A Propensity-Matched Cohort Assessing Impact of a Neutralizing Monoclonal Antibody in Mild-to-Moderate Coronavirus Disease 2019

• Published in: Journal of intensive care medicine 2023-06, Vol.38 (6), p.511-518
• Main Authors: Abbas, Malak, Farhat, Nada, Hammoud, Zainab, Dickey, Curtis, Shuayto, Ali, Chen, Nai-Wei, Hsaiky, Lama M, Sims, Matthew, Sengstock, David, Schramski, Joseph, Shamoon, Zafar
• Format: Article
• Language: English
• Subjects: Antibodies, Monoclonal - therapeutic use; COVID-19; Humans; Retrospective Studies; SARS-CoV-2
• ISSN: 0885-0666; 1525-1489
• DOI: 10.1177/08850666231155822

Background: Severe acute respiratory syndrome coronavirus 2 (SARS-CoaV-2) is responsible for the coronavirus disease 2019 (COVID-19) pandemic. In randomized clinical trials, patients who were treated with the anti-spike monoclonal antibody bamlanivimab had fewer COVID-19-related hospitalizations or emergency department (ED) visits than the control group. Methods: A retrospective cohort was assembled across a multisite healthcare system between November 20, 2020 and March 31, 2021. Ambulatory COVID-19 patients treated with bamlanivimab (n = 209) were propensity score matched without replacement (1:1) to a pool of 1024 eligible control patients who received similar care without bamlanivimab. The primary endpoint was all-cause mortality or admission at 30 days. Secondary endpoints included hospitalization, critical care admission, oxygenation requirements, and infusion-related reactions. Propensity score matching (PSM) analysis was used to assess the effect of bamlanivimab infusion on the composite endpoint and secondary endpoints. Results: A total of n = 209 matched patients were included in each arm of the study. The absolute standardized difference (stddiff) was calculated and indicated a balance between the groups. Almost all variables had a stddiff of less than 0.10, except for respiratory rate (RR) (stddiff = −0.11). For the primary composite endpoint of the matched cohort, 10.1% (n = 21) of patients in the intervention group were hospitalized or deceased within 30-day postbamlanivimab infusion versus 27.8% (n = 58) in the control group (adjusted odds ratio [aOR]: 0.29, 95% confidence interval [CI]: 0.17 to 0.51, P