Protocol for the creation of a core outcome set for clinical trials of interventions to improve sleep in people with cognitive impairment: The Sleep in Cognitive Impairment Core Outcome Set (SCICOS) study.

Introduction

There are approximately 55 million people living with dementia worldwide and, with a rapidly growing and ageing population, the global prevalence of dementia is expected to increase to 139 million by 2050¹. Despite the recent emergence of disease-modifying therapies^2–4, there is no curative treatment. Clinical management, therefore, remains largely focused on addressing modifiable risk factors to prevent development and progression of prodromal dementia, and on reducing the symptom burden in more advanced cases.

Sleep disturbance affects up to 70% of people with cognitive impairment⁵ and contributes to worsening cognitive outcomes⁵, impaired daytime functioning and deteriorating quality of life⁶. The increased carer burden caused by sleep disturbance in people with dementia has been shown to lead to sub-optimal care⁷ and early institutionalisation⁸, which increases the economic burden of dementia care⁹.

There is also increasing evidence that sleep disturbance may play a role in precipitating the onset and accelerating the progression of cognitive decline¹⁰. At a biological level, for example, there is evidence that sleep disturbance may increase accumulation within the cerebrospinal fluid of the beta-amyloid and tau proteins that are biomarkers of Alzheimer’s disease¹¹. Although the recently published Dementia Prevention, Intervention and Care: 2024 report of the Lancet Commission¹² has reaffirmed that additional research is required to definitively clarify the effect of sleep on cognitive decline, interest is nevertheless growing in sleep as a modifiable risk factor for cognitive impairment¹³.

Despite the importance of sleep as a treatment target throughout the natural history of dementia, to potentially slow disease progression in prodromal stages and to reduce the burden of symptoms in more advanced cases, the evidence base supporting clinical practice is limited¹⁴. One review notably concluded that “the literature available to guide effective treatment of sleep in patients with Alzheimer’s disease is woefully insufficient”⁶. Systematic reviews of interventions to improve sleep in people with cognitive impairment have been limited by small sample sizes with wide heterogeneity, both in the manner in which sleep is measured and in the outcome measures reported, limiting data synthesis^15,16.

A 2022 scoping review of the measurement of sleep in people with mild cognitive impairment (MCI) and mild dementia found that, after synonymous terms had been combined, 165 separate sleep outcomes were reported across all included studies¹⁷. This heterogeneity limits comparability of data, which precludes data synthesis and inhibits the development of a robust evidence base to guide clinical practice. Furthermore, 62 (37.6%) of the 165 sleep-related outcomes were reported in only one included study, raising the possibility of positive-finding publication bias.

In addition to this heterogeneity in the outcome measures reported, the aforementioned scoping review¹⁷ demonstrated significant variability in the methods used to measure sleep. Of the 188 studies included in the review, 131 used subjective measures of sleep such as questionnaires or sleep diaries. Objective measures of sleep were used less frequently, with polysomnography being used in 88 and medical-grade actigraphy being used in 37 of the included studies. It is, perhaps, unsurprising that most of the studies used subjective measures of sleep, given their low cost, ease of administration and the relatively low burden imposed on participants^17A. However, discrepancies between subjective and objective measures of sleep in people with cognitive impairment are well documented^17B–17C. While the subjective experience of sleep is an important outcome measure in its own right, subjective measures of sleep are undoubtedly susceptible to recall bias, which is likely to be particularly problematic among people with cognitive impairment^17D. Polysomnography is widely recognised as the gold standard method of measuring sleep^17E. However, it is relatively expensive, requires expertise for interpretation, and the more invasive nature of the assessment – generally requiring overnight stays in sleep laboratories whilst wearing multiple attachments to measurement equipment – may be unsuitable for people with more advanced cognitive impairment^17D. There are several other methods of measuring sleep objectively, each with its own associated advantages and disadvantages^17F and varying degrees of accuracy compared to gold standard polysomnography^17G. Acknowledging that there is no direct one-to-one relation between the different methods of measuring sleep, subjective and objective measures likely make their own uniquely valuable contributions to sleep quantification^17H. The lack of concordance between them, however, further limits opportunities to compare data from clinical trials that use different methods of measuring sleep.

A core outcome set (COS) is therefore urgently required to improve the coherence and utility of research data in the field. A COS is an agreed standard or minimum set of measurements and outcomes that all studies in a particular field of research are recommended to report¹⁸. This study aims to involve key stakeholders, including people living with cognitive impairment and their caregivers, in the creation of a COS for clinical trials of interventions to improve sleep in people with cognitive impairment – the Sleep in Cognitive Impairment Core Outcome Set (SCICOS). The SCICOS will increase clinical trial efficiency, reduce the risk of reporting bias, improve data synthesis and help build a robust evidence base for interventions that improve the lives of people living with cognitive impairment and their caregivers.

Methods

The SCICOS has been registered on the Core Outcome Measures in Effectiveness Trials (COMET) database (see: https://www.comet-initiative.org/Studies/Details/3250) and will be created in accordance with the Core Outcome Set - STAndards for Development (COS-STAD) recommendations from the COMET Initiative¹⁹. These recommendations will, however, be adapted to facilitate participation of people living with cognitive impairment and their caregivers, as suggested by the CO-research Involvement and Engagement in Dementia (COINED) model²⁰. The SCICOS study design will therefore involve a multi-stage mixed methods approach in four steps, as follows (see Figure 1):

Figure 1. Flowchart of the SCICOS study design.

SCICOS = Sleep in Cognitive Impairment Core Outcome Set.

Step 1: A systematic review will be undertaken of randomised and non-randomised clinical trials involving both pharmacological and non-pharmacological interventions to improve sleep in people with cognitive impairment (MCI and dementia). The systematic review will produce a list of all the outcome measures reported in those trials. A protocol for the systematic review has been published elsewhere²¹ and has been registered with the International Prospective Register of Systematic Reviews (PROSPERO), registration number: CRD42024556750.

Step 2: Qualitative semi-structured interviews will be conducted among a convenience sample of 20 – 25 people living with cognitive impairment and their caregivers to examine what matters most to them when determining the effectiveness of an intervention to improve sleep. The interviews will therefore endeavour to ascertain the outcome measures that people living with cognitive impairment and their caregivers deem most important in any clinical trial of an intervention designed to improve sleep. It is important to garner the opinions of people with lived experience of the medical condition being examined in this manner, as they may indicate additional outcomes of importance or emphasise different priorities compared to those suggested by clinicians or researchers²². These participants will be drawn randomly from patients and, if required, their caregivers attending memory clinics at the Mercy University Hospital in Cork, Ireland. They will also be drawn from the Patient and Public Involvement (PPI) panels of Dementia Trials Ireland (DTI), a Clinical Trials Network funded by the Health Research Board of Ireland. Prior informed written consent will be obtained from all participants. A bespoke interview topic guide will be used (see Table 1 in extended data). Responses will be recorded and transcribed verbatim. Qualitative data analysis software will be used to analyse the transcripts in accordance with the reflexive thematic analysis approach espoused by Braun and Clarke²³.

Step 3: Collating the results of the systematic review and the qualitative semi-structured interviews. Any additional outcome measures suggested by the qualitative semi-structured interviews (described in step 2 above) will be collated with the list of outcome measures produced by the systematic review (described in step 1 above) to produce a preliminary ‘long list’ of outcome measures. These initial outcome measures will be categorised by domain and core area in accordance with the OMERACT Filter 2.0 Framework²⁴. Outcome measures within the same domain that are synonymous or similar may be combined to avoid unnecessary duplication. The categorisation and combination of outcome measures in this manner will be undertaken independently by two researchers and any disagreements will be resolved by the principal investigator, as recommended by the COMET Handbook²⁵.

A previous scoping review examining studies involving the measurement of sleep in people with MCI and mild dementia has demonstrated that many outcome measures appear in only one (37.6%) or two (58.8%) studies¹⁷. As any outcome measure that appears so infrequently is unlikely to be considered a core outcome measure, we will exclude from further consideration in the modified Delphi process any outcome measure that:

Step 4: A modified Delphi process will be employed to reach consensus regarding the outcome measures that should be included in the SCICOS²⁵. There is no consensus on the recommended sample size for a Delphi panel²⁵. We aim to include sufficient lay and professional participants to ensure comprehensive representation of different perspectives at all stages of the process²⁵. Different modes of participation in the modified Delphi process will be offered to professional and lay participants in order to facilitate involvement of people living with cognitive impairment and their caregivers²⁶. Fully informed prior written consent to participation will be obtained from all participants. Demographic details, including age, sex and stakeholder group (either professional or lay participants) will be collected.

Study status

The study has commenced and is currently on-going.

Discussion

In recent years, due to the growing burden of dementia¹ and an increasing preponderance of evidence implicating sleep disturbance not just in its symptomatology but also perhaps in its pathophysiology^10,11, there is growing research interest in the treatment of sleep disturbance in people with cognitive impairment^13,17.

Most clinical trials of interventions to improve sleep in people with cognitive impairment have been limited by small sample sizes^15,16. The synthesis of data from these trials has been limited in turn by wide heterogeneity both in the methods used to measure sleep and in the outcome measures reported, precluding meta-analysis¹⁶ and impeding the development of a reliable evidence base to guide clinical practice. Given the challenges inherent in conducting clinical research among people with cognitive impairment³², it is important to ensure that future research is reliable and coherent, thus increasing the likelihood that the research will lead to clinically meaningful insights. The SCICOS will improve the utility and efficiency of future clinical trials by prescribing a standard minimum set of outcome measures, and making recommendations regarding the methods used to measure them, thus reducing risk of reporting bias and facilitating meta-analysis.

The SCICOS will be developed in accordance with the COS-STAD recommendations¹⁹. Clinicians and researchers from around the world with established expertise in sleep and/or dementia will be invited to participate to optimise generalisability, credibility and relevance. The COS-STAD recommendations will be adapted to ensure meaningful involvement of people living with cognitive impairment and their caregivers throughout the process of developing the SCICOS. A 2020 Cochrane review of pharmacotherapies in dementia³³ found that clinical trials do not always succeed in taking account of what matters most to the end-users of the tested interventions. In ancillary qualitative research among caregivers of people with dementia, it was noted that their highest priority outcome was uninterrupted sleep at night yet this was not included as an outcome measure in any of the included trials. The process for developing the SCICOS includes people living with cognitive impairment and their caregivers as key stakeholders.

As a practical consequence of the adaptations made to the study design to facilitate meaningful involvement of people living with cognitive impairment and their caregivers, data collection from this key stakeholder group will be limited to Ireland. This may potentially introduce bias, as the level of importance attributed to outcome measures may vary across different countries and cultures. Nevertheless, by involving people living with cognitive impairment and their caregivers throughout the process of creating the SCICOS, this study will ensure that the voices of this key stakeholder group will be heard in all relevant future clinical trials.

Conclusion

There is growing interest in sleep disturbance as a modifiable risk factor for cognitive impairment. Sleep disturbance remains an important treatment target in more advanced dementia to alleviate the associated symptom burden and carer strain. Existing literature regarding the role of sleep in cognitive impairment is beset by wide heterogeneity in the methods used to measure sleep and in the outcome measures reported. Creating a COS for clinical trials involving interventions to improve sleep in people with cognitive impairment will increase comparability of data and help build the evidence base for interventions that can improve the lives of people living with cognitive impairment and their caregivers.

Ethics and consent

The SCICOS study will be conducted in accordance with the Declaration of Helsinki and General Data Protection Regulations (GDPR). Ethical approval was granted for each phase of this study by the Clinical Research Ethics Committee of the Cork Teaching Hospitals (CREC) on 30^th July 2024, reference numbers: ECM 4 (c) 06/02/2024 & ECM 5 (7) 20/02/2024 & ECM 4 (dd) 30/07/2024.

Fully informed written consent will be obtained from all participants prior to their participation in the study. In order to facilitate the process of obtaining fully informed consent, each prospective participant will be provided with a comprehensive Participant Information Leaflet that is appropriate to their personal circumstances. Separate versions of the Participant Information Leaflet will be provided to prospective participants from each stakeholder group, namely professional and lay participants. Within the latter stakeholder group, separate versions of the Participant Information Leaflet will be provided to people living with cognitive impairment and their caregivers. Copies of each version of the Participant Information Leaflet and Consent Form are provided in the extended data.

Dissemination

The SCICOS will be reported in accordance with the Core Outcome Set-STAndards for Reporting: the COS-STAR Statement³⁴. Everybody who participated in the process of developing the SCICOS will be provided with a copy of the final report. The SCICOS will be disseminated through submission for publication in a peer reviewed medical journal and promulgation at relevant clinical conferences.