Outcomes of a “hospital at home” programme for the supervised home recovery of COVID-19 patients in Singapore

Abstract

Background

The Singapore General Hospital COVID-19 Virtual Ward is a “hospital at home” (HaH) programme for the supervised home recovery of higher-acuity COVID-19 patients from the hospital and the community.

Objective

To describe how an existing HaH programme was redesigned so that COVID-19 patients could be remotely monitored at home and report the outcomes of the first 100 patients in this Virtual Ward.

Methods

Patients received an admission package comprising instructions and equipment for home monitoring, and uploaded their parameters into a clinical dashboard via a secure messaging app. Medical staff conducted video or telephone consultations daily. Patients were discharged according to time-based criteria, although some required SARS-CoV-2 PCR testing, which were conducted at home by a third-party medical provider. De-identified data of the first 100 patients were analysed by demographic details, indication for enrolment into the Virtual Ward, and the need for subsequent inpatient readmission.

Results

Of the first 100 patients admitted into the Virtual Ward, 58 were female, mean age was 63.1 years old (23–95 years), and 76 were fully vaccinated. There were 77 hospital referrals and 23 community referrals. The number of days of inpatient hospitalisation avoided was 717 days (average of 7.9 days per patient). Three hospital referrals (3.9%) were readmitted, while seven community referrals (30.4%) required subsequent hospitalisation.

Conclusion

The Virtual Ward programme demonstrates that selected COVID-19 patient can safely recover at home with remote medical support and monitoring, thereby expanding hospital capacity.

Keywords

Hospital at home hospital to home sars-cov-2 singapore telemedicine

Introduction

In early September 2021, Singapore experienced a surge in the number of COVID-19 patients.¹ This surge was driven by the Delta variant of the SARS-Cov-2 virus, which proved to be 97–100% more infective than the original Alpha variant^2–4 and was the main epidemic strain driving the wave of infections in both the vaccinated and unvaccinated communities in Singapore.

Although the majority of COVID-19 patients (98.4%) were asymptomatic or suffered only mild symptoms, approximately 7% were admitted into hospitals⁵ because of more severe symptoms or pre-existing medical conditions that exclude them from recovering at home or at the designated COVID-19 Treatment Facilities (CTFs). Unsurprisingly, the occupancy for COVID-19 isolation beds rose from 62% in July 2021 to 85% for October 2021,⁵ placing significant pressure on our healthcare resources.

The Department of Family Medicine and Continuing Care (FMCC) and the Population Heath Office and Integrated Care Office (PHICO) have been the drivers for transitional care services in Singapore General Hospital (SGH). For example, the SGH @ Home programme provides acute and sub-acute treatment in a patient’s residence for conditions that would normally have required continued hospital admission. Such “hospital at home” (HaH) programmes facilitate the early discharge of suitable patients without compromising on their care, and have been associated with reductions in mortality, readmission rates and cost, and increases in patient and carer satisfaction.⁶ We describe how the SGH @ Home programme was redesigned to support the safe recovery of COVID-19 patients in their homes, and report the outcomes of the first 100 patients in this “COVID-19 Virtual Ward”.

Methods

The development of the COVID-19 virtual ward

The pivot from SGH @ Home patients to COVID-19 patients was made at the request of the Ministry of Health (MOH) in mid-September 2021. The primary aim of this new COVID-19 Virtual Ward programme was to facilitate early discharge of hospitalised COVID-19 patients to their homes, so as to free up bed capacity in our COVID Isolation wards for the more seriously ill. By 1^st November 2021, the programme was expanded to include newly-diagnosed COVID patients from the community, so as to avoid hospitalisation completely. The focus would be on patients who were not eligible for the existing Home Recovery Programme (HRP) or CTFs, due to their advanced age (70 years and above), presence of mild or moderate symptoms, and/or because they were not fully vaccinated against COVID-19.

The COVID-19 Virtual Ward was launched on 6th October 2021 and the first patients were admitted into the programme from SGH. Community-based referrals were only enrolled into the Virtual Ward from 1^st November 2021. The primary team for the Virtual Ward has the capacity to care for up to 100 patients simultaneously, and includes a Consultant Physician, three Resident Physicians, as well as a Nurse Clinician and five Staff Nurses. SpeedDoc, a private 3rd party medical care provider and the existing partner in the SGH @ Home programme, was engaged to provide on-call services after office hours.

Inclusion/exclusion criteria

The detailed inclusion and exclusion criteria for the Virtual Ward can be found in Table 1. The inclusion criteria were designed to be broad so as to encourage referrals, and each patient was evaluated on a case-by-case basis. Correspondingly, the exclusion criteria was quite limited, excluding patients with active medical issues that required inpatient care (eg requiring daily intravenous antibiotics) or those with isolation requirements that could not be supported at home or in the community facilities (eg haemodialysis).

Table 1.

Inclusion and exclusion criteria for the COVID-19 virtual ward.

Inclusion criteria	Exclusion criteria
• age ≥18	• end-stage renal failure on haemodialysis
• No severe symptoms of COVID-19 infection (eg shortness of breath, chest pain)	• active cancer requiring close inpatient treatment or monitoring
• assessed by the patient’s primary physician to be medically fit for transfer to the COVID-19 virtual ward	• cognitive impairment (moderate to severe)
	• medical issues requiring daily inpatient treatment (e.g. intravenous antibiotics)
• functional status: Capable of carrying out activities of daily living (ADLs) independently, or otherwise has a dedicated caregiver if assistance is required for ADLs	• persistent fever of 38.0^OC or higher, for 3 or more days
• access to a home Internet connection and a smart device or phone, to facilitate telemonitoring and teleconsultations	• residents of intermediate or long-term care facilities (eg nursing homes)
• home and social environment suitable for home recovery

Referral process and admission into the COVID-19 virtual ward

Suitable patients were referred to the FMCC for enrolment into the Virtual Ward. Patients that were hospitalised in SGH were referred by the primary physician, while community cases were referred by MOH’s Case Management Task Group (CMTG), which co-ordinates the national level administration of COVID patients in the community. Each day, CMTG would provide the Virtual Ward with a list of patients who were not suitable for home recovery nor the CTFs; typically, these were patients aged 70 years old and above, with PCR-positive COVID-19 infection and who were not fully vaccinated. The staff from CMTG were non-medical staff, and generally did not have access to patients’ medical records.

The patient’s electronic medical records (EMR) were reviewed where available, and our medical staff conducted a phone or video consultation with the patient in the ward or at home to assess their suitability. Aside from their medical status, patients must have been willing to recover at home, be able to care for themselves, and assist in the daily monitoring of their vital signs. Patients that required assistance with ADLs would have had a willing and responsible caregiver at home to be considered for this programme. The same inclusion-exclusion criteria were applied to both hospital-based and community-based referrals, and community referrals that were deemed unsuitable for the Virtual Ward were arranged for inpatient hospitalisation by CMTG.”

Once informed consent for this programme was received, arrangements are made for hospitalised patient to be discharged and transported back to their home. A home monitoring package was delivered to all patients enrolled in the Virtual Ward; the package included a digital thermometer, a pulse oximeter, and written information about the Virtual Ward. A digital blood pressure machine and glucometer were also issued to patients that required these parameters to be monitored. Once at home, an “admission assessment” was conducted by a doctor and a nurse via video consultation, and the patient and caregivers were instructed on the remote monitoring requirements as well as the escalation protocols.

Remote monitoring and daily check-ins

From Monday to Friday, the medical staff performed a daily telephone and/or video consultation with each patient to monitor their general condition. During the check-ins, patients were asked about their general condition and evaluated for any changes to their symptoms. On the weekends and public holidays, staff from SpeedDoc would perform the daily check-ins.

The DrCovid + application

The DrCovid + application was a chatbot hosted on the secured text-messaging application Telegram Messenger (Telegram Messenger LLP, London, United Kingdom). It served as a clinical dashboard for the automated collection, organisation and display of patients’ vital signs during their time in the Virtual Ward. Patients submitted their vital signs via Telegram, and the DrCovid + dashboard would then collate these parameters under the correct patient’s details, display the vital signs graphically to the Virtual Ward staff during the daily review, and trigger an immediate alert if the readings were outside of the normal range. The parameters submitted via DrCovid + still had to be electronically transcribed into the patients’ records by the ward staff, as the application was not integrated into the SGH EMR. Nonetheless, the time savings were substantial as staff need only perform a check-in once a day for the majority of patients, and the DrCOVID + enabled the Virtual Ward to scale up the number of patients that could be admitted and safely monitored at any one time.

Escalation and return to hospital

Patients were empowered to identify symptoms and signs that indicated a worsening of their COVID infection, and instructed to contact the Virtual Ward hotline if these occurred. These included:-

• Worsening symptoms such as chest pains or increased breathlessness

• Unstable vital signs, such as pulse oximetry of less than 95%, pulse rate more than 100 beats/minute, blood pressure below 90/60 mmHg, and/or fever of 38°C and above for three consecutive days

The medical staff would assess the patient via video consultation, and decide if the patient required a return to the hospital for further assessment. For other medical issues that were unrelated to the COVID-19 infection but still warranted an in-person assessment, the doctors and nurses from the Virtual Ward or from SpeedDoc would be deployed to perform a home visit.

Discharge criteria

There was a time-based discharge criterion from the Virtual Ward. Fully vaccinated patients were discharged without the need for a PCR test on Day 10 of their infection, or on Day 14 if they were unvaccinated or only partially vaccinated. Immunocompromised patients were only discharged upon reaching Day 21 of infection, regardless of vaccination status; however, they could be discharged earlier if they had a negative PCR test performed before Day 21.

Data collection

This study adhered to the tenets of the Declaration of Helsinki. Patient information including age, gender, vaccination status, comorbid health conditions and length of stay in the COVID-19 Virtual Ward was obtained from electronic health records. This data was recorded by administrative staff as part of an on-going audit of the outcomes of the Virtual Ward. The existing data set for the first 100 patients was de-identified and provided to the study team for analysis. Institutional Review Board approval was waived for this study as de-identified data was used.

Statistical analysis

Descriptive analysis was performed with Microsoft Excel Version 14.

Results

One hundred patients were admitted to the Virtual Ward from 8^th Oct 2021 to 7^th Dec 2021 (Table 2). Of these 100 patients, 58 (58.0%) were females, the mean age was 63.1 years old (range 23–95 years), and 76 (76.0%) were fully vaccinated against COVID-19 whilst the remaining were either unvaccinated or had not completed their vaccination. 56% of the patients were immunosuppressed and would have required an isolation period of 21 days or longer (Table 3). There were 77 (77.0%) hospitalised patients enrolled in the Virtual Ward, while the other 23 (23.0%) were enrolled directly from the community. The total number of days of inpatient hospitalisation avoided was 717 days, with an average of 7.9 days saved per patient. While most patients completed their home recovery without incident, 10 patients (10.0%) required a return to hospital for further management. There have been no deaths in this cohort, even in those that were transferred to hospital.

Table 2.

Demographic and COVID-19 profile of the first 100 patients in the COVID-19 virtual ward.

	Total (n = 100)	Hospital-based referrals (n = 77)	Community-based referrals (n = 23)
Gender
Male	42	32	10
Female	58	45	13
Ethnicity
Chinese	69	50	19
Malay	21	19	2
Indian	9	7	2
Eurasian	1	1	0
Age (years old)
Mean	63.1	57.1	83.2
Range	23–95	23–89	63–95
Vaccination status
Fully vaccinated	76	59	17
Partially vaccinated	11	9	2
Unvaccinated	13	9	4
Severity of COVID-19 infection
Pneumonia	18	18	0
Acute respiratory infection	80	58	22
Asymptomatic	2	1	1
Days of COVID-19 infection at time of enrolment
Mean	9.5	11.1	4.1
Range	2–40	4–40	2–9
Days of inpatient hospitalisation avoided
Mean	7.9	8.1	4.6
Range	1–16	1–16	1–13

Table 3.

Concurrent medical conditions and risk factors for severe COVID-19 complications in 77 inpatients enrolled into the COVID-19 virtual ward.

Medical Conditions / Risk Factors	Number of patients
Active Cancers with recent / on-going chemotherapy	30
Post-renal transplantation	17
Other solid organ transplants	3
Other conditions requiring systemic immunosuppression	9
Poorly-controlled diabetes mellitus	6
Pregnancy	4
Cardiac Disease	2
Others	6

Hospital-based admissions

Of the 77 hospital inpatients that were enrolled into the Virtual Ward, the mean age was 57.1 years old (range 23–89 years), and 59 (76.6%) had completed COVID-19 vaccination. 18 patients (23.4%) had been hospitalised for pneumonia, while 58 (75.3%) had symptoms of an acute respiratory infection, and 1 (1.3%) was asymptomatic. On average, patients had been hospitalised for 7.7 days (2–16 days) before their admission into the Virtual Ward; patients with pneumonia had been hospitalised for a mean of 10.1 days compared to the 7.0 days for those with mild or no symptoms.

Prior to their admission to the Virtual Ward, 19 (24.7%) patients had already received treatment with remdesivir (Veklury, Gilead Sciences, California, USA), 9 (9.1%) were treated on the combination monoclonal antibody therapy of casirivimab and imdevimab (REGEN-COV, Regeneron Pharmaceuticals Inc, New York, USA), 3 (3.9%) with sotrovimab (GlaxoSmithKline plc, London, England) and 15 (19.5%) had received a combination of these treatments. Of the 18 patients (23.4%) with pneumonia, all were systemically immunosuppressed and most had received either remdesivir (9 patients), casirivimab-imdevimab (2 patients), or a combination of remdesivir and either casirivimab-imdevimab or sotrovimab (6 patient) before their transition to the Virtual Ward. Only one of the patients with pneumonia declined an escalation of treatment and was treated symptomatically.

The majority of these patients in the Virtual Ward completed their home recovery without incident and avoided an average of 8.1 days (1–16 days) in hospital, although three patients (3.9%) required a return to hospital. The first patient had undergone renal transplantation two decades prior, and developed cellulitis of the upper limb during home recovery, while a second patient, also post-renal transplantation, developed generalised weakness due to a hyponatremia of 117 mmol/L. The third patient was receiving treatment for lymphoma and returned to hospital for an unexplained fever and rash.

Community-based admissions

From 1^st November 2021, CMTG referred 23 community-based patients to the Virtual Ward. None of patients were considered to be immunosuppressed. In this group, the mean age was 83.2 years old (range 63–95 years), and 17 had been fully vaccinated. On average, patients were on Day 4 of COVID-19 infection when they were enrolled into the Virtual Ward, and the mean number of hospitalisation days avoided was 4.6 days (1–9 days).

While the majority completed home recovery without complication, there were 7 (30.4%) patients that required inpatient hospitalisation for further management. Of these seven, two progressed to COVID pneumonia, two developed chest pains requiring further assessment, one had a hypertensive urgency, one developed a new-onset fever, and one patient had hyponatremia that resulted in a fall at home.

Discussion

The COVID-19 pandemic has resulted in millions of patients infected, and shortages in hospital capacity (particularly isolation and intensive care beds), equipment and staffing in numerous countries. One way for hospitals to expand their inpatient capacity will be for HaH programmes to manage COVID-19 patients at home. Our SGH COVID-19 Virtual Ward has demonstrated that a well-selected group of medically-complex patient with COVID-19 can safely recovery at home with remote medical support and monitoring, while making substantial savings in hospital bed-days.

Over the past three decades, HaH programmes have become more common worldwide because of concerns over the cost and availability of inpatient hospital care. Hospitalisation is also associated with physical and mental deconditioning, adverse events and even mortality.^7,8 Conversely, there remains concern that patients in HaH programmes do not receive care that is of equal quality to inpatient care, especially if urgent, life-saving treatment is required. Review of individual HaH programmes have not always shown consistent benefits, possibly due to the variety of conditions that are managed, and the widely differing populations, healthcare systems and reimbursement models involved.^9–12 Nonetheless, meta-analyses have demonstrated associations with reductions in mortality, readmission rates, and increases in patient and carer satisfaction,⁶ at similar or lower costs compared to inpatient care.¹³

The initial patients in our Virtual Ward were predominantly SGH inpatients with only acute respiratory symptoms, and were hospitalised because of immunosuppression or other systemic medical issues that excluded them from conventional home recovery or the CTFs. In the absence of the Virtual Ward, these patients would have spent up to 21 days in isolation before being discharged from the hospital, even though they may have already received treatment and were otherwise recovering well. Likewise, the 23.4% of patients with pneumonia would also have been excluded from home recovery, yet the Virtual Ward allowed them to recuperate at home without incident. Furthermore, the three patients who did require a return to hospital did so due to medical issues that were not directly related to their COVID-19 infection.

As the Virtual Ward expanded to include community-based referrals, the demographic profile of our patients began to change. On average, these patients were older, not fully vaccinated, and were enrolled earlier in their COVID-19 infection compared to the inpatient referrals. None would have received COVID-19 specific treatment. By definition, this a population at greater risk of COVID-19 related complications, and this is reflected in the much higher percentage that required a transfer to hospital for inpatient management of COVID-related complications and other medical issues. This percentage may have been reduced if every patient had been physically assessed by our medical staff prior to Virtual Ward enrolment, but that would have been logistically impractical given the number of referrals and their distribution throughout Singapore. Likewise, restricting the inclusion criteria any further (in terms of age or vaccination status) would have removed the Virtual Ward as an option for many of the patients that did eventually complete their supervised home recovery without incident.

A review of the literature^14–18 found a small number of HaH programmes and “virtual wards” managing COVID-19 patients since the pandemic began in 2020. The number of patients in these programmes ranged from 23 to 1477 individuals, typically enrolled over a 2–3-month period. Similar to our programme, some centres selected mild-to-moderately severe COVID-19 patients from the hospital, while other centres would admit patients directly from the emergency department or from family physicians. All the centres found their programmes to be safe and efficacious, and were able to substantially increase their inpatient hospital capacity with just a small dedicated medical team for the HaH programme. In fact, one HaH programme in Barcelona, Spain was staffed by senior doctors whom were recently retired, or who were immunosuppressed and otherwise restricted from working in the hospitals; these doctors volunteered to run the programme whilst working from home.¹⁷

We observed a hospitalisation rate of 10% in our first 100 patients, and the rate was even lower (3.9%) if we only considered our hospital-based referrals. These figures are similar to the 4.8%–7.6% reported by the majority of comparable COVID-19 HaH programmes. In contrast, our community-based patients had a hospitalisation rate of 30.4%, which is coincidentally the same as another HaH programme¹⁴ from Australia (7 out of 23 patients, 30.4%). Similar to our community referrals, they enrolled COVID-positive outpatients, many whom were elderly (median age 75 years old), and the majority (70%) were enrolled within a week of their symptom onset. Only one centre¹⁸ reported a death in their programme, out of the 917 patients admitted into their HaH over 3 months; this was a 75-year-old with pneumonia who passed away suddenly while at home.

Our study on the COVID-19 Virtual Ward has some limitations. Firstly, the findings reported here are based on a small sample at the time of writing, and involves a heterogeneous group that includes both hospital-based and community-based referrals. As we continue to enrol more patients directly from the community, it may be that the overall percentage of patients who will deteriorate and require hospitalisation will increase. Secondly, the Virtual Ward is successful because of the widespread availability of internet-connected phones and tablets, and patients and/or caregivers who are familiar with teleconferencing software. The DrCOVID + app also automates the data collection and monitoring of vital parameters. For less technologically-savvy patient populations, or in health systems without the easy access to such technology, it will be challenging for a Virtual Ward programme to scale up. Finally, all Singapore residents have their COVID treatment fully paid for by the state if they have undergone COVID vaccination. However, the Singapore government has “vaccine-differentiated” measures for patients who choose not to be vaccinated; such patients who become infected with COVID-19 will have to pay for part of their own treatment.¹⁹ While services such as the Virtual Ward remain highly subsidised, the out-of-pocket costs are not covered by government or by private insurance. There will be a financial disincentive for this group of patients to be in the Virtual Ward, and more patients may choose to remain hospitalized in order to make use of standard healthcare financing.

Conclusion

The early experience with the SGH COVID-19 Virtual Ward demonstrates that selected COVID-19 patients with mild-to-moderate symptoms and underlying medical conditions can be safely monitored and managed at home. This will help relieve some of the pressure on inpatient hospital resources, in particular isolation beds, while allowing patients to recover in a familiar environment without compromise in their care. With the necessary resources and technology tools, this Virtual Ward can be expanded to manage a much greater pool of patients from the hospital and the community.

Footnotes

Author contributions

M.W.J.T. conceptualised the study and drafted the initial manuscript. M.G.A.A, J.b.A, R.M.T, S.L, W.H.T. participated in the study design and data collection. L.L.L provided critical input. All authors approved the final manuscript as submitted.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval

Approval by the Singhealth Centralised Institutional Review Board was waived for this study as only de-identified data was used. This study was completed in accordance with the Helsinki Declaration.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Informed consent

Informed consent was not sought for the present study because only de-identified data was used for this study.

Data availability

The datasets generated and analysed during the current study are available from the corresponding author.

ORCID iDs

Michelle Woei Jen Tan

Rachel Marie Towle

References

COVID-19 situation report website. Available at: https://www.moh.gov.sg/covid-19/testing/situation-report-pdf

Duong

. Alpha, beta, delta, gamma: What’s important to know about SARS-CoV-2 variants of concern? CMAJ 2021; 193(27): E1059–E1060.

Campbell

Archer

Laurenson-Schafer

, et al. Increased transmissibility and global spread of SARS-CoV-2 variants of concern as at June 2021. Eurosurveillance 2021; 26(24): 2100509.

Zhang

Deng

Zhang

, et al. Transmission dynamics of an outbreak of the SARS-COV-2 Delta variant B.1.617.2 — Guangdong province, China, May–June 2021. China CDC Weekly 2021; 3(27): 584–586.

Update on hospitals’ capacity and manpower situation in managing the covid-19 surge. Available at: https://www.moh.gov.sg/news-highlights/details/update-on-hospitals-capacity-and-manpower-situation-in-managing-the-covid-19-surge

Caplan

Sulaiman

Mangin

, et al. A meta-analysis of hospital in the home. Med J Aust 2012; 197(9): 512–519.

Brennan

Leape

Laird

, et al. Incidence of adverse events and negligence in hospitalized patients: results of the harvard medical practice study I. N Engl J Med 1991; 324(6): 370–376.

Stockwell

Landrigan

Toomey

, et al. Adverse events in hospitalized pediatric patients. Pediatrics 2018; 142(2): e20173360.

Langhorne

Taylor

Murray

, et al. Early supported discharge for stroke patients: a meta-analysis of individual patients’ data. Lancet 2005; 365: 501–506.

10.

Ram

Wedzicha

Wright

, et al. Hospital at home for patients with acute exacerbations of chronic obstructive pulmonary disease: systematic review of evidence. BMJ 2004; 329: 315–319.

11.

Shepperd

Doll

Angus

, et al. Admission avoidance hospital at home. Cochrane Database Syst Rev 2008; 4: CD007491.

12.

Shepperd

Doll

Broad

, et al. Early discharge hospital at home. Cochrane Database Syst Rev 2009; 1: CD000356.

13.

Shepperd

Doll

Angus

, et al. Avoiding hospital admission through provision of hospital care at home: a systematic review and meta-analysis of individual patient data. CMAJ, 2009, 180(2), pp. 175–182.

14.

Lwin

Burgess

Johnston

, et al. Hospital-in-the-home experience of first 23 COVID-19 patients at a regional NSW hospital. Intern Med J 2020; 50(10): 1271–1273.

15.

Ferry

Moloney

Spratt

, et al. A virtual ward model of care for patients with covid-19: retrospective single-center clinical study. J Med Internet Res 2021; 23(2): e25518.

16.

Pericas

Cucchiari

Torrallardona-Murphy

, et al. Hospital at home for the management of COVID-19: preliminary experience with 63 patients. Infection 2021; 49(2): 327–332.

17.

Sitammagari

Murphy

Kowalkowski

, et al. Insights from rapid deployment of a “virtual hospital” as standard care during the covid-19 pandemic. Ann Intern Med 2021; 174(2): 192–199.

18.

Nogués

Sánchez-Martinez

Castells

, et al. Hospital-at-home expands hospital capacity during COVID-19 pandemic. J Am Med Dir Assoc 2021; 22(5): 939–942.

19.

“Calibrated adjustments in stabilisation phase.” Available at: https://www.moh.gov.sg/news-highlights/details/calibrated-adjustments-in-stabilisation-phase_8Nov2021