Hospital-at-Home care for acute heart failure: Feasibility and safety pilot

In the Netherlands, chronic heart failure (HF) affects approximately 1–2% of the population, with a much higher prevalence of 21% among individuals aged 85 and older [1, 2]. The Netherlands Heart Foundation estimates that 20–30% of Dutch citizens will develop HF in their lifetime [3]. Since HF patients are frequently hospitalised during episodes of decompensation, this places increasing pressure on hospital capacity and costs [4].

To address this issue, there is growing support for shifting hospital-level care to patients’ homes through so-called Hospital-at-Home (HaH) initiatives [5]. From July 2022, Deventer Hospital (DZ) introduced a version of HaH for acute decompensated heart failure (ADHF) patients: DZThuis (i.e. DZHome). Guided by the principle of ‘treatment at home when possible, in hospital when necessary,’ DZThuis patients are admitted to home care if their condition allows. There, they receive furosemide via portable continuous infusion pumps while being closely monitored. Once sufficient clinical stability is achieved, they transition back to routine outpatient care.

Previous studies have demonstrated the benefits of HaH concepts. Qaddoura et al., in a comparison of in-hospital and HaH care for ADHF patients across European settings, found no difference in all-cause mortality or time to hospital readmission [6]. However, feasibility studies on HaH initiatives in the Netherlands are scarce and do not report clinical outcomes [7].

Hence, this retrospective study aims to assess the feasibility of safely treating this patient group in a HaH setting. Additionally, we conducted an exploratory comparison with the outcomes of standard in-hospital treatment for ADHF patients.

This pilot study demonstrated the feasibility of DZThuis, a hybrid HaH model. No significant differences were found in mortality, mid-term readmissions, or complications in this small group compared with a historical in-hospital patient group. This was observed despite a higher prevalence of comorbidities in the DZThuis group, consistent with findings from Qaddoura et al. in their systematic review [6]. Interestingly, a recent observational study by Sankaranarayanan et al. on a telehealth-aided virtual ward for ADHF patients reached a similar conclusion, even proposing that home treatment may be more beneficial for patients [8].

Notably, as in a recent study by Achanta et al. [9], approximately 10% of patients required transition to in-hospital care, typically for intravenous antibiotics or inotropes. We did not classify this as readmission, as in our view this is not a failure of home therapy but part of the concept of hybrid care.

One relevant but non-significant finding was a longer treatment duration in the DZThuis group. Similar trends have been reported by Mendoza et al. and Tibaldi et al., who observed prolonged total treatment durations in Hospital-at-Home (HaH) models [10, 11]. This may be attributed to the higher prevalence of comorbidities among DZThuis patients. Practical challenges that postponed adjustments in clinical management, such as laboratory results coming available in the afternoon rather than the morning, could also have played a minor role.

The average dosage of furosemide was higher in the DZThuis group, which probably also caused an increase in creatinine levels. A possible explanation is the absence of a standardised protocol for diuretic dosing and euvolaemic criteria, leaving these decisions to the discretion of the supervising cardiologist. Possibly, the more distant care made it harder to assess when a patient was completely decongested. Also, in the light of the 2022 STRONG-HF study [12], treatment regimens may have become more intense over time. Another possible, albeit speculative, explanation is lower adherence to salt and fluid restrictions in the home setting, as nurses reported observing food remnants, such as pizza boxes and soda bottles, in patients’ homes.

This study has limitations inherent to its retrospective observational design and relatively small sample size. The higher disease burden in the DZThuis group suggests potential confounding and selection bias, as the control group was formed post hoc. Non-HF related readmissions were not recorded. Additionally, the data for the in-hospital group were collected one year prior to the DZThuis group, during which time prescribing practices evolved, including greater use of SGLT‑2 inhibitors in the DZThuis group (Fig. S1 of Electronic Supplementary Material). We did not prospectively register eligible patients for home treatment who were not included in DZThuis, nor the reasons for eventual hospital treatment. This figure is estimated to be around 40%, in line with a previous research letter by Haywood et al., which estimated that 1 in 3 ADHF patients may be eligible for HaH [13]. Although our pilot study successfully demonstrated the feasibility and suggest the safety of HaH treatment, it did not thoroughly examine cost-effectiveness. Nevertheless, even without precise cost data, it was evident that the current set-up was not cost-effective. With only 53 admissions over a span of 15 months, the average number of patients receiving home treatment at any given time was 1.4, while a hospital nurse dedicated an entire shift and was on call 24/7. To address this, Deventer Hospital is actively working to make the approach more cost-efficient. Agreements have been made with home nursing care organisations to train nurses to deliver daily care under the supervision of Heart Failure physician assistants. These assistants will conduct weekly visits and consult with a cardiologist to ensure hospital-level care. To make this possible, a treatment protocol is also being developed to improve the quality and consistency of the care during admission. Future patients will continue to be monitored. Safety will be reassessed and efficacy judged with more included patients. Since maintaining a good quality of life is as important as survival for many chronically ill patients [14], assessing quality of life was initially a goal of the study. However, insufficient data were collected for this purpose. Future research should investigate the difference in quality of life between DZThuis and in-hospital care, as well as patient satisfaction with treatment.

The findings from this pilot study demonstrate that it is feasible to treat ADHF patients at home with intravenous diuretics. In DZThuis, the Hospital-at-Home initiative from Deventer Hospital, there was no significant difference in total mortality and time until readmission of heart failure patients compared with the in-hospital group, despite a higher rate of patients with comorbidities in the first. Furthermore, no significant differences were seen in the rate of complications in this small pilot study with historical controls. However, larger studies are required to draw more definitive conclusions regarding safety and efficacy.