Heart failure (HF) is the leading cause of hospital readmission in developed countries.1 It is a particular concern for patients ≥ 65 years of age, who comprise approximately 80% of the population with HF.2 The total annual cost of HF in the US is estimated to be $30.7 billion, with about two-thirds attributable to HF-related hospitalizations.3

Rates of rehospitalization within 30 days of discharge can reach 20%–30%.4 Early rehospitalization is attributed to underlying disease exacerbation5 and several other medical derangements.6 Up to 65% of 30-day readmissions are for diverse medical circumstances7 that often occur early in the hospital-to-home transition.8 Early postdischarge readmissions are strongly linked to the quality of inpatient care.9,10 Although they may be partly due to incomplete patient stabilization, they are frequently caused by poor discharge coordination and failure to ensure good quality postdischarge care. Indeed, up to 75% of such cases are considered to be preventable.11

Several interventions have been proposed and tested for reducing the social and economic burdens of early readmissions.12,13 However, previously tested interventions show only modest ability to reduce all-cause or unplanned 30-day postdischarge readmissions among patients hospitalized for HF, with absolute reductions of around 2%-3% (relative reductions of 10%-18%).12,14 Moreover, some studies exclude HF patients admitted to noncardiology wards, although these patients are usually elderly, frail, and have several comorbidities, and thus show the highest readmission rates.15

In the present study, we evaluated a STructured multidisciplinary outpatient clinic for Old and frail postdischarge patients hospitalized for HF (the “STOP-HF-Clinic”), which was established with the aim of reducing 30-day readmission rates and facilitating transition to primary care. We assessed the efficacy of this STOP-HF intervention in terms of the enrolled patients’ 30-day readmission rates compared with their readmission risk calculated using the CORE-HF risk score. As a secondary objective, we examined the impact of the STOP-HF-Clinic intervention within the official readmission data registry of the Catalan Health Service (CatSalut), which provides medical care to 7.5 million people in Catalonia, Spain.

A total of 518 patients attended the STOP-HF-Clinic from February 2014 to June 2016. These patients’ demographic and clinical characteristics are shown in Table 1. The mean patient age was 82.3 ± 8.3 years, 25% were ≥ 88 years old, and 57.1% were women. Common comorbidities included diabetes, anemia, and renal failure. Hemoglobin levels were < 10g/dL in 55 patients (10.6%) and < 9g/dL in 13 patients (2.5%). The median time from discharge to first STOP-HF-Clinic visit was 5 days (Q1-Q3, 3-6 days), and the mean number of visits per patient within 30 days was 3.1 ± 1.2.

The need for intravenous therapy and adjustment of HF treatment in the 30-day STOP-HF-Clinic are displayed in Table 2. One-third of patients required at least one infusion of 40mg intravenous furosemide. Infusions of intravenous furosemide were required once by 104 patients, twice by 39 patients, 3 times by 21 patients, 4 times by 11 patients, 5 times by 4 patients, and 6 times by 3 patients. One-sixth of patients required intravenous ferric carboxymaltose 1g, with up to 93 infusions needed. Twenty-one patients required red blood cell transfusion. Oral HF treatments were routinely initiated or adjusted in the STOP-HF-Clinic, with some patients requiring substantial dose uptitration. Angiotensin-converting enzyme inhibitors/angiotensin receptor blockers were discontinued in 10.5% of patients due to renal worsening or hypotension. Those who continued treatment had a mean angiotensin-converting enzyme inhibitor/angiotensin receptor blocker dose increase of 15% after the 30-day follow-up. During the 30-day follow-up period, 12 patients (2.3%) died. At the end of the intervention, 54% of patients were transitioned to their general practitioner and 30% to a primary care specialist and 16% continued follow-up with hospital HF specialists.

The predicted CORE-HF all-cause 30-day readmission risk was 26.5% ± 5.3%, with 29.5% of patients having a CORE-HF risk of ≥ 30%. The observed all-cause 30-day readmission rate was 13.9% (n=72), and the observed HF-related 30-day readmission rate was 7.5% (n=39). Overall, the STOP-HF-Clinic intervention achieved a 47.5% relative reduction in all-cause 30-day readmission risk and an absolute reduction of 12.6% compared with the predicted CORE-HF readmission risk score.

Multivariable logistic regression analysis revealed that all-cause 30-day readmission was independently associated with age (OR, 1.04; 95% confidence interval [95%CI], 1.00-1.08; P=.03), Charlson index (OR, 1.19; 95%CI, 1.06-1.34; P=.005), length of index hospital stay (OR, 1.03; 95%CI, 1.00-1.06; P=.03), and NT-proBNP (OR, 1.34; 95%CI, 1.01-1.78; P=.045). On the other hand, HF-related 30-day readmission was independently associated with only the Charlson index (OR, 1.22; 95%CI, 1.05-1.41; P=.008) and NT-proBNP (OR, 1.43; 95%CI, 1.00-2.05; P=.05).

We next compared readmission rates within the STOP-HF Referral Area (∼250 000 people) with readmission rates within the entire CatSalut registry (∼7.5 million people) during 2 time periods: pre–STOP-HF (2012-2013) and post–STOP-HF (2014-2015) (table of the supplementary material). Within these 2 periods, 9267 all-cause and 6686 HF-related 30-day readmissions were documented after a total of 65 131 index HF-related admissions. Figure 1 illustrates the decline in 30-day readmission rates observed in the STOP-HF Referral Area during the 2014-2015 time period compared with the 2012-2013 time period (P < .001), with no significant changes in the rest of the CatSalut area (P=.11). The all-cause 30-day readmission rates during the 2012-2013 period in the 2 studied populations are shown in Table 3 (P=.88). On the other hand, during the 2014-2015 period, all-cause 30-day readmission rates were significantly lower in the STOP-HF Referral Area (P < .001), which was mainly driven by reduced HF-related readmissions.

Figure 1.

Thirty-day readmission rates with and without the STOP-HF intervention. Thirty-day readmission rates in the STOP-HF Referral Area vs the CatSalut area before STOP-HF (2012-2013) and with STOP-HF (2014-2015). CatSalut, Catalan Health Service; HF-related, heart failure recurrence; non–HF-related: chronic noncirculatory disease and readmissions due to a complication of the index admission; STOP-HF, structured multidisciplinary outpatient clinic for old and frail postdischarge patients hospitalized for heart failure.

(0.09MB).

Actuarial curves of probability after HF hospitalization within the CatSalut area and the STOP-HF Referral Area are shown in Figure 2. Compared with the rest of the CatSalut area, the STOP-HF Referral Area showed a significant decline in HF readmissions during the 2014-2015 period, with the 2 curves following divergent paths after day 10.

Figure 2.

Actuarial curves of the probability of 30-day readmission with and without STOP-HF. Probabilities of 30-day readmission in the STOP-HF Referral Area vs the CatSalut area before STOP-HF (2012-2013) and with STOP-HF (2014-2015). P values reflect comparisons between study groups. CatSalut, Catalan Health Service; STOP-HF, structured multidisciplinary outpatient clinic for old and frail postdischarge patients hospitalized for heart failure.

(0.14MB).

The concept that “hospitalization begets further hospitalization” is certainly applicable to HF.21 Because many HF readmissions are considered preventable, novel strategies are needed to reduce rehospitalizations. We have developed a comprehensive patient-centered model: the STOP-HF-Clinic. Our present data show that implementation of the STOP-HF-Clinic intervention was associated with a significant ∼50% reduction in all-cause 30-day readmission, mainly driven by reduced HF-related readmissions. Moreover, the STOP-HF intervention profoundly impacted all-cause readmission rates in the STOP-HF Referral Area compared with the CatSalut registry.

Rather than focusing on only one aspect of patient care, the STOP-HF-Clinic integrated a number of interventions, including quality of medical management, early reassessment, health literacy, and care transition. Despite growing interest in remote monitoring of these patients to reduce readmissions, structured telephone support interventions and home telemonitoring have not shown greater ability to reduce 30-day all-cause or HF-related readmission rates.22 Recent reports from the BEAT-HF trial, which aimed to explicitly adapt the care transition approach in combination with remote telemonitoring, failed to demonstrate reduced all-cause 30-day and 180-day readmission rates.23 On the other hand, our face-to-face early intervention reduced both all-cause and HF-related 30-day readmissions in our population.

Compared with most prior studies, our present real-life, prospective, all-comers study enrolled patients who were older and more frail and vulnerable in terms of medical complexity (mean age, 82 years; Barthel score, 70; Charlson index, 6; preserved left ventricular ejection fraction and high NT-proBNP).24–31 Previous studies have commonly excluded patients with advanced renal insufficiency or severe cardiovascular disease22 but elderly individuals with prevalent renal dysfunction and prominent congestion are most prone to early readmissions.32 The prospect of an increasingly aging frail population increases the need for solutions suitable for this patient group. Such measures will likely require the cooperation of multidisciplinary teams, such as in the STOP-HF-Clinic, rather than a focus on any one aspect of patient care.

It appears that early follow-up within 7 days of discharge is critical for reducing readmissions.21,33,34 In the STOP-HF-Clinic, we offered systematic medical contact at a median of 5 days postdischarge. During this transitional period, ineffective communication, low health literacy, and adherence issues contribute to readmissions. However, intervention programs often fail to act during this time frame,35 sometimes due to a lack of coordination with the medical provider, revealing a gap in transitional care. This period of postdischarge vulnerability has been described as “post-hospital syndrome”36 and is related to factors such as age, cognitive impairment, frailty, and polypharmacy. The STOP-HF intervention likely acted to prevent ‘post-hospital syndrome’.

The STOP-HF-Clinic provided a quick therapeutic intervention to promote disease stability. Indeed, the reduced readmission rate was likely affected by the high number of intravenous infusions of both furosemide and ferric carboxymaltose. Although intravenous loop diuretics are the standard-of-care for inpatient management in acutely decompensated HF, here we extended intravenous furosemide infusion to the outpatient setting during the 30-day postdischarge period among patients with refractory HF and congestion. This is not a widespread practice, although it has shown benefits as part of transitional care in some case series, promoting symptom improvement and avoidance of emergency department transfers and readmissions.37–40 Moreover, the CONFIRM-HF trial showed that iron replacement with ferric carboxymaltose significantly reduced the risk of hospitalization due to worsening heart failure, regardless of functional class severity, and had particular benefits in patients with diabetes or renal impairment,41 such as in our cohort, which comprised over 50% diabetic patients and 80% patients with renal insufficiency.

It seems clear that transition care interventions can reduce readmissions. As Comín-Colet et al. noted in a recent review article,42 a paradigm shift in the management of chronic diseases has taken place in recent years. This new approach is based on the development of a multidisciplinary model that provides integrated care to patients with HF throughout the duration of the disease, ensuring a successful follow-up and transition of care to different health care settings depending on the progress of the condition. In our case, the STOP-HF Clinic is included in a specialized HF unit that fulfills the standards recommended by the Spanish Society of Cardiology.43

Moreover, readmission risk assessment tools may help to appropriately target the delivery of these interventions to at-risk patients. Here, we predicted readmission risk using a validated risk score (the CORE-HF calculator). Although our cohort substantially differed from the population used to derive this score,18 we believe than the CORE-HF tool was appropriate given the high-risk population studied. However, it is notable that a reliable readmission predictive model for the current “real-life” population is not yet available.

Regarding our secondary endpoint, the present study was designed as a natural experiment rather than a conventional clinical trial. Our analysis included all patients within the STOP-HF Referral Area, regardless of their participation in the STOP-HF-Clinic. Their course was compared with that of a control group comprising the patients in the rest of the CatSalut area. This procedure minimized the characteristic selection bias of clinical trials, whose patient profile is often distinct from that of the general population and more pragmatically reflects the efficacy of the intervention. Our data revealed that the STOP-HF intervention had remarkable benefits during the post–STOP-HF period (2014-2015) compared with the pre–STOP-HF period (2012-2013) in terms of indicators relevant to the health care system and the patients (all-cause and HF-related 30-day readmissions).

The STOP-HF-Clinic included early follow-up, HF nurses tasked with medication reconciliation, education and patient self-care empowerment, staff assigned to follow up on postdischarge test results, immediate availability of intravenous treatments and patient treatment titration, and partnerships with community physicians. This intervention resulted in an ∼50% reduction in the all-cause 30-day readmission rate after an index hospitalization for HF, which was mainly driven by a reduction in HF-related readmissions. Our results with the STOP-HF-Clinic in an elderly and frail comorbid population were better than with previously reported strategies. Our present data support the value of the STOP-HF-Clinic, an approach that could be implemented elsewhere to reduce the global burden of HF readmissions.

None declared.