Intensity of public health and social measures are associated with effectiveness of SARS-CoV-2 vaccine in test-negative study

Abstract

The intensity and duration of exposure can influence vaccine effectiveness (VE). For "leaky" vaccines such as SARS-CoV-2 vaccines, which reduce but do not entirely prevent infections, repeated or prolonged exposures may increase breakthrough infection likelihood. To test this hypothesis, we conducted a systematic review and meta-analysis of 76 test-negative design studies reporting VE against SARS-CoV-2 infection or severe disease. Exposure intensity was approximated using Oxford COVID-19 Government Response Tracker indices: Stringency Index (SI), Containment and Health Index (CHI), and Government Response Index (GRI). Based on 1,419 VE estimates, pooled VE against infection was significantly higher in settings with higher index values (lower exposure intensity): 82% (95% CI: 80-83%) in high-SI settings versus 39% (95% CI: 35-43%) in low-SI settings. Similar patterns appeared for other indices and severe disease outcomes. These associations persisted in meta-regression models adjusting for viral variant, vaccine type, time since vaccination, prior infection status, and enrollment criteria. Correlation analyses showed moderate-to-strong positive correlations between VE estimates and exposure indices (Spearman's correlation: 0.50-0.62). These findings establish exposure intensity as a critical effect modifier of SARS-CoV-2 VE, demonstrating the leaky nature of COVID-19 vaccines and explaining heterogeneity in real-world effectiveness estimates. Future VE evaluations and vaccination strategies should account for exposure intensity to ensure accurate, context-specific estimates.

Keywords: COVID-19; SARS-CoV-2; pre-existing immunity; test-negative design; vaccination; vaccine effectiveness.

Figure 1.

Selection of studies for the systematic review

Figure 2.

Association between vaccine effectiveness (VE) and public health indices. Scatterplots illustrating the relationship between VE and the Stringency Index (SI), Containment and Health Index (CHI), and Government Response Index (GRI) for different outcomes (infection and severe disease). Correlation coefficients (r and ρ) are shown for the overall dataset and stratified by variant periods (pre-Delta and Delta, late-Delta, Omicron BA.1/BA.2, Omicron BA.4/BA.5). The figures demonstrate positive correlations between VE and higher index values, indicating that stricter public health measures are associated with increased VE.

Figure 3.

Impact of public health indices on pooled vaccine effectiveness (VE). Meta-analysis results comparing VE across tertiles of the Stringency Index (SI), Containment and Health Index (CHI), and Government Response Index (GRI). The pooled VE is significantly higher in the highest tertile of each index, suggesting that stricter control measures, and consequently lower exposure levels, are associated with improved VE against both infection and severe disease.