The purpose of this study is to assess whether convalescent plasma therapy is associated with reduced 30-day all-cause mortality in a population of Veteran inpatients with non-severe coronavirus disease 2019 (COVID-19).

The convalescent plasma (CP) of individuals who have recovered from COVID-19 may help reduce symptoms and mortality when given to ill patients infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Clinical trials are needed to provide definitive evidence of the benefits and harms of CP in the treatment of COVID-19. While clinical trials are underway, observational research plays a critical role in providing preliminary scientific evidence to guide clinical practice.

This observational study aims to estimate the association between CP therapy and mortality after 30 days in a sample of SARS-CoV-2-positive patients with non-severe illness. Patients include those hospitalized at a facility within the Veterans Health Administration (VHA) of the Department of Veterans Affairs (VA), the largest integrated healthcare system in the United States. Using VA electronic health record data, the study is designed to emulate a target trial of convalescent plasma transfusion and mortality. The primary exposure of interest is the use or non-use of COVID-19 convalescent plasma, and the primary outcome is 30-day all-cause mortality.

A nested trial design will be used, in which patients will be followed from each day in which eligibility criteria are met (while hospitalized and after date of first positive SARS-CoV-2 test) for 30 days. The association between CP therapy and mortality will be assessed using pooled logistic models to estimate 30-day cumulative incidence curves, risk differences, and hazard ratios. Inverse probability weighting derived from propensity scores will be employed to minimize confounding by indication and other biases.

Inclusion Criteria:

• US Veterans aged 21-80 years old
• Hospitalized between May 1, 2020 and November 17, 2020 with a SARS-CoV-2 positive test, at a VA Medical Center where convalescent plasma had been administered to at least one patient and remained a current practice at that VA Medical Center
• SARS-CoV-2 positive test within 7 days before or after hospital admission
• Minimum oxygen saturation (measured within the past day) >=90%
• Vitals (pulse, respiration, temperature, systolic blood pressure) and acute labs (hemoglobin, platelet, white blood cells) measured within the past 2 days
• Albumin, Alanine Aminotransferase (ALT), creatinine measured within the past 30 days
• Weight measurement recorded in the past 2 years

Exclusion Criteria:

• Prior intubation, ventilation, high flow oxygen, extracorporeal membrane oxygenation (ECMO), dialysis, or vasopressors during current hospitalization
• Record of prior treatment with CP
• Received long-term care in a domiciliary or nursing home in the past 90 days
• First CP recipient at a site
• Less than 30 days of follow-up

Veterans with non-severe COVID-19 who tested positive for SARS-CoV-2 infection during admission at a VA medical center or who were admitted as inpatients to a VA medical center within 30 days after a positive test for SARS-CoV-2 infection

Data on SARS-CoV-2 testing, COVID-19 treatment, comorbidities, and other COVID-19-related information are available through the VA COVID-19 Shared Data Resource data domain within the VA Corporate Data Warehouse (CDW). Veterans Health Administration (VHA) researchers can request access to VHA data sources and data access tools for a research protocol approved by a VA Research and Development (R&D) Committee and reviewed by an Institutional Review Board (IRB). Information about requesting access to http://vaww.vhadataportal.med.va.gov/DataAccess/ResearchAcces (internal to VA).

Because the study design allowed a patient enter a trial more than once, the sum of unique patients assigned to plasma and non-plasma groups (n=5,044) exceeds the number of eligible patients identified in the electronic health records (n=4,755).

30 days

No adverse event data were collected for this retrospective observational EHR-based study