BACKGROUND/RATIONALE:
Remdesivir is a prodrug of an inhibitor of the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) RNA polymerase and one of the first drugs studied for treatment of coronavirus disease 2019 (COVID-19). Randomized trials have yielded conflicting results about the impacts of remdesivir therapy on survival and length of hospital stay among people hospitalized with COVID-19. The Adaptive COVID-19 Treatment Trial (ACTT-1) found that remdesivir shortened time to recovery from a median of 15 days to 11 days (rate ratio 1.32, 95% CI 1.12-1.55).1 Remdesivir treatment in ACTT-1 was also associated with a statistically non-significant reduction in mortality at 28 days (11.4% vs. 15.2%, hazard ratio 0.73, 95% CI 0.52 to 1.03). Based largely on these results, the United States Food and Drug Administration (FDA) issued an emergency use authorization (EUA) for remdesivir treatment for hospitalized adults in May 2020 and formally approved its use in October. In contrast, the World Health Organization (WHO) Solidarity trial found that remdesivir did not reduce length of hospital stay or improve survival compared to standard of care (rate ratio for death by 28 days 0.95, 95% CI 0.81 to 1.11),2 and the WHO does not recommend remdesivir treatment for people with COVID-19.
Observational studies can provide useful information about remdesivir effectiveness in routine clinical practice, and further inform recommendations for use.

OBJECTIVE(S):
Our overall objective was to use national data available in the Veterans Health Administration (VHA) to conduct a cohort study to characterize remdesivir use and effectiveness among Veterans hospitalized with COVID-19 during the EUA period from May-October 2020. Specific aims included:

Aim 1: Characterize hospitalized Veterans with COVID-19 and identify veteran and facility-level factors associated with variation in remdesivir use

Aim 2: Determine associations between remdesivir treatment and patient outcomes, including 30-day all-cause mortality and length of hospital stay

METHODS:
We used data in VHA's COVID-19 Shared Data Repository (CSDR) and Corporate Data Warehouse (CDW) to identify Veterans with a first admission to a VHA facility with laboratory-confirmed COVID-19 from May 1 to October 8, 2020. We included patients with a positive Polymerase Chain Reaction (PCR) result for SARS-CoV-2 between 14 days before and 5 days after admission (n=7,243). We then excluded those with: 1) no primary care visits in VHA in the 2 years prior to admission because they lacked data on key risk adjustment variables (N=740); 2) admission to hospice care in inpatient setting on first day of hospitalization (N=119); and 3) no valid values for alanine amino transferase (ALT), aspartate amino transferase (AST), or estimated glomerular filtration rates (eGFR) during hospital stay (N=486) because VHA limited remdesivir availability to patients with ALT and AST less than 5 times upper normal limit and eGFR greater than 30 ml/minute. This left 5,898 patients in the initial cohort. We used two data sources to identify exposure to remdesivir: 1) bar-code medication administration data, and 2) a list obtained from Pharmacy Benefits Management (PBM) of Veterans with an approved remdesivir EUA request, which we used to validate medication administration data under emergency use. We then compared characteristics of patients who received and did not receive remdesivir during hospitalization.

We used two methods to quantify risk-adjusted associations between remdesivir treatment, time to death within 30 days, and time to hospital discharge as a competing event with death. Our primary approach involved propensity score matching of patients initiating remdesivir to control patients who had not initiated remdesivir by the same hospital day, with stratification according to receipt of dexamethasone or other corticosteroids at time of match. We used Cox regression models to estimate differences in treatment outcomes in the matched cohort. An alternate approach used marginal structural models with inverse probability of treatment weights (IPTW), using stabilized weights by hospital day. Risk adjustment variables included demographics, comorbidities, and laboratory values, vital signs, mechanical ventilation, and ICU admission by hospital day.

FINDINGS/RESULTS:
Compared to 3,524 (59.7%)patients who never received remdesivir during hospitalization, the 2.374 (40.3%) remdesivir recipients were older (mean age 67.8 vs. 67.0, p=0.03), more likely to be White (59.6% vs. 54.4%, p<0.001), more likely to have chronic pulmonary disease (37.4% vs 32.0%, p<0.001), and more ill at admission based on ICU care and vital signs. Propensity score-matched analyses included 1,172 remdesivir recipients matched to 1,172 controls (total N= 2.344). Remdesivir recipients and matched controls had similar mean age (66.6 vs. 67.5 years), male sex (93.9% vs. 93.9%), comorbidity, dexamethasone use (47.7% vs. 47.7%), intensive care unit stay (20.7% vs. 19.1%),mechanical ventilation(5.9% vs. 3.8%), laboratory values, and vital signs. Standardized differences were less than 10% for all measures.
Remdesivir treatment was not associated with 30-day mortality [12.2% for remdesivir recipients vs.10.6% for controls, log rank p=0.26, adjusted hazard ratio (aHR) 1.06, 95% confidence interval (CI) 0.83 to 1.36]. Results were similar for people receiving and not receiving dexamethasone at remdesivir initiation (aHR 0.93, 95% CI 0.64 to 1.35, for dexamethasone recipients and aHR 1.19, 95% CI 0.84 to 1.69, for non-recipients). Remdesivir recipients had longer time to hospital discharge compared to matched controls (median 6 days from remdesivir initiation to discharge compared to 3 days from corresponding hospital day to discharge for controls, p<0.001). Many patients receiving remdesivir were discharged 5 or 6 days after initiating remdesivir, in association with completion of 5-day remdesivir courses. Results were similar in analyses using marginal structural models. The hazard ratio for death within 30 days in these analyses was 0.98 (95% CI 0.71-1.35) for remdesivir recipients relative to controls, and the hazard ratio for hospital discharge was 0.72 (95% CI 0.53-0.97).

Findings in this observational study complement results from prior randomized trials. ACTT-1 indicated that remdesivir shortened time to clinical recovery and could potentially hasten discharge, but the trial excluded patients who were thought likely to be discharged with 72 hours. Our study suggests that remdesivir treatment actually increased time to hospital discharge as it was administered in routine clinical settings, possibly because clinicians were keeping patients in hospital to complete 5-day remdesivir courses. Results may be subject to unadjusted confounding by indication and limitations in available data.

IMPACT:
Routine use of remdesivir may be increasing utilization of scarce hospital beds during a pandemic without leading to clear improvements in survival. There is need for further research to identify optimal indications for remdesivir use and quality improvement projects to ensure that patients are not kept in hospital solely to receive remdesivir.

PUBLICATIONS:

Journal Articles

1. Ohl ME, Miller DR, Lund BC, Kobayashi T, Richardson Miell K, Beck BF, Alexander B, Crothers K, Vaughan Sarrazin MS. Association of Remdesivir Treatment With Survival and Length of Hospital Stay Among US Veterans Hospitalized With COVID-19. JAMA Network Open. 2021 Jul 1; 4(7):e2114741. [view]