INTRODUCTION — Systems for outpatient, inpatient, and transitional care for patients with heart failure (HF) must aim to comprehensively address the complex medical and social issues that contribute to high rates of hospitalization in this population. Strategies include optimizing evidence-based drug and device therapies, addressing causes and precipitants of HF, treating comorbidities, and improving management of care. The clinician, working with others in a care team, must provide appropriate coordination of care, manage the patient between office visits, ensure adequate patient education to support the acquisition of self-care skills, address social determinants of health, and create safe transitions of care between settings. This includes reducing disruption to care during and following hospitalizations [1,2].
HF care involves more than focused medical management of an isolated syndrome, since patients with HF typically have multiple medical conditions, often treated by many clinicians. Patients with HF are typically older adults with complex drug regimens for HF and multiple concurrent diagnoses with resulting polypharmacy [3]. Moreover, self-management is particularly important in HF, but HF patients may have trouble with balance, eyesight, hearing, and/or cognitive function that makes self-management difficult and may increase the risk of readmission [4]. (See "Heart failure self-management".)
This topic will review outpatient- and inpatient-based management strategies aimed at reducing the risk of hospitalization and rehospitalization for patients with HF. The following issues are discussed separately:
●Patient self-management of HF. (See "Heart failure self-management".)
●Management of chronic HF (with reduced ejection fraction, mid-range ejection fraction, or preserved ejection fraction). Management of other causes of HF (including valve disease) is discussed in individual topic reviews. (See "Overview of the management of heart failure with reduced ejection fraction in adults" and "Treatment and prognosis of heart failure with mid-range ejection fraction" and "Treatment and prognosis of heart failure with preserved ejection fraction".)
●Management of acute heart failure. (See "Treatment of acute decompensated heart failure: General considerations" and "Treatment of acute decompensated heart failure: Specific therapies" and "Treatment of acute decompensated heart failure in acute coronary syndromes".)
EPIDEMIOLOGY — HF is one of the most common causes of hospitalization and many patients who are hospitalized for HF are readmitted within 30 days of discharge. An estimated 26 million people worldwide have HF [5]. Approximately 6.2 million Americans and 15 million Europeans are living with HF [6,7]. In 2016, the HF hospitalization rate in the United States was 2.8 per 1000 for all ages and 13.4 per 1000 for those over age 64 [8], making HF among the top 10 causes of hospitalization for the full population and among the top three for those over 65 [8]. Furthermore, almost one-fifth of those who are hospitalized with HF are readmitted within 30 days [9].
In the United States, both hospitalizations and rehospitalizations for patients with HF have been declining, with greater decreases in admissions than in readmissions. For instance, admission rates fell 30.5 percent (95% CI 29.3-31.6) between 1999 and 2011, while readmission rates declined 9.7 percent (95% CI 8.5-10.8) over the same period [10,11]. Similar patterns have been seen for hospitalizations for all causes in the United States. From 2010 to 2013, hospital referral region-level admission rates declined from 30.8 to 27.1 per 100 person-years at risk, while readmissions declined from 17.2 to 16.1 per 100 hospital discharges [12]. However, since passage of the Affordable Care Act in 2010 in the United States, which included penalties for higher-than-average readmission rates for HF, United States readmissions for HF have fallen faster than for other conditions [13]. Importantly, many readmissions after HF hospitalization are not for HF [14].
Mortality for patients with HF has been increasing [15] since at least 2006 [16], which has raised concern about unintended effects of readmission reduction, but hospital risk-standardized rehospitalization rates and mortality rates are only weakly and negatively correlated [17-19].
OUTPATIENT STRATEGIES — Outpatient care requires coordinated management of medical and social issues.
Key components of outpatient care — Outpatient care for HF should generally include the following components:
●A process for identifying patients with HF. (See "Heart failure: Clinical manifestations and diagnosis in adults", section on 'When to suspect heart failure'.)
●Clinical assessment:
•Confirm presence/type of HF (eg, physical examination and echocardiography). (See "Heart failure: Clinical manifestations and diagnosis in adults".)
•Assess functional capacity (eg, initial and subsequent six-minute walk). (See "Determining the etiology and severity of heart failure or cardiomyopathy".)
●Psychosocial care:
•Determine individual needs and address social determinants of health (eg, apply validated tool to assess concurrent depression, social support needs, and caretaker’s coping). (See "Screening for depression in adults" and 'Address social determinants' below.)
•Develop and document a plan personalized to the patient’s culture and language.
●Coordinated care – Whenever possible, patients should have access to a care team including physicians (primary care provider, cardiologist, and others to treat comorbid conditions), nurses, care coordinators, social workers, pharmacists, physical therapists, nutritionists, and others as needed. Care should include frequent, coordinated contact with providers, with regular in-person contact (in clinic or home visits) when possible, supplemented by telephone or video contact, as needed.
●Medical management:
•Treat with drug and device therapies according to evidence-based guidelines. (See "Overview of the management of heart failure with reduced ejection fraction in adults" and "Initial pharmacologic therapy of heart failure with reduced ejection fraction in adults" and "Secondary pharmacologic therapy in heart failure with reduced ejection fraction (HFrEF) in adults" and "Treatment and prognosis of heart failure with mid-range ejection fraction" and "Treatment and prognosis of heart failure with preserved ejection fraction".)
•Develop a treatment plan to manage comorbid and related conditions.
•Provide routine preventive care (eg, immunizations, treatment of risk factors, prevention of thromboembolism).
•Develop personalized treatment plans (eg, personalized exercise program, well-defined medication regimen).
●Comprehensive patient education particularly focused on self-care (see "Heart failure self-management"):
•Assess patient capacity for self-care (eg, formal assessment of patient ability to self-care, health literacy, cognitive function, and screen for depression).
•Provide appropriate information and support for self-care (eg, ensuring understanding of the causes and consequences of HF, purpose of medicines, medicines to avoid).
•Develop action plan and provide clear instructions to patients and caretakers (eg, daily weight monitoring and recording in personal diary, personalized actions to take when weight or symptoms change). (See "Heart failure self-management", section on 'Heart failure action plan'.)
●Palliative care:
•Discuss and assess palliative care needs (eg, routine discussion and offering of advance directives and reassessment and documentation of palliative care options). (See "Legal aspects in palliative and end-of-life care in the United States" and "Palliative care for patients with advanced heart failure: Indications and systems of care" and "Palliative care for patients with advanced heart failure: Decision support and management of symptoms".)
●Early postdischarge support if hospitalized. (See 'Inpatient and transitional care strategies' below.)
Structured programs — The key components listed above are typically beyond the expertise or time of any individual clinician to meet. Accordingly, major society guidelines and position statements for management of HF have endorsed disease management [20] or have noted the importance of multidisciplinary care [21-23] to achieve comprehensive outpatient care of the patient with HF. Disease-management programs are structured programs typically providing a multidisciplinary, integrated approach to care for patients with chronic conditions. Disease-management programs facilitate the development of an individualized care plan, provide patient education and support, and coordinate care among health care providers, patients, and patients’ support systems.
Disease management programs include a wide variety of care models, including nurse-led case management, multidisciplinary HF clinics, programs promoting self-care, telephone-based case management, and telemonitoring. The efficacies of these approaches differ, as described below. The generalizability and validity of the available evidence are difficult to ascertain given the wide variety of interventions tried and personnel involved and the fact that the evaluators and developers are often the same and blinding is not possible. Furthermore, as with all complex interventions of practice improvements, local context is critical for disease management, and the quality of implementation may be as important to success as the strategy itself [24].
Address social determinants — Systems of care for patients with HF should include strategies to screen for and address social determinants of health (SDOH). SDOH are defined by the World Health Organization as “the circumstances in which people are born, grow up, live, work and age” [25] and are a substantial contributor to long-term morbidity and mortality among patients with HF apart from any medical treatment. Adverse social determinants (including limited access to food, housing, and transportation; reduced education and health literacy; lack of social support; and financial challenges) add stress to patients’ lives, take time and resources away from self-care of medical conditions, compete with medical needs, and reduce access to high-quality health care. It is well established that disability, race, ethnicity, insurance status, household income, adverse neighborhood characteristics, and other markers of social challenges are associated with higher readmission rates and mortality for patients with HF [26-29].
Conversely, a report by the Robert Wood Johnson Foundation found that addressing SDOH was a key component of Accountable Care Organizations that succeeded in achieving cost and quality targets [30]. The 2020 American Heart Association (AHA) Scientific Statement on Social Determinants of Health recommended that interprofessional collaborative practice teams implement interventions to address SDOH [31]. The statement includes specific recommendations in the areas of employment, access to care, health care coverage, food insecurity, environment, race, social support, and health literacy.
At a minimum we recommend the following strategies:
●Ensure the full clinical team is trained in principles of SDOH, including what they are, how they influence health, how they can be identified and measured, and how they can be addressed within the health system. Include social workers in the care team.
●Incorporate routine screening for SDOH into clinical care (ideally integrated into the electronic health record) using a standardized tool such as the Accountable Health Communities Social Needs Screening Tool [32], the OCHIN Social Determinant of Health Screening Tool (available in multiple languages [33]), or the THRIVE screener [34].
●Maintain an updated list of community resources for referrals for patients screening positive; a number of commercial entities now exist to provide such connections, including Aunt Bertha, Healthify, NowPow, Unite Us, and others [35]. Merely screening without such referrals is ineffective.
●Make easily available to clinicians the resources that help ensure social determinant-sensitive health care, such as the USD $4 prescription formulary at local pharmacies, patient education materials that include strategies for low-cost but healthy food (ie, frozen vegetables, inexpensive low-sodium food, best options at fast food establishments, nutrition apps for diet management), and patient education materials for exercise that can be incorporated into daily life without additional expense.
●Establish referral connections to mental health, cardiology, and other providers that are accessible for patients with social determinant challenges such as lack of insurance, disabilities, or lack of fluency in English.
●Ensure patient materials are available in the main languages spoken by patients in the practice, at an appropriate literacy level.
Specific models of outpatient care — Strategies for outpatient HF care include a variety of structured, integrated programs (known as disease management) as well as outpatient cardiologist care, outpatient infusion centers, and hospital at home.
In general, the more comprehensive, multidisciplinary, and in-person the outpatient intervention, the more evidence there is for effectiveness. In summary, effective outpatient HF care models include:
●Nurse-led case management, including drug titration. (See 'Nurse-led titration of drugs' below.)
●Multidisciplinary HF clinic care provided by HF specialist teams (including, at a minimum, an HF cardiologist and HF nurse). (See 'Multidisciplinary heart failure clinics' below.)
●Programs emphasizing self-care education and support. Self-management interventions are discussed separately. (See "Heart failure self-management", section on 'Benefit of self-management interventions'.)
Observational studies have found associations between outpatient cardiologist care and lower mortality rates, although a causal relationship has not been established. (See 'Outpatient cardiologist care' below.)
By contrast, outpatient programs limited to telephone contact or remote data gathering (telemonitoring) were the least likely to be effective [36]. (See 'Telephone-based case management' below and 'Telemonitoring' below.)
More effective models
Nurse-led titration of drugs — Nurse-led case management including titration of drugs is an effective means of improving outcomes for patients with HF. A 2015 Cochrane review specifically examined HF disease-management programs that included nurse-led titration of HF-related medications, which could occur either in an HF clinic setting or via home visits. A total of seven randomized controlled trials (RCTs) with 1684 subjects were included. Collectively, patients exposed to disease management with nurse-led titration had fewer hospitalizations (risk ratio [RR] 0.80, 95% CI 0.72-0.88, high-quality evidence) and lower mortality (RR 0.66, 95% CI 0.48-0.92, moderate-quality evidence) [37].
Multidisciplinary heart failure clinics — Care by a multidisciplinary HF clinic appears to reduce hospitalizations. Trials of multidisciplinary HF clinic intervention have emphasized care by a multidisciplinary team (including, at a minimum, a cardiologist and nurse specialist, with some teams also including a dietician, pharmacist, and/or social worker/case manager), which contrasts with the approach of some trials of nurse-led clinics [38]. An increasing number of multidisciplinary HF clinics have been developed, often in concert with dedicated infusion centers capable of providing intravenous diuretic or inotropic treatment to patients on an outpatient basis.
A meta-analysis of 10 high-quality RCTs found that multidisciplinary HF clinics, typically involving HF nurses, cardiologists, geriatricians, psychologists, dieticians, and physiotherapists, reduced all-cause readmission rates (RR 0.80, 95% CI 0.66-0.97) as well as all-cause mortality (RR 0.80, 95% CI 0.67-0.97) [39].
Similarly, a meta-analysis that included eight RCTs of multidisciplinary disease management programs found reduction in mortality (RR 0.67, 95% CI 0.54-0.83) and an uncertain effect on readmission rates (RR 0.85, 95% CI 0.71-1.01) [40]. However, this meta-analysis also included a separate set of seven RCTs of clinic-based care, and for this group of studies the clinical benefit was uncertain (RR for mortality 0.87, 95% CI 0.68-1.10 and RR for readmission 0.90, 95% CI 0.72-1.12).
Programs promoting self-care — These programs are discussed separately. (See "Heart failure self-management", section on 'Promoting effective self-care'.)
Outpatient cardiologist care — Observational studies have reported better patient outcomes when a cardiologist participated in outpatient HF management [41], particularly when a cardiologist and primary care provider were both involved with care [42,43], compared with management solely by a generalist (general medicine or primary care provider). These studies report an approximately 20 to 40 percent reduction in mortality, or mortality combined with hospitalization, with management by both a cardiologist and a generalist as compared with care solely by a generalist.
Improved survival may be due to greater guideline adherence when a cardiologist is involved [41,43-49], as illustrated by the following studies:
●A retrospective study of 10,599 HF patients discharged from the emergency department in Ontario, Canada, found that patients seen by both cardiologists and primary care physicians versus seen solely by primary care physicians in the first 30 days postdischarge had a lower mortality rate at one year (hazard ratio [HR] 0.79, 95% CI 0.64-1.00) [43]. Mortality was lower among patients who were seen by a primary care physician compared with those who saw no physician at all (HR 0.75, 95% CI 0.64-0.87) in the first 30 days. There was a nonsignificant trend toward higher mortality with sole cardiology care as compared with combined cardiology and primary care (HR 1.41, 95% CI 0.98-2.03).
Patients receiving combined cardiology and primary care were more likely than those receiving solely primary care to undergo assessment of left ventricular function (57 versus 29 percent) and more likely to receive standard HF medications, including angiotensin-converting enzyme (ACE) inhibitors, angiotensin II receptor blockers, beta blockers, and spironolactone. However, information on left ventricular function was not available, so differences in appropriateness of care for HF with reduced ejection fraction were not assessed.
●In an earlier study from the Kaiser system of 403 patients with new-onset HF, cardiology care was an independent predictor of lower risk of death or cardiovascular hospitalization at 24 months (HR 0.62, 95% CI 0.42-0.93) [41]. Patients with a low ejection fraction who were seen by a cardiologist were significantly more likely than those seen only by a primary care physician to have been treated with ACE inhibitors (91 versus 71 percent) or beta blockers (38 versus 21 percent).
However, the role of differences in patient populations seen by cardiologists and generalists versus by generalists alone is uncertain. Although patients seen by cardiologists have more severe cardiac disease [41,43], they also tend to be younger and to have fewer comorbidities [41,42]. They may also, therefore, have fewer contraindications to medical therapies. It remains uncertain whether the improvement in guideline adherence is due to cardiologist involvement or to a patient cohort with fewer contraindications.
Less effective models
Telephone-based case management — Most disease management studies primarily involve telephone-based management, typically by a nurse; these studies have yielded mixed results. One meta-analysis found that of 47 RCTs of disease management, 40 involved some type of telephone follow-up. Among 26 low-quality RCTs involving community case management (largely but not entirely by telephone), the RR for mortality was 0.78 (95% CI 0.68-0.90) and for readmission the RR was 0.92 (95% CI 0.83-1.01). Meta-analyses of studies restricted to telephone support have largely found less benefit. For instance, one meta-analysis of nine RCTs of telephone support found no mortality benefit (RR 0.82, 95% CI 0.62-1.08) and no readmission benefit (RR 0.86, 95% CI 0.64-1.15). However, five of the six largest RCTs of disease management involved the telephone-based case management approach [50-54], some of which were successful:
●A randomized trial of 174,120 patients in the United States assigned patients to telephone-based disease-management support ("enhanced support") or usual care [50]. Patients in the enhanced support arm had a 10.1 percent reduction in annual hospital admissions compared with those in the control group. The study was sponsored and conducted by a for-profit company that owns the intervention and was designed as a quality improvement study rather than a research study, meaning that the decision about publication followed the study rather than preceded it.
●The DIAL trial (Randomized Trial of Phone Intervention in Chronic Heart Failure) randomly assigned 1518 Argentinean patients with HF to either routine care or an intervention consisting of an explanatory booklet plus periodic telephone contact by a specialized nurse over the course of one year [51]. The nurse performed telephone evaluation and enforced self-management education, adjusted medications, and recommended nonscheduled medical or emergency visits. The intervention resulted in a lower rate of death or hospitalization for HF compared with usual care at a mean of 16 months (26.3 versus 31 percent) [51]; a small benefit was sustained three years after the intervention ended (55.7 versus 57.5 percent) [55]. This benefit was related mostly to a reduction in admission for HF.
●A randomized trial of 1069 patients with HF assigned two-thirds of the patients to an intervention consisting of home scales and a disease manager who provided telephone management over 18 months [52]. Intervention patients, compared with controls, had increased survival (mean survival of 526.9 versus 450.5 days) and improved New York Heart Association functional class, but there was no difference in overall utilization (hospitalizations, outpatient visits, medications, and procedures) or costs.
In contrast, the following two studies involving nurse-led telephone-based care found no benefit:
●The largest study of chronic disease management reported to date evaluated eight Medicare Health Support pilot programs around the United States, which randomly assigned over 240,000 fee-for-service Medicare patients, most with HF, to usual care or participation in disease-management programs sponsored by Medicare [54]. These were commercial disease-management programs, not directly involved with the patient’s clinical practice, and included nurse-led telephone-based chronic care services (symptom monitoring and management, medication counseling, health education, self-care coaching). Approximately 75 to 96 percent of patients in the intervention groups participated in the interventions. However, evaluators identified little improvement in self-care practices in the intervention group. After 18 months, evaluators found no significant difference between intervention and control groups in hospitalizations, readmissions, emergency department utilization, or mortality. Possible reasons for lack of success included the lack of integration of the disease-management program into the beneficiary’s primary health care team, the infrequency of phone contact (on average 80 days between contact), contacts not coinciding with the time of disease flare or hospitalization, and the large number of comorbid illnesses in this patient population.
●The Israel Heart Failure Disease Management Study (IHF-DMS) was a randomized trial of 1360 patients with HF who were randomly assigned to usual care (by primary care practitioners and consultant cardiologists) or disease management consisting of care coordination, patient education, symptom monitoring, medication titration, and telemonitoring of weight, blood pressure, and heart rate [53]. Most care was by telephone calls by nurses between six-month HF clinic visits. There was no significant difference between the two groups in time to death or HF-related hospitalization (adjusted HR 0.91, 95% CI 0.79-1.05).
Telemonitoring — The goal of telemonitoring is to detect and intervene on signs of HF decompensation to avoid clinical deterioration or readmission. However, trials of telemonitoring have not established the benefit of the general approach of telemonitoring.
●Definition and goals – Telemonitoring is a specific type of chronic disease management in which patients are provided with equipment, such as blood pressure monitors and scales that electronically transmit their measurements to clinicians. Upon receipt of this information, health care teams review it and respond with interventions that include appropriate changes to specific medications (eg, diuretics) or closer follow-up in clinic. Telemonitoring may be a standalone intervention, but it can also be combined with other disease-management strategies. Telemonitoring should be distinguished from direct telephone calls with clinicians (ie, "telehealth"). The challenges of telemonitoring include ensuring that the data are submitted (strategies include self-reporting versus automated data transmission) and that they are appropriately reviewed and addressed.
Implantable monitors (eg, pulmonary artery pressure monitor) function similarly to telemonitoring and are discussed elsewhere. (See "Treatment and prognosis of heart failure with preserved ejection fraction", section on 'Device-based therapies'.)
●Evidence – The available data suggest that telemonitoring may not have a large effect on reducing readmissions or mortality among patients recently hospitalized with HF. However, these complex interventions require acquisition and transmission of health care data as well as an effective response system; sending data alone is not a meaningful intervention. Thus, the trials of these complex systems are challenging to interpret; no two studies evaluated the same set of interventions, and the reasons for negative trial results are often unclear (eg, ineffective measurement, ineffective response system, or both).
The 2018 Heart Failure Society of America white paper on remote monitoring noted that routine use of external remote patient management devices is not recommended [56].
The trials that evaluated telemonitoring include:
●In a trial of 552 patients with HF who were recently discharged after admission for HF, patients were randomly assigned to usual care or to a complex intervention that included remote scales, electronic pill bottles that generated remote alerts to clinicians, and notable use of financial incentives for adherence based on behavioral economic theory [57]. Despite more than 3700 alerts generated by the complex intervention, the rates of readmission or death were similar between the two groups (64 versus 63 percent in the complex intervention group had at least one event; adjusted HR 0.90, 95% CI 0.72-1.12). The effect was similar in patients with HF with reduced ejection fraction and HF with preserved ejection fraction. Notably, a clinician response to an alert for a change in weight was only documented in 34 percent of such alerts.
●A meta-analysis of 37 RCTs including 9582 patients reported on 24 studies with all-cause mortality and 17 studies with all-cause hospitalization. Telemonitoring with noninvasive devices reduced all-cause mortality (RR 0.81, 95% CI 0.70-0.94) but not all-cause hospitalizations (RR 0.94, 95% CI 0.85-1.03) [58].
●The BEAT-HF trial enrolled 1437 patients with HF at six sites in the United States [59]. Participants were randomly assigned to health coaching telephone calls and telemonitoring or usual care. No differences in rates of readmission for any reason (51 versus 49) or mortality at 180 days were found. Although 180-day quality-of-life scores were slightly better in the intervention group compared with the usual care group, data were available for only 55 percent of participants.
●The Tele-HF trial enrolled 1653 patients with HF from 33 sites in the United States [60]. Subjects were randomly assigned to telemonitoring using an automated telephone-based system that collected daily information about symptoms and weight or to usual care. There were no significant differences between the two treatment groups for the primary endpoint (ie, death or all-cause readmission) or for the secondary endpoint (ie, HF readmission).
Models with unclear benefit
Outpatient infusion centers — Evidence for outpatient infusion centers is limited. Initial enthusiasm for intermittent parenteral inotropic or vasodilator therapy 20 to 30 years ago has been tempered by later negative evidence, such as the FUSION II trial (911 patients) that demonstrated no benefit of intermittent outpatient nesiritide in avoiding death or hospitalization in 12 weeks (HR 1.03, 95% CI 0.82-1.3) [61] and a systematic review of intermittent outpatient dobutamine (310 patients) that showed no mortality benefit (odds ratio 1.47, 95% CI 0.98-2.21) [62]. The 2013 American College of Cardiology/AHA guidelines for management of HF recommend against long-term use of positive inotropic drugs except for palliation [22].
There is insufficient evidence to determine whether outpatient intravenous diuresis is an effective method of reducing hospitalizations in patients with HF [38,63,64]. A handful of small case series have demonstrated feasibility of outpatient intravenous diuresis for patients with acute decompensated HF, potentially avoiding hospitalizations [64-68]. Only one of these identified a comparison cohort (patients admitted to an inpatient setting for <48 hours for diuresis); the 106 patients undergoing ambulatory infusion experienced three fewer days in the hospital in the subsequent 180 days compared with 143 patients briefly hospitalized for diuresis [64].
Hospital at home — An approach to avoiding hospitalization even for acutely ill patients is the "hospital at home" model, which provides patients who would otherwise be admitted to the hospital with acute care-level care at home. A hospital at home program would likely require substantial resources and experience with this approach.
●An RCT of hospital at home for patients with HF was performed by a group with a well-established, sophisticated hospital at home program. In this single-blind trial, 101 patients 75 years or older were randomly assigned to receive hospital care or hospital at home care. There was no difference between the two groups in six-month mortality or total hospital readmissions, but the mean time to first additional admission was longer for patients receiving hospital at home care (84.3 days versus 69.8 days) [69].
●A Cochrane review of 16 RCTs of hospital at home with a total of 1814 adult participants with varied medical conditions found possible reduction in likelihood of living in residential care at six months’ follow-up and possible positive effects on patient satisfaction [70]. However, hospital at home probably had little or no effect on mortality at six months’ follow-up (RR 0.77, 95% CI 0.77-0.99; moderate certainty evidence) and little or no effect on the likelihood of being transferred (or readmitted) to hospital (RR 0.98, 95% CI 0.77-1.23; moderate certainty evidence).
INPATIENT AND TRANSITIONAL CARE STRATEGIES — Among the variety of tested strategies in inpatient and transitional settings, programs limited solely to support in the inpatient setting (education, medication reconciliation, and care coordination) are generally less likely to be successful than those that span the inpatient-outpatient divide. Similarly, programs focused on only one aspect of discharge support (eg, only medication reconciliation or only sending discharge summaries to outpatient clinicians) are less likely to be successful.
In hospital — Strategies initiated during hospitalization for HF include evidence-based medical care, inpatient cardiologist care, self-care education, and communication with the outpatient team.
Documentation of evidence-based therapies — Evidence-based medical care for HF reduces readmissions, but there is little evidence that processes of care designed to increase adherence to such therapies (eg, medication reconciliation, documentation at discharge) are incrementally effective [71,72]. As examples:
●An observational study of patients hospitalized with HF at 150 hospitals found that one-year cardiovascular readmission risk was not associated with hospital-level process measures that included documentation of angiotensin-converting enzyme inhibitor, angiotensin II receptor blocker, beta blocker, or smoking cessation counseling [71].
●A cross-sectional study of 3476 acute care hospitals found that hospitals with higher adherence to clinical process measures for HF had a 0.2 percent lower 30-day readmission rate [73]. However, this reduction was closely linked to baseline performance: hospitals already performing well experienced no improvement, while those beginning with scores in the 25th to 50th percentile reduced readmission rates by an average of 0.5 percent.
Inpatient cardiologist care — Several observational studies have suggested lower mortality for patients cared for by cardiologists while hospitalized for HF compared with those cared for by internists or family practitioners [74-78]. As examples:
●A study of 7516 patients with acute decompensated HF at three urban hospitals examined mortality between 2001 and 2010 for patients with HF cared for by cardiologists or not [78]. Patients cared for by cardiologists (19 percent) had lower 60-day mortality (5.4 versus 7.0 percent; hazard ratio [HR] 0.70, 95% CI 0.52-0.96) and lower 30-day readmission rates (HR 0.76, 95% CI 0.66-0.89). A Canadian study of 7634 patients hospitalized from 2004 to 2005 found very similar results [77].
●A propensity-matched study found lower mortality (odds ratio [OR] 0.37, 95% CI 0.29-0.47) and infections (pneumonia, sepsis, and urinary tract infection) with care by cardiologists compared with care by noncardiologists [79].
Self-care education — Self-care education should be initiated in hospital and continued after discharge. The role of self-care education is discussed separately. (See "Heart failure self-management" and "Heart failure self-management", section on 'Promoting effective self-care'.)
Communication with the outpatient team — Communication with the outpatient team is an important part of transitional care, although studies of the effect of discharge summary transmission have reported mixed results. One observational study of 1246 discharge summaries from 45 hospitals found an association between discharge summary transmission to outpatient clinicians and lower readmission risk (OR 0.53, 95% CI 0.32-0.90); in addition, summaries including more content recommended by the Transitions of Care Consensus Conference were associated with lower readmission risk (OR 0.67, 95% CI 0.46-0.97) [80]. Other studies, however, have not found benefit from discharge summary transmission. A multisite observational study of 1078 general medicine patients found no relationship between receipt of a discharge summary within two weeks of discharge and a composite outcome of emergency department visit, readmission, and death [81]. Also, there was no apparent benefit to direct communication with physicians. Similarly, an observational study comparing 1039 patients rehospitalized within 30 days of discharge with 981 nonrehospitalized patients found no association between transmission of a discharge summary to an outpatient physician and readmission [82].
Since these are observational studies, it is uncertain whether timely discharge summaries and clinician communication may be more likely for patients with the most complex clinical issues and highest risk of readmission. Additionally, the timeliness of discharge summaries may only be one component of their effectiveness, which might also be affected by the completeness and quality of the content and the availability of support systems to facilitate appropriate response to issues raised by discharge summaries.
On discharge — Comprehensive discharge support is designed to reduce readmissions, although trials have not clearly shown a benefit.
Improved discharge support, also termed "hospital-based case management" or "transitional care interventions," includes some or all of the following activities (see "Hospital discharge and readmission"):
●Medication support, including appropriate medication reconciliation, patient education, and assurance that patients can obtain medications. (See "Hospital discharge and readmission", section on 'Medication reconciliation'.)
●Patient education, including definition of HF, precipitants of worsening disease, warning signs, and home self-care activities. (See "Heart failure self-management".)
●Follow-up planning.
●Coordination of information between settings.
●Prompt postdischarge follow-up (by phone/e-mail, in the office, or via home visit).
●Multidisciplinary team-based care, including clinicians, nurses, social worker, care coordinator, pharmacist, physical therapist, and/or others as needed.
The 2010 Heart Failure Society of America, 2013 American College of Cardiology/American Heart Association, and 2016 European Society of Cardiology HF guidelines recommend the use of discharge support [20-22]. Since publication of these guidelines, two large trials were published that did not show an effect of comprehensive discharge planning.
●In a trial that randomly assigned 5647 patients hospitalized with HF with reduced ejection fraction (HFrEF) to a hospital and postdischarge quality improvement intervention (eg, medication reconciliation, risk assessment, self-management education, goal setting, access to postdischarge resources) or to usual care, rates of the combined outcome of all-cause mortality and HF rehospitalization were similar between the groups (39 versus 40 percent, respectively) [83]. In addition, there was no significant change in a composite HF quality-of-care score between the two groups. Of note, many centers in the trial were already participating in a program to improve the care of patients with HFrEF.
●An RCT comparing a program of discharge support with usual care in 2494 adults hospitalized with HF found reduction in emergency department visits at 30 days but no improvement in composites of clinical outcomes (all-cause readmission, emergency department visit, or death at three months; or all-cause readmission or emergency department visit at 30 days) [84]. The program of discharge support included nurse-led self-care education, a structured hospital discharge summary, and a family physician follow-up appointment less than one week after discharge; high-risk patients also received structured nurse home visits and HF clinic care. One limitation of this trial is that less than 40 percent of patients in the intervention group received nurse home visits and HF clinic care.
Discharge support interventions have variable effect on cost. For example:
●An RCT of 223 patients with HF compared a standard discharge procedure with a one-hour, one-on-one teaching session with a nurse educator before discharge [85]. Patients in the intervention group had a lower risk of rehospitalization or death (risk ratio 0.65, 95% CI 0.45-0.93) 180 days after discharge. This intervention was estimated to save USD $2823 per patient.
●However, a program that reduces readmissions does not necessarily reduce direct costs. In a prospective study of a nurse-led transitional-care program for patients with HF being discharged from a hospital in the United States, thirty-day readmissions, lengths of stay, and 60-day direct costs were compared with those for HF patients concurrently discharged from other hospitals within the same health care system [86]. The program was associated with a 48 percent decrease in 30-day readmissions but had little impact on direct costs to the health care system over 60 days and, under the existing Medicare reimbursement system, had a negative impact on hospital revenue.
Studies of individual components of discharge support have not shown a clear reduction in HF readmissions. Examples include:
●An observational study of 16,791 patients hospitalized with HF at 150 hospitals found that meeting the discharge instruction requirement set by the Centers for Medicare and Medicaid Services (CMS) had no relationship to one-year cardiovascular readmission [71].
●Similarly, an observational study of 2222 hospitals found no association between hospital-level performance on the CMS discharge instruction measure and 30-day readmission rates for patients admitted with HF [87].
Studies of discharge support in the general medical population (not limited to HF) are discussed separately. (See "Hospital discharge and readmission", section on 'Multiple interventions'.)
Early follow-up — Although prospective data are lacking, observational data suggest that an early follow-up appointment may reduce the risk of HF readmission. The value of an early appointment may depend on the quality of care at the visit, which may be affected by the information available to the provider. An appointment may be necessary but not sufficient to reduce risk. Ideally, assessment should take place within one to two weeks after discharge to reassess fluid status, review and adjust medications, plan additional testing if needed, and provide further patient education [88]. Since this strategy might not be feasible for all of the many patients hospitalized for HF, particular focus for early follow-up should be targeted to patients who are at increased risk for early decompensation (eg, low systolic blood pressure, prolonged QRS duration, renal dysfunction, elevated natriuretic peptide level, elevated troponin levels, hyponatremia, or other comorbidities).
An observational study of 30,136 patients with HF found that hospitals in the lowest quartile of early postdischarge physician follow-up (that is, hospitals whose patients were least likely to have follow-up within seven days) had a 3 percent higher 30-day readmission rate than the other 75 percent of hospitals [89].
An observational study of 98,730 Medicare patients discharged to home care after a hospitalization for HF found that patients who received both four home-care visits in the first week (including one within 48 hours of discharge) and a physician office visit in the first week had lower readmission rates than those with neither [90]. However, neither intervention on its own was effective.
Finally, an observational study of 39,249 patients hospitalized or seen in the emergency department for HF over 12 years in Alberta, Canada, found that those who followed up within 14 days with a physician they had seen before had significantly lower readmission rates than those who had no follow-up (HR 0.94, 95% CI 0.89-0.99) [91]. There was no benefit to seeing an unfamiliar physician versus no follow-up (HR 1.05, 95% CI 0.97-1.15).
Postdischarge home visit programs — In-person, at-home care, typically by nurses, is among the most effective types of transitional care management following hospitalization for HF. A meta-analysis of six RCTs with a total of 1176 subjects involving 1 to 12 home visits by nurses found that they reduced both all-cause mortality (incident rate ratio 0.78, 95% CI 0.62-0.98) and all-cause readmission (risk ratio [RR] 0.65, 95% CI 0.49-0.86) [39]. An additional 11 RCTs with a total of 2891 subjects of nurse-led case management, including home visits and telephone support, found no mortality benefit (RR 0.86, 95% CI 0.71-1.05) but a significant reduction in readmissions (RR 0.77, 95% CI 0.63-0.95) [39].
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Heart failure in adults".)
SUMMARY AND RECOMMENDATIONS
●Heart failure (HF) is one of the most common causes of hospitalization, and many patients who are hospitalized for HF are readmitted within 30 days of discharge. Systems for outpatient and transitional care for patients with HF must aim to comprehensively address the complex medical and social issues that contribute to the high rates of hospitalization in this population. (See 'Introduction' above and 'Epidemiology' above and 'Address social determinants' above.)
●Key components of HF disease management programs – The components of outpatient HF disease management programs typically include (see 'Key components of outpatient care' above):
•A process for identifying and assessing patients with HF. (See "Heart failure: Clinical manifestations and diagnosis in adults", section on 'When to suspect heart failure'.)
•Psychosocial care, including addressing social determinants of health. (See 'Address social determinants' above.)
•Coordinated care that includes a care team composed of physicians (primary care provider, cardiologist, and others to treat comorbid conditions), nurses, care coordinators, social workers, pharmacists, physical therapists, dieticians, and others as needed.
•Medical management, including adherence to evidence-based practice guidelines for medications and other therapies. (See "Overview of the management of heart failure with reduced ejection fraction in adults" and "Treatment and prognosis of heart failure with mid-range ejection fraction" and "Treatment and prognosis of heart failure with preserved ejection fraction".)
•Comprehensive patient education particularly focused on self-care. (See "Heart failure self-management".)
•Palliative care. (See "Legal aspects in palliative and end-of-life care in the United States" and "Palliative care for patients with advanced heart failure: Indications and systems of care" and "Palliative care for patients with advanced heart failure: Decision support and management of symptoms".)
•Early postdischarge support if hospitalized. (See 'Inpatient and transitional care strategies' above.)
●More effective models – Outpatient HF care models that have been shown to be most effective in reducing hospitalizations include:
•Nurse-led case management including drug titration. (See 'Nurse-led titration of drugs' above.)
•Multidisciplinary HF clinics. (See 'Multidisciplinary heart failure clinics' above.)
•Programs emphasizing self-care education and support. (See "Heart failure self-management", section on 'Benefit of self-management interventions'.)
●Less effective models of care – The evidence to support the routine use of outpatient programs limited to telephone contact or remote data gathering (telemonitoring) is relatively weak, perhaps because their effectiveness is related to the overall system into which the patient information flows. (See 'Telephone-based case management' above and 'Telemonitoring' above.)
●Models with unclear benefit – The routine use of comprehensive discharge support for hospitalized patients with HF does not clearly reduce hospital readmissions. (See 'Models with unclear benefit' above.)
●Inpatient processes of care – Among the variety of tested strategies in inpatient and transitional settings, programs limited solely to support in the inpatient setting (education, medication reconciliation, and care coordination) are generally less likely to be successful than those that span the inpatient-outpatient divide. Similarly, programs focused on only one aspect of discharge support (eg, only medication reconciliation or only sending discharge summaries to outpatient clinicians) are less likely to be successful. (See 'Inpatient and transitional care strategies' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Leora Horwitz, MD, MHS, who contributed to earlier versions of this topic review.
