Effectiveness of COVID-19 mRNA Vaccines Against COVID-19–Associated Hospitalization — Five Veterans Affairs Medical Centers, United States, February 1–August 6, 2021

Abstract

this report was posted as an MMWR Early Release on the MMWR website (https://www.cdc.gov/mmwr).COVID-19 mRNA vaccines (Pfizer-BioNTech and Moderna) have been shown to be highly protective against COVID-19associated hospitalizations (1-3). Data are limited on the level of protection against hospitalization among disproportionately affected populations in the United States, particularly during periods in which the B.1.617.2 (Delta) variant of SARS-CoV-2, the virus that causes COVID-19, predominates (2). U.S. veterans are older, more racially diverse, and have higher prevalences of underlying medical conditions than persons in the general U.S. population (2,4). CDC assessed the effectiveness of mRNA vaccines against COVID-19associated hospitalization among 1,175 U.S. veterans aged ≥18 years hospitalized at five Veterans Affairs Medical Centers (VAMCs) during February 1-August 6, 2021. Among these hospitalized persons, 1,093 (93.0%) were men, the median age was 68 years, 574 (48.9%) were non-Hispanic Black (Black), 475 were non-Hispanic White (White), and 522 (44.4%) had a Charlson comorbidity index score of ≥3 (5). Overall adjusted vaccine effectiveness against COVID-19-associated hospitalization was 86.8% (95% confidence interval [CI] = 80.4%-91.1%) and was similar before (February 1-June 30) and during (July 1-August 6) SARS-CoV-2 Delta variant predominance (84.1% versus 89.3%, respectively). Vaccine effectiveness was 79.8% (95% CI = 67.7%-87.4%) among adults aged ≥65 years and 95.1% (95% CI = 89.1%-97.8%) among those aged 18-64 years. COVID-19 mRNA vaccines are highly effective in preventing COVID-19-associated hospitalization in this older, racially diverse population of predominately male U.S. veterans. Additional evaluations of vaccine effectiveness among various age groups are warranted. To prevent COVID-19-related hospitalizations, all eligible persons should receive COVID-19 vaccination.

COVID-19 Vaccine Effectiveness by Product and Timing in New York State

“…10,11 This potential bias comes in exchange for large sample sizes and outcome numbers that exceed those of other nations' studies and of smaller, more controlled designs. 6,7,13,14 Our findings are further strengthened by broad age representation across time-cohorts and the consistency of changes in calendar time.…”

The Mask-Wearing Bias In The Estimates Of Vaccine Efficacy

“…In this note I argue that the discrepancy in estimated vaccine efficacy seen in various countries can be explained by a confounder not usually recognized as such 12 . Especially in the United States, there appears to be a strong correlation between mask wearing (and adherence to other infection-control measures) 3 and vaccine acceptance [17].…”

“…At the center of the debate are the data from studies by Pfizer and collaborators [1,2] and Israeli researchers [3,6] showing that the efficacy of the Pfizer BNT162b2 COVID19 vaccine decreases after about 4-6 months. The need for boosters is questioned by other authors and regulators [7,8,9] who aver that the effectiveness of vaccines, especially against severe disease, appears to still be high in the United States [10,11,12]. Which view is correct has important consequences for the public health decisions currently being made all over the world.…”

“…That is, in both arms the pattern of mask wearing is the same as the actual pattern in the vaccinated population. 12 Then the number of cases in the vaccine arm would be expected to be n uv p uv + n mv p mv . In the control arm we would expect different proportions of disease, with the resulting case count n uv p uu + n mv p mu .…”

“…The respective proportions of mask-wearers among the vaccinated and unvaccinated are q(m|v) = n mv n uv + n mv (12) q(m|u) = n mu n uu + n mu (13) Using these definitions and ( 11), F = (1 − q(m|u))p uu + q(m|u)p mu (1 − q(m|v))p uu + q(m|v)p mu (14) Note that this expression depends only on the ratio p mu /p uu , which is directly related to the effectiveness of masks (absent vaccination). In fact, similarly to V E, we can define the mask effectiveness…”