Multidisciplinary heart failure management programmes

• Top
• Next

Introduction

In the early 21st century, heart failure (HF) guidelines and other consensus documents recommend a multidisciplinary approach to the management of people with this diagnosis.^1–8 This trend towards, and aspiration to, the involvement of a range of health care professionals in the care of a particularly vulnerable patient group does not appear controversial. It seems intuitive that this would deliver best care, and is an approach welcomed by patient groups. But how robust is the evidence for this approach, and which components of these multidisciplinary services confer benefit and in what health care contexts? To what extent are they then applicable to other populations and health care environments? The literature is extensive, diverse, and complex, which explains the repeated meta-analyses undertaken, and both published and planned Cochrane reviews.

In this chapter we explore the background to, and evidence for, different models of multidisciplinary working, and conclude by arguing for a more consistent implementation of care for those with HF.

HF management programmes tend to focus on the care of patients who have been admitted to hospital; this group has a high readmission rate and subsequent mortality, so there is much to be gained by improving care. Worryingly, 60% or more of HF patients continue to present in advanced HF, as acute admissions. This failure to make an early robust diagnosis and implement best care in the community is then frequently confounded by a failure to optimize management in the hospital setting.⁸ This is not intended to underestimate the difficulty, or time involved, in making a robust diagnosis in patients who are often, though not invariably, elderly, and who frequently suffer from a range of other conditions or comorbidities. These problems are further confounded by the lack of access to natriuretic peptide testing and echocardiography, both in primary care and in acute hospital settings. Following stabilization, the introduction of drugs—angiotensin-converting enzyme (ACE) inhibitors, β-blockers, and aldosterone antagonists—that prolong life and reduce hospital readmissions can be time consuming, and the optimization of these drugs more so. Too many patients leave hospital without an adequate diagnosis or even the introduction of basic life-enhancing medication,^8,9 and optimization of the medical therapy is unusual. Hence the inevitable cycle of inadequate treatment and early readmissions, with psychological and physical deterioration and high mortality, is established. Yet the length of hospital stay in the United Kingdom is long relative to Europe and the United States, suggesting that care is not well organized. These failures are confounded by other factors including insufficient discharge planning or poor follow-up. In this context poor patient self-care behaviour, including increased risk of noncompliance with medication and lifestyle change, or lack of symptom recognition,^10,11 should be no surprise. The management of patients with HF can be further complicated by behavioural, psychosocial, and financial considerations. Collectively these factors may contribute to 50% of HF exacerbations³ and over one-third of hospital readmissions. HF is one of the commonest causes of either admission or readmission to hospital in adults over 65,⁴ with about 2% of the NHS budget spent on HF-related care in the United Kingdom, though in absolute terms the United Kingdom spends less per patient on HF care than France, Germany, or the United States (see Fig. 54.1)¹² where outcomes are often better. Irrespective of these differentials, HF care is regarded as expensive across Europe, the United States, and Australia, with a high proportion of the costs attributable to hospitalization.⁵

• View full-sized figure
• Download figure as PowerPoint slide

Fig. 54.1

Absolute expenditure for heart failure in France, Germany, the United Kingdom, and the United States.

Data adapted from Bundkirchen A, Schwinger RHG. Epidemiology and economic burden of chronic heart failure. Eur Heart J Suppl 2004;6(D):D57–60.

Inadequately diagnosed and managed, a diagnosis of HF continues to carry an unacceptably high mortality.^8,13 For these reasons, in recent years there has been a growing interest in the development of more effective strategies for the care of patients with HF, including the use of HF management programmes. These programmes are designed to improve outcomes through a multidisciplinary approach with structured follow-up including patient education, outpatient optimization of medical treatment, psychosocial support, and better access to care when needed. The multidisciplinary teams may include any of a wide number of health care professionals, though inclusion of a HF nurse is usual, and in some HF management programmes may be the only intervention. The content and structure of HF management programmes vary from one health care setting to another, but they appear to be most effective where tailored to meet local needs and available infrastructure.¹⁴ With a few important exceptions,^15–18 inpatient care is not formally optimized (and rarely described) before recruitment, and the focus of interventions tends to be on subsequent outpatient care with an emphasis on reducing readmissions. Whilst this is an important consideration, an opportunity to reduce early (inpatient) and subsequent mortality and further reduce the readmission rate is often lost. These differences may explain the wide range of mortalities for the total cohorts reported from different studies (see Table 54.1), where similarly high-risk patients have been the focus of a range of interventions. Support for this thesis comes from the demonstration that one of the predictors of long-term outcome for patients who have been admitted to hospital with HF, from the 2009 National HF audit for England and Wales, is the number of (HF) drugs a patient is taking at the time of discharge (see Fig. 54.2) and better outcome for those whose management includes cardiologists and care within specialist wards.⁸

• View full-sized figure
• Download figure as PowerPoint slide

Fig. 54.2

Cumulative survival of heart failure patients over the course of a year. Patients on only one core drug have a worse prognosis compared to heart failure patients on two, whereas those taking three, or more core drugs for heart failure have the best survival.

Copyright © 2010, Re-used with the permission of The Health and Social Care Information Centre.

Attempts to understand and to interpret the diverse range of studies relating to HF programmes for patients who have been admitted to hospital, should recognize the considerable impact the quality of inpatient care for HF patients will have on immediate and long-term outcomes, and the extent to which any subsequent intervention confers benefit. Papers often provide little detail of this, but where standard inpatient HF care is minimal it will likely serve to exaggerate the benefits of the intervention under scrutiny. In contrast, good standard inpatient clinical care may make it more difficult for any intervention to demonstrate additional benefit (though should be reflected in mortality outcomes for the overall cohort). Similarly, in the literature there is often a dearth of information describing the standard availability of, and access to, HF and other services across the wider health communities. Yet this context may also exert important differential influences on the short- and longer-term patient outcomes, and the ability to demonstrate benefits of any intervention. A further complexity that impacts on an individual’s response to a diagnosis of HF, including any advice, is the illness beliefs of that person—a complexity we should recognize but which is for the most part beyond the scope of this chapter.^19,20

• Top
• Previous

Evidence base for multidisciplinary heart failure management programmes

One of the first published reports of multidisciplinary HF management programmes was a pilot study performed in the early 1990s by Rich and colleagues, exploring the feasibility of a multidisciplinary HF management programme in elderly patients.²¹ The subsequent study was a randomized controlled trial in 282 patients, powered to show a reduction in survival at 90 days without readmission. The randomized intervention—nurse-led education with review of medication by a geriatric cardiologist and intensive follow-up at home—did not achieve significance for the primary endpoint, but a range of secondary endpoints relating to readmissions were significant in favour of the intervention, and this appeared both to be cost-effective and to improve quality of life when compared with usual care. Other studies of home-based interventions and other multidisciplinary interventions followed, with numerous publications involving many thousands of patients, but only a relatively small proportion of these were randomized controlled trials, rarely blinded, and some more statistically robust than others.²²

Stewart’s work from Adelaide, Australia, examined the impact of a nurse-led intervention, with access to multidisciplinary input, on patients with recurrent hospitalization for HF.²³ The home visit was undertaken within 2 weeks of leaving hospital and included a thorough clinical assessment, review of medications, and identification of factors likely to provoke readmission. This resulted in a high level of additional input from others including GPs, cardiologists, pharmacists, and a range of social support. This was followed by telephone calls at 3 and 6 months. No detail is given of inpatient care, but reference to short admissions and the relatively high overall subsequent mortality might suggest no formal programme of inpatient optimization. The intervention significantly reduced the combined primary endpoint of the frequency of unplanned readmissions plus all-cause out-of-hospital deaths at 6 months, but there was no impact on out-of-hospital deaths or all-cause mortality at 6 months. The unplanned readmission rates were similar for the two groups, but overall unplanned bed usage was significantly less in the intervention group. Following a hospital admission the background level of care from a range of professionals was high, and services much more accessible than in the United Kingdom and some other European countries. Thus, typical follow-up for the patients receiving usual care included both a cardiology outpatient appointment and an appointment with their primary care physician within 14 days, and subsequent regular review by both. The study thus raised a number of interesting questions as to its reproducibility in other, less well-resourced, health care systems. An interesting finding was a trend towards increased elective bed usage in the intervention group, with most of this due to surgical intervention that had earlier been delayed when the patients were unstable. If we assume this contributed to the well-being of these patients, the finding illustrates the hazards of employing reduced bed usage as a surrogate for good care for patients with HF. Stewart had earlier reported a very similar study²⁴ where the intervention had also conferred benefit on their rather unusual combined primary endpoint of frequency of unplanned readmissions plus all-cause out-of-hospital deaths at 6 months.^23,25 The combining of these study cohorts and long-term follow-up suggest that the early intervention, delivered in the context of a well-resourced and well-supported health care community, continued to deliver considerable late benefit to the patients including reduced mortality (Fig. 54.3), lower readmission rates, and related cumulative bed day usage, while being highly cost-effective.²⁶ In discussing these late benefits the authors make an important point, which may be key to an effective multidisciplinary grouping: as the improved outcomes are observed by those involved in the ongoing care of the patients, good practice is reinforced and perpetuated over time.

• View full-sized figure
• Download figure as PowerPoint slide

Fig. 54.3

All-cause mortality, and the impact of a home-based intervention (HBI), on mortality in a group of patients hospitalized for heart failure and then randomized to either usual care or HBI with all patients followed up for a minimum of 7.5 years.

From Inglis SC, Pearson S, Treen S, Gallasch T, Horowitz JD, Stewart S. Extending the horizon in chronic heart failure: effects of multidisciplinary, home-based intervention relative to usual care. Circulation 2006;114(23):2466–73, with permission.

From Scotland came further evidence of the effectiveness of nurse-led interventions, supported by the local HF cardiologists, in reducing HF readmissions.²⁵ These patients were not required to receive any predetermined standard of inpatient care, but following recruitment (from the cohort undergoing echocardiography, in whom echocardiography confirmed that the HF failure was due to left ventricular systolic dysfunction, LVSD) were randomized either to intervention or to standard care under the admitting physician and subsequently their GP. Those randomized to intervention received input from the HF nurse during that admission and then a home visit within 48 h and subsequent intensive input with visits at 1, 3, and 6 weeks and then at 3, 6, 9 and 12 months. The nurses additionally made 10 phone calls at prespecified time-points and encouraged further contact from the patient and their carers. The nurse input combined education with encouraging the patient’s awareness of symptoms, and psychological support. The HF nurses also aimed to up-titrate drugs, using written protocols. The primary endpoint was a combined endpoint of all-cause mortality or hospital admission for HF at 12 months, for which a significant difference was reported in favour of those receiving the intervention; though, had the very different inpatient mortalities (during the index admission) not been included, statistical significance would not have been reached.²² At 12 months HF readmissions and related bed usage were both significantly reduced, but there was no impact on all-cause admissions, all-cause hospital bed usage, or all-cause mortality, which was of the order of 30% for the total cohort, and similar to that of Stewart et al.²⁴

McDonald’s group, from Ireland, report a rather different approach, initially as a pilot and subsequently as a randomized controlled trial.¹⁶ In this study all patients with HF received inpatient optimization of care, including echocardiography, and medical therapy (then defined for those with impaired systolic function as diuretics, digoxin, and ACE inhibitors at maximally tolerated doses). They were required to satisfy predefined clinical and stability criteria before discharge. Those randomized to the active intervention received three or more consultations with a nurse specialist and dietician while inpatients, and subsequently were telephoned within 3 days of leaving hospital and weekly thereafter. Both these telephone calls and the HF clinic appointments at 2, 6, and 12 weeks allowed for clinical reassessment and reinforcement of the educational messages. In clinic appointments, intravenous infusions of diuretics could be administered when necessary. In contrast, following optimization of inpatient care, including predischarge clinical stability, the control group were referred back to their primary physician for any ongoing care. The study achieved statistical significance for the primary combined endpoint of death or HF readmission at 12 weeks, with the intervention conferring benefit. Although this paper is not without its own statistical quirks,²² arguably its greatest interest lies in the very low mortality at 3 months (7.1%) for the entire cohort, which it can be argued reflects the strategy of inpatient optimization of care.

In this context the mortality rates we reported from a randomized controlled trial of a low-intensity, nurse-led, self-management intervention, in patients who were hospitalized for HF due to LVSD and recruited when stable following a routine strategy of cardiology-led inpatient optimization of care, are of interest.^17,18 It is of note that the patients in this cohort were elderly and high risk, with multiple comorbidities, and referred from primary care populations noted to have a year-on-year excess cardiovascular mortality, when compared with the United Kingdom, London, and similar Primary Care Trusts.

This study was designed to explore the possibility that a low-intensity, nurse-led, self-management intervention using a problem-solving strategy would reduce HF readmissions. Following recruitment, those randomized to the intervention were visited by a nurse twice while in hospital and then again at home within 10 days of leaving hospital. Thereafter the nurse phoned the patients once, but was available by phone should the patient initiate further telephone contact.

In all other respects there was no difference in the HF care between the two randomized cohorts which included a hospital-wide, protocol-driven, shared-care strategy involving cardiologists and the admitting physicians, aimed at early diagnosis and subsequent inpatient optimization of ACE inhibitors, diuretics including aldosterone receptor antagonists, and, for some, initiation of β-blockade. Patients were allowed home when they were clinically and biochemically stable without treatment changes for 48 h. Thereafter all patients had early and continued cardiology review. The broader context was that when the study was designed and undertaken there were no HF nurses employed either by the hospital or by the local community, and there were no GPs with a special interest in HF. Provisional results, reported elsewhere, demonstrated no difference in the readmission rates at either 3 or 12 months between the two randomized groups. The all-cause mortality for the total cohort was relatively low at both 3 months¹⁷ (and similar to MacDonald’s study), and again at 12 months (17.6%). We would suggest this reflects the early inpatient optimization of care for all.¹⁸

Interestingly, using post-hoc subset analyses, we demonstrated a differential response between those admitted with a pre-existing diagnosis and those admitted with a new diagnosis (incident HF). In the latter group the randomized intervention conferred benefit in terms of reduced HF readmissions and bed usage at 12 months (see Fig. 54.4), and importantly there was no mortality penalty, but rather a trend towards improved mortality.¹⁸

• View full-sized figure
• Download figure as PowerPoint slide

Fig. 54.4

The impact of a low-intensity self-management intervention on event-free survival for patients admitted to hospital with incident heart failure (due to left ventricular systolic dysfunction). Results based on post-hoc subset analysis of a randomized controlled blinded trial.

From Zaphiriou A, Mulligan K, Hargrave P, et al. Improved outcomes following hospitalisation in patients with a new diagnosis of heart failure: Results from a randomised controlled trial of a novel, nurse-led self-management intervention Heart 2006;92:A119–A120.

Thus far we have explored in some detail just a few publications from an extensive literature. Elsewhere papers argue for the benefits of nurse-led HF clinics,²⁷ strategies around patient education,^28,29 the use of nurses and or pharmacists to effect the up-titration of drugs using agreed protocols and treatment algorithms³⁰ as hospital outpatient led initiatives,^{16,27,31–33} or community-based approaches including programmes where home visits are key.^{23–26,29,34} Many of the studies use phone calls to reinforce care and provide increased access to additional care. The phone calls can provide reassurance and allow patients the opportunity to discuss symptoms, treatment, and side effects where contact with a programme has already been established.^{16,18,24,25,32,35} Phone calls may be less well received where there has been no prior contact, or where the patients are elderly or have some degree of mental impairment. The literature may be especially vulnerable to publication bias here, in favour of studies with positive outcomes. Patients who are elderly, immobile, or less mentally alert may derive particular benefit from home visits, but this can be a much more costly intervention than clinic-led care. Increasingly in some health care systems this community care is being devolved to other health care professionals who may have little or no specific training in the area of HF.

Remote monitoring is undoubtedly an emerging model for delivering components of HF care, either to a large group of individuals who may not have access to traditional programmes, or as an adjunct to other programmes of HF care. The subject is dealt with in detail elsewhere, but involves daily monitoring of symptoms and signs measured by patients, family, or caregivers at home, so allowing patients to remain under close supervision³⁶ in their own home. A range of equipment may be installed in the home allowing variables such as blood pressure, heart rate, oxygen saturation, weight, symptoms, and medication compliance to be recorded and transmitted. This then allows a designated health care worker to remotely advise on changes in care. The success of remote monitoring is highly dependent on patient education and the ease with which the equipment can be used, both by those setting it up and by those endeavouring to use it. Central to its effectiveness is patient, and or carer, education, which is usually time consuming, and more difficult in those with cognitive impairment. Currently there appears to be no consensus regarding which variables are most helpful to monitor. New equipment, additional monitoring parameters, and increasingly sophisticated technologies are under development.³⁶ Nonetheless, themes discussed earlier re-emerge from the telemonitoring literature, namely that the extent of the benefit appears greatest where baseline care is less well developed. This may explain in part the differences in outcomes between the Trans-European Network Initiative (TENS-HMS)¹³ and those of the Home-HF study.³⁷ ‘Usual care’ from the Trans-European Network Initiative reported the highest 12-month mortality of 48%, but it was also high both in those receiving home telemonitoring, at 29%, and in those receiving nurse telephone support, at 27%. In contrast, in the Home-HF study, 6 month mortality rates were 9.8% for the intervention group and 5.4% in the usual care group, though the study had not been powered to demonstrate a mortality difference. The message from this may be simple and appears to be supported by details of those patients studied: namely, that baseline HF care for the patients being recruited from three hospitals in north-west London was probably better than that received by their counterparts recruited from hospitals in the United Kingdom, Germany and the Netherlands. The home telemonitoring from TENS-HMS did not reduce readmissions, but the mean duration of stay was shorter by 6 days for those randomized to home telemonitoring, whereas Home-HF reported no reduction in days alive and out of hospital but avoided readmissions via the emergency department and was cost neutral.

The HF literature exploring the impact of various interventions ranges from tiny studies to large, well-devised randomized controlled trials with an accompanying range of outcomes, so the extensive literature also includes systematic reviews and meta-analyses.^22,38,39 All recognize the heterogeneity of the included studies (even in the absence of descriptors of standard care) and so exclude and group the studies in a range of different ways.^22,38,39 McAllister and Holland both endeavoured to report on whether multidisciplinary strategies improve outcomes for HF patients, with the main emphasis on mortality and readmissions. McAllister’s group concluded that where follow-up included a specialized multidisciplinary team both HF admissions and mortality were improved, whereas interventions designed to enhance self-care and telephone follow-up or telemonitoring reduced admissions but had no impact on mortality. In contrast, Holland’s group found that it was the postdischarge interventions which combined education with self-management strategies that reduced mortality alongside readmissions. As ever, the answer lies in the detail, much of which goes unpublished in the interest of brevity.^38,39

So, can the limitations of meta-analyses and structured reviews be overcome by larger studies? The Coordinating study evaluating Outcomes of Advising and Counselling in Heart Failure (COACH),³² aimed at examining the effect of education and an intense support programme by HF nurses on top of frequent visits with a cardiologist, is of considerable interest in this context, and poses some fundamental questions which challenge earlier accepted conclusions relating to HF care. As so often, there is no detail surrounding the quality or norm for inpatient care or whether inpatient optimization is usual. Of note is a reduction from NYHA class III or IV on admission to NHYA class II/III for the vast majority by discharge, though curiously no patients were rendered asymptomatic despite high levels of prescribing of ACE inhibitors/angiotension receptor blockers (83%), diuretics (95%), and β-blockers (66%). These figures of course do not tell us anything about dose levels or timing of up-titration. All patients saw a cardiologist on at least four occasions and additionally as dictated clinically. On this background of usual care there was a three-way randomization to a control group, a basic support group (where there was a nurse-led programme of structured education and outpatient visits, but no home-based intervention), or an intensive nurse-led support programme where contacts with the nurse were monthly over the 18-month study period, and included two home visits and some multidisciplinary input from a physiotherapist, dietitian, and social worker. This multicentre study recruited and randomized more than 1000 patients from across the Netherlands with a mean age 71 ± 11 years, of whom 38% were women, and then followed them up over an 18-month period. Although there was no significant difference in the primary endpoint of time to death or rehospitalization and the number of days lost to death or hospitalization, the all-cause mortality for the entire cohort (with no significant differences between the groups) was relatively low at around 21–22% at 12 months. The COACH authors conclude that the patients in the control group were already well managed, making it more difficult to improve the outcome with the added intervention. We concur with that conclusion, noting that the mortality rate at 12 months is substantially lower than that of many published studies,^23–25 but in keeping with our own.¹⁸ The authors further note that hospital admissions may be beneficial (where care is carefully and appropriately targeted at the patients needs), rather than deleterious, noting that in the intervention groups a nonsignificant trend to reduction in mortality was accompanied by more frequent, shorter hospital admissions.

So what might we conclude? A critical reading of the literature and recognition of its limitations should not in any way discourage, but rather support, the establishment of effective and responsive multidisciplinary teams for the care of people with known or suspected HF. Encouragingly, these aims are consistent with the patients’ priorities that emerged from the Health Care Commission within England (see Box 54.1).

To date the emphasis of these groupings has been community based but a number of studies and emerging audit data argue powerfully, in addition, for an improvement in hospital care where best outcomes can be achieved when patients are proactively identified, and ongoing care, including early echocardiography, is led by a (HF) cardiologist, and ideally within a specialist ward where a range of health care professionals will be involved to ensure rapid stabilization and inpatient optimization of treatment. As the patient recovers opportunities for education arise and should be embraced—and this is the ideal time for first contacts with those who will be involved in their care when they leave hospital. This often involves a HF nurse specialist who may subsequently see the patients in clinics or their own homes, but might equally be a number of others. Discharge planning should include plans for primary care and cardiology review, rehabilitation, β-blocker up-titration and ongoing opportunities for individuals to understand their condition. It is important that the patient knows whom to contact and how to do so in the early days following a hospital admission when they may be most vulnerable. For individual patients, input from others may make the difference between ongoing improvement and an early, unplanned readmission. The nature of this input may be diverse—certain needs will be identified before the patient goes home but others may only become apparent in the days after leaving hospital and this argues for an early home assessment which, wherever possible, should be inclusive of spouse or carer.

And so we see the emerging need for an established multidisciplinary team, ideally led by the HF cardiologist who has been involved with the care of the patient during the index admission, working closely with the GP and HF nurse specialist, but with access to many others. The skill in building this team is to establish mechanisms of support, mutual trust, and learning so that the patient’s changing needs may be met in a timely fashion and delivered in a cost-effective manner across the different health care communities. Table 54.2 summarizes important components of a multidisciplinary HF team. This team will include people who may never directly meet many of the patients but their expertise can be drawn upon. It is essential that when discussions occur that they include a clinician who knows the patient well.

Thus, rather than an overriding obsession with length of stay the HF community needs to focus on the quality of care during an index HF admission and thereafter within the patient’s community, recognizing that for some timely and planned readmissions may be part of a health care strategy that delivers a much lower mortality rate than we have seen hitherto, and an improved quality of life. Timely access to a range of expertise through a multidisciplinary team is essential for the patient, with the HF cardiologist, a HF nurse specialist, and the GP ideally at the centre of this care delivery (Fig. 54.5). This will allow specialist input when necessary alongside the input of the generalist, which is also critical given the wide range of comorbidities that confound and complicate the problems of this patient group. Over time, continuity of care and (wherever practicable) self-management will reduce the intensity of support needed by some and allow individual patients to re-establish their sense of normality.

• View full-sized figure
• Download figure as PowerPoint slide

Fig. 54.5

Model of an integrated heart failure service. Like Table 54.3, this emphasizes the need for multidisciplinary working across the different health care domains.

Adapted from McDonagh TA. Lessons from the management of chronic heart failure. Heart 2005;91(Suppl 2):ii24–7, with permission.