Challenges for the design of randomised trials for interventions for people with multiple long-term conditions: Recruitment and outcome selection

Abstract

The burden of multiple long-term conditions is significant for the individual and society. Exploring interventions to alleviate disease progression, accumulation of long-term conditions and symptom burden is crucial. This commentary debates two important considerations for intervention design and subsequent recruitment approaches.

Keywords

multiple long term conditions trials

Introduction

Appropriately designed and well conducted randomised controlled trials are the gold standard for generating evidence for the efficacy, safety, and cost effectiveness of healthcare interventions to inform decision making. Two key aspects of trial methodology and design are the processes for trial participant identification and reporting of appropriate outcome measures.

Despite the high healthcare burden and need, to date, intervention trials targeting people with multiple long-term conditions (MLTCs) have shown limited improvement in patient outcomes or health economic gains.¹ There is clearly a pressing need to develop acceptable and effective interventions against the backdrop of an aging population and the health inequalities associated with living with MLTCs.

As a result, the design and delivery of trials to identify clinically effective and cost-effective interventions for the effective management of MLTCs has been identified as a priority for global research by healthcare policy groups and national research funders including the National Institute of Health Research (NIHR) in United Kingdom and the National Institute of Health (NIH) in United States.

We raise two specific design issues for trials of interventions for MLTCs¹: population identification and related trial recruitment and² the selection of outcome measures. We do so based on our experience of an ongoing NIHR funded research programme – PERFORM: Personalised Exercise-Rehabilitation For people with Multiple long-term conditions. The PERFORM programmes aims of develop and evaluate an exercise-based rehabilitation intervention specifically designed to target the people with MLTCs (https://fundingawards.nihr.ac.uk/award/NIHR202020).

Population selection

In the development phase of our programme of research we considered the challenges of the heterogeneity of a population with MLTCs and chose to be ‘disease agnostic’ in our approach, that is, we sought to avoid a scenario of selecting only those with a key index condition plus another related long-term condition or ‘comorbidity’ e.g. the combination of cardiovascular disease plus diabetes. To inform the design of the PERFORM intervention we undertook two parallel pieces of research work which we believe is novel and worthy of consideration by the research community. Firstly, we identified clusters of long-term conditions using large UK datasets that had a significant impact on HRQoL, hospitalisations and mortality.² Secondly, given our intervention focused on exercise as a key component, we conducted an overview of existing systematic reviews to identify which of a predefined list of 45 long term conditions was exercise an effective intervention,.³ These results have directly influenced our patient selection for randomised controlled trial to assess the clinical and cost-effectiveness of the PERFORM intervention.

Outcome selection

A core outcome dataset (COSmm) has been proposed for multimorbidity research.⁴ While this COSmm set proposes ‘domains’ (e.g. HRQoL, mental health) it was not designed to define specific outcome measures (e.g. EQ-5D or Short-Form-36 for HRQoL). As a result, there is currently inconsistency in key primary and secondary outcome measures reported in MLTCs trials making direct comparison of intervention results and quantitative data pooling challenging. Furthermore, some of the core domains listed in the COSmm are infrequently reported e.g. shared decision making and prioritisation. A specific outcome dataset (COSMOS) for MLTCs has been developed for low-and-middle income countries.⁵ The four domains are: adherence to treatment, adverse events, out-of-pocket expenditure, and quality of life. Whilst the brevity and generic approach of COSMOS is attractive it may miss many of the aspects relevant to interventions for MLTCs. In the case of our PERFORM programme, two key missing COSMOS outcome domains are symptom burden and physical function. There are a number of core outcome sets proposed for single disease focused exercise-based rehabilitation, which describe overarching domains with the symptom burden measures being disease specific. The challenge for MLTC researchers and clinicians is the selection of outcomes reflecting the burden of often heterogeneous complex set of symptoms experienced by individual patients with MLTCs whilst also having the sensitivity to detect changes post interventions and not over burdening trial participants. The recent COVID-19 pandemic challenged researchers to consider a much broader range of outcome measures to reflect the wide range of symptoms reported. Equally there is a tension between the selection of a generic measure that is broadly applicable to a range of long-term conditions versus the nuances and likely greater sensitivity of disease specific measures.

Conclusions

Whilst MLTCs is a key global research priority, there is a currently little or no formal methodological recommendations for the conduct and reporting of randomised trials of interventions for people living withMLTCs. We have sought to provide an overview of the specific challenges of trial participant identification and trial outcome selection and drawing on the experiences of our ongoing PERFORM research programme evaluating an exercise-based rehabilitation intervention for people with MLTCs. In Table 1 we summarise the central challenge of each of these two issues and propose potential solutions for researchers and clinicians.

Table 1.

Participant identification & outcome collection in intervention trials in MLTC–challenges and potential solutions.

Challenge	Potential solutions (and examples)
Participant identification/selection & recruitment
In contrast to trials that classically focus on a narrow disease-focused population, by definition, trials of MM need to recruit a heterogeneous population that includes people with two or more LTCs and potentially a wide range of different LTCs	To ensure recruitment to target, recruit patients from general practice/primary care and/ from a wide range of secondary/specialist care specialist settings (e.g. from musculoskeletal/diabetes /renal/cancer services) (e.g. PERFORM)
	To maximise the likelihood of intervention benefit, select participants with specific MLTCs on basis of their specific LTCs with prior evidence that they are likely to benefit from any chosen intervention (e.g. PERFORM)
	To maximise the likelihood of intervention benefit, select specific MLTC participant with a combination of LTCs with high unmet needs in terms of their poor HRQoL/risk of mortality
	To maximise the likelihood of intervention benefit, exclude MLTC participants who do not demonstrate a significant symptom burden against which the intervention is targeted (e.g. Exclude those with a high existing exercise capacity given their limited potential for benefit) [e.g. PERFORM]
	To reduce heterogeneity, could recruit to a core LTC (e.g. patients with diabetes mellitus plus the presence of one or more other LTC) but has major caveat that trial will lack applicability to the wider MLTC population
	To ensure appropriate representation of trial population by stratifying inclusion on the basis of prevalence of LTCs and/or combination/clusters of LTCs
	To improve the evidence base (e.g. undertake studies within a trial – SWATs) on strategies to test different approaches to trial recruitment in MM trials [e.g. PERFORM]
Outcome assessment
In contrast to trials that classically focus on a narrow disease-focused populations, the people recruited in trials of MLTC will have a very wide range of symptoms and functional needs. Thus, instead of evaluating outcomes in the context of one specific condition, outcomes need to be assessed (& interpreted) in the context of the overall burden of MLTC.	To include a wide range of outcomes that might pick up the range of health needs of different MM patients but consider caveat of participant completion burden [e.g. PERFORM trial collecting data on the burden of discreet symptoms to include pain, fatigue, psychological well-being, function etc )]
	To use generic/disease agnostic outcome measures (e.g. EQ-5D) albeit it is recognised that such outcomes may lack responsiveness to change with introduction of an intervention [e.g. [20]], in PERFORM our PPIE panel preferred the EQ-5D over the SF-36 as our primary outcome measure reflecting HRQoL, with a range of secondary outcomes identified that addressing a number of key symptoms reported by MLTC participants
	To develop core outcome data set(s) (COS) for trials in MLTC exercise-based trials to complement existing outcome datasets
	To use and develop psychometrically validated outcome measures for MLTCs to understand the overall symptom burden in collaboration with those with lived experience of MLTC

The PERFORM team

¹Rachael A Evans, ²Sharon A Simpson, ¹James Manifield, ³Hannah Gilbert, ³Amy Branson, ³Shaun Barber, ³Ghazala Waheed, ⁴Emma McIntosh, ³Gwen Barwell, ¹Zahira Ahmed, ⁵Sarah Dean, ⁶Patrick Doherty, ⁷Nikki Gardiner, ⁸Colin Greaves, ⁸Paulina Daw, ⁹Tracy Ibbotson, ⁹Bhautesh D Jani, ¹⁰Kate Jolly, ⁹Frances S Mair, ⁹Cristina Vasilica, ¹¹Paula Ormandy, ¹²Susan M Smith, ^2,13

¹Department of Respiratory Sciences, University of Leicester, Leicester, United Kingdom

²MRC/CSO Social & Public Health Sciences Unit, School of Health and Wellbeing, University of Glasgow, Glasgow, United Kingdom

³Clinical Trials Unit, University of Leicester, Leicester, United Kingdom

⁴Health Economics and Health Technology Assessment, School of Health and Wellbeing, University of Glasgow, Glasgow, United Kingdom

⁵University of Exeter Medical School, Exeter, United Kingdom

⁶Department of Health Science, University of York, York, United Kingdom

⁷Department of Cardiopulmonary Rehabilitation, University Hospitals of Leicester NHS Trust, Leicester, United Kingdom

⁸School of Sport, Exercise and Rehabilitation Sciences, University of Birmingham, Birmingham, United Kingdom

⁹General Practice and Primary Care, School of Health and Wellbeing, University of Glasgow, Glasgow, United Kingdom

¹⁰Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom

¹¹School of Health and Society, University of Salford, Manchester, United Kingdom

¹²Discipline of Public Health and Primary Care, Trinity College Dublin, Dublin, Ireland

¹³Robertson Centre for Biostatistics, School of Health and Wellbeing, University of Glasgow, Glasgow, United Kingdom

Footnotes

Acknowledgements

This study was funded by the National Institute for Health and Care Research (NIHR; Personalised Exercise-Rehabilitation FOR people with Multiple long-term conditions (multimorbidity)—NIHR202020). This study was supported by the NIHR Leicester Biomedical Research Centre. S.J. Singh is a NIHR Senior Investigator. The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the National Institute for Health Research, Health Education England, or the Department of Health.

ORCID iD

Sally J Singh

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the National Institute for Health and Care Research (NIHR; Personalised Exercise-Rehabilitation FOR people with Multiple long-term conditions (multimorbidity)—NIHR202020).

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: SJS is Clinical Lead for National Respiratory Audit Programme—Pulmonary Rehabilitation.

References

Smith

Wallace

O'Dowd

, et al.. Interventions for improving outcomes in patients with multimorbidity in primary care and community settings. Cochrane Database Syst Rev. 2021 Jan 15;1(1):CD006560.

Krauth

Steell

Ahmed

PERFORM research team , et al.. Association of latent class analysis-derived multimorbidity clusters with adverse health outcomes in patients with multiple long-term conditions: comparative results across three UK cohorts. EClinicalMedicine. 2024;74:102703.

Dibben

Gardiner

Young

HML

, et al. PERFORM research team. Evidence for exercise-based interventions across 45 different long-term conditions: an overview of systematic reviews. EClinicalMedicine. 2024;72:102599.

Smith

Wallace

Salisbury

, et al. A core outcome set for multimorbidity research (COSmm). Ann Fam Med. 2018;16:132-138.

Vidyasagaran

Ayesha

Boehnke

, et al. Core outcome sets for trials of interventions to prevent and to treat multimorbidity in low- and middle-income countries: the COSMOS study. medRxiv [Preprint]. 2024 Jan 30:2024.01.29.24301589.