dementia are varied in their names, models and set-up. A previous work package of the AQUEDUCT research programme developed a Best Practice Model, a measure of best practice, and a Resource Kit for 'Teams Managing Crisis in Dementia' (TMCDs) designed to help teams improve their practice. A second work package tested the feasibility of evaluating the Resource Kit in a randomised controlled trial in Work Package 3.
During this randomised controlled trial, the AQUEDUCT research team will measure the extent to which the TMCDs use 'best practice' before and after a 6 month period, using the measure of best practice developed in Work Package 1. Information on quality of life, and NHS and social care service use will be collected for people with dementia and carers who are referred into the TMCD during the study period. Information about TMCDs will also be collected, such as number of hospital admissions of people with dementia relevant for the team's catchment area, and staff absenteeism.
At the end of the study, TMCD staff, people with dementia and carers will fill in a questionnaire or take part in an interview about their experience of the Resource Kit.
Lay Summary of Results:
Reflections on what was and what was not successful in the programme From the start of the programme, NHS Staff, stakeholders, Service Users and carers were very enthusiastic about the relevance of the programme and the research team met with a great deal of openness and interest when engaging with participants. Recruitment for the three Work Packages (WP) was excellent and both WP1 and WP2 recruited to target. Involved Teams managing Crisis in Dementia (TMCDs) and other clinical services were keen for the programme to be a success in further developing crisis working and to learn from innovative practices in other services. Indeed, on occasions when the research team brought people from different services together this resulted in sharing of knowledge and practice approaches among attendees. From the beginning, the good response to WP1 produced a rich source of data which fed into the development of the trial intervention, ensuring this was evidence based and grounded in practice.
Patient and public involvement (PPI) in AQUEDUCT was successful with PPI members working alongside the researchers at various stages of the research and supporting the team in various activities throughout. These have been discussed more fully earlier but included co-facilitating consensus discussion groups and being members of reviewing teams. One challenge was ensuring the correct governance was in place for PPI members, particularly when acting as co-researchers. Many PPI members were learning to use technology such as iPads and computers and using different software to complete governance checks was at times difficult for them. The research team worked closely with the Volunteering Manager at the Sponsor Trust and with the Trust Research and Innovation Department to develop processes that were acceptable to different Trusts in the study. The process was lengthy and at times delayed their involvement. Despite these challenges, PPI members made a significant contribution to the research bringing new insights into discussions and creating a deeper level of empathy with carers and SUs than otherwise might have been the case.
The relocation of the Chief Investigator to the University of Nottingham, early in the programme, led to a change of Clinical Trials Unit from Bangor University to the Clinical Trials Unit at the University of Nottingham (NCTU). This presented a challenge, and the new trials unit’s subsequent involvement was not successful, reflecting differing perspectives on the use of routine data as primary outcome measures. Consequently, NCTU ceased involvement prior to the commencement of the trial. Work originally planned as the responsibility of the NCTU was successfully managed by the AQUEDUCT research team which made additional demands upon the research team with the Programme Manager responsible for drafting the RCT protocol and work in recruiting alternative statistician support.
The arrival of the COVID-19 pandemic in early 2020 towards the end of the feasibility study (WP2) provided a major challenge to the research team, TMCDs, PPI members, People with Dementia (PwD), their carers, and other stakeholders. The AQUEDUCT research method had been planned to place considerable reliance on the input of NHS staff; the programme involved contacts with groups at high risk from COVID; and the trial outcome measures were admissions to hospital which was disrupted during the initial phase of the pandemic. The work was thus seriously impacted by COVID-19 and there was as a result a real risk of non-completion of the RCT. It was a success of the AQUEDUCT programme, that despite these challenges, and with continued support from NIHR, the team was able to adapt the trial design in a manageable and safe way that allowed the work to be completed. This success provided helpful lessons as to how research design can be constructively modified in the context of major external changes.
Equality, diversity and inclusion
The issues of equality, diversity and inclusion were addressed at different levels and stages throughout the study. These are in PPI and the research process, access to study information and geographical and population representativeness of the study. In undertaking research with PwD, the study was focused upon a group of people often under-represented in research.In terms of population representativeness, 30 NHS Trusts from across England engaged with WP1. These Trusts reflected a wide geographical spread, from County Durham in the north to Devon in the south, from Shropshire in the west to Suffolk in the east, as well as demographic variability encompassing both urban and rural locations, across more and less affluent areas. Similarly, the trial in WP3 engaged with 24 Trusts across different geographical and sociodemographic areas.
In the research process, the trial was designed in conjunction with both PPI and clinical staff reference group members to minimise the demands on participants, reflecting the twin unique circumstances of research undertaken at a time of personal crisis and overshadowed by the impact of COVID-19 on people and the NHS. Furthermore, the study PPI members contributed to ensure their voice was heeded, working in a range of roles such as co-researchers, advisers and appointment of staff.
Regarding access to information, as well as information made available to research, policy and practitioner groups through media, such as journal articles and local, national and international conferences, the study team produced newsletters and worked with local PPI groups and wider focused not for profit organisations to publicise the work of the study for interested members of the public who would otherwise lack access to information.
Despite these efforts, we recognise that the design of the research may not have permitted the inclusion of salient groups of people. Members of the team had undertaken research with ethnic minorities and on the issues surrounding uptake of services for PwD by members of the South Asian community and are aware of their lower uptake of services and use of services at a later stage in their illness. 70–72 Since the sample of PwD was from referrals to the TMCDs this may have excluded groups such as these who may have been unable to participate because of more advanced illness and or who sought help elsewhere. This is something that we were unable to address in this programme but should be a focus of future research, such as whether there are differential approaches to a crisis by PwD and their carers from ethnic minorities and if care pathways to TMCDs differ by ethnic minority groups. Again, it has been observed that dementia research has tended to focus on people with carers or close support and neglected those who are more isolated. Here too, recruiting people from service engaged groups, in this case crisis services, may have excluded more isolated individuals undergoing crisis.
Limitations
The programme examined the work of TMCDs based in England and while it may also be representative of the other three devolved nations in the UK, it has not been tested there and may also not be applicable internationally. It is also possible the programme excluded some groups in the analysis who were less likely to take up community mental health services such as ethnic minorities and people living alone, and that these groups may well have required a different approach to managing crises to engage with services. This was not explored in the study.Some limitations have been discussed previously for WP1. There may have been some methodological limitations in the scoping survey (WP1.1) associated with lack of access to appropriate managers, variation in their direct knowledge of services, and some respondents excluding themselves from the survey due to design issues. In WP1.3, there was a lack of full psychometric testing and statistical analysis of the Best Practice Fidelity Tool primarily reflecting to the small numbers in the field testing and its application in the intervention.
There are likely to have been limitations upon the RCT which were imposed by the context in which it was undertaken, the immediate aftermath of COVID-19. In this context, a feasible primary outcome had to be chosen, psychiatric hospital admission rates, which would not have been the chosen outcome in the absence of the pandemic. However, first, it may be the case that patterns of admissions both prior to the intervention and subsequently were changed in ways that differed from usual care due to the pandemic and influenced the results. Second, it is possible that there were changes in the operation of complementary and substitute services in study localities in response to COVID-19 such that probabilities of admissions to hospital of PwD were reduced. Finally, it is possible that a more complex and sophisticated method of standardising admission rates may be required, taking account not just of population but bed numbers and substitute services, if admissions are to be employed as an outcome measure in studies.
Recommendations for future research
The programme has highlighted the limited knowledge base regarding the most appropriate support for people addressing crises in dementia. Arising from WP1, there are questions regarding the operation of TMCDs. These include: whether crisis responses are better managed as a role within community mental health team for older people (CMHTOPs) or as specialist services; the most effective linkages between primary, secondary care and social care in addressing crises; and staff mix within crisis teams. Arising from the best practice tool (BPT) is work to further examine its structure and psychometric properties. From the trial there are questions regarding the issue of continuity of care and the effective delivery of supportive engagement with PwD and carers. More generally, there would appear to be room for further consideration of the differential form and impact of dementia crisis and crisis management strategies in different sub-groups in the population. These could include minority ethnic populations, likely to access services at a later stage of dementia, people living alone, and the impact of rurality.Implications for practice
Findings from the systematic review showed a lack of good evidence on impact to guide practice for those practitioners caring for PwD during a crisis and the scoping survey highlighted the variability in arrangements for crisis management in dementia across respondents and lack of use of a care pathway by nearly half of teams. Further qualitative work (WP1.2) also highlighted variability in teams and the need for a better defined and broader role to respond to crises, a clearer definition of a crisis and care pathway which linked with other services and understood by all stakeholders. This was confirmed by a strong positive interest in the research by TMCD staff committed to improve their practice. This suggests that a national evidence-based intervention offering standards and guidance would be welcomed by practitioners. In a policy context advocating care at home standards to improve practice and maintain community tenure have immediate salience for practitioners, managers, commissioners, and PwD and their carers. 10Prior to the AQUEDUCT research programme, there were no practice standards for TMCDs. With the development of the 50 statement BPT, TMCDs can measure their current level of performance against best practice standards and develop strategies to improve specific aspects of their practice. The 50 standards could provide national level benchmarking data about practice and variation in TMCD services which would be useful for both policy makers and service planners, providing a basis for quality standards, service development and audit, even if the impact on outcomes is more subtle than that found in the trial.
Conclusions from the programme
The programme was devised as an integrated whole with a clear linear logic pathway flowing from multi-source evidence scoping, to tool design and refinement, design and validation of intervention, through to testing the intervention in a randomised controlled trial (RCT.) Bringing these results together and reflecting upon research findings and process we may conclude:The literature review, scoping survey and qualitative work with team staff, PwD and carers and other stakeholders produced a picture of the operation and scale of TMCD. This mapping of the different structures and ways of working of services managing crises for PwD and their carers, and the perspectives of these different stakeholders on the nature of crises and what is helpful has value for commissioners and managers.
The aim of producing an evidence-based model of best practice via the Best Practice Resource Kit for practitioners was successfully achieved, as evidenced by the development of the BPT with its 50 Standards has potential wider utility and could be employed in quality improvement as in the Memory Services National Accreditation Programme (MSNAP) in England. 36 However, the RCT found no evidence that the use of the Resource Kit impacted psychiatric hospital admission rates, according to the geographical postcode region in which the TMCD operate, or other secondary outcomes.
The pragmatic randomised trial of the Best Practice Resource Kit found no evidence of effects on psychiatric hospital admission rates or of effects on other secondary outcomes (such as the self-reported psychometric changes observed in quality of life (QoL), satisfaction or staff well-being within the TMCD). The intervention was implemented successfully, and qualitative evidence suggests that it was well received, and people felt engaged with it. There are lessons from the study regarding outcome measure choice and processes of data collection.
Data collection in crisis is inherently difficult due to the circumstances of people in highly stressful situations. Research in crisis teams, which are frequently short-term interventions, makes engagement and follow-up more difficult. Furthermore, reflecting the circumstances of the COVID pandemic, the necessary choice of a routine data source (psychiatric hospital admissions) as the primary outcome (perhaps more distant from the intervention than ideal) and collecting secondary outcomes of well-being and satisfaction remotely is likely to have affected sensitivity to the effects of the intervention and response numbers, respectively.
The unique context of the COVID pandemic and its legacy raised lessons for research in relation to adaptability since there was a real risk of non-completion of the RCT. For example, health economic data, initially planned for inclusion in the AQUEDUCT programme, were not collected due to the COVID pandemic and the feasibility study showing that that it was not feasible to collect cost-related data.
The adaptation of the trial design enabled us to successfully complete the study. This provided helpful lessons as to how research design can be constructively modified in the context of major external changes beyond the control of the research team.
A related success was the continued involvement of PPI and reference groups during the pandemic by adapting the modes of contact. Just before the start of lockdown, when it was becoming clear that COVID-19 would make direct contact infeasible, the research team met remotely with PPI and clinical staff reference groups and arranged communication via telephone, digital platforms, e-mail and newsletters, thus maintaining the health and safety of all concerned.
Finally, it is evident that the salience of dementia crisis care can be seen in work published since the AQUEDUCT programme commenced with studies confirming the importance of clarifying operational processes in TMCDs and identifying key components of intervention in care at home. 23,30–32 The AQUEDUCT programme, by developing quality standards of best practice for TMCDs, has contributed to and complemented this knowledge base
REC name
West Midlands - Coventry & Warwickshire Research Ethics Committee
REC reference
21/WM/0004
Date of REC Opinion
9 Mar 2021
REC opinion
Further Information Favourable Opinion