Service innovations for people with multiple long-term conditions: reflections of a rapid evaluation team

Health service context

As noted in the Background, it is the norm and not an exception to have more than one long term health condition. 23 This begs an important question as to why living with multiple long-term conditions is so often framed as something special or exceptional that requires particular attention in policy, research and practice. As a respondent to our interviews noted:

So we need to think of this in terms of mainstream services. I guess most people over 40 will have one condition, so most over 50 or 55 will fall into your categories [the NICE definition of multimorbidity]. So we’re not looking about sort of niche or special services. This needs to be how business as usual can adopt and help this patient cohort I think.

(Report Supplementary Material 4 presents the overall findings from the stakeholder interviews.)

The development of the healthcare system in England has followed various critical paths, including from pre-NHS times, when as early as 1920 the Dawson Report37 set out a blueprint for integrated preventative, primary and secondary health care, recognising the already siloed nature of service provision. Our findings suggest that the ambition in the Dawson Report is arguably still to be achieved, as the historical development of NHS (and many other countries’) health services has adapted to reflect the constant expansion in what is technologically possible in terms of treatment and has led to greater specialisation within the hospital sector; kept primary, mental health and secondary care largely separate; and tended to deal with patients according to a single presenting health condition. 38

Clinical specialism and generalism

Clinical specialisation is an important and much-needed aspect of our health system. Patients rightly expect to be cared for by specialists when this is necessary, and the help required is deemed to be beyond the scope of primary or secondary generalist care. However, there is increasing evidence that current models of hospital care organised on a disease-specific basis do not serve well the large and rising number of patients with multiple long-term conditions. Hence there are associated calls for a revival of medical generalism. 39 This builds on analysis by the Royal College of Physicians’ Future Hospital Commission of 2013,40 which set out a vision of hospitals where care would be organised to meet the needs of the largest cohort of patients: people with multiple long-term conditions.

The original Royal College of Physicians’ proposal40 was for a medical division led by a chief of medicine who would be responsible for all medical services in a hospital, with care organised so that specialist input and services would come to patients in an acute care hub where they would be stabilised before a co-ordinated acute and community plan of care was put in place. Following a 3-year programme of piloting ideas in the Future Hospital Programme, the Royal College of Physicians set out an updated vision for medical care based on principles for improving future health and care, noting, ‘[t]he … vision of improving patient care through enhanced access to specialist medical care closer to home and earlier in hospital pathways was realised in part’ (page 9). 41 This would seem to suggest that achieving major change to how the care of people with multiple long-term conditions would be organised within hospitals proved too hard to effect, yet also points to other ways in which the needs of such a patient population might be met, such as via specialist outreach into community locations.

NIHR-funded research by Vaughan et al. 39 examined models of medical generalism used in small hospitals from patient, professional and service perspectives. They concluded that while the case-mix of smaller hospitals was dominated by patients whose conditions were amenable to generalist care, there was no evidence to change models of care per se (echoing the conclusions of the Royal College of Physicians’ Future Hospital Commission), but rather to better match hospital staff to meet local needs and seek to ensure more collaborative environments.

The siloed and single disease focus of the healthcare system was highlighted by almost all interviewees in this study. This is reflected in, or reflects, the highly specialised organisation of clinician training and careers whereby doctors, nurses and other healthcare professionals are expected to opt for a particular specialism or training pathway at a relatively early stage. This then translates to the ways in which services are organised and patients are treated. Even within the patient group charity sector, organisations are often single-condition specific, although there are some umbrella groups such as National Voices and The Richmond Group that attempt to find commonalities across charities and represent broader patient and carer needs and expectations.

When multimorbidity matters

A recurring theme in interviews for this study, and in our analysis of the BRACE portfolio of evaluations was that long-term conditions are not of themselves a problem until they cause a particular concern for a patient, including when patients seek help to manage and live with the conditions. One interviewee commented: ‘[F]or most people multimorbidity it’s fine as it is. You don’t need to do anything different […] You could have loads of conditions, but actually be managing just fine and you don’t need a different approach to care’.

The heterogeneity of people living with multiple long-term conditions, and the combinations and severities of those conditions, was emphasised by interviewees. To make this more tractable when thinking about organising health care, one respondent referred to a pyramid, which recalled the Kaiser Permanente Triangle of Care42 whereby the top represents the most severe needs (with or without end-of-life care), the middle tier represents those needing secondary care and/or enhanced primary care support, and the bottom tier represents those managing well with self-care and routine family medicine input. The Triangle of Care drew on Wagner’s Chronic Care Model43 in its use of strong patient self-management as the foundation of effective care, supported by community and acute care diagnostic and clinical support where necessary.

What an examination of the BRACE portfolio revealed

Our cross-analysis of the 10 evaluations within the BRACE portfolio bears witness to how the needs of people with multiple long-term conditions are not typically foregrounded when service innovations are designed and planned for implementation. The topics identified for evaluation by BRACE that could have broken the single-condition-focus mould included the early implementation of Primary Care Networks;1 hospitals managing general practice;2,9 the impact of telephone-first (Telefirst) access to general practice;4 and the use of AI with sensors in home-based social care. 7

In practice, BRACE rapid evaluation did not find that these service innovations were taking direct account of the needs and priorities of people with multiple long-term conditions. For example, serving patients with multiple long-term conditions did not emerge as a theme in the Primary Care Networks evaluation,1 albeit these new networks have proactive and better co-ordinated care as a key policy objective. The Primary Care Network structure and resources arguably offer a receptive context for designing services to meet at least some of the specific needs raised by living with multimorbidity, whether at the self-management or the enhanced primary care level. Indeed, some of the interventions planned for and by Primary Care Networks have the potential to improve the care of people with multiple long-term conditions.

In the BRACE second phase evaluation of hospitals managing general practice,2 the research team for that project specifically looked for differential impacts of this new organisational model on people living with multiple long-term conditions. They found evidence that it might slightly reduce use of emergency hospital care, but this is as much in patients with single or no long-term conditions as in patients living with multiple long-term conditions. 9

Similarly, in the BRACE study of telephone-first access to general practice,4 the researchers built in an explicit focus on seeking evidence of inequalities in access related to multiple long-term conditions yet did not find any. 4 They did however identify that differences between general practices in how they organise care appear to affect people’s access to appointments, suggesting that in the case of telephone-first innovations it is important for attention to be given first to improving care arrangements in the less well performing practices and then to monitoring carefully for differential impacts on different patient groups. 4

The BRACE study of AI and sensors for home care7 found that problems with planning and implementing the technology at a local level meant that it was not yet impacting on anyone’s (social) care. 7 This echoed the BRACE studies of hospitals managing general practice, which found that there is a need to keep considering the needs and experiences of people with multiple long-term conditions over the longer term, for early evaluation is likely to be less granular and more likely to reveal more general barriers to progress that affect all users, not specifically those with more complex needs.

Bridging the gaps

In addition to there being a single-disease focus in the healthcare system, service provision is often siloed at the individual patient level. Mutual recognition and communication between specialty teams is too often poor and it is then left with the patient’s GP, if not the patient and their family, to fill and bridge the gaps. It was also noted that clinicians need to have knowledge of different referral pathways and services to meet patients’ wider needs, together with confidence to talk about areas beyond their own specialism.

An example of how the needs of people with multiple long-term conditions could be better assured within health service provision was highlighted by a PPI participant in the stakeholder workshop we undertook for this study. They proposed that GPs and other healthcare professionals should be willing to discuss more than one condition per appointment, rather than keep to the ‘one thing at a time’ approach that patients had encountered in primary care, where a holistic and comprehensive approach to care is arguably most expected and needed. 44 Siloed teams often have distinct ways of working (sometimes even within a single organisation), making integration of care at the patient level challenging. In addition, IT infrastructure is often not connected across systems or teams, and especially when crossing healthcare organisations, making sharing patient data and care co-ordination challenging. 45

The role of funding and incentives

A further factor that can entrench (or reduce, if used appropriately) siloed working within the health system is funding mechanisms and incentives. Interviewees pointed out that care for people living with multiple long-term conditions is typically not recognised in the way that services are commissioned and incentivised, as funding streams in both secondary and primary care sectors reflect unifocal services and specialties. For instance, in general practice there is a desire to offer a multifocal and holistic approach to care, but funding is made to practices partly on a single condition basis, as in payments for care for specific conditions such as diabetes or hypertension, and with no funding directly to pay for care co-ordination. In addition, general practices are not measured on their ability to offer continuity of care, which research consistently shows to be associated with improved patient outcomes,46 but instead on the number of patients seen. One interview respondent noted: ‘it’s [multimorbidity is] not a focus, … if it was something that we were specifically measured against and harassed against, we’d measure it and we’d do something about it.’

We return to the issue of how best to measure what matters for people living with multiple long-term conditions in Theme 4 later in this report.

A further issue raised in study interviews was a lack of funding to develop and deliver innovations for people with multiple long-term conditions and particularly for sustained resources beyond the pilot stage of a new service development. In most BRACE evaluations, meeting the needs of people living with multiple long-term conditions did not emerge as a theme unless we explicitly sought to raise it. This is very evident from the summary of our cross-project analysis set out in Report Supplementary Material 2.

The nature of service innovations for people with multiple long-term conditions

In a similar vein, interviewees were for the most part unable to recall any service innovations that were specifically for people with multiple long-term conditions:

I could honestly say I know of virtually no service innovations for people with multimorbidity, but I do know of lots of service innovations for older people or people in later life which are effectively multimorbidity interventions in groups of people who are expected to benefit more or have more problems.

Other interviewees described innovations for people with multiple long-term conditions as ‘patchy’, ‘variable and bitty’ and ‘evolving’, with good practice located in pockets but not consistent or widespread. When asked about service innovations for people with multiple long-term conditions, respondents cited a wide variety of examples. But these were often not specific to people with multiple long-term conditions and were rarely connected by respondents to the specific care needs of people with multiple long-term conditions, unless prompted by the interviewer. Innovations mentioned that could be linked to people with multiple long-term conditions are listed in Table 5 in Theme 5 below, alongside the care need(s) they seek to address.

Interestingly, interviewees generally considered relatively simple (but not costless) things like longer appointment times, asking patients if they wanted to see the same clinician as previously, and specific clinics for people experiencing frailty as a result of multiple long-term conditions, as being ‘innovative’. One interviewee commented that basic administrative support for care needs to be subject to sustained modernisation and attention, for example, ensuring that records are up to date, avoiding duplication of services and improving data sharing for care co-ordination. This echoes work by the King’s Fund that explored the impact of NHS administration on patient care. 47 While not ‘exciting’ as a service innovation, giving careful attention to improving NHS patient administration holds significant potential to address some of the needs of people with multiple long-term conditions by helping them feel safe, with well-co-ordinated and organised care and clarity about what is to happen next and why. 48 As the King’s Fund report notes in its ‘key messages’:

For people who live with long-term conditions, use multiple health and care services or who have additional needs, for example, people with a disability, their experience of admin can play a critical role in their overall experience of care. 47

Context

Health research in many ways mirrors the single-specialty and siloed nature of the healthcare system and how care is organised at the individual level, as explored in the previous section of this report. This is unsurprising as most UK empirical health research is conducted within the NHS, often led by clinician-academics who are trained and work within single specialties or sectors. Even in the more generic area of health services research, academics often have some form of sector or specialty focus to their research interests, following the way in which calls for research are scoped, programmes of study crafted and career reputations built. Where researchers do undertake work that is more generic in nature, it is often on organisational issues such as integrated care, care co-ordination or partnership working, rather than the needs of people living with multiple long-term conditions per se.

The Chief Medical Officer for England, Professor Sir Chris Whitty, along with other senior national medical leaders, wrote in 2020 of the need to shift the paradigm of medical training, practice and research to reflect that, ‘[t]reating each disease in a patient as if it exists in isolation will lead to less good outcomes and complicate and duplicate interactions with the healthcare system.’18 He went on to note that:

Medical science is also disease based. Clinical trials still often exclude people who have more than one condition. Good vertical integration exists from bench to bedside for a single condition or disease, but there is little or no horizontal integration between diseases that often coexist. This will require an intellectual shift and rethinking some elements of our research, training, and practice in virtually every discipline. 18

Reproduced from Whitty et al. 18 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY-NC 4.0) licence, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/

The authors of an editorial examining how research funders might respond to the challenges of multimorbidity27 drew together the overarching health system focus on single specialties with the need for holistic research that attends to the majority needs of people living with multiple long-term conditions:

More people now live with several health problems at an earlier age and health inequalities have increased, but most clinical services remain geared towards management of single diseases. Research systems have reinforced these distortions, with a focus often on single condition research, typically excluding people with comorbidities as trial participants. 27

It seems from our work for the overarching study reported here that the health research community, much as with the broader healthcare system and its organisation of services, has not yet made significant progress towards achieving the changes recommended in health research policy. Just as the need for better integrated care is frequently highlighted as a central policy priority, so is the need for researchers to attune their intellectual time and energy to the complex overlapping needs of people of all ages living with many different combinations of long-term conditions, rather than exclusively studying treatment or service innovations designed with a single condition or speciality in mind. In both cases – health services delivery and research – it seems the challenges presented by multimorbidity are profound in respect of behavioural and cultural change, meaning that it is easier to keep commissioning and doing research that reflects the status quo of professional specialities and often siloed healthcare organisation and provision.

Research policy ambitions

The overarching strategy of the NIHR ‘Best Research for Best Health: The Next Chapter’25 has ‘improving the lives of people with long-term conditions’ (page 30) as an area of strategic focus and acknowledges the requirement for action:

The needs of people living with MLTC are not well served either by clinical services or by science, both of which are organised vertically around single conditions. NIHR is taking steps to address these needs through research, working in partnership with the Academy of Medical Sciences, the MRC, Wellcome and a number of medical research charities, but we know that more needs to be done. 25

Reproduced from NIHR 2021. 25 p. 30 The text above includes minor formatting changes to the original text.

The NIHR is investing in studies to map and understand trajectories and clusters of long-term conditions using data science and AI and acknowledges the need for more research to understand better the specific needs of people with long-term conditions and their carers, and how to get evidence-based interventions into practice. Other initiatives have included offering additional funding for research teams to extend current studies to accommodate the needs of people with multiple long-term conditions; research calls (including ones that embrace the full suite of NIHR programmes) with an explicit focus on services to meet such needs; resource for innovative methodological approaches that can help examine how care interventions work (or not) for people with multimorbidity; and global health research programmes to learn about the topic from an international perspective.

The NIHR published a strategic framework for multiple long-term conditions research in 202049 setting out priorities and ‘a pathway to foster the cultural changes’25 required. This included commitments to (1) review all funding processes to ensure they actively support multimorbidity-focused applications; (2) ensure that studies do not unjustifiably exclude people with multiple long-term conditions; (3) encourage team science across health, social care and public health research and (4) deliver research that helps the health system to implement changes that are required to meet the needs of people with long-term conditions. That a bespoke strategy for multiple long-term conditions research has been established in the UK is clearly important. How far it is feeding through into research calls, study applications and the work of local health research teams remains to be seen.

What an examination of the BRACE portfolio revealed

Our cross-analysis of 10 BRACE evaluations revealed that the needs of people with multiple long-term conditions were seldom a specific focus of the service innovations we were asked to evaluate (see Table 1 and Report Supplementary Material 2). Some innovations evaluated by BRACE may in practice be more advantageous for people with multiple long-term conditions than for other patients. For example, vertical integration of hospital and primary care in one case study was set up following a national Vanguard project aimed at redesigning services for people with multiple long-term conditions, but we found no specific impacts related to the needs of this population within the scope of our two BRACE vertical integration studies. 2,9,34 In the digital-first primary care evaluation, we found the innovation to be disadvantageous for people with multiple long-term conditions because algorithms for accessing care implicitly assumed single condition issues. 8 BRACE’s evaluation of Telefirst general practice showed that people with multiple long-term conditions enjoyed speedier access to healthcare professionals as a result of the innovation but so did other patients and to a very similar extent. 4

The BRACE evaluation of WHHs was not concerned with multiple long-term conditions, but the study revealed parallels between integrating these acute care services designed to meet women’s diverse health needs across the life course and integrating care for people with multiple long-term conditions10 (Box 2). A survey responded to by 12 WHHs revealed that reproductive health care for women is not well integrated currently, and that a one-stop shop for the patient or a co-ordinated hub-and-spoke arrangement would be valuable to enable multiple issues to be addressed in the same appointment. Eleven of the 12 responding hubs had providing ‘holistic’ care as an objective (to women, within the scope of women’s reproductive health services, in this case).

These examples from BRACE’s portfolio of evaluations point to the significant potential of finding ways in which evaluations of innovations for single service areas or conditions can offer important insights for people with multiple long-term conditions if there is sufficiently consistent and focused attention to this priority issue. Later in this report, we explore how such attention might be better assured within broader health research practice.

Bridging the gaps

The example of the BRACE WHHs evaluation10 illustrates how, with focused attention to the needs of people living with multiple long-term conditions, a study of a single service or innovation designed for a specific population group can be mined for insights for the overarching priority of meeting the needs of people living with multiple long-term conditions. In the view of the research team for the overarching study, this seems to require the development of a set of criteria (perhaps in the form of a checklist) to apply to a study to explore implications for people with multiple long-term conditions. We suggest that there might in future be a requirement that all research proposals should explicitly consider the implications for people living with multiple long-term conditions, much as research proposals to NIHR are already required to consider PPI and engagement, and equality, diversity and inclusion. These latter two examples bear witness to the need to make an issue critically important for it to start to form a core part of research priorities, planning and practice.

A further area that appears to us to need attention within research funding would be to find ways in which multiple long-term conditions can be understood, profiled and applied as being about the needs of people across their life course. Hence any application of multiple long-term conditions criteria to research commissioning needs to include projects about children and young people’s conditions and services, those of women in pregnancy, people living with disabilities, and those people of working age with perhaps one condition but likely to acquire others within a few years. Explicit attention to younger age groups would generate data and insights that could help services to address age-related issues and inequalities.

On average, people experiencing socioeconomic deprivation are more likely than others to be living with multiple long-term conditions and they start doing so at younger age. 50 This also points to the need for more longitudinal studies of how people can be supported to manage their long-term conditions at whatever age these emerge, and what innovations might be needed to enable them to expand such self-management into having several conditions. We suggest that more use of large-scale and longitudinal studies – such as Biobank and the ZOE cohort studies – could prove fruitful here, pointing to the need for more use of data sources across different research traditions.

In the interviews, it was pointed out that there is potential for much more secondary analysis of health research trials data, examining aspects of implementation of innovations that may not have been central to the core study. An example was given of an Extended Stroke Rehabilitation Service (EXTRAS) trial led by Helen Rodgers of Oxford,51 where health economics analysis showed that having sufficient levels of allied health professional-delivered reviews – which included identification of rehabilitation needs, joint goal setting, and action planning – made a significant difference to the health-related quality of life of stroke survivors52 and led to less use of social care resource longer term. One interviewee commented that these were ‘gold-dust findings but largely ignored as they were secondary analysis’, yet were about what really matters to patients and carers. In our view, this points to the need for more emphasis within research commissioning on implementation studies to understand core features and contextual factors for adoption, and for this to occur across different types/disciplines of research funding programmes.

The issue of having core data to understand the profile and needs of the population living with multiple long-term conditions was cited by several respondents to the study interviews. Some interviewees were unaware of any data collected specifically or routinely about or for people with multiple long-term conditions and others were unsure about what data could be used to show whether a service innovation or more integrated approach to care was benefiting people with multimorbidity.

Across the portfolio of evaluations undertaken by the BRACE, our experience was that routinely collected data, such as the national General Practice Patient Survey and Hospital Episode Statistics, can be used to investigate some differential impacts for people with multiple long-term conditions compared with other patients (e.g. on the use of and satisfaction with some NHS services). The BRACE evaluation of telephone triage4 is an example (Box 3).

We conclude that ways in which the gap between the NIHR research priority for people with multiple long-term conditions and actual research practice might be bridged include:

• having clarity of definition of who constitutes a person with multiple long-term conditions

• how they can be identified within NHS patient registers and data sets

• what data sources are available to researchers to understand the needs of this population

• which clusters of conditions are most common.

National Institute for Health and Care Excellence has undertaken important work in this area, developing guidelines for optimising the care of people with multimorbidity12 and this should arguably be made available to researchers when planning studies, through links in research commissioning briefs. Furthermore, the NIHR strategy about research for people with multiple long-term conditions could be more strongly profiled and encouraged as a core consideration in health services and other research calls. 25

Speaking evidence to policy and power

The relative lack of evaluation evidence about whether and how service innovations address the needs of people with single or multiple conditions was suggested by Marmor53 to present a particular risk of ‘policy fads and fashions’ being adopted. Some of the stakeholders we interviewed asserted that too many innovations advocated by policy-makers for people with multiple long-term conditions are not sufficiently evidence based, giving the example of virtual wards where evidence of effectiveness remains uncertain. 54,55 This highlights a profound risk related to the relative absence of research that is properly and consistently focused on the needs of the many NHS users who are living with multiple long-term conditions. Namely, where a service innovation is identified, such as pulse oximetry at home (which was rapidly evaluated jointly by BRACE and other teams during the COVID-19 pandemic) or virtual wards, they may be deemed too quickly to ‘work’, without sufficient attention being paid to understanding the results of early evaluation and using these to build more extensive trials or implementation studies.

A point made strongly to us in the interviews was that it can appear to be more attractive to policy-makers and NHS leaders to adopt a ‘shiny new’ innovation that lacks a deep and extensive evidence base, rather than exploring how to implement in a sustained manner less glamorous interventions already known to be effective and important for patients and carers, such as improving patient communication and administration;47,48 implementing multidisciplinary teams at scale so that all people living with complex multiple conditions can have such support when they need it;56 or ensuring meaningful regular medicines reviews with people and integrated communication with their team of health and social care professionals. 57

Context

All BRACE evaluations had at least one research question relating to patient or service user experience or impact in relation to the particular service innovation. It was therefore unsurprising that the analysis undertaken in our overarching study underlines the importance of this feature. This echoes the ‘engaged, informed individuals and carers’ dimension (or wall as it was termed) of the House of Care model, which is one of the two conceptual approaches underpinning this study. In exploring the theme of user engagement and self-management, we have focused particularly on two components of the SELFIE framework, highlighting what we have learnt from our overarching study of the BRACE portfolio of evaluations in the areas of technology and service delivery especially as they relate to integrated care.

A foundation of the House of Care is people with long-term conditions who are ‘fully engaged in articulating their needs and capacities, deciding on priorities, agreeing goals, and jointly developing a plan for achieving these’. 30 Such involvement is asserted by the developers of that model to lead to people having more confidence and engagement in their own care. Activities such as structured individual or peer support and education related to health conditions, social prescribing and health coaching may help patients to better manage their own conditions. Such enabling care is described by the Health Foundation58 as including ‘supporting people to recognise and develop their own strengths and abilities to enable them to live an independent and fulfilling life’ (p. 6). Yet Ham et al. 59 describe how far we are from the ideal scenario presented in the Wanless report,60 which outlined factors needed for a high-quality health service, where the public were fully engaged in taking responsibility for their health. The NHS Long Term Plan21 renews calls for increased support for shared responsibility for care management between providers and patients.

Our analysis of the BRACE portfolio of evaluations sheds light on the complexities of supporting self-management of care. Table 4 summarises points from the BRACE evaluations that, based on our cross-analysis, appear relevant to the issue of supporting people with multiple long-term conditions to self-manage their care. Thus, the evaluations of COVID oximetry at home5 and in care homes,6 which was intended to support people to self-manage in one specific regard (presenting at hospital with severe respiratory problems due to COVID), found that pulse oximeters were generally considered to be easy to use by people outside healthcare settings – many of whom would have been living with multiple long-term conditions. But the evaluation of AI plus home sensors in social care7 revealed service user concerns that the support for self-management that the technology might provide would be at the cost of less in-person social care, and they did not want that.

Building on these insights from the BRACE portfolio, one of our interviewees articulated the potential role of technology to support shared responsibility:

I would like to think that we’ll be able to use technology so that patients will be … able to monitor their own disease progression and maintenance much better. And actually hand back some of the responsibility for management to the patient. I think one of the challenges over the last few years is that we’ve tried to professionalise care, which has meant that we’ve actually disempowered people from looking after themselves.

Remote monitoring to enable care for people with multiple long-term conditions

This connection between the theme of enabling or self-managed care and the role of technology emerged from our cross-project analysis of the BRACE portfolio of evaluations (in particular pulse oximetry at home,5 social care sensors and AI7 and digital-first primary care8) and led us to undertake an additional rapid evidence review of remote monitoring to support care for people living with multiple long-term conditions. Box 4 summarises the findings from that review. We highlight in the following paragraphs what this literature reveals about how innovations such as remote monitoring can support the care for people with more than one long-term condition.

We found few studies that related specifically to remote monitoring for people with more than one long-term condition: only four studies in the UK met our eligibility criteria (see Report Supplementary Material 3 presenting the evidence review). There was some evidence in the literature assessed in our rapid review of remote monitoring innovations being a means to support self-management of care, improve mental health and well-being, and enhance quality of life for people with multiple long-term conditions. For example, in one study, participants with heart failure, uncontrolled hypertension and diabetes reported that a remote monitoring intervention that collected data via a phone-based app and Bluetooth-enabled weighing scales, blood pressure and blood glucose monitors made it easier to take readings and provided users with more oversight of their condition. 61

There was little evidence in respect of whether remote monitoring can improve clinical outcomes related to specific conditions or hospital utilisation outcomes. The Ware et al. study, for example, despite positive findings regarding patient empowerment, did not find evidence that the interventions had an impact on quality of life, mental health or physical health. 61 Such results call for more research specific to innovations for people with multiple long-term conditions, and research designed with the involvement of service users and carers. But such results also raise questions about how decisions are made to pilot or adopt technologies related to self-management. Are user benefits sufficient to justify adoption, or must such technologies show improvements in clinical outcomes or hospital utilisation measures?

Several studies in the literature we reviewed for this overarching study mentioned issues about health inequalities. They focused on inequalities in access to digital technologies and digital capabilities by age, geography, education or socioeconomic status. 62,63,69,70 People with particular vulnerabilities or health conditions may also face additional challenges in accessing and using technology. For example, a qualitative study of older adults with mild cognitive impairment found that although individuals report benefits from remote monitoring technology, such as increased feelings of security and independence, they reported wanting more educational material (including about technical issues and connectivity of devices) and more consideration of their specific needs (e.g. about forgetfulness). 62

Several of the studies we reviewed excluded people without smartphones, computers or internet access,63–65 which indicates the potential for remote monitoring to exclude certain populations that may already face issues with health inequalities. This represents a wider issue related to telemonitoring and limits the degree to which telemonitoring might close health equality gaps where vulnerable populations lack key facilitators of remote monitoring such as internet access. There are also selection biases in some studies, in that the people who participated in prospective studies had already agreed to use technology66 and so may be expected to have more favourable than average views of technology.

Applying lessons from the review of evidence on remote monitoring

Our rapid review of evidence on remote monitoring highlighted factors related to the design and use of technological innovations to support the care needs of people with multiple long-term conditions which were also present in our BRACE portfolio of evaluations. Several studies emphasised the need for greater patient and carer engagement in the design and development of software applications and digital healthcare interventions,63,67–69 which can help ensure that technologies meet user needs, are easy to use, and can be integrated into existing routines and healthcare pathways. These findings mirror what we learnt in the BRACE evaluation of digital-first primary care8 highlighted in Box 5.

People living with multiple long-term conditions may need additional support in using technology such as remote monitoring. We have already mentioned the presence of inequalities in access to digital technologies and digital capabilities by age, geography, education or socioeconomic status. 63,66–68 It is critical to account for how technological innovations such as remote monitoring have the potential for widening disparities in healthcare provision. Yet, in our rapid evidence review, several studies excluded people without smartphones, computers or internet access. 63–65

In one study within the BRACE portfolio, we evaluated the use of pulse oximetry within care homes6 (Box 6) and found that people with dementia or learning disabilities may require a different approach to using oximetry. Staff reported that occasionally people with dementia would be confused and not understand what was happening or become agitated in the process of taking an oximetry reading. While the evaluation focused on the usefulness of using pulse oximetry within care homes and did not explore the specific effects of the diagnostic on people with multiple long-term conditions, our findings may have some relevance for people with more than one condition and including cognitive impairments.

Finally, the use of remote monitoring technologies needs to be built proactively into care pathways from the outset. Many interventions included regular engagement between patients and healthcare providers, which provided both clinical oversight and person-to-person contact. 61,69–71 More input from healthcare providers in remote monitoring appears to be beneficial to user experience and increases the effectiveness of interventions bringing into question the potential for remote monitoring as a direct substitute for any in-person aspects of care. Kroenke et al. 71 suggest a stepwise approach to interventions, in which a more resource-intensive intervention can be deployed if self-management is not producing the desired outcomes, which can increase the cost-effectiveness of the intervention for the overall population.

This again mirrors findings from a study in the BRACE portfolio of work: the rapid evaluation of AI for digital social care work7 (Box 7). We concluded in that study that service users and frontline social care staff both needed to be involved in decision-making and implementation related to new technology. New roles and staff training to accompany technological implementation are needed if technological innovations are to help people living with multiple long-term conditions to co-ordinate and manage their own care.

Taken together, our findings from a rapid evidence review and cross-project analysis of the BRACE portfolio indicate that relatively little work has been done in respect of innovations specific to people living with more than one long-term condition. Though not specific to this population, the BRACE evaluations highlight the complexity of interventions to improve self-management and in particular the multiple considerations needed to be taken into account, perhaps even more so for people with multiple conditions, when embedding technology to support self-management activities. These include involvement of patients and their carers to develop innovations; considerations related to access for disadvantaged populations including populations with specific needs (e.g. learning disabilities or cognitive impairments); and the intricacies of implementation within complex organisations. While we see widespread recognition of the potential to better leverage technology to improve patient care, including that for people with multiple long-term conditions, we find that implementation of such technological innovations needs to account better for the breadth of sociotechnical issues that arise, including those related to broader systems, care pathways, and staff roles that interact with the technology.

Context

This theme directly relates to the ‘information and research’ segment of the SELFIE framework and is an important support to the ‘responsive commissioning of health services’ ‘wall’ of the House of Care model. If what gets measured is what matters,72 then the experience of people living with multiple long-term conditions and their carers clearly does not yet matter sufficiently within the NHS. Reflecting the single-specialty or -condition focus of how health services are organised, routine measurement of inpatient and outpatient activity, patient experience and outcomes usually happens within these condition-specific boundaries. Quality and finance reports from NHS trusts are typically organised around clinical services or directorates, with data set out for each department or specialty and not in respect of how patients – particularly those with complex or multiple long-term conditions – experience their care within and across different departments and specialties, and across into primary care, mental health, social care and other services. As Crump et al. 73 noted in their research designing and testing a survey to capture users’ experience of care co-ordination:

Understanding user experience is an important way for providers to assess how well the full package of care is meeting users’ needs. But measuring user experience of receiving integrated care from multiple sources can be hampered because of the limited availability of tools for capturing user and carer experience of care co-ordination across organisational boundaries.

Reproduced from Crump et al. 73 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, 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 above includes minor additions and formatting changes to the original text.

The context of siloed specialties and services has of itself led to this lack of available assessment tools that start from and focus on the actual experience of patients and carers making their way through the health system. The NHS Patient Survey Programme collects data on patient experience by organisation for five separate services (maternity care, adult inpatient, community mental health, children and young people’s services, and urgent and emergency care) and questions are focused on a single episode of care (e.g. on the adult inpatient experience that includes at least one overnight stay, and for adults using A&E department or urgent care centre services). Similarly, individual trusts usually administer their Friends and Family Test and other patient experience surveys by ward, department or clinical specialty. While this approach makes sense in respect of performance management of specific areas of a healthcare organisation, it is fundamentally based on a professional or institutional perspective of what needs to be understood, rather than being curious about the lived experience of patients and their carers across health and social care services and organisations and what works or not.

Measures of professional performance in primary care (and associated incentives or penalties) are similarly focused largely on single diseases and their treatment, as evidenced by the quality and outcomes framework (QOF) for GPs, where measures largely ignore combinations of diseases, care integration or factors that matter to patients living with multiple long-term conditions, as discussed further in Theme 5. It is likewise difficult to use routine NHS data to monitor the extent to which patients are able to have continuity of care from their GP, despite research evidence and guidance favouring continuity. 46,74

The importance of continuity of care as a measure that particularly matters to people with multiple long-term conditions was underlined by the interviews we undertook, where the metrics most frequently cited as being highly relevant were:

• continuity of care (especially in complex or poorly controlled conditions, not as important for people with fewer well-controlled conditions)

• experiences of care co-ordination, fragmentation and communication (e.g. ‘Do you think your clinician knew what was happening in other areas of your care?’, whether GP and specialist can both access a patient’s care records, having a comprehensive care and support plan across all services being used, and whether the patient’s preferences have been taken into account)

• impact of services received on carers.

One interviewee commented that a particular challenge of trying to collect information from carers about their experiences of service co-ordination or fragmentation is that the system is not always aware of their existence and even where their name and role are logged in patient records, little appears to be done with this information. The importance of the carer’s (and in this study, social carer’s) role in interpreting and profiling the importance of the patient’s lived experience (including care co-ordination and integration) was underlined by Reinhoudt-den Boer et al. 75 when reporting on a longitudinal study in the Netherlands of the care trajectories of people with multiple problems:

Formal caretakers seem to have a key role in initiating integrated care for this client group [people with multiple health problems]. This approach requires formal caretakers who can build strong trust relationships with CWMPs [clients with multiple problems], can organize shared reflexivity to unravel the complexity of CWMPs’ situations, and can take on supportive, compassionate and confrontational roles (coaching).

Reproduced from Reinhoudt-den Boer et al. 75 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, 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 above includes minor additions and formatting changes to the original text.

Other researchers have argued for a shift from examining care co-ordination or integration towards exploring care fragmentation, applying the logic that there is a need to start from understanding what is not working for patients and why not, before setting out to determine local or policy solutions. This turns much of the ‘integrated care’ literature on its head, challenging researchers to reorientate their attention, methods and analysis to the experience of care fragmentation (Nolte et al. ). 76 Crump et al. 73 developed this approach in the UK context, designing and testing a user-reported measure of care co-ordination and noting its uniqueness:

[this tool] will be able to offer some insights to purchasers and providers about how patients experience their efforts to coordinate services across boundaries of provision […] and also provide some information about how well services are supporting patients and service users in achieving their own life goals. In this respect, the survey tool has the potential to occupy a role not currently filled within an English context. 73

Reproduced from Crump et al. 73 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 above includes minor additions and formatting changes to the original text.

That such a tool needed to be developed and tested reveals the systematic lack of attention to the experience within and across health services of users who live with multiple long-term conditions. As revealed in our work for this report, so many measures in the health system are focused on single conditions, or sometimes clusters of conditions, such as the QOF in general practice, Primary Care Network Directed Enhanced Services specifications and measures, or the Getting it Right First Time programme focused on evidence-based improvements to NHS care.

BRACE evaluations cross-analysis findings

That there are few measures of care co-ordination and integration that are holistic or truly take the patient or service user perspective and are applied within policy and practice was echoed by our analysis of the 10 BRACE evaluations. It was only in our Type 2 evaluations (see Table 1) where an explicit focus was placed on the experience of people living with multiple long-term conditions that methods were designed to explore this. Examples here were the evaluation of the extent of any inequalities in access to Telefirst general practice by people with multiple long-term conditions,4 and the impact of hospitals managing general practice on the use of hospital and other health services by people living with multiple long-term conditions. 9

For the other BRACE studies, the perspectives and measures that were of interest to evaluation commissioners were almost always at the specific service innovation or organisational level, as with the early implementation of Primary Care Networks1 (e.g. their stage of development, impact on professionals, ability to establish new services required by the Primary Care Network contract), pulse oximetry at home5 (e.g. implementation of this addition to COVID-19 services, impact on hospitalisation, experience of staff supervising care) or Mental Health Trailblazers3 (e.g. impact on pupils’ mental health, parents/families and schools, mental health service provision as a result of the programme).

It is of note that in the BRACE team’s experience of rapid evaluation practice, it proved easier to do this using quantitative methods, and in particular analysis of routine healthcare utilisation data, than to adopt a more qualitative approach. This was evidenced in the evaluation of Telefirst general practice, where the national General Practice Patient Survey was analysed along with data from the national Understanding Society survey to triangulate dimensions of socioeconomic and other potential disadvantage with patterns of general practice access4 (see Box 3). However, in the BRACE evaluation of the impact of digital-first primary care on people with multiple long-term conditions and their carers,8 it proved so hard to secure ethical approvals and access to patients for the qualitative research within the timeframe of the rapid evaluation that professionals’ views of impact on users had to be adopted as a pragmatic and imperfect proxy approach to gaining these insights.

In the second phase of the hospitals managing general practices evaluation,9 the research team was able to interview patients as well as professionals, and then compare these insights with analysis of routine primary care experience data and secondary care utilisation data, reaching conclusions about how people with multiple long-term conditions experienced this form of general practice compared to other patients (Box 8). The results indicate that changes in patients’ use of acute hospital services when hospitals started to run the general practice where they are registered were similar regardless of how many long-term conditions the patient was living with; for example, there was a slight reduction in non-elective hospital activity for all patients. 9

Challenges of adopting measures that matter to patients and carers

The challenges faced by the BRACE team were mirrored by comments made in our interviews with national and regional stakeholders. For example, some interviewees were unaware of any data collected specifically for people with multiple long-term conditions, and certainly not any collected in a systematic or routine manner. Others were unsure what data could be collected to show if an innovation intended to benefit people with multiple long-term conditions was actually doing so. For example, one interviewee explained:

I am very sceptical about how we measure integration and how we monitor integration [...] In response to a single condition, you can probably track the impact of that through RCTs [randomised controlled trials] or other uh, you know, experimental designs to see how effective that innovation is. […] If you are looking at a pathway innovation as opposed to a tech innovation, as soon as you start to drop in multiple factors into an individual, it becomes incredibly difficult to find out what has been the single thing that has changed the outcomes for that individual or that system. And that’s part of the reason it’s so hard to measure it at an integrated level.

Other interviewees mentioned data that are collected that are not specific to people with multiple long-term conditions but which include and are relevant to much of this population, including access-related outcomes (e.g. attendances to services, number of contacts, duplications of contacts, referrals); health outcome data; patient experience feedback; and patient activation measures. Some interviewees commented that there were a lot of data already available, but these needed to be better linked to be able to track at a patient level and hence allow population health analyses.

An interviewee noted that one of the main challenges with existing NHS data is that there is no system identifier for multiple long-term conditions to make this patient group more tangible. Furthermore, it was asserted that while routine utilisation data within secondary care tend to be more readily accessible to researchers, many people with multiple long-term conditions receive much of their care from primary, community health and social care services, so broader improvements in data collection outside hospitals and linkage across sectors would be important for this population.

Suggested measures and approach

In the interviews we explored stakeholders’ thoughts about measures that matter to patients and carers. Patient-reported outcome measures (PROMs) were thought to be potentially important in monitoring the experiences of people with multiple long-term conditions. Examples given included quality of life, well-being measures, the degree of understanding of their conditions, and describing and assessing outcomes that matter to the individual patient (e.g. living and dying where you want, meeting goals that were set by the patient in collaboration with their informal and professional carers).

Clinical outcome measures (e.g. physical function, illness-specific dimensions of health) were also considered useful to collect as they can act as proxy measures to assess overall care. Other information suggested as worth considering included data examining health inequalities, staff measures related to a service innovation (e.g. staff satisfaction) and economic analyses. Two national-level interviewees commented that there is much quantitative data available already, but the use and triangulation of more qualitative data could improve understanding of impacts on patients and carers. This clearly echoes the finding from our cross-project analysis of the BRACE portfolio that it can prove more difficult to use qualitative (as opposed to quantitative) methods to explore the outcomes and experiences of people living with multiple long-term conditions.

The issue of measures is fundamental to understanding and researching integrated and co-ordinated care as well as the needs of people with multiple long-term conditions. The cross-analysis of BRACE evaluations highlighted the importance of measures that matter most to patients and service users, the themes of these being their experience of care; continuity of care; and the impact on carers. National Voices has developed and updated ‘I Statements’ from its work with patients, concisely capturing what it is patients want from integrated care. 77 Based on our cross-project analysis and stakeholder engagement work in interviews and workshops, we suggest in Box 9 the questions to consider in the development and use of measures that matter to patients living with multiple long-term conditions.

Stony ground

Service innovation in health and care refers to new ways of delivering services. This may be an entirely new approach or technology, or the application of an existing approach or technology in a setting, or for a patient group, where it has not previously been used. The focus of service innovation is on improving any or all of the following: health and care service quality, users’ experience of care, efficiency, and cost-effectiveness. Innovation is a process of establishing, implementing, sustaining, scaling and spreading the new approach or technology before any outcomes are achieved. The success or otherwise of an innovation is inevitably affected by the context(s) into which it is introduced. An innovation landing on stony ground, that is in a difficult context, will not be taken up by service providers and users. 78 Marjanovic et al. 79 identify six groups of key contextual factors influencing innovation in health care, as listed in Figure 2.

FIGURE 2.

Key contextual factors influencing innovation in health care. Reproduced from Marjanovic et al. 79 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, 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 figure above includes formatting changes to adapt the original text into a figure.

Copyright © 2024 Sussex et al.

Weak or absent drivers of innovation under these headings represent barriers to innovation of all kinds in the NHS including, but not limited to, that which might benefit people living with multiple long-term conditions. The lack of sufficient or appropriately skilled staff with time to familiarise themselves with and adopt innovations, let alone help to create them, is a major constraint, we have found. 80 This may be manifested in too few healthcare staff relative to patient demand for their services, or the available skill mix not matching service needs. In our interviews, the point was made that heavily loaded healthcare professionals are likely to lack the ‘head space’ to think about innovating and changing how they work in order to better support patients with multiple long-term conditions. Multidisciplinary teams are desirable to support many patients, but especially those with multiple conditions. Our stakeholder engagement work highlighted that the staffing level and mix needed to build multidisciplinary teams, and even the physical space in clinics, etc., to house such teams, may be insufficient.

Overall, the overarching study has revealed several barriers to innovation or to the implementation of innovation for people living with multiple long-term conditions. Some of the barriers cut across more than one of the innovation factors identified by Marjanovic et al. ,79 but others relate to single factors, for example, a lack of opportunities for active PPI and engagement with innovation. The degree to which barriers are specific to service innovation for people with multiple long-term conditions or obstruct health service innovation more generally also varies. A lack of drivers of the kind listed above is a barrier for any innovation. In the following paragraphs, we concentrate on those barriers that seem specific to, or at least particularly obstructive to, innovation for people with multiple long-term conditions.

Focus on single conditions

As discussed earlier in this report, the healthcare system and training of healthcare professionals to work in that system are designed around treatment pathways for single diseases or conditions. However, it may be that clinicians are less likely to refer to other services needed by a patient with multiple conditions if there is no clear pathway to do so. Furthermore, patients’ own expectations may reflect an understanding of the care on offer that is focused on single conditions. It was suggested by a PPI participant that patients may lack the confidence to press for multiple conditions to be considered at the same time when consulting a healthcare professional who evidently wishes to focus on one condition at a time.

Faced with clinical and patient expectations like these, there may seem to potential innovators to be little reason to seek an approach that is more accommodating of multiple long-term conditions. Indeed, such an approach may not be considered at all. An example of how a lack of attention to multiple long-term conditions appears to have been a barrier to service innovation for this population was revealed in the BRACE evaluation of digital-first primary care8 (see Box 3) where we found that digital-first primary care algorithms suited people with a straightforward single condition issue but did not allow for wider contextualisation by the patient.

The review of evidence on evaluations of remote monitoring technologies for people with multiple long-term conditions demonstrated that there are nevertheless a few innovations that are aimed at this group (see the discussion in Theme 3). The barrier of single-disease focus can evidently be overcome. In the next thematic section of this report below, we consider how the priority accorded to research for the large part of the population living with multiple long-term conditions can be augmented.

Lack of a clear understanding of what matters to people living with multiple long-term conditions

It is widely accepted that patients’ needs should be the focus of innovation in health care, and that it is essential to involve patients and the public in developing innovations, but it is also recognised that this does not always happen. 81 A national-level stakeholder we interviewed reinforced this in respect of health technology innovation. They expressed the view that it can be difficult to identify innovators willing to work with the healthcare system to create and design technologies in response to system and patient needs, rather than developing a product and then finding a healthcare need that product might address.

None of the innovations that BRACE was commissioned to evaluate were specifically aimed at people living with multiple long-term conditions. However, our rapid evidence review found several studies that described steps that were taken to meet the needs of people with multiple long-term conditions by including them and their carers in the design process. For example, Doyle et al. 82 describe a process of consulting with patients, as well as clinicians and other stakeholders, to design bespoke features for a remote monitoring intervention that would cater specifically to people with multiple long-term conditions. Through this process, Doyle et al. found that patients may be prone to focusing on one condition at the expense of others, which led them to implement a ‘flower’ user interface, presenting multiple facets at once to provide unobtrusive prompts for conditions that were not being monitored. Two studies also discussed the need to include a patient’s entire care network when designing remote monitoring interventions, with some providing an option to share data with carers and family members. 62,82 Other studies emphasised the need for greater patient and carer engagement in the design and development of apps and digital healthcare interventions to help to ensure that technologies meet user requirements. 63,67–69

In the interviews we conducted, participants mentioned innovations they felt to be relevant, but rather than being specifically aimed at people with multiple long-term conditions the innovations merely happened to be used predominantly by those people. The innovations mentioned were rarely discussed in relation to the specific care needs of people with multiple long-term conditions, unless prompted by the interviewer.

Failure to sustain implementation of innovations beyond pilots

Challenges relating to the adoption of innovations have been long known and much researched. 83,84 Such research includes highly diverse literature on implementing and sustaining innovations, and points to the ‘iterative, complex and multidirectional’ process of implementation whereby both formal and informal decisions by individual users comprise the ongoing process of assimilating innovations within organisations. 85 Understanding the innovations themselves, as well as under what conditions a service innovation may or may not be successful, may be critical to its scale-up and spread within and across organisations.

After prompting, interviewees mentioned a number of innovations that were considered to have been targeted at, or had proved useful to, people with multiple long-term conditions. These innovations are listed in Table 5 alongside the care need(s) they address. This is not intended to represent a comprehensive list of recent innovations beneficial for people with multiple long-term conditions but is indicative of what these expert informants to our overarching study actively had in mind. This suggests that well-evidenced service innovations suitable to meet the needs of people with multiple long-term conditions exist. But they are not yet implemented at sufficient scale or with sustainable funding across the NHS. Shaw et al. 86 echo this when noting that, while health and care technologies may often succeed on a small scale, they often do not achieve widespread use or become routine practice in other locations or settings. Thus, a major challenge is to progress beyond pilots (which may themselves have multiple purposes), to find ways to roll out and sustain across the NHS long term service change to benefit people with multiple long-term conditions. 87

Reviews of healthcare innovations from the implementation science literature highlight the multilevel factors that impact their implementation. 83,88,89 In their review of evidence-based health innovations, Chaudoir et al. 89 classify implementation factors at the innovation, patient, provider, organisational, and structural/environmental levels. Robert et al. ,83 in their review of implementations of technological innovations in the UK context, note the importance of various contextual factors in implementation, such as the role of power and politics, that is the dominant role of clinicians in advocating for adoption/implementation and the quality of interprofessional relationships in implementations of technological innovations in the UK context. 83,88,89

Financial incentives focus on single conditions

Financial incentives for providers of secondary care in the NHS are focused on a combination of overall budget constraint, via block payments, and incentivising increased activity in the form of individual episodes of specialty-based care, via case mix-adjusted activity-based payments. 90 There are no specific incentives for meeting the needs of patients with multiple conditions. In the absence of such incentives, the drive to innovate services to help those patients is weaker than it need be.

In NHS primary care, the financial arrangements for GP practices reward larger patient list sizes, the quantity of some condition-specific services (such as vaccinations) and, by means of the QOF, the achievement of a long list of pre-specified service quality targets. All the indicators that are used in the QOF to determine payments to GP practices concern the provision of individual services and none is related to addressing needs resulting from multiple long-term conditions (with the possible exception of the palliative care indicator: ‘the contractor establishes and maintains a register of all patients in need of palliative care/support irrespective of age’). There are no indicators of care co-ordination or continuity of care across conditions, for example. Box 10 lists the categories of indicators under the QOF used to incentivise GPs: all focus on single conditions. 91

Barriers in summary

The range of barriers to innovation aimed at helping people living with multiple long-term conditions is broad. Nevertheless, there are such innovations – for example, as revealed in our focused review of evidence on remote monitoring for patients with multiple long-term conditions – but they are relatively rare given the large number of people living with multiple long-term conditions. Barriers to innovation arise from contextual factors in the healthcare system. As discussed earlier in Themes 1 and 2, there is a general tendency in healthcare systems to focus on single conditions, and hence the problems of living with multiple simultaneous conditions are relatively overlooked. An insufficient focus on gaining a clear understanding of what matters to people contributes to that relative neglect. We find, unsurprisingly, that finance and resources also have an influence: successful innovation requires sustained devotion of resources into the future, not just during initial piloting; and financial incentives do not yet stimulate innovation for people living with multiple long-term conditions.

In the next and final theme, we reflect on how such barriers might be overcome, by making the needs of people living with multiple long-term conditions a priority in planning and delivering health care.

Explicit focus at all levels in the NHS on people living with multiple long-term conditions

In the last 10 years, NHS policy in England has been significantly influenced by two national policy documents: NHS England’s ‘Five Year Forward View’92 and the ‘NHS Long Term Plan’. 21 The ‘Five Year Forward View’ led to the piloting of five types of ‘Vanguard’ models of care (primary and acute care systems, multispecialty community providers, urgent and emergency care, acute care collaboratives and enhanced health in care homes) across a total of 50 locations in England. Local evaluations were funded for each Vanguard scheme and the NIHR Policy Research Programme commissioned a national evaluation. MacInnes et al. 93 concluded from the national evaluation that only the Enhanced Health in Care Homes model had been scaled and spread across health services in England, despite extensive spending on the setting-up and piloting of the Vanguards, and their overall assessment underlined the challenges of sustaining innovative care models beyond a pilot phase:93

Our study suggests that the general scaling up of the Vanguard programme was difficult to achieve. We found little evidence that the service changes introduced locally had any significant influence on subsequent service developments. Lack of leadership or direction near the end of the Vanguard programme, lack of ongoing resources, and limited success in providing real-time monitoring and evaluation may all have contributed to the failure to straightforwardly scale and spread most of the Vanguard models.

Reproduced from MacInnes et al. 93 in accordance with the terms of the CC BY 4.0 license (https://creativecommons.org/licenses/by/4.0/), which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access page (https://us.sagepub.com/en-us/nam/open-access-at-sage). The text above includes minor formatting changes to the original text.

Some of the Vanguard models focused on care integration, and there were examples within these of people with multiple long-term conditions being a particular focus because of the complexity of their needs and because, in the most severe cases, of their exceptionally high use of healthcare services. One of these examples is the ‘Greenvale’ case study site in the BRACE phase one evaluation of hospitals running GP practices,2 which had, prior to the start of that ‘vertical integration’, been a pilot site for a ‘Five Year Forward View’ ‘Vanguard’ new model of care. In that example, care was reorganised for patients with complex and multiple conditions who were the heaviest users of healthcare resources.

In the words of the ‘The Government’s 2022–23 mandate to NHS England’, which at the time of writing is the most recently published annual set of objectives from the UK Government for the NHS in England: ‘The NHS Long Term Plan continues to be the guiding strategy for the NHS’ (paragraph 2.8). 94 The NHS Long Term Plan21 recognises at one point (on page 100) the contribution to rising NHS costs of people with multiple conditions, but otherwise all discussion of care provision is in terms of single disease areas – cancer, cardiac, etc. – with no proposals or strategies set out to better tackle the care issues raised by living with multiple long-term conditions. The Government’s mandate to NHS England makes no mention of serving people who are living with multiple long-term conditions, and none of the metrics it is using to measure NHS progress refers to them. 94

Thus, the policy context within which the NHS in England operates does not currently give prominence to promoting or supporting the care of people with multiple long-term conditions, and it is within this context that BRACE evaluations have been undertaken. At an operational level, the NICE has published a clinical guideline (NG56) for the NHS on ‘Multimorbidity: Clinical Assessment and Management’. 12 This guides healthcare professionals to:

Consider an approach to care that takes account of multimorbidity if the person requests it or if any of the following apply:

NICE [2016, paragraph 1.2.1; © NICE (2016) ‘Multimorbidity: Clinical Assessment and Management’. Available from www.nice.org.uk/guidance/ng56. All rights reserved. Subject to Notice of Rights. Reproduced in line with the licence (www.nice.org.uk/re-using-our-content/uk-open-content-licence). The text above includes minor formatting changes to the original text.

NG56 makes it clear that multimorbidity should not be overlooked. But, to put this in context, there are hundreds of NICE guidelines focused on clinical assessment and management of single conditions. Given the thousands of combinations of long-term conditions that are possible, it is not reasonable to expect guidelines to focus on specific combinations of conditions beyond those that occur frequently. Stokes et al. 95 found no clear high cost (in terms of hospital use) combinations of long-term conditions as possible targets for intervention. Using NHS data from England, they found 63,124 unique combinations of conditions for hospital patients with multiple long-term conditions in 2017–8; only seven of these unique combinations contributed any more than 1% of total cost of secondary care for patients with multiple morbidities, and the highest share of hospital costs for a single combination of conditions was 3.2% (diabetes plus hypertension). So, rather than guidelines for particular combinations of conditions, we propose that what might be more feasible would be for guidelines focused on single conditions to also prompt healthcare professionals always, by default, to consider whether the patient has other conditions and how care might best be delivered in the light of that. The need for this can be seen from the work of Guthrie et al. ,96 who investigated how multimorbidity was accounted for in three exemplar NICE guidelines for common long-term conditions (type 2 diabetes, depression and heart failure). They found that: ‘[c]omorbidity was rarely accounted for in the clinical research questions that framed the development of the exemplar guidelines and was rarely accounted for in treatment recommendations.’96

In the interviews we undertook, several stakeholders, both at national and at regional (Integrated Care System) level, commented that people with multiple long-term conditions represent a large proportion of patients and acknowledged that they have particular care needs, for example, for personalised and holistic care, and for care to be communicated and co-ordinated across conditions, services and sectors. But, in the words of one interviewee: ‘some of the national frameworks just don’t align to help the incentives that need to happen to make multimorbidity more joined up at the local level’.

National interviewees were asked specifically about their thoughts on the role of the (regional) Integrated Care Systems and Integrated Care Boards in driving service improvement for people with multiple long-term conditions. Answers predominantly concerned setting a strategic vision for integrated working and linking services together with joint funding that was collaborative rather than competitive, and sharing learning across the locality. Respondents underlined the importance of robust needs assessments for people with multiple long-term conditions with services commissioned and provided based on these assessed needs. The provision of funding and support to foster innovation for people with multiple long-term conditions was thought to be part of the role of new Integrated Care Systems. Indeed, interviewees at Integrated Care System level referred to examples of a regional initiative to fund a ‘complex care service’ and another to develop ‘personalisation’ of care. The value of national support and impetus for innovation was evident from the findings of the BRACE evaluation of WHHs,10 for example. In that study, interviewees highlighted that support at a national policy level had helped to drive change and develop and embed WHHs but that there was, unhelpfully, a lack of associated national funding or guidance for the policy (at the time of fieldwork).

Service design with patients and carers

Our interviews elicited that Integrated Care Systems and the service providers within them need to be clear about what it is they aim to achieve for people with multiple long-term conditions. When asked about what a system for service development that prioritises people with multiple long-term conditions would look like, a national interviewee proposed that there would be frameworks for decision-making developed with NICE, NIHR and the Royal Colleges, and with a lot of patient and carer involvement. This interviewee drew a parallel with work led by NHS England to produce consensus guidelines for PEoLC. 97 Palliative care focuses on the patient, not the individual conditions they may be living with, and as such can be seen as an example to learn from. Integrated Care Boards have a statutory duty to commission palliative care services; guidance on how they are to do this, including links to resources and good practices, has been developed by NHS England to support that statutory duty. Indeed, Integrated Care Boards are required by statute to take the NHS England guidance into account and that guidance in turn requires that:

Every ICB [Integrated Care Board] should commission PEoLC [Palliative and End of Life Care] services that meet people’s needs, aligning to the commitments within the Ambitions for Palliative and End of Life Care: A national framework for local action 2021–2026. 97

page 5

The framework referred to in the statutory guidance was developed and published jointly by 34 organisations, including a large number of patient and carer charities along with Royal Colleges, health and social care professional and provider associations, and NHS England, Health Education England and Public Health England. 98 It is notable that the first of the six ‘ambitions’ that make up the framework is that ‘each person is seen as an individual’ and that another ‘ambition’ is that ‘care is co-ordinated’. Box 11 summarises what we see as the key elements of successful co-design and promotion of such guidance. Providing guidance co-designed with patient and carer groups, even when backed by a statutory requirement to follow it, does not guarantee that the desired care will be available. But it is clear about what is wanted. Similarly, clear and forceful guidance developed with strong patient and carer input might well help to give priority to serving people with multiple long-term conditions.

At the level of an individual service, as distinct from an overall approach to be taken to providing health care in an area, some of the studies we found in our evidence review of remote monitoring describe steps being taken to meet the needs of people living with multiple long-term conditions. For example, Doyle et al. 82 describe a process of consulting with patients, clinicians and other stakeholders such as experts and academics, to design a remote monitoring intervention that would cater specifically to people living with multiple long-term conditions. This contrasts with what BRACE researchers found in the evaluation of AI with home sensors in social care,7 where it was evident that different stakeholders were seeking to resolve a number of potentially different problems and seeking a number of different outcomes, with an apparent lack of a shared understanding of what success would look like.

Another study in our evidence review highlighted a way to gather feedback from people living with multiple long-term conditions, their carers and healthcare providers. Craven et al. ,68 in their qualitative study to inform the development of a remote measurement technology, used a method they called ‘universal points of care (UPOC)’ as a way to elicit and analyse care pathways using scenarios. This method includes asking patients and clinicians how they would go about sharing data, communicating and selecting treatments across different conditions. We consider that similarly inclusive approaches to planning provision of other healthcare services would help lead to care that better supports people with multiple long-term conditions.

Financial incentives

The subthemes of explicit focus and the co-designing of services with patients and carers, presented in the preceding paragraphs, respond to some of the barriers to developing and implementing appropriate services that were set out in Theme 5. These barriers concerned the single-condition focus evident in NHS organisation; the training and ways of working of healthcare professionals, and even the expectations of patients; and the lack of a clear understanding of what matters to people living with multiple long-term conditions. We also noted within Theme 5 that interviewees had pointed out that financial incentives to primary care and secondary care providers in the NHS are aligned with care for single conditions, and that none reward co-ordination of care across conditions for someone living with more than one of them.

We propose that incentives that signal the importance of measures to support specifically patients with multiple long-term conditions would help direct more of the scarce management time in secondary care (where funding goes to the provider organisation, i.e. the NHS trust or NHS foundation trust) and GP/practice management time (where funding goes to the practice) towards that end. For example, specific QOF indicators for care co-ordination for patients with multiple conditions, and hence associated payments to GP practices, might help to overcome the very practical difficulties of giving more staff time to achieve that. The existence of such incentives would also signal the seriousness with which co-ordinating care for patients with multiple long-term conditions is taken.

We note that while financial incentives might prove helpful, they do not appear to be a necessary condition for better serving patients with multiple long-term conditions. The BRACE evaluation of WHHs10 – see Box 2 in the discussion of Theme 2 – found that where these hubs had been established, this was due to local initiative rather than a specifically funded or financially incentivised development. WHHs exist to co-ordinate care for women for their reproductive health, which although different from the issue of multiple long-term conditions has parallels with that, particularly concerning the desirability of better co-ordination of care and taking account of wider health needs and inequalities. The experience with WHHs implies that improved care co-ordination is prioritised in some places (and for this patient group), without there being targeted funding or incentives.

Challenging the narrative that multimorbidity is inevitable

Reflecting the nature of the portfolio of service innovations that BRACE has evaluated, our attention in this report is mainly on how healthcare services are, or could be, provided. But achieving priority for patients who are already living with multiple long-term conditions may need to be associated with making ‘multiple long-term conditions’ explicitly part of the public health agenda by including strategies and activities aimed at risk identification – underpinned by data analytics – and prevention, or at least delayed onset, of people acquiring multiple long-term conditions. More prioritisation of, and hence investment in, primary prevention and secondary prevention would, in the words of one of the national-level stakeholders we interviewed, ‘challenge the narrative that multimorbidity is inevitable when actually so much of it isn’t’. Such prioritisation and investment would at the same time be fully aligned with the desire to reduce health inequalities, as people in more deprived groups of the population acquire multiple long-term conditions years younger than do others. 14,50,99

Research team

The research team for the overarching study was a small, mixed-methods, multidisciplinary group (applied health services research, health services management, health economics, statistics). Team members included two senior researchers, two mid-career researchers and a number of early career researchers. The research team had a majority of female members. All team members’ contributions have been recognised, and all were given opportunities to attend analysis workshops. Throughout the overarching study and as an integral part of our research method, the research team held frequent meetings to discuss project management, data collection, analysis and write-up, with additional workshops to shape the interpretation of findings.

Wider involvement

Throughout the evaluation, the team sought the advice of external academic experts experienced in aspects of the study (‘critical friend’ colleagues at the University of Birmingham and RAND Europe who were not otherwise involved in this study), the BRACE Steering Group and members of the BRACE PPI Panel.