Home monitoring of coronavirus disease 2019 patients in different phases of disease

SARS-CoV-2 caused a global pandemic with substantial morbidity and mortality and stressed healthcare systems by necessitating reduced face-to-face consultations and creating shortages of beds and staff. To mitigate these challenges, the WHO urged implementation of home monitoring. Prior to COVID-19, remote monitoring showed benefits in other diseases (e.g., heart failure). During the pandemic, home monitoring rapidly expanded across three phases of COVID-19 care: prehospital monitoring to avoid observational admissions and detect early deterioration; early discharge to virtual wards with oxygen and medications; and post-discharge monitoring of recovery. This narrative review discusses home monitoring across these phases and its impact on safety, outcomes, and health-system burden.

Prehospital phase: UK Oximetry@home provided pulse oximeters to suspected/confirmed COVID-19 patients to track oxygen saturation and symptoms, enabling escalation when needed; baseline populations were largely mild (SpO2 ≥95% in 85%) and low-risk, and outcomes (presentations and mortality) were low, supporting safety and leading to national rollout. In a large US retrospective cohort (9278 invited; 5364 activations, mean age 46), activation was associated with lower hospitalization (2.4% vs 3.9%), ICU admission (0.3% vs 1.1%), and 90-day mortality (0.2% vs 0.6%); after propensity matching, hospitalization odds ratio was 0.68 (95% CI 0.54–0.86). The Mayo Clinic high-risk remote monitoring program showed lower hospitalization (13.7% vs 18%), shorter admissions, lower mortality (0.5% vs 1.7%), and cost savings ($1259 per patient over 30 days) after matching. Australian virtual hospital models monitoring low-risk patients reported very low admission and no deaths, supporting safety; targeted programs (e.g., pregnant women with mild disease) also appeared safe with early detection of deterioration. Hospital care at home (virtual wards): Early discharge programs sending recovering patients home with oxygen (commonly up to 3–4 L/min), dexamethasone and antibiotics showed reduced hospital days based on days of oxygen use at home (e.g., ~6–11 days), though estimates may be inflated compared with in-hospital tapering. An unblinded RCT (n=62) found a non-significant increase of 1.7 hospital-free days with home monitoring; design issues and crossover to home oxygen in controls likely diluted differences. UK evaluations of virtual wards noted shorter stays in monitored patients (10.3 vs 11.9 days) compared with contemporaneous admitted patients, but potential confounding from unmeasured baseline differences. Across studies, ED presentations were 3–12%, readmissions 3–9%, and no fatalities were reported; patient satisfaction was high. Recovery monitoring: Post-discharge monitoring for patients without ongoing in-hospital treatments showed mixed effectiveness. US programs reported reduced ED utilization and hospitalization, but a national UK analysis (~140,000 patients) comparing hospitals with and without COVID Virtual Wards found longer hospital stays and no difference in readmissions, with interpretation limited by incomplete enrollment and administrative artifacts. Longitudinal home spirometry studies showed ongoing FVC recovery (e.g., ~19% improvement over 6 months), and telerehabilitation improved physical outcomes. Requirements and tools: Effective programs combined symptom assessment with vital signs (especially SpO2), used triage by nurses/clinicians with 24/7 availability, leveraged phone/video for assessment, and employed thresholds for escalation. The centrality of oxygen saturation stems from silent hypoxemia in COVID-19; however, systematic reviews noted heterogeneity and limited evidence that SpO2 adds benefit beyond standard monitoring. A range of sensors (patches, wearables, in-ear devices) and apps/text-based systems supported data capture, education, and automated alerts.

Narrative review of home monitoring for COVID-19 across prehospital, early discharge (virtual ward), and recovery phases. The authors synthesize findings from observational cohorts, one randomized controlled trial, and systematic reviews, and summarize key studies in a comparative table. No formal systematic search strategy, meta-analysis, or predefined inclusion criteria are reported; evidence spans different pandemic waves and healthcare contexts.

- Prehospital monitoring appears safe and is associated with improved outcomes in several studies. In a large US cohort, activation reduced hospitalization (2.4% vs 3.9%), ICU admission (0.3% vs 1.1%), and 90-day mortality (0.2% vs 0.6%); adjusted OR for hospitalization 0.68 (95% CI 0.54–0.86). A high-risk program showed lower hospitalization (13.7% vs 18%), shorter admissions (about 6.7 vs 8.2 days), lower 30-day mortality (0.5% vs 1.7%; OR 0.31, 95% CI 0.12–0.78), and ~$1,259 cost savings per patient over 30 days. - Virtual wards enabling early discharge with home oxygen and medications were safe (ED presentations 3–12%, readmissions 3–9%, no fatalities) and may reduce in-hospital days by roughly 1.7–11 days depending on study and calculation; an RCT showed a non-significant 1.7-day increase in hospital-free days, likely underpowered and affected by crossover. - Recovery-phase monitoring without ongoing inpatient treatments provides support and may reduce ED visits and rehospitalizations in some settings, but a large UK analysis found longer hospital stays at hospitals with virtual wards and no difference in readmissions, with results limited by incomplete program uptake and data artifacts. - Oxygen saturation is the pivotal metric due to silent hypoxemia; thresholds for concern varied (92–95%). Systematic reviews indicate home monitoring is safe and can identify deterioration, but definitive added benefit of SpO2 over standard monitoring remains unproven. - Patient satisfaction with home monitoring is high; tele-rehabilitation improves physical recovery. - Effectiveness and cost-effectiveness vary by patient selection (often younger, low-risk in early programs), monitoring intensity, and pandemic phase; targeting high-risk populations may enhance value.

The review indicates that home monitoring across COVID-19 care phases can enhance safety, facilitate early detection of deterioration, and potentially reduce hospital utilization and costs, particularly in prehospital and early discharge settings. However, much of the evidence is observational, often involving younger, low-risk, digitally capable patients, introducing selection bias and limiting generalizability. Pandemic phase and evolving treatments/vaccination influence disease burden and the relative benefits of monitoring, complicating cross-study comparisons. To improve efficiency and impact, programs should prioritize high-risk groups (e.g., immunocompromised, transplant, dialysis) and incorporate validated risk prediction and escalation algorithms using symptoms and vital sign trends. Operationally, scalable 24/7 monitoring centers, integration with primary and hospital care, judicious use of wearables, and automated digital tools can optimize workload. The mixed findings in large administrative analyses highlight the importance of complete program enrollment, accurate status tracking, and robust study designs to evaluate real-world effectiveness.

Home monitoring programs for COVID-19 have been deployed across prehospital, early discharge (virtual ward), and recovery phases with strong safety profiles and high patient satisfaction. In the prehospital phase, monitoring can reduce hospitalization through early detection and intervention. Early discharge with home oxygen and medications likely shortens in-hospital stays but can extend time under hospital oversight due to slower oxygen tapering. Recovery monitoring mainly offers patient support, with modest effects on utilization. Future systems should be adaptable to fluctuating caseloads, focus on high-risk populations, integrate with monitoring for other conditions, and be ready for future pandemics.

- Evidence base dominated by retrospective observational studies, often without control groups or with propensity matching; limited randomized evidence and small sample sizes in RCTs. - Selection bias toward younger, healthier, digitally literate patients; activation was higher among young, female, and white patients, potentially reducing observed intervention needs and cost-effectiveness. - Heterogeneity in program design, monitoring intensity, escalation thresholds (e.g., SpO2 92–95%), and healthcare system context across pandemic waves limits comparability and precludes meta-analysis in reviews. - Potential overestimation of hospital-day reductions when calculated from at-home oxygen days due to slower tapering outside the hospital. - Administrative and data artifacts (e.g., patients registered as admitted while in virtual wards; incomplete enrollment in hospitals with virtual wards) complicate interpretation of large-scale analyses. - Limited evidence that adding SpO2 to symptom monitoring improves outcomes beyond standard care; further validation of risk algorithms and sensor utility is needed.