Using Virtual Care to Facilitate Direct Hospital Admissions in Outpatients with Worsening COVID-19 Infection

Background

Virtual care has emerged as an important method in delivering health care, for patients both with and without COVID-19 infection.¹ Although most outpatients with COVID-19 infection will recover at home, 4–8% will experience clinical deterioration requiring presentation to the emergency department (ED) for admission to hospital.^2,3 Recognizing an already overburdened ED at our institution, we identified an opportunity for virtual care to help triage and facilitate direct hospital admission in outpatients with COVID-19 and clinical deterioration. Doing so would allow outpatients to bypass the ED thereby improving patient flow, reducing redundant medical assessments (and health care worker exposure), minimizing handover, and improving patient-centered care.

The implementation of virtual care during the COVID-19 pandemic has been well described in the literature,^1,4–6 but the evaluation of such a model, specifically as it relates to facilitating direct hospital admission, has not been described to our knowledge. We report a 15-month cohort where a patient-centered virtual care program was used to bypass the ED and facilitate direct hospital admission for outpatients with worsening COVID-19 infection.

Methods

Sunnybrook Health Sciences Centre (SHSC) is a 678-bed academic health sciences center located in Toronto, Ontario, Canada. The first case of COVID-19 in Canada was diagnosed at our institution in January 2020.⁷ In March 2020, the COVID-19 expansion to outpatients (COVIDEO) virtual care program launched to provide education, counseling, and follow-up care for outpatients diagnosed with COVID-19 at SHSC.⁸ The model was staffed primarily by infectious diseases physicians, nurse practitioners, and family physicians.

The program consisted of an initial standardized assessment with follow-up assessments scheduled based on the patient’s risk of developing worsening disease. Telephone was the main virtual care modality used in this program. Pulse oximeters were proactively delivered to select high-risk patients, (>65 years of age, significant comorbidities, or cardiorespiratory symptoms)³ and all virtual encounters documented in the electronic medical record (EMR).

Within the first month of COVIDEO launching, the opportunity for this program to facilitate direct admissions and bypass the ED was identified. Criteria for direct admissions were developed through collaborative discussions between the COVIDEO team, the Division of General Internal Medicine, and COVID-19 inpatient ward team. Outpatients deemed appropriate for direct admission were patients with an oxygen saturation consistently <92% without severe respiratory distress. Oxygen saturations <92% have been associated with increased morbidity and mortality in the setting of pneumonia and is a generally accepted threshold for initiating oxygen supplementation.⁹

Outpatients excluded from this pathway included those for whom private transportation to hospital was logistically impossible or unsafe, as well as those with fulminant deterioration requiring immediate transfer by emergency medical services. Other reasons for direct admission were discussed with the admitting hospitalist on a case-by-case basis. To facilitate the rapid identification of patients who would potentially require hospitalization, patients were advised to contact the on-call COVIDEO physician through a pager (available 24 h a day, 7 days a week), in the event of worsening symptoms (such as chest pain or shortness of breath) or oxygen saturation persistently <92%.

The decision to directly admit a patient was shared between the COVIDEO physician and admitting hospitalist, and was based on whether in-hospital diagnostic tests or treatments (such as intravenous fluids or supplemental oxygen) were needed, whether these treatments could be delivered as an outpatient, and whether the patient could be safely and reliably supported at home. Once a patient was deemed appropriate for this pathway, telephone handover from the COVIDEO on-call physician and admitting hospitalist was given and COVIDEO EMR documentation reviewed. The hospitalist then informed the COVID-19 ward to ensure a bed was available for hospital admission.

Patients were advised by the COVIDEO on-call physician to arrange private transportation to hospital utilizing infection control best practices (such as performing hand hygiene before entering the vehicle, wearing a mask, and sitting in the backseat with windows rolled down), and to remain in their vehicle outside a designated hospital entrance. Upon arrival, the patient (or family) called the COVID-19 ward to inform the nursing team leader of their arrival.

A nurse wearing appropriate personal protective equipment would meet the patients at their vehicle and transport them to the COVID-19 ward. Upon arrival to the ward, the patient was assessed by the hospitalist and admission orders were completed. If no hospital beds were available, the patient was advised to seek care at the closest ED or call paramedic services. This situation was encountered only a few times during the study period.

To assess the impact of the intervention, demographic information, reason for direct admission, treatments employed in hospital, need for ICU transfer, length of hospital stay, and clinical outcome were collected for each patient directly admitted to hospital. Descriptive statistics were used to characterize the cohort. This study met criteria for exemption of research ethics review based on our institutional process for confirming that the project involved quality improvement and not human subject research.

Results

From March 1, 2020, to May 31, 2021, 9,116 patients were managed by the COVIDEO program. During this period, 164 (1.8%) outpatients followed by the COVIDEO program were hospitalized at SHSC due to clinical deterioration, of whom 83 (50.6%) were directly admitted to the COVID-19 ward. Clinical characteristics of the 83 patients directly admitted to hospital are outlined in Table 1. Ten (12.0%) of these patients required transfer to ICU, occurring a median time of 4 days from hospital admission. No patients required transfer to ICU within the first 48 h of hospitalization. The overall mortality rate in this cohort was 3.6% (3/83).

The number of patients directly admitted to hospital fluctuated based on each wave of the pandemic, with few direct admissions occurring between waves (Fig. 1). Twenty-six direct admissions occurred in April 2021 that corresponded to the peak of the third wave in the province.