Integrated systems: Population health management

Published date
Last updated
Authors
David Attwood
Sarah Zaidi
Topic tags
Commissioning
Demographics
Digital Health
Frailty indexes
Integrated care
Transfer of Care

This resource outlines key considerations to guide systems wishing to use Population Health Management (PHM) approaches to support effective, frailty-attuned proactive and personalised care. It aligns with the three shifts as set out in the Ten Year Health Plan for England, although the themes will be relevant to colleagues from across the four nations. The document provides top tips for using a PHM approach and case studies of areas successfully using these methods. This resource has been developed by Dr David Attwood and Dr Sarah Zaidi on behalf of the British Geriatrics Society.

Introduction and top tips

This first section aims to help colleagues to understand what Population Health Management is and why it is relevant to frailty care. This section also provides top tips and principles for systems wishing to implement a Population Health Management approach to frailty care. 

What is population health management?

Population Health Management (PHM) is a data-driven approach to improving health outcomes and reducing disparities in a geographic area. PHM aims to move from a largely reactive system that responds when an individual becomes unwell towards a more proactive system, aiming to prevent illness, reduce the risk of hospital admission and address inequalities. It involves:

  1. Identifying high-risk populations

  1. Segmenting them into actionable cohorts

  1. Delivering tailored, evidence-based interventions

PHM is delivered as a partnership between healthcare and other public services including local authorities, emergency services, the voluntary sector, housing associations, social services, patient groups, communities and the public. PHM uses healthcare data from clinical systems as well as non-healthcare data from sources such as local authorities and social services. By collecting and analysing data, population groups can be identified and the needs of those groups and individuals can be anticipated. 

The aim of PHM is to enable systems to focus their providers and local services to act as early as possible to keep people well, targeting the right type of support and evidence-based interventions where it will have the greatest impact. Primary care, and other community-based services, have a unique understanding of the needs of the communities they serve, and PHM seeks to build on this insight. Populations living with frailty are commonly in most frequent contact with community health services, primary care services, adult social care, care sector providers and often adult mental health (including dementia) teams. 

Standardised datasets and systematic use of data can support ICSs, and in turn their local providers, to understand the needs of their local population on an ICS-wide footprint covering population sizes of up to one million people. In addition, this significantly strengthens effective neighbourhood model working, both across multi-neighbourhood place footprints covering population sizes of approximately 250,000-500,000 people and in smaller neighbourhood localities covering population sizes aligned to Primary Care Network (PCN) footprints. 

Segmentation of the population with frailty and possible interventions

Figure 1. Segmentation of the population with frailty and possible interventions

Frailty and population health management

Frailty is a growing health concern in the UK and is increasingly acknowledged as a priority in national strategy and publications. Reports such as the Chief Medical Officer for England’s Health in an Ageing Society1 and the Health Foundation’s Health in 20402 emphasise the urgency of redesigning services for older adults with complex needs.

Statistics bear this urgency out – half of hospital beds are occupied by people aged 75 and over and 75% of these have been diagnosed with frailty. According to latest national hospital data tools using the Hospital Risk Frailty Score (HRFS) to identify hospitalised patients with frailty across England, up to 70% of hospital beds may be occupied by people with frailty. One in 25 of those with mild frailty who are admitted to hospital will die in hospital. For people with moderate frailty, 1 in 17 admitted to hospital will die in hospital and 1 in 5 of those with severe frailty. 1 in 7 people with frailty will be readmitted to hospital within 28 days of being discharged. 40% of those aged over 75 who are admitted to hospital are in their final year of life and a third of them will spend more than a month in hospital.

Prevalence of frailty is not evenly distributed across England. Coastal and urban areas, which are usually more deprived, have a higher prevalence of frailty than inland rural areas which are usually more affluent. Higher deprivation drives higher multimorbidity, causing increased prevalence of earlier onset frailty in populations.3

Frailty significantly drives demand and cost, consuming substantial NHS workforce and resources. It is, however, preventable and modifiable. Early identification and coordinated care can improve quality of life, reduce admissions and address inequalities.

Frailty is an ideal candidate for a PHM approach. Defined as a state of reduced physiological reserve, frailty can be prevented and reversed, when identified early enough. Frailty exists on a spectrum – from mild to severe – and therefore people on the spectrum are identifiable and the population with frailty can be segmented. Because frailty is both preventable and modifiable, there are a range of evidence-based interventions available across health and care systems.

PHM for frailty pursues two objectives:

  • Prevent or delay the onset of frailty through healthy ageing interventions

  • Manage established frailty to halt, delay or reverse decline.

The segmentation of the population with frailty and possible interventions are outlined in Figure 1. 

Through all tiers of mild, moderate and severe frailty, holistic personalised assessment can support the delivery of beneficial interventions such as:

  • Multifactorial falls risk assessment (FRA)

  • Structured medication review (SMR)

  • Personalised care and support planning (PCSP) which may include advance care planning (ACP)

  • Shared decision making (SDM).

 

Case study: Population identification and segmentation in Plymouth with the Pathfields Tool

Systems cannot implement effective PHM strategies or drive these consistently at scale if they cannot easily identify people in their populations at risk of living with frailty (includes all living with dementia), nor stratify frailty severity (stage) accurately in the first place. This is essential to proactively target evidence-based interventions, to those who may benefit most.

The Pathfields Tool3 is a clinician-led solution embedded within primary care IT systems (SystmOne). It is applied across the registered population aged 65 years and over. Within this 65+ population, individuals are identified as being at higher risk of undiagnosed frailty if they have one or more of the following characteristics:

• Age 90 years or over
• A recorded diagnosis of dementia
• Residence in a care home
• Housebound status or a recent home visit recorded
• Observed mobility problems during consultations

When a patient in one of these groups consulted a clinician, saving their record triggered a prompt to code frailty status using four discrete categories: not frail, mild, moderate, or severe.

The Pathfields Tool utilised the pictorial representations from the Rockwood Clinical Frailty Scale to guide the assessment, supplemented by the clinician’s longitudinal knowledge of the patient:

Pathfields%20tool.png

Published early comparative data in a single practice (n=4,552 65+ year olds) demonstrated the Pathfields Tool identified frailty in 75.9% of cases, compared with eFI, which detected it in 47.3%.3 The tool was used effectively by multiple clinicians and feedback was positive, noting that the tool was easy to use and did not add extra time to their daily routines.

In addition, the tool systematically identified mild frailty in the practice population, enabling the deployment of earlier interventions that may delay, halt, or even reverse progression.3,[i],[ii],[iii] The Tool has been deployed to all SystmOne practices in the Plymouth area.

However, it is important to acknowledge the tool's limitations:

  • Published data was based on development and testing within a single setting. Further research in diverse clinical environments is needed to validate its effectiveness and broader applicability in frailty identification and progression tracking.

  • It is only available for use in SystmOne.

Case study: Mid and South Essex Athena PHM Data tool

To overcome common issues and challenges that often hamper frailty identification and segmentation in populations, Mid and South Essex ICS used linked aggregated datasets from multiple providers across the system including primary care, all community providers, acute hospital, mental health services and social care. This also included incorporating data about wider determinants of health (such as BMI, smoking status, social isolation, alcohol misuse and polypharmacy) as well as age, ethnicity, gender and deprivation deciles

The key development with the data-gathering methodology for population segmentation in Mid and South Essex, was moving towards gathering linked aggregated datasets from multiple sources across all parts of the health (including emergency services), mental health and social care sectors, therefore capturing a more inclusive and comprehensive breadth of datasets, from all providers, teams and services system wide.

This has helped to overcome previous pitfalls causing inaccurate estimates (often under-estimations) of frailty within populations, caused by large gaps in data feeds.

This enabled the ICS to segment its whole 1.34 million population into 10 different segments – to more accurately identify its frailty population and stratify risk. It revealed that around 118,000 people across the ICS are likely to be living with frailty or dementia or approaching the end of life (just under 9% of the whole population).

To identify people with frailty, the datasets assimilate coded dataset markers gathered from all providers and sectors (health, social care, mental health) of the system such as:

  • Clinical frailty scores (CFS)
  • Other mild, moderate and severe coded frailty stages from provider record systems
  • Social care package needs, care home resident status, supported living, having an unpaid carer.
  • All dementia diagnosis coded datasets, cognitive impairment/ cognitive decline codes, coded episodes of delirium, acute confusion as well as recorded persistent functional impairments in activities of daily living (ADLs)
  • Risk of falls, previous falls, fragility fractures
  • High frailty prevalence long term conditions such as heart failure, COPD, Parkinson’s disease, Stroke, peripheral vascular disease.
  • Other End of life risk proactive indicators (such as recorded Gold Standard Framework (GSF)stage status and other long-term condition/ disease specific End of life (EOL) proactive indicator symptom/sign codes).

Segments 7 to 10 are those most relevant to people living with frailty.

  • Segment 7 (complex cases) aligns to all who are at rising risk of frailty, but who do not have any very mild frailty as yet, which includes groups such those with mild cognitive impairment, those with learning disability, in the absence of any other phenotypic frailty markers.
  • Segment 8 includes all with markers that would indicate very mild, mild and moderate stages of frailty or dementia, in the absence of any other (end of life proactive indicator) markers, that would put them at greater risk of being within last year(s) of life.
  • Segment 9 (End of life) includes all with severe and very severe stages of frailty or dementia or any adults with other end of life clinical or symptomatic indicators (who may not have any markers of frailty)
  • Segment 10 (Mortality) includes all who actually died

The data is viewable and can be interrogated, either ICS wide or broken down by place geography, local authority footprint, by PCN neighbourhood locality and even down to individual GP practice registered population. 

Advantages of this model include:

  • It detects segment drift to understand which populations move across different segments, enabling development of predictive modelling strategic approaches to proactively target populations for preventative interventions.
  • It shows service utilisation and demand across all sectors and services for each population segment.
  • It is accessible to all providers (including system leads and commissioners). This has widened the use of PHM data insights beyond just primary care and into wider health, mental health and social care sectors across the whole system too.  This helps to mobilise a system-wide strategic and operational effort towards PHM. 

Impact:

  • Actionable PHM data drove significantly increased frailty scoring ICS Wide (more than 20,000 more patients with CFS scores of 4 or above identified) with many more receiving proactive care interventions within 2 years
  • Collaboratively delivered by adult community health teams (including virtual ward and hospice services), community dementia teams and primary care worked more efficiently together to share the delivery, in a far more effective united effort.
  • A clear blueprint proactive care framework for neighbourhood delivery was set across all 25 neighbourhoods in the ICS. (See Case Study three)

More information about the Athena Platform PHM data tool development can be found at: https://www.ardengemcsu.nhs.uk/showcase/news-events/news-events/athena-improves-care-for-people-with-frailty-dementia-and-life-limiting-conditions-in-mid-and-south-essex/

Key considerations of frailty-attuned population health management

This section sets out key considerations that systems looking to implement a population health management approach should take into account when planning such an approach.

  1. To identify and stratify adults living with frailty at population level, systems should implement fit-for-purpose population segmentation using aggregated linked datasets from all providers and services across health, mental health, social care and VSCE sectors across systems. This improves data gaps and accuracy for stratifying fluid populations with frailty, enabling innovation towards predictive modelling strategies, better dynamic detection of frailty stage transitions, to improve targeting the right preventative and proactive approaches for each person.
  2. To improve practitioner validation of frailty stage, systems should embed use of validated frailty staging tools (e.g. CFS) from all providers, teams and services, system wide. Frailty identification and staging using validated tools (such as CFS) must be conducted routinely as part of the expected standard for the assessment and management of older people with frailty in both routine and urgent care and ideally mandated for all providers and teams across the system. Including for acute hospitals, all community-based services, primary care, mental health, social care and VCSE organisations including hospices. Frailty stage diagnostic codes must be recorded in all electronic health record systems to improve data capture.
  3. To improve consistent frailty proactive care delivery, systems should embed a clear consistent proactive care framework for all levels of frailty for all providers. This would help to drive fit-for-purpose PHM by all system partners. All the highest impact evidence-based, preventative and proactive interventions must be embedded within all provider contracts. This should be accompanied by a frailty-specific minimum dataset of mandated key performance indicators (KPIs), encompassed within a national performance reporting framework.
  4. Systems should develop integrated models to drive high quality shifts from hospital to community. This is key to better outcomes for the population with frailty, the workforce and the system. These need to be led by community services with neighbourhood health models at the heart. The current separate frameworks, contracts and models for urgent and emergency care (UEC) must unite to become one efficient, higher value model, aligned to and embedded within developing neighbourhood health models.
  5. Systems should develop interoperable shared IT infrastructure across all key providers/teams and services, using shared digital frailty-specific proactive care delivery tools that can share CGA, including ACP datasets, live in real time and in chronological patient journey (patient story) order, accessible to and used by all teams. The system-wide use of these tools can dramatically improve impact, reducing unplanned hospital admissions within one to two years. They improve the visibility of a fluctuating population of people with frailty and improve care coordination as well as helping to unite and align efficient shared capacity in proactive care delivery of evidence-based interventions. The same shared digital record tools should be used by both front door frailty and same day emergency care (SDEC) hospital teams and all urgent care teams in the community, as well as in routine neighbourhood care. This would enable system-wide delivery and reporting of proactive care KPI metrics.
  6. Systems should prioritise dedicated system wide workforce frailty education and training with training delivery strategies for all sectors. This is essential to increase required frailty knowledge, competencies amongst more of the workforce, in all parts of health, adult mental health and social care and to develop more skilled and confident front-line staff and practitioners, especially with required Tier 2 and Tier 3 frailty skills. Training should also focus on promoting a cultural shift away from reactive care towards personalised and proactive practices. This would help to improve workforce competency, confidence and culture change in all frontline staff, across all sectors, adult specialties and service directorates.
  7. Systems should develop comprehensive frailty-specific proactive care performance dashboards with agreed KPI metrics to be reported from all neighbourhood teams and urgent emergency care (including acute hospital) providers, teams and services. These are essential for systems to capture and monitor performance as well as measure impact. This helps to inform and drive continuous improvement and Quality Improvement (QI) effectively.
  8. Systems must establish formal partnerships between senior clinical and managerial leadership for frailty. Clinical leaders should have strong frailty subject matter expertise, experience of care in the community and relationships with teams involved in day-to-day care. Clinical leaders with frailty expertise, should hold positions of sufficient authority within the organisational hierarchy to influence strategic commissioning, resource allocation and service redesign, working alongside managerial leaders responsible for system performance and transformation. Leadership structures must remain closely connected to frontline delivery. This requires embedding practising clinicians with current experience of neighbourhood and community care within decision-making forums, alongside managerial leaders driving transformation, so that strategy remains grounded in operational reality.

Practical tips for mobilising improved PHM segmentation and performance reporting capability

These tips have been informed by the experiences of the authors in Devon and Mid & South Essex and highlight some key requirements and steps that could improve PHM data segmentation methodology, aiding progress in identifying and stratifying populations with frailty more accurately.

Top tips:
  1. Work towards ensuring data warehousing aggregates linked datasets across multiple (ideally all) providers and sectors across an ICS and includes work towards gathering data about wider determinants of health as well, where possible.
  2. Obtain, as a matter of priority, ICS-wide information-sharing agreements (ISAs) that all providers, across all relevant parts of health, mental health, social care and the voluntary sector, including hospices, sign up to in order to start gathering linked datasets.
  3. Ensure people with clinical and subject matter expertise (frailty, dementia, end of life care) from a range of professional backgrounds and all settings work with business intelligence analysts early in the process to correctly define and identify coded datasets that can identify population segments with greater accuracy.
  4. Work on an aligned frailty (including dementia) minimum dataset and adult end of life minimum dataset to define the data points that need to be captured across providers and services and achieve improved code alignment across providers. This also facilitates automated performance reporting.
  5. Roll out validated frailty identification and staging tools (such as the CFS) as a priority for all teams and providers across a system and start to use them at every patient interaction to ensure that this is recorded in a coded format within electronic health records. Teams that serve adults with high frailty prevalence should be prioritised for this roll-out. This includes: district nursing; community long term condition teams for respiratory, heart failure, Parkinson’s disease, diabetes, stroke; all older people’s mental health and dementia teams; continence teams; falls prevention services; tissue viability nursing; urgent community response; adult virtual wards for frailty, heart failure and respiratory; intermediate care services; community hospitals; continuing healthcare assessment teams; and adult social care teams, in addition to primary care and acute hospitals.
  6. Use common digital tools for delivery of proactive comprehensive geriatric assessment (CGA) and population-wide electronic frailty registers, adopted by multiple providers across a system to improve data quality, code consistency and data capture.
  7. Consider starting smaller e.g. at place or neighbourhood level, where it is more likely that more providers are using the same system, if the electronic patient record digital landscape locally is particularly complex and challenging on an ICS-wide footprint.
  8. Test early PHM data insights with frontline teams in a locality to validate early findings and to identify any data quality or data feed glitch issues.
  9. Implement widespread training and education for all staff in all sectors in order to enable effective PHM, for real-world frailty populations and their real-world frequent interactions with multiple providers, teams and services.
  10. Provide clarity to all teams on delivery of the high-impact evidence-based proactive interventions required, which improve outcomes the most, such as high-quality comprehensive geriatric assessment (CGA) which must include:
    1. Structured medication review (SMR) addressing polypharmacy risks.
    2. Cognition and delirium screening using validated tools, e.g, 4AT.
    3. Multi-factorial Falls risk assessments and prevention (FRA)
    4. Advance care planning (ACP)

Population heath management and the three shifts

This section attempts to place PHM within the three policy shifts as set out in the Ten-Year Health Plan for England. Although the three shifts are specific to England, health systems across the UK have similar priorities and we would expect this to be relevant to colleagues from all four nations.

Sickness to prevention

This chapter will focus on prevention of ill-health for people living with frailty, describing key strategic requirements that can significantly improve consistent delivery of targeted, more preventative approaches, at scale across systems to drive both improved population and system outcomes.

This includes:

  • The case for why prioritising system-wide delivery of best practice proactive frailty care from all multi-agency community teams and acute hospital teams, with a sense of urgency, is critical to mobilise shifts towards more preventative PHM.
  • Determining the key strategies needed to mobilise scalable, consistent proactive and preventative care delivery by multiple teams and providers across the frailty spectrum. This ranges from early-stage (pre-frailty/mild frailty) to those at later stages (moderate to severe frailty) requiring tertiary prevention, including personalised ACP and proactive palliative and end of life care (PEOLC).
  • Exploring whether shared frameworks, including clear ‘branding’, can provide clarity for systems on the key deliverables needed (and the reasons why), and driving ‘everybody’s business’ system-wide adoption and consistent delivery of personalised, proactive frailty care.

Proactive frailty care is not often perceived as a vital part of the prevention agenda, whether that be publicly, politically, within NHS and Department of Health and Social Care policy or within the healthcare and social care professions. Prevention often has a narrow view, focusing on a specific single organ disease or long-term conditions. Many professionals do not see frailty as a diagnosis in its own right, which contributes to ongoing misunderstanding of its potential for prevention of frailty and prevent accelerated progression to enhance survival. Frailty can be positively influenced by primary, secondary and tertiary prevention.

Primary prevention influences factors that are associated with the earlier onset of many diseases and long-term conditions. These include lifestyle factors (smoking, alcohol, inactivity, loneliness), social determinants (poverty, housing, employment, education) and environmental factors (green space, air quality, built environment). These are mostly public health measures which need to be addressed at a national level. However, if tackled, these will lead to healthier biological ageing for more people in the population, delaying the onset of frailty.

Secondary prevention for single long-term conditions (LTCs) is currently embedded in the healthcare system as single pathways, often through Quality and Outcomes Frameworks (QOFs) in primary care. However, it is more common now for people to be living with multiple LTCs and the system has yet to catch up. A more holistic and efficient approach to management of multiple LTCs (multi-morbidity), including screening for rising risk of frailty, remains poorly articulated.

Tertiary prevention for frailty aims to optimise function, quality of life and independence and can often stabilise or reverse aspects of frailty This is achieved by reducing avoidable crisis episodes such as unplanned hospital admissions and preventing avoidable harms, preventing accelerated deterioration of frailty (often causing premature mortality).

Unlike many other LTCs, frailty lacks clearly embedded preventative care models being prioritised both within policy and within contracts to support system-wide delivery. Few high priority measurement or system incentives exist for evidence-based interventions for frailty such as system-wide CGA delivery or advance care planning. This leads to inconsistency and significant under-delivery of known high-impact proactive personalised interventions and actions, at scale, from multiple providers across systems. High impact, evidence-based, frailty-attuned proactive tertiary prevention includes:

  1. High-quality Comprehensive Geriatric Assessment (CGA)
  2. Structured Medication Reviews (SMR) reducing harmful polypharmacy
  3. Cognitive assessments using validated tools such as 4AT to detect delirium and multi-factorial falls risk assessments
  4. Personalised care and support planning including Advance care planning
  5. Supporting individuals nearing the end of life to achieve their preferences and priorities, respecting their autonomy, values and preserving their dignity. This can enable more people to spend more time at home in their last years of life and to die comfortably in the place of their preference.

Tertiary prevention for frailty is not just clinically important. It is also a strategic system enabler that should be urgently mobilised in order to provide benefits to the whole population.

Benefits include:

  • A focus on a relatively small cohort (8-12% of most populations) with the highest impact on system demand
  • Potential to deliver significant measurable improvements in both population outcomes and system demand quickly, if prioritised and delivered consistently at scale across systems
  • Strong alignment with ICS objectives of reducing inequalities, improving outcomes and delivering value

Tertiary prevention can:

  • Reduce adverse events such as falls, fractures and delirium
  • Reduce unnecessary ED attendances and unplanned hospital admissions
  • Reduce current long stays in hospital
  • Improve quality and experience of care
  • Reduce premature admission to care homes and long-term loss of independence
  • Reduce premature mortality
  • Enable people to live and die well, in line with what matters most to them

Primary and secondary prevention offer essential long-term gains to benefit the wider population but may take many years, decades or even generations to achieve desired impacts. Tertiary prevention, targeted at a smaller population group of people living with frailty and dementia now, is the most immediate lever for improving outcomes and reducing both system demand pressures and costs in the immediate future.

While this has obvious advantages for those living with frailty, it is also beneficial for the whole system. Targeting frailty at scale enables ICSs to significantly reduce hospital admissions, reduce premature mortality and improve quality of life, including improving independence and experiences of care. 

Case study: Mid and South Essex ICS FRAIL plus 7 High Impact Proactive Actions Framework

In 2022, Mid and South Essex ICS recognised that one of the key challenges to overcome was that ICB executive directors, place-based delivery leads, providers and services lacked clarity and understanding on what best practice frailty-attuned care needed to look like for improved PHM.

To provide a clear framework for the ICS, the Mid and South Essex Ageing Well stewardship programme started to work on the development of a simple framework that could help its providers to create a blueprint for driving both neighbourhood health delivery as part their place-based strategies, plus help improve delivery of frailty attuned care in Urgent and emergency care (UEC)

The Framework was called FRAIL Plus, which stands for Find, Refer (or Review), Assess, Intervene and Listen to, and incorporates seven key proactive actions including:

  1. Identification of Frailty using Clinical Frailty Scoring (and Dementia/Delirium using 4AT)
  2. Delivering Comprehensive Geriatric Assessment (CGA)
  3. Structured Medication Review (SMR) to reduce avoidable harms/risks from polypharmacy
  4. Falls Risk Assessment (FRA) and prevention
  5. Personalised Care and Support Planning (PCSP) – including identifying unpaid carers and assessment of their needs if need be
  6. Early Identification of any who may be nearing end of life (EOL) (e.g all CFS > 7, or any adults with other End of life (EOL) proactive indicators (e.g. GSF PIG/SPICT indicators)
  7. Advance Care Planning (ACP) and anticipatory palliative care

These deliverables are the core part of the ICS Tertiary prevention strategy - to be targeted towards around 9% of the whole population, representing approximately 118,000 people identified by the MSE PHM data segmentation tool (as detailed in case study two). The delivery of this framework is facilitated and captured (reportable) using a digital care delivery tool that has been rolled out to all providers teams and services across the ICS.

See case study five for information about results achieved within less than two years of embedding this framework, using a digital CGA and ACP tool as an enabler to help drive measurable delivery at scale across the ICS.   

Analogue to digital

Effective frailty care is dependent on the delivery of evidence-based, proactive inputs known to improve outcomes in people with frailty. These inputs need to be captured and measured to reveal performance and to drive continuous Quality Improvement (QI) that can impact positively on both population and system.

Shared digital platforms, records systems and tools that enable integrated care delivery across multi-agency teams and providers and across care settings, ideally in real time, is essential for adults living with frailty, especially as they often have rapidly changing multi-agency needs and high use of healthcare services. Digital enablers can increase capacity across teams and services in the delivery of personalised proactive care, improving vital continuity of care and maximising efficiency.

In this section we will examine:

  • The impact of shared digital proactive care delivery tools e.g. electronic Comprehensive Geriatric Assessment (e-CGA), personalised care and support plans (PCSPs) and Advance Care Plans (ACPs) in aligning and uniting capacity of multidisciplinary teams to provide proactive frailty care and improved continuity of care. In addition, this enables teams to record, coordinate and evidence delivery of best practice, at scale across systems.
  • The benefits of system-wide electronic frailty population registries, enabling identification, staging and longitudinal tracking of individuals living with frailty providing shared case management across multiple providers, teams and services – supporting the goal to make ‘every contact count’. This replaces previously siloed, fragmented and mostly reactive crisis-driven care (which is often hospital-centric) with a system that unites all teams in an ‘everyone’s business’ ethos in the delivery of frailty-attuned proactive and preventative care in the community.
  • How population-wide electronic frailty registers also facilitate collating proactive care key performance indicators  (KPIs)  from multiple providers, teams and services. This enables and informs more meaningful performance dashboard developments, which can effectively drive continuous quality improvement (QI) across ICSs.
  • Importance of defining  frailty-specific minimum datasets and key performance indicators (KPIs) necessary to monitor, evaluate and improve frailty-attuned care across all providers.
  • Recommendation of the optimal digital electronic patient record (EPR) functionalities required that can enable effective, efficient multi-provider integration for delivery and to provide improved care continuity.

Older people with frailty usually encounter almost every provider / team / service across a system and are the population group most likely to interact with the health service including social care sector. This population has one of the most frequently changing multi-agency needs of all population groups, explaining why it generates high demands falling on all parts of the system. Efficient high-quality assessments, with essential continuity of care and personalisation are almost impossible if agencies cannot view the same data to see the latest story and status of the person in real time.

Reactive and fragmented care delivered in silos results in a lack of capacity to meet demand. It also misses opportunities to improve outcomes and prevent deterioration which may be caused by inappropriate polypharmacy not being addressed in time to prevent an adverse event (such as a fall causing a hip fracture or delirium), absent or delayed ACP and unnecessary hospital admissions.

In order to move away from mostly reactive, usually fragmented, misaligned approaches, towards more successful proactive, holistic and personalised frailty care, systems must:

  • Align all providers and teams that serve older adults (across both hospital and community) in frailty-attuned care to meet population needs and prevent missed opportunities
  • Measure what matters, using shared digital tools that support consistent delivery, data capture and reporting of evidence-based frailty attuned care inputs across the whole system
  • Ensure real-time shared care across all providers involved in care – not just specialist frailty services.
  • Enable seamless continuity of care across transitions of care, including between community teams, mental health, primary care, social care, single points of access for urgent and emergency care (UEC), urgent community response (UCR), hospital at home, emergency services and acute hospital teams

This enables the vision for neighbourhood health models to become a reality. Without this, teams are often working in isolation, lacking required information for higher quality assessments and personalised decision making. This can lead to diagnostic delays, misdiagnosis, duplication which results in inefficiencies, lack of capacity, lack of alignment and missed opportunities with consequent poorer outcomes for people with frailty and ever worsening ‘revolving door’ pressures across all parts of the health and care system.

Systems that have used digital CGA tools have demonstrated positive impacts. They have been effective in mobilising proactive care across multiple agencies to align high quality CGA delivery with required data-sharing on assessments in real time across different multi-agency teams, providers and organisations. Mobilising shared care capacity and better continuity with effective collaboration between teams has resulted in more aligned, consistent and measurable delivery of proactive evidence-based interventions in a more efficient and higher-value joint effort.

Key features and benefits of such a system include:

  • Shared, structured proactive care templates aligned with the CGA framework used by multiple providers and teams. This helps to align the quality of personalised care and high impact proactive evidence-based actions so that this is more consistently delivered.
  • Embedded SNOMED CT7 aligned codes on proactive care KPI metrics covering identification (CFS), CGA, 4AT, SMR, multifactorial Falls Risk assessment (FRA), personalised care and support planning (PCSP) and Advance Care Planning (ACP). This standardises coded datasets matched to evidence based interventions, to enable automated performance reporting  of frailty specific KPI metrics, making the delivery of frailty attuned proactive care from multiple services and providers to become measurable and every contact can be made to count.
  • Two-way data exchange with read and write capability live in real time across electronic patient records, accessible by all involved providers/teams in real time.
  • Shared functionality for electronic prescribing, live updating of Personalised Care and Support Planning, including sharing Advance Care Plans
  • Live care coordination via digital inter-team referrals and communication – systems that support contemporaneous data-sharing in chronological patient journey order, rather than split data feeds or retrospective multidisciplinary working.
  • Real-time visibility of population needs and activity to enable personalised, anticipatory and adaptive care.
  • High impact for reducing unplanned hospital admissions, ED attendances, reducing polypharmacy-related harms, improving palliative and end of life performance (reducing hospital admissions in the last months of life and reduced proportion of population deaths in hospital) and improving survival.

Table 1: Key shared digital enablers
Frailty population identification and stratification using real-time linked integrated datasets from multiple providers and all sectors.  This improves capability to identify and segment frailty populations more accurately and dynamically and is better able to keep pace with detecting transitions across frailty stages. Identification of frailty should be conducted by all providers, teams and services with frontline staff that serve adults, using tools like the Clinical Frailty Scale (CFS).
ICS-wide electronic frailty population register platforms that all providers / teams / services serving adults can access and input into.  This can help systems to see and track their frailty population far more accurately than previously possible. Improving visibility to all providers system-wide can help prioritise the delivery of the right care for each person. System-wide frailty registers also enable systems to capture and report delivery of frailty proactive care metrics to measure performance from every part of the system.
Integration of teams delivering frailty proactive care. Shared teams in neighbourhood-aligned hubs and integrated neighbourhood teams (INTs) with integrated urgent care (UEC single point of access, UCR, Hospital at Home (Virtual Wards) and acute hospital same day emergency care (SDEC).
Frailty Skilled shared teams staffed by sufficient numbers of Tier 2 and Tier 3 frailty skilled practitioners capable of delivering high quality CGA. This should include Advance Care Planning (ACP), in complex patients across all care settings.
Shared IT infrastructure – one shared record system used across all parts of the primary and community system and all parts of UEC acute care.  This enables far more efficient digital MDT working, to keep pace with frequently changing holistic needs-as opposed to relying purely on more inefficient, static and time-consuming periodic MDT “meetings” across the whole spectrum of routine care and whilst responding to higher acuity needs during urgent care episodes.
Single shared CGA-based assessment and personalised care planning digital tools used by all adult providers, teams and services during both routine and urgent emergency care (UEC).  This can help to improve consistency on the quality of delivery of proactive interventions. It can also improve efficiency, avoiding duplication and gaps in care delivery, increase shared workforce capacity to meet demand and supports better personalisation and continuity of care. As well as enabling automated reporting of frailty proactive care KPI metrics from multiple teams- so that delivery can be measured and can effectively drive continuous improvements
Shared reporting framework and performance dashboards capturing and reporting frailty proactive care delivery from all providers, teams and services.  This allows PHM for frailty to become a reality and helps to capture more meaningful metrics of evidence-based interventions that can intelligently inform systems, places, providers, services and neighbourhoods on their areas of focus to drive continuous improvements.
Clinically led, evidence-based design. Practising clinicians embedded in leadership roles.

The Ten-Year Health Plan specifically mentions progress towards a single care record for all patients. A priority group to focus on would be populations with complex needs, including older adults living with frailty and dementia.

Shared care multi-provider CGA digital tools can provide a frailty specific holistic electronic single care record that follows the person wherever they go and whichever healthcare professionals they see.

Digital enablers demonstrate that driving a successful and high-quality left shift from hospital to community, in order to improve  outcomes for the population and the system is not necessarily dependent on having more or new services.

Often this simply requires the right tools and digital infrastructure that can unite teams, to make existing services work better together with shared ownership for delivery of evidence-based interventions with much better personalisation and continuity of care - all of which underpin higher quality care for frailty populations to deliver successful outcomes.

Case study: Pathfields PCN – iCGA 3.0 Tool

Once frailty was identified using the Pathfields Tool, it became clear that CGA would need to be delivered at neighbourhood scale.

However, there were significant challenges. Workforce capacity was limited, specialist frailty expertise was scarce and CGA was labour intensive and administratively heavy. Clinical information sat across multiple systems with no shared assessment or visible care plan. The population need had been identified, but traditional approaches were not viable.

In response, the team, working with Target Health Solutions, developed iCGA 3.0, a digital CGA embedded within SystmOne, the core primary care clinical system and most complete longitudinal patient record. The aim was to deliver high quality CGA within existing workforce constraints.

Key features:

  • Full clinical picture at a glance: Diagnoses, long term conditions, medications and recent results are pulled into a structured CGA screen with alerts for uncontrolled conditions, abnormal results and high risk medicines. Clinicians can address multiple overdue issues in a single consultation.
    • Making unmet need visible: Structured searches identify patients on high risk medicines, care home residents whose admission status does not reflect their stated preference not to be hospitalised, and individuals with escalating healthcare use over the previous 6 to 8 weeks.
    • Real time admission alerts: Enable targeted community in reach for people living with frailty.
    • One assessment, one shared plan: A single CGA with read write access allows a multi provider integrated neighbourhood team to complete one assessment and generate one care plan visible to community, hospice, out of hours GP, ambulance, emergency department and hospital teams.
    • Rapid care plan sharing: Care plans are auto generated and shared electronically in minutes, with minimal administrative burden.

Outcomes:

  • Improved patient continuity
    • Improved staff experience through shared working
    • Reduced high risk prescribing, saving an estimated £75 per patient annually
    • 25.4% reduction in A&E attendances in people aged 75+ (41% reduction in A&E attendances from care homes)

This model shows how digital infrastructure enables a multi provider integrated neighbourhood team across a PCN footprint to deliver CGA at scale and reduce avoidable hospital use.

Case study: Mid and South Essex ICS – Fr EDA Tool and electronic Frailty Registry (e-Fra CCS)

Mid & South Essex ICS developed the Fr EDA (Frailty, End of Life, Dementia Assessment) tool and electronic Frailty Care Coordination system registry (e-Fra CCS) to support unified proactive frailty care by primary care, all community providers, including community older peoples mental health and dementia services  and hospice teams , frailty virtual ward services  and potentially some acute hospital teams, across the whole ICS footprint.

Key features of Fr EDA digital CGA +ACP tool include:

  • Embedded in existing provider and team electronic patient record (EPR) systems, (SystmOne), enabling real-time collaboration across all primary care practices across 25 PCNs, all community health providers, intermediate care teams, all community mental health including dementia teams, virtual wards, UCR, urgent care single point of access (UEC SPOA) frailty UEC hotline, and hospices (including 111 GP clinical assessment teams and some teams in the acute hospital trust) across the whole ICS.
  • Structured on the domains of CGA, thereby enabling frontline staff in multiple different teams, providers, care settings and situations to deliver CGA-based assessment in a matter of minutes. Providing an evolving holistic view (a “walking CGA”) that tracks the person’s story and their changing needs – even if that person moves across different teams, providers, care settings and even geographies in the ICS.
  • Delivers the seven frailty-attuned proactive actions of the ICS MSE FRAIL Plus Framework (as detailed in case study three), including facilitating earlier end of life (EOL) needs identification, Advance care Planning (ACP) and anticipatory palliative care interventions
  • Links to education weblinks and resources to guide staff – such as for example polypharmacy Screening Tool Of Older People's Prescriptions and Screening Tool to Alert to Right Treatment (STOPP/START) guidance tools, Fracture Risk assessment tool (FRAX), (as well as free falls prevention including strength and balance resources for the public), End of life care Gold Standards Framework Proactive Identification Guidance (GSF PIG) guidance  and more.
  • Live real-time data-sharing, communication and instant inter-team referrals functionality across multiple providers, teams and services across the ICS.
  • Improves personalisation and continuity of care during both routine care and urgent care delivery from multiple teams and services, (including virtual wards for UEC)
  • Saves staff time for assessments, supports higher quality diagnosis and management, plus multi-agency working, live in real time.

ICS-wide population level and system impacts (2023–2025) include:

  • 50% reduction in falls rates per 1,000 population aged over 65.
  • 54% reduction in unplanned admissions from care home residents.
  • Greater than15% improvement in dementia diagnosis rates.
  • 50% reduction in total population experiencing more than three unplanned hospitalisations in their last 90 days of life, alongside at least 6% fewer total population deaths in acute hospitals
  • In highest Fr EDA adopting neighbourhoods, 17% fewer deaths in their total population occur in hospital, compared with current England average.

Limitations of these digital tools and opportunities for greater impact include:

  1. Only available to providers and teams using SystmOne electronic record system (albeit 100% of all primary and community-based teams including 111 clinicians use this as their digital electronic record system).
  2. Requires more front-line staff to engage in Frailty training to maximise impacts.
  3. Requires improved uptake by Acute hospital- especially Front Door Frailty teams and all staff providing care in the first 72 72 hours of hospital care, including hospital discharge teams for maximal impact.
  4. Financial incentives for providers could help drive more universal adoption.

Hospital to community

This section will assess new care models that aim to provide urgent emergency and acute hospital level care for older people living with frailty in their own homes, and consider key factors (including digital record sharing and digital care delivery enablers) that could help to maximise their potential to successfully shift delivery of care from hospital to community settings.

This section will explore:

  • How neighbourhood models (eg, Integrated Neighbourhood Teams) linking to urgent care innovations (urgent community response (UCR), Hospital at Home (Virtual Wards), Tier 3 frailty skilled clinicians providing Silver Triage for Urgent and Emergency care (UEC) Single Points of Access) can deliver timely, proactive care in a person’s usual place of residence.
  • Approaches required for optimising integration with acute hospital interfaces such as Front Door Frailty and Same Day Emergency Care including short-stay frailty units to support reductions in length of stay and bed occupancy in acute hospitals.
  • Understanding whether both system-wide workforce training programmes and the adoption of integrated data tools and shared digital systems can improve frailty fluency, care consistency and cross-provider workforce competencies across all professional roles.

There is good evidence of proactive frailty interventions help to reduce unnecessary hospital admissions and time spent in hospital, as well as support improved outcomes, stability and wellbeing in the community.

Despite this, delivery of such best practice interventions remains inconsistent. Lack of frailty identification and understanding about what frailty-attuned proactive care should be constrains many systems in mobilising consistent strategic delivery.

There is currently no minimum standard for frailty identification or high quality CGA. In addition, outside of the framework of Enhanced Health in Care Homes (EHCH) national service specification for Primary Care Networks, there are currently no national contracts mandating delivering of CGA for either primary care, wider community teams, dementia teams or hospital teams. There are also no national performance dashboards collating evidence of CGA delivery to hold systems, providers and services to account, nor is there a focus on the digital and data enablers needed to achieve this.

New models of care for delivering Urgent and Emergency care (UEC) services for people living with frailty in their own homes (such as UEC Single Points of Access (SpoA), Urgent Community Response (UCR), Hospital at Home (also known as Virtual Wards) and Discharge to Assess (D2A) have been shaped by separate NHS England national operating frameworks and contracts over the last few years. However, reporting requirements are not consistent across England and do not always specify that frailty proactive care interventions are reported as mandatory performance metrics.

Acute hospital services such as Front Door Frailty (FDF) interface models (including same day emergency care (SDEC) for frailty) are recommended but not yet mandated within acute trust contractual requirements. In addition, these services often lack joined up data-sharing and electronic records systems with the wider primary and community parts of the system to improve personalised management, continuity of care and enable a seamless, efficient and effective shift from hospital to community. The distinct frameworks and contracts within UEC have an operational, activity-based focus and are therefore not properly connected. The currently separate national frameworks have been mostly translated by systems and providers into separate services, each with its own structure, remit, patient record, criteria and different performance frameworks.

As a result, delivery is often focused on meeting mostly activity-based contractual targets of separate service models, rather than focusing a priority on providing holistic, evidence-based care. This has led to fragmentation, inefficiency and likely a less cost-effective way to achieve the hospital to community shifts.

Services frequently operate in silos, offering mostly reactive care by incomplete multidisciplinary teams with minimal frailty skills and lacking input from senior clinical decision makers with required Tier 3 frailty skills and community care experience.

This leads to problems such as inefficient duplicated assessments by separated teams driving capacity challenges, avoidable ED attendances caused by insufficient frailty knowledge and skills to manage higher acuity needs well in the community (especially for people at moderate to severe frailty stages), as well as lack of continuity, both between services and after discharge.

Virtual wards often operate under separate condition-specific frameworks (e.g. heart failure, respiratory, frailty) and many systems have multiple separate pathways with specific inclusion and exclusion criteria.

Up to 76% of people admitted to hospital with heart failure are also living with frailty.8 This is significant, given recent evidence that major causes of hospitalisations and death in people with heart failure and frailty are more often non-cardiac related. These hospitalisation and deaths are more commonly attributed to problems like falls, acute kidney injury and electrolyte disturbance,9,10 much of which may have been driven by medication induced adverse events, especially for those with moderate -severe frailty.11

Specialist teams overseeing care for those in disease-specific virtual wards may lack the skills and knowledge to care for people who are also living with frailty. Teams delivering hospital-level acute care for older people must have generalist holistic proactive care and frailty skills to ensure that the model of care fits the reality of the population to meet the more complex nature of their UEC needs

It is crucial that new UEC models work together to offer better efficiency, value, affordability and impact and deliver frailty-attuned proactive personalised care. The population with frailty frequently transition between routine and high-acuity urgent care needs, meaning any effective frailty team or model of care needs to be capable of spanning both, ensuring seamless support before, during and after a crisis.

The key to success for any new urgent care models is to prioritise delivery of frailty-attuned proactive care, not just reactive elements. Otherwise, these services will fail to reach their optimal potential for delivering better outcomes at individual, population and system level.

The current policy landscape creates opportunities to strengthen population health management for frailty through better neighbourhood-level integration.

The Ten-Year Health Plan promotes primary and community care leadership. It advocates the use of digital enablers such as a single patient record to achieve the vital integration needed to unite care delivery across teams and services, achieving personalisation and continuity of care.

4 top enablers for improving hospital to community shifts include:

  1. Integrate the current separate UEC models to become one joined up model (UEC SPoA, UCR, Virtual wards/Hospital at home, Acute hospital front door frailty including SDEC services) and align them to neighbourhood health models
  2. Ensure all are staffed with Tier 2 frailty skilled practitioners with immediate access to senior Tier 3 frailty skilled clinicians with strong community care experience.
  3. Ensure all use the same shared read-write IT digital record system, including digital CGA and ACP proactive care delivery tools.
  4. Mandate reporting of Frailty attuned proactive care metrics from all UEC models caring for older people with frailty CFS scores of 5 and above.

Case study: Hull & East Riding

An entirely community-based team of frailty skilled clinicians including full-time community-based geriatricians, GPs with extended roles (GPwERS) in Frailty, and community-based advanced practitioners including AHPs with frailty skills provide:

  • Silver triage UEC “call before convey” support to 999 ambulance teams, usual PCN and community team staff (including professionals supporting care homes).
  • Urgent Acute care Frailty Virtual Ward (Hospital at home) support
  • Oversight and proactive care in bed-based intermediate care units, including care home settings.

More than 80% of calls to the UEC Frailty Silver Triage line are successfully managed in the community – avoiding unplanned emergency hospital care.

The team provide best practice proactive frailty-attuned care – both virtually and in person – harnessed by using the same digital EPR system (SystmOne), as the rest of the usual primary care and community provider system in that area. This aids improved personalisation of care and higher quality of decision-making.

Conclusion – case for change for frailty-attuned population health management

Mobilising effective PHM for frailty should be one of the most urgent priorities nationally and for local systems. Tackling this properly is the key to unlocking progress that could enable the shift towards an overall more preventative model of health care, as per vision of the Ten-Year Health Plan, to become a reality. It is long overdue for prioritising a comprehensive system-wide approach to frailty, with its own unique approach to best practice prevention and management. This is a vital part of effective PHM and the key to reducing current unsustainable overwhelming pressures across health and care systems.

Frailty-specific preventative and proactive care is currently deprioritised and not fully appreciated as one of the most vital arms of prevention today. Without urgent action, population outcomes and system pressures will continue to worsen

While data should be able to tell the true story of the population with frailty, decades of paucity of the required datasets from all providers means that there is very limited information available about who the population living with frailty is and what type of care they have been receiving.

However, there is evidence that these historic gaps can be addressed and that outdated approaches to health and care delivery for this population group can be overcome.

The health and care system is extremely stretched and depleted. It is therefore important that change is not left any longer. If it is, mobilising the required initiatives and innovations to deliver sustainable progress may become impossible.

Mobilising tertiary prevention for frailty at scale, now, is the critical bridge to enable systems to move towards an overall preventative model to benefit the whole population. To achieve success will require significantly reforming our thinking and our approach. Care in the community, continuity and personalisation of care are crucial components of achieving successful shifts for better outcomes, experience and quality of care.

To help achieve this,

  1. Improved methods using linked aggregated data from all multiagency sectors are needed to provide more sophisticated population health data insights, to ensure that more is known about the population living with frailty.
  2. Effective digital care delivery tools and record sharing innovations are required to help unite multi-agency teams, creating more shared care capacity to drive scalable and more sustainable delivery of more consistent frailty attuned proactive care, system wide- to meet population needs and demand, both feasibly and effectively. As well as facilitating essential personalisation and continuity of care.
  3. There is an urgent need to develop a national and contractual performance reporting framework of frailty-specific proactive care KPI metrics. This should be mandated for all adult teams, services and providers delivering routine and urgent care, to significantly drive increased delivery of evidence-based frailty-specific interventions which make the biggest positive difference to this population and to the system as a whole.
  4. This could provide much needed clarity for systems, commissioners, providers and the frontline workforce, vital for achieving the aims and ambitions of the Ten-Year Health Plan. The evidence and case studies show that population health management using effective PHM data and digital shared care delivery tools adopted by multiple providers, can start to drive success and start to improve  outcomes for older people living with frailty (within relatively short timescales).

These can collectively help to drive more successful “left shifts” away from mostly hospital-centric reactive and often fragmented care towards more proactive, coordinated and preventative care in the community, as well as enable systems to improve sustainability of their wider health and care economy.

However, for this to become a reality, population health management for this population must be prioritised as part of national policy, in contractual services and within commissioning and national reporting frameworks for all providers and services providing care for older adults.

This must be accompanied by the necessary funding allocations, vital digital infrastructure and data-sharing enablers required, as well as ensuring high priority to mobilise essential comprehensive workforce training, aimed at all professional and practitioner roles across health and social care.

References

  1. Whitty C, 2023. Chief Medical Officer’s Annual Report 2023: Health in an Ageing Society. Available at: https://assets.publishing.service.gov.uk/media/6674096b64e554df3bd0dbc6/chief-medical-officers-annual-report-2023-web-accessible.pdf (accessed 22 September 2025)
  2. Watt T, Raymond A, Rachet-Jacquet L, Head A, Kypridemos C, Kelly E and Charlesworth A, 2023. Health in 2040: Projected patterns of illness in England. Available at: https://www.health.org.uk/sites/default/files/upload/publications/2023/Projected%20patterns%20of%20illness%20in%20England_WEB.pdf (accessed 22 September 2025)
  3. Sinclair D, Maharani A, Chandola T, Bower P, Hanratty B, Nazroo J, O’Neill TW, Tampubolon G, Todd C, Wittenberg R, Matthews FE and Pendleton N, 2022. ‘Frailty among Older Adults and Its Distribution in England’, The Journal of Frailty & Ageing. 11:2, 163-168
  4. Hale, M., et al. (2019). Frailty, inequality, and resilience. Clinical Medicine, 19(3), 219-223.
  5. Living well in communities with frailty. Evidence for what works. Healthcare Improvement Scotland ihub. 2018
  6. Attwood D, Hope SV, Spicer SG, Gordon AL, et al. Does proactive care in care homes improve survival? A quality improvement project. BMJ Open Qual. 2024 Jun 4;13(2):e002771. doi: 10.1136/bmjoq-2024-002771. PMID: 38834371; PMCID: PMC11163642.
  7. NHS England, 2025. SNOMED CT. Available at: https://digital.nhs.uk/services/terminology-and-classifications/snomed-ct (accessed 17 October 2025)
  8. Beezer J and O’Neil H, 2019. ‘Heart failure, older people and frailty.’ The Pharmaceutical Journal. Available at: https://pharmaceutical-journal.com/article/ld/heart-failure-older-people-and-frailty#fn_7 (accessed 22 December 2025)
  9. S, Pellicori P, Zhang J, Weston J, Squire IB, Clark AL. Effect of frailty on treatment, hospitalisation and death in patients with chronic heart failure. Clin Res Cardiol. 2021 Aug;110(8):1249-1258.
  10. Abel AAI, Samuel NA, Cuthbert JJ, Brown OI, Pellicori P, Kazmi S, Cleland JGF, Johnson MJ, Clark AL. Hospital admissions in the last year of life of patients with heart failure. Eur Heart J Qual Care Clin Outcomes. 2024 Mar 1;10(2):168-175.
  11. Woodford H, Pollock L Appropriate management of heart failure in older people with frailty | The BMJ  2024;387: e078188