NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Guideline Centre (UK). Multimorbidity: Assessment, Prioritisation and Management of Care for People with Commonly Occurring Multimorbidity. London: National Institute for Health and Care Excellence (NICE); 2016 Sep. (NICE Guideline, No. 56.)
Multimorbidity: Assessment, Prioritisation and Management of Care for People with Commonly Occurring Multimorbidity.
Show detailsO.1. Organisation of Care
O.1.1. Research Question
What is the clinical and cost effectiveness of alternative approaches to organising primary care compared with usual care for people with multimorbidity?
O.1.2. Why is this important
The guideline committee felt that primary care was well suited to managing multimorbidity, but agreed that this was often challenging partly because of how primary care is currently organised. However, there was inadequate high-quality research on alternative approaches to organising care for people with multimorbidity. Trials should be undertaken to examine the impact of different strategies on important clinical outcomes, quality of life and cost effectiveness. The committee believed that no single trial could likely address this research need, because there are many plausible interventions and many defined populations in which such interventions might be of value.
Large, well designed trials of alternative ways of organising general practice based primary care for people with multimorbidity would be of value in defined patient groups (for example, people with multimorbidity who find it difficult to manage their treatment or care or day-to-day activities, people with multiple providers or services involved in their care, people with both long-term physical and mental health problems, people with well-defined frailty, people frequently using unscheduled care, people prescribed multiple regular medicines, and people who are housebound or care home residents).
Such trials should have clear identification and justification of the planned target population, careful piloting and optimisation, and well-described interventions. They need to be sufficiently powered to provide evidence of clinically important effects of interventions on outcomes that are relevant to patients and health and social care services (for example, quality of life, hospital and care home admission, mortality).
O.1.3. Proposed format of research recommendations
| Criterion | Explanation |
|---|---|
| Population | Adults with multimorbidity living in the community in their own homes or in residential care and nursing homes. Within this broad group, researchers should specify which patients the intervention will focus on and the rationale for this focus. Plausible rationales include place of residence (care home, housebound, non-housebound), frailty and poor function, polypharmacy, high unscheduled care or emergency hospital admission rates, and combined physical and mental health morbidity (recognising that many people will fit in more than one category). People in the terminal phase of illness should be excluded. |
| Intervention | The intervention should be clearly tailored to the population being targeted, but could contain different blends of components including for example: Patient-level for example, methods for identifying and engaging suitable patients (depending on the targeted population); self-management in interventions for example, PRISMS Practitioner level for example, education and training around management of people with MM and polypharmacy (with emphasis on identification of treatment burden, shared decision-making around medication management; case-finding for anxiety and depression) Practice level for example, case management + key worker; tailored mechanisms for access (for example, accelerated urgent access, or directed access to a known clinician wherever possible); longer appointment time; proactive follow-up/call and recall; offering alternative means of consultation; discussion of patients at clinical meetings – with all multi-disciplinary team members; creation of new, general practice based multidisciplinary teams for targeted patients (for example, including pharmacists, specialist nurses [for example, geriatric or psychogeriatric], social workers) approaches such as regular involvement of a pharmacist in medicines management; liaison with specialist care. The potential components would therefore include changes to:
|
| Comparator(s) | Routine NHS care. |
| Outcome | Essential outcomes are quality of life and mortality to allow the evaluation of cost-effectiveness. It is important that adequate time is given to implement the intervention in each practice, since complex interventions involving reconfiguring existing systems typically require time to bed in, so too early evaluation may be falsely negative. Minimum follow-up should be one year, although two years would be preferable. |
| Study Design | Appropriate development and optimisation of the intervention, with evaluation of the final intervention in a cluster randomised controlled trial with practice as the unit of randomisation, with an associated economic evaluation and parallel process evaluation of implementation. |
| Timeframe | Is there a timeframe in which the study needs to be completed? For example to inform a guidance review, or whether it is anticipated that the technology could be superseded before the results of any study are anticipated. |
O.1.4. Potential criteria to support prioritisation of key research recommendations
| Potential Criterion | Explanation |
|---|---|
| Importance to patients or the population | This is of high importance of the population A large proportion of the population have multimorbidity and this is even more common in people who are older and frail. New models of delivery of care are required. |
| Relevance to NICE guidance | Good quality evidence would inform updates of NICE guidance |
| Relevance to the NHS | This area is of importance to the NHS as a whole. |
| National priorities | Improving care of older people is a national priority and many older and frail people are multimorbid. |
| Current evidence base | There is very little evidence available on alternative ways of organising general practice services |
| Equality | Multimorbidity is more common at a younger age in less affluent population. |
| Feasibility | The proposed research is feasible but requires significant funding and organisation to ensure appropriate piloting and involvement of all relevant personnel. Mercer et al ( BMC Medicine (2016) 14:88 DOI 10.1186/s12916-016-0634-2) have conducted a feasibility study which demonstrated plaubile benefit and value for money for a larger trial for an intervention of longer consultation, relationship continuity, practitioner support and self- management support. |
| Other comments |
O.2. Holistic assessment in the community
O.2.1. Research Question
What is the clinical and cost effectiveness of a community holistic assessment and intervention for people living with high levels of multimorbidity?
O.2.2. Why is this important
There was low quality evidence to indicate potential benefit from community assessments based on the principles of comprehensive geriatric assessment in older people. However, the studies were conducted outside the UK and were not aimed at all adults living with multimorbidity. The guideline committee believed that there was some evidence that holistic assessment and intervention in the community may be of benefit for older people, but that the evidence was of low quality and not adequate to inform strong recommendations.
Large, well-designed trials of holistic assessment and intervention in people with multimorbidity would be of value in defined patient groups in the community (for example, people in nursing homes, people who are housebound, people of all ages with well-defined frailty, people with high levels of multimorbidity or polypharmacy).
Such trials must be rigorous, with clear identification and justification of the planned target population, careful piloting and optimisation, and well-described interventions. They need to be sufficiently powered to provide evidence of clinically important effects of interventions on outcomes that are relevant to patients and health and social care services (for example, quality of life, hospital and care home admission, and mortality).
The guideline committee believed that no single trial could likely address this research need, since there are many plausible interventions and many defined populations in which such interventions might be of value. The committee believed that assessment should follow the principles of Comprehensive Geriatric Assessment or the Standardised Assessment of Elderly People in Europe (STEP) tool, and that interventions would likely involve a multidisciplinary team.
O.2.3. Proposed format of research recommendations
| Criterion | Explanation | |
|---|---|---|
| Population | People living in the community with high levels of multimorbidity. Inclusion: Adults with multimorbidity living in the community in their own homes or in residential care and nursing homes. Within this broad group, researchers should specify which patients the intervention will focus on and the rationale for this focus. Plausible rationales include place of residence (care home, housebound, non-housebound), frailty and poor function, polypharmacy, high unscheduled care or emergency hospital admission rates, and combined physical and mental health morbidity (recognising that many people will fit in more than one category). Exclusion: People in the terminal phase of illness, People in hospital. | |
| Intervention | The intervention should be clearly tailored to the population being targeted. Although there are many plausible intervention sub-components, the framework of Community Holistic Assessment and Interdisciplinary Intervention will include the principal components of holistic structured assessment and interdisciplinary care planning: A structured holistic assessment, based on the principles of Comprehensive Geriatric Assessment (CGA) and the Standardized Assessment of Elderly People in Primary care in Europe (STEP tool). An example of some of the subcomponents that should be considered for inclusion in the assessment is shown in the table below. A holistic individualised plan for treatment, rehabilitation, support and long term follow up. This is to include realistic shared goals, reviewing and optimising medications and specialist appointments, clear follow-up and review arrangements, and a communication plan with the patient and between professionals. The plan will result from an interdisciplinary discussion including the GP. Possible holistic assessment subcomponents | |
| The Patient | Perspectives, attitudes, values and priorities. The perceived burden of treatment. Ideas, concerns and expectations. | |
| Medical/physical state | Co-morbid conditions and disease severity | |
| Medication Review | ||
| Nutritional status | ||
| Primary preventive issues | ||
| Assessment of frailty if appropriate | ||
| Problem list | ||
| Mental Health | Cognitive function | |
| Depression and/or anxiety | ||
| Functional capacity | Activities of daily living | |
| Gait and balance | ||
| Activity/exercise status | ||
| Social circumstances | Social network including visitors or daytime activities | |
| Informal support available from family or friends | ||
| Eligibility for being offered care resources | ||
| Environment | Home comfort, facilities and safety | |
| Use or potential use of telehealth technology, aids and adaptations | ||
| Transport facilities | ||
| Accessibility to local resources | ||
| Comparator(s) | Comparator The comparator would be treatment as usual with no extra assessment or multidisciplinary intervention beyond that which happens in usual care. | |
| Outcome | Outcomes Critical outcomes Health related quality of life Health and social care costs including GP time Admission to a care home (for those not already resident in a care home) Cost effectiveness Important outcomes Mood (measures of anxiety/depression) Function (measures of activities of daily living physical and cognitive function) Hospital admission Mortality Other resource use (for example, social care) | |
| Study Design | Structure Interventions will require development and careful piloting before evaluation in a phase 3 trial, following the principles of the MRC Framework for the development and evaluation of complex interventions. To avoid contamination within each practice setting it would be preferable to perform a cluster randomised trial with GP practices clustered to active and comparator groups. Consideration should be given to stratifying the sample according to age, care setting and frailty. It should be specified whether there is to be a single intervention or repeated interventions with a specified follow-up period. A comprehensive process evaluation will be undertaken that ideally clearly describes the following: The context in which the intervention is implemented The intervention intended to be delivered and actually delivered The population to which the intervention was intended to be delivered and actually delivered, including an assessment of the reach of the intervention (that is, the proportion of eligible patients who accept or receive the intervention, and reasons why not) A systematic examination of the success and failure of implementation and of effectiveness will be required. This process evaluation is likely to require mixed methods, and should follow the principles of the MRC guidance on process evaluation of trials of complex interventions. | |
| Timeframe | The minimum follow-up period should be one year, although two years would be preferable. Longer follow-up would be desirable, but may be constrained by resources (although extended routine data follow-up should be feasible for at least some outcomes such as mortality and hospital or care home admission, and should be planned for if possible). | |
O.2.4. Potential criteria to support prioritisation of key research recommendations
| Potential Criterion | Explanation |
|---|---|
| Importance to patients or the population | High. Growing numbers of people have multimorbidity in the context of ageing and frailty. |
| Relevance to NICE guidance | High: the research is essential to inform future updates of key recommendations in the guideline. |
| Relevance to the NHS | High: There has been increasing interest in delivering proactive holistic to older frail people and this model of care has already been rolled out in areas such as attempts to reduce unplanned hospital admissions. Evidence of the cost and cost effectiveness of these interventions is highly relevant to NHS. |
| National priorities | Improving care for older people is a core component of the NHS operating framework. |
| Current evidence base | There are no UK based studies examining these programmes in people in the community but health economic modelling indicated that such interventions may be cost effective.. |
| Equality | Studies should include both older and frail people who are often omitted from studies but also younger less affluent groups who suffer from multimorbidity earlier than their more affluent counterparts. |
| Feasibility | The proposed research can likely be carried out within a realistic timescale. Pilot studies to establish the intervention may be required (for example as a single grant with staged funding). The expense needed to resolve the question would be warranted. |
| Other comments | None. |
O.3. Stopping drugs
O.3.1. Review question
What is the clinical and cost effectiveness of stopping preventive medicines in people with multimorbidity who may not benefit from continuing them?
O.3.2. Why is this important
There is good evidence from randomised controlled trials of the medium term (2–10 years) benefit of medicines recommended in guidelines for preventing future morbidity or mortality, including treatments for hypertension, hyperglycaemia and osteoporosis. However, there is much less evidence about the balance of benefit and harm over longer periods of treatment. It is plausible that harms outweigh benefits in some people with multimorbidity (for example, because of higher rates of adverse events in older, frailer people prescribed multiple regular medicines, or because the expected benefit from continuing a preventive medicine is reduced when there is limited life expectancy or high risk of death from other morbidities). These people are unlikely to have been eligible or included in published trials showing initial benefit from preventive medicines. The systematic review undertaken by NICE in 2015 did not find any randomised controlled trials of stopping antihypertensive medicines in people with multimorbidity. The review found 1 small randomised controlled trial of stopping statins in people with a life expectancy of 1 year, but the committee did not consider this provided enough evidence to make a recommendation. The review found several randomised controlled trials of stopping bisphosphonates (although not clearly in populations with multimorbidity) and a recommendation was made for to this, but no randomised controlled trials were found of stopping calcium and/or vitamin D. Recommendations based on robust evidence on the clinical and cost effectiveness of stopping preventive medicines in people with multimorbidity who may not benefit could have significant budgetary implications for the NHS. No ongoing trials have been identified.
The guideline committee considered that 1 or more large, well-designed trials of stopping preventive medicine in people with multimorbidity would be of value in defined patient groups in the community (for example, people in nursing homes, people who are housebound, people with well-defined frailty, people with high levels of multimorbidity or polypharmacy, people with limited life expectancy). Discontinuation could either be complete (all relevant medicines) or partial (for example, reduced intensity of hypotensive or hypoglycaemic treatment). Such trials have to be sufficiently powered to provide evidence of clinically important effects of interventions on outcomes that are relevant to patients and health and social care systems (for example, quality of life, hospital and care home admission and mortality). The committee believed that given the existing evidence, it would be of greater value to evaluate the effects of stopping discrete medicines or drug classes, rather than stopping all preventive medicines at the same time. The committee also believed that no single trial could likely address this research need, since there are many medicines that could be stopped and many defined populations in which this might be of value.
O.3.3. Proposed format of research recommendations
| Criterion | Explanation |
|---|---|
| Population | People with multimorbidity who plausibly may not benefit from continued preventive medications, for example people in nursing homes, people who are housebound, people with well-defined frailty, people with high levels of multimorbidity or polypharmacy, people with limited life expectancy. |
| Intervention | Stopping or reducing the intensity of the targeted medications according to a well-described protocol suitable for routine use in primary care settings. Researchers will have to justify their choice of medications to target and whether the aim is complete cessation or reduced intensity aiming for a particular intermediate outcome target. It may be appropriate to carry out a pilot study to examine recruitment and feasibility before going to full trial. |
| Comparator(s) | Continuation of preventive medications at the same intensity. |
| Outcome | Critical outcomes for all studies are total mortality and quality of life. Important outcomes will vary depending on the targeted medication, but could include disease specific outcomes including disease specific mortality, hospitalisation, institutionalisation, and intermediate outcome control. Follow-up should be an absolute minimum of one year, and ideally a minimum of two years. |
| Study Design | Patient randomised controlled trial, consider use of equivalence or non-inferiority design depending on the primary outcome. |
| Timeframe | There is an urgent need for this type of research to inform future guidance including any update of NICE multimorbidity guideline. |
O.3.4. Potential criteria to support prioritisation of key research recommendations
| Potential Criterion | Explanation |
|---|---|
| Importance to patients or the population | High. Growing numbers of people are taking drugs for long-term prevention based on evidence that these drugs are effective in the first 3-10 years of treatment. Whether observed benefits are maintained as people age, become multimorbid, become frail or are co-prescribed many drugs for other conditions is unclear. Very large numbers of people are affected, and if net benefit is not maintained, then this would have significant impact on large numbers of vulnerable people. |
| Relevance to NICE guidance | High: the research is essential to inform future updates of key recommendations in the guideline. |
| Relevance to the NHS | High: the cost of preventive drugs is considerable, and if there are populations in which the drugs cause net harm, then this may have significant additional impact on the NHS. |
| National priorities | Improving care for older people is a core component of the NHS operating framework; improving care for older people with frailty, including more appropriate prescribing, is supported in specialty guidelines (e.g. BGS Fit for Frailty) and NHS England guidance (e.g. Safe, compassionate care for frail older people using an integrated care pathway). Also improving care for care home residents is a core component of NHS England New Models of Care Programme & 2011 BGS Working Party Inquiry. |
| Current evidence base | Limited. Randomised controlled trials of preventive medications typically focus on effectiveness in the first 2 to 10 years of treatment, and are largely carried out in younger, fitter and less multimorbid populations. For most conditions, there is at least some more recent evidence in relation to starting preventive medications in older populations, but there is uncertainty regarding continuing benefit. This is a particular issue in some groups of people with multimorbidity, including older people with frailty, care home residents and people receiving palliative care. These groups are typically at greater risk of medication-related side effects and may gain less benefit (or experience harm) from continuing preventive treatment. |
| Equality | The research recommendation focuses on particular populations who may not benefit from continued antihypertensive treatment. |
| Feasibility | The proposed research can likely be carried out within a realistic timescale, but pilot studies to evaluate feasibility including recruitment may be required (for example as a single grant with staged funding). The expense needed to resolve the question would be warranted. The main ethical issues would be around the potential vulnerability of the populations selected, including consideration of obtaining informed consent in the context of cognitive impairment/dementia. |
| Other comments | None. |
O.4. Predicting life expectancy
O.4.1. Review Question
Is it possible to analyse primary care data to identify characteristics that affect life expectancy and to develop algorithms and prediction tools for patients and healthcare providers to predict reduced life expectancy?
O.4.2. Why is this important?
Many people take preventive medicines which are likely to offer small benefits because of reduced life expectancy from other causes. Medicines and other treatments may therefore be adding to treatment burden without adding quality or length of life. The ability to identify people with reduced life expectancy could provide healthcare professionals and people with information that could inform decisions about starting or continuing long-term preventive treatments. Conversely younger people with multimorbidity and reduced life expectancy may benefit from additional preventive treatments. Because this information would be used most often in a primary care setting, the committee considered that a tool derived from information within primary care databases would be most useful.
O.4.3. Proposed format of research recommendations
| Criterion | Explanation |
|---|---|
| Population | Adults (aged 18 years and above), although in practice the focus is likely to be on middle aged and older adults in whom reduced life expectancy is more common. |
| Risk tools | Multivariable risk tools comprised of variables available within routine primary care databases to identify people at risk of a reduced life expectancy. Validated thresholds for varying levels of risk of mortality should be provided. |
| Target condition or reference standard | All cause mortality after 12-months, and within the normal time horizon under which preventative treatment decisions are made (typically 3-10 years, as appropriate). |
| Outcomes | Statistical outputs should include discrimination data, where the sensitivity and specificity of the tool at validated thresholds should be reported with their respective confidence intervals (i.e. not AUC data alone). Calibration data should also be reported, including the Hosmer-Lemeshow statistic and calibration plots and the adjusted and unadjusted pseudo R2. The tool should prioritise high specificity (i.e. the ability of the tool to correctly identify people who are not at risk of a reduced life expectancy; true negatives). This is because the primary use of this tool will be to guide decisions on initiating or withdrawing treatment in people who may not live long enough to experience the benefit, and there may be the harm of withdrawing medication in people unnecessarily. However, the tool should also demonstrate at least moderate sensitivity (i.e. the ability of the tool to correctly identify people who are at risk of reduced life expectancy; true positives), in order to add meaningfully to clinical practice. Ideally, the tool should demonstrate accuracy in a broad population of people including in both younger and older adults, but this will depend on the technical performance of the tool in different age groups which should be specifically examined separately. |
| Study Design | Development and external validation of a risk tool. External validation should ideally be geographical as well as temporal, and should be prioritised compared to validation within the same study population (i.e., split half validation). Studies conducted within the UK would be preferable, in order to be most applicable to the population and settings in the NHS. |
| Timeframe | To inform update of Multimorbidity guideline |
O.4.4. Potential criteria to support prioritisation of key research recommendations
| Potential Criterion | Explanation |
|---|---|
| Importance to patients or the population | High. |
| Relevance to NICE guidance | High: the research is essential to inform future updates of key recommendations in the guideline. |
| Relevance to the NHS | High: the cost of preventive drugs is considerable, and if there are populations in which some people have significantly reduced benefit, then this may have significant additional impact on the NHS. |
| National priorities | Improving care for older people is a core component of the NHS operating framework. |
| Current evidence base | Limited. A number of studies have developed and validated risk tools to predict mortality, however the evidence base is significantly limited. The systematic review undertaken by NICE in 2015 found that risk tools are evaluated by a very small number of studies, insufficient for meta-analysis. Studies report limited data on the performance of the tool; in particular, a significant number of studies do not report the sensitivity and specificity of the tools, which is the only data that can indicate the clinical implications of using the tool in practice (i.e. the number of true/false positives and negatives). The systematic review undertaken by NICE in 2015 also found that few studies report a validated threshold for the tool, which can be used to indicate risk of reduced life expectancy. Finally, the systematic review undertaken by NICE in 2015 found that no existing risk tool was able to demonstrate both adequate specificity and sensitivity for predicting mortality was available and validated in England.. |
| Equality | The purpose of using this tool would be to ensure that treatment is tailored to individual needs, and there are no specific equality issues. |
| Feasibility | The proposed research can likely be carried out within a realistic timescale. Large databases of primary care data already exist and are used to develop prediction tools. |
| Other comments | None. |
- Research recommendations - Multimorbidity: Assessment, Prioritisation and Manage...Research recommendations - Multimorbidity: Assessment, Prioritisation and Management of Care for People with Commonly Occurring Multimorbidity
Your browsing activity is empty.
Activity recording is turned off.
See more...