Using Survey Data to Estimate the Impact of the Omicron Variant on Vaccine Efficacy against COVID-19 Infection

Jesús Rufino, Carlos Baquero, Davide Frey, Christin Glorioso, Antonio Ortega, Nina Reščič, Julian Roberts, Rosa E. Lillo, Raquel Menezes, Jaya Prakash Champati, Antonio Fernádez AntamedRxiv 2022 · DOI

Abstract

Data collected in the Global COVID-19 Trends and Impact Surveys (UMD Global CTIS), and data on variants sequencing from GISAID, are used to evaluate the impact of the Omicron variant on the prevalence of COVID-19 among unvaccinated and vaccinated populations. In South Africa, we observe that the prevalence of COVID-19 in December (Omicron-dominant) among the unvaccinated population is comparable to the prevalence during the previous Delta wave (August-September). However, among the vaccinated, the prevalence in December is much higher, indicating a significant reduction in vaccine efficacy. For instance, efficacy dropped from 0.81 to 0.30 for those with two doses. This study extends these findings to 24 other countries, finding a significant negative correlation between Omicron prevalence and vaccine efficacy.

Introduction

Throughout the COVID-19 pandemic, public health officials relied heavily on official case counts. However, these numbers are often delayed and limited by testing capacity. When the Omicron variant emerged in late 2021, its rapid spread and potential for immune evasion required immediate quantification. This research utilizes "participatory surveillance"—large-scale, self-reported survey data—to estimate how Omicron reduced vaccine protection against infection compared to the Delta variant.

Methodology: Adaptive Symptom Proxies

The researchers introduced a Random Forest classifier to identify COVID-19 cases from survey symptoms. Unlike static rules (e.g., fever + cough), the model was trained on a "ground-truth" subset of respondents who reported both symptoms and definitive test results. This allowed the model to learn evolving symptom signatures in real-time.

Comparison of symptom-based prevalence proxies with official confirmed cases.

The study integrated three primary data streams:

  1. UMD Global CTIS: Over 100,000 daily responses globally on symptoms and vaccination status.
  2. GISAID: Genomic sequencing data to track variant prevalence.
  3. Official Case Counts: Used to validate the survey-based proxies.

Vaccine efficacy ($V_E$) was calculated as $V_E = 1 - P_V/P_U$, where $P_V$ and $P_U$ are the prevalence among vaccinated and unvaccinated populations, respectively.

Results: The Erosion of Protection

The most striking finding was the divergence in prevalence trends between vaccinated and unvaccinated groups during the Omicron wave. In South Africa, while the peak prevalence for the unvaccinated in December was similar to the Delta peak in August, the prevalence among the vaccinated spiked dramatically. This divergence is a clear indicator that the variant was bypassing existing vaccine protection rather than just being more transmissible across the board.

COVID-19 prevalence trends by vaccination status and dosage in South Africa.

Quantitative Impact

The drop in efficacy was severe, particularly in South Africa:

  • Two-Dose Efficacy: Plummeted from 0.81 during the Delta period (Aug-Sep 2021) to 0.30 under Omicron (Dec 2021).
  • One-Dose Efficacy: Fell from 0.51 to just 0.09.
  • Global Correlation: Across 24 countries, a strong negative Pearson correlation (approx. -0.6) was found between Omicron prevalence and measured vaccine efficacy.

Discussion and Limitations

The study demonstrates that symptom-based proxies must be dynamic to account for viral evolution. However, several limitations remain. The data does not distinguish between primary series and booster shots, nor does it fully decouple variant impact from natural waning immunity over time. Crucially, this study focuses on infection prevalence; vaccines remained highly effective against severe disease and hospitalization during the Omicron wave.

Future research will incorporate vaccination timing and booster status to provide a more granular view of immune resilience against emerging variants.

Summary updated

Abstract

Abstract Data collected in the Global COVID-19 Trends and Impact Surveys (UMD Global CTIS), and data on variants sequencing from GISAID, are used to evaluate the impact of the Omicron variant (in South Africa and other countries) on the prevalence of COVID-19 among unvaccinated and vaccinated population, in general and discriminating by the number of doses. In South Africa, we observe that the prevalence of COVID-19 in December (with strong presence of Omicron) among the unvaccinated population is comparable to the prevalence during the previous wave (in August-September), in which Delta was the variant with the largest presence. However, among vaccinated, the prevalence of COVID-19 in December is much higher than in the previous wave. In fact, a significant reduction of the vaccine efficacy is observed from August-September to December. For instance, the efficacy drops from 0.81 to 0.30 for those vaccinated with 2 doses, and from 0.51 to 0.09 for those vaccinated with one dose. The study is then extended to other countries in which Omicron has been detected, comparing the situation in October (before Omicron) with that of December. While the reduction measured is smaller than in South Africa, we still found, for instance, an average drop in vaccine efficacy from 0.53 to 0.45 among those vaccinated with two doses. Moreover, we found a significant negative (Pearson) correlation of around − 0.6 between the measured prevalence of Omicron and the vaccine efficacy.
BibTeX
@misc{rufino2022using,
  author = {Jesús Rufino and Carlos Baquero and Davide Frey and Christin Glorioso and Antonio Ortega and Nina Reščič and Julian Roberts and Rosa E. Lillo and Raquel Menezes and Jaya Prakash Champati and Antonio Fernádez Anta},
  title = {{Using Survey Data to Estimate the Impact of the Omicron Variant on Vaccine Efficacy against COVID-19 Infection}},
  year = {2022},
  doi = {10.1101/2022.01.21.22269636},
  howpublished = {medRxiv},
}
Figure