Long-term effectiveness of COVID-19 vaccines against infections, hospitalisations, and mortality in adults: findings from a rapid living systematic evidence synthesis and meta-analysis up to December, 2022

BACKGROUND

Synthesising evidence on the long-term vaccine effectiveness of COVID-19 vaccines (BNT162b2 [Pfizer-BioNTech], mRNA-1273 [Moderna], ChAdOx1 nCoV-19 [AZD1222; Oxford-AstraZeneca], and Ad26.COV2.S [Janssen]) against infections, hospitalisations, and mortality is crucial to making evidence-based pandemic policy decisions.

METHODS

In this rapid living systematic evidence synthesis and meta-analysis, we searched EMBASE and the US National Institutes of Health's iSearch COVID-19 Portfolio, supplemented by manual searches of COVID-19-specific sources, until Dec 1, 2022, for studies that reported vaccine effectiveness immediately and at least 112 days after a primary vaccine series or at least 84 days after a booster dose. Single reviewers assessed titles, abstracts, and full-text articles, and extracted data, with a second reviewer verifying included studies. The primary outcomes were vaccine effectiveness against SARS-CoV-2 infections, hospitalisations, and mortality, which were assessed using three-level meta-analytic models. This study is registered with the National Collaborating Centre for Methods and Tools, review 473.

FINDINGS

We screened 16 696 records at the title and abstract level, appraised 832 (5·0%) full texts, and initially included 73 (0·4%) studies. Of these, we excluded five (7%) studies because of critical risk of bias, leaving 68 (93%) studies that were extracted for analysis. For infections caused by any SARS-CoV-2 strain, vaccine effectiveness for the primary series reduced from 83% (95% CI 80-86) at baseline (14-42 days) to 62% (53-69) by 112-139 days. Vaccine effectiveness at baseline was 92% (88-94) for hospitalisations and 91% (85-95) for mortality, and reduced to 79% (65-87) at 224-251 days for hospitalisations and 86% (73-93) at 168-195 days for mortality. Estimated vaccine effectiveness was lower for the omicron variant for infections, hospitalisations, and mortality at baseline compared with that of other variants, but subsequent reductions occurred at a similar rate across variants. For booster doses, which covered mostly omicron studies, vaccine effectiveness at baseline was 70% (56-80) against infections and 89% (82-93) against hospitalisations, and reduced to 43% (14-62) against infections and 71% (51-83) against hospitalisations at 112 days or later. Not enough studies were available to report on booster vaccine effectiveness against mortality.

INTERPRETATION

Our analyses indicate that vaccine effectiveness generally decreases over time against SARS-CoV-2 infections, hospitalisations, and mortality. The baseline vaccine effectiveness levels for the omicron variant were notably lower than for other variants. Therefore, other preventive measures (eg, face-mask wearing and physical distancing) might be necessary to manage the pandemic in the long term.

FUNDING

Canadian Institutes of Health Research and the Public Health Agency of Canada.

How well do covariates perform when adjusting for sampling bias in online COVID-19 research? Insights from multiverse analyses

COVID-19 research has relied heavily on convenience-based samples, which-though often necessary-are susceptible to important sampling biases. We begin with a theoretical overview and introduction to the dynamics that underlie sampling bias. We then empirically examine sampling bias in online COVID-19 surveys and evaluate the degree to which common statistical adjustments for demographic covariates successfully attenuate such bias. This registered study analysed responses to identical questions from three convenience and three largely representative samples (total N = 13,731) collected online in Canada within the International COVID-19 Awareness and Responses Evaluation Study ( www.icarestudy.com ). We compared samples on 11 behavioural and psychological outcomes (e.g., adherence to COVID-19 prevention measures, vaccine intentions) across three time points and employed multiverse-style analyses to examine how 512 combinations of demographic covariates (e.g., sex, age, education, income, ethnicity) impacted sampling discrepancies on these outcomes. Significant discrepancies emerged between samples on 73% of outcomes. Participants in the convenience samples held more positive thoughts towards and engaged in more COVID-19 prevention behaviours. Covariates attenuated sampling differences in only 55% of cases and increased differences in 45%. No covariate performed reliably well. Our results suggest that online convenience samples may display more positive dispositions towards COVID-19 prevention behaviours being studied than would samples drawn using more representative means. Adjusting results for demographic covariates frequently increased rather than decreased bias, suggesting that researchers should be cautious when interpreting adjusted findings. Using multiverse-style analyses as extended sensitivity analyses is recommended.