Public Health Benefits of Applying Evidence-Based Best Practices in Managing Patients Hospitalized for COVID-19

Abstract

Background: As coronavirus disease 2019 (COVID-19)-related mortality remains a concern, optimal management of patients hospitalized for COVID-19 continues to evolve. We developed a population model based on real-world evidence to quantify the clinical impact of increased utilization of remdesivir, the effectiveness of which has been well established in hospitalized patients with COVID-19.

Methods: The PINC AI healthcare database records for patients hospitalized for COVID-19 from January to December 2023 were stratified by those treated with or without remdesivir ("RDV" and "no RDV") and by supplemental oxygen requirements: no supplemental oxygen charges (NSOc), low-flow oxygen (LFO), and high-flow oxygen/non-invasive ventilation. Key vulnerable subgroups such as elderly and immunocompromised patients were also evaluated. The model applied previously published hazard ratios (HRs) to 28-day in-hospital mortality incidence to determine the number of potential lives saved if additional no RDV patients had been treated with remdesivir upon hospital admission.

Results: Of 84 810 hospitalizations for COVID-19 in 2023, 13,233 no RDV patients were similar in terms of characteristics and clinical presentation to the RDV patients. The model predicted that initiation of remdesivir in these patients could have saved 231 lives. Projected nationally, this translates to >800 potential lives saved (95% confidence interval, 469-1126). Eighty-nine percent of potential lives saved were elderly and 19% were immunocompromised individuals. Seventy-six percent were among NSOc or LFO patients.

Conclusions: This public health model underscores the value of initiating remdesivir upon admission in patients hospitalized for COVID-19, in accordance with evidence-based best practices, to minimize lives lost because of severe acute respiratory syndrome coronavirus 2 infection.

Keywords: COVID-19 model; SARS-CoV-2; data science; dexamethasone; mortality; real world data; remdesivir.

Figure 1.

Potential lives saved in the PHD Population hospitalized for COVID-19 if remdesivir were initiated in the model cohort. Base case and scenarios 1–4 results are shown. Stacked bar percentages show relative impact of remdesivir treatment stratified by baseline supplemental oxygen requirements. Scenario 1: apply HR derived from the full omicron period; scenario 2: exclude patients on baricitinib, tocilizumab, or oral antivirals; scenario 3: include patients on IMV; scenario 4: compare RDV + DEX vs DEX monotherapy (excluding prior nonremdesivir treatments). Abbreviations: DEX, dexamethasone; HFO, high-flow oxygen; HR, hazard ratio; IMV, invasive mechanical ventilation; LFO, low-flow oxygen; NIV, noninvasive ventilation; NSOc, no supplemental oxygen charges; PHD, PINC AI healthcare database; RDV, remdesivir.