Changing patterns of SARS-CoV-2 infection through Delta and Omicron waves by vaccination status, previous infection and neighbourhood deprivation: a cohort analysis of 2.7 M people
Mark A. Green; Daniel J. Hungerford; David M. Hughes; Marta Garcia-Fiñana; Lance Turtle; Christopher Cheyne; Matthew Ashton; Gary Leeming; Malcolm G. Semple; Alex Singleton; Iain Buchan (2022). BMC Infectious Diseases, 22(1). DOI: 10.1186/s12879-022-07878-2
Abstract
Background Our study examines if SARS-CoV-2 infections varied by vaccination status, if an individual had previously tested positive and by neighbourhood socioeconomic deprivation across the Delta and Omicron epidemic waves of SARS-CoV-2.
Methods
Population cohort study using electronic health records for 2.7 M residents in Cheshire and Merseyside, England (3rd June 2021 to 1st March 2022). Our outcome variable was registered positive test for SARS-CoV-2. Explanatory variables were vaccination status, previous registered positive test and neighbourhood socioeconomic deprivation. Cox regression models were used to analyse associations.
Results
Originally higher SARS-CoV-2 rates in the most socioeconomically deprived neighbourhoods changed to being higher in the least deprived neighbourhoods from the 1st September 2021, and were inconsistent during the Omicron wave. Individuals who were fully vaccinated (two doses) were associated with fewer registered positive tests (e.g., individuals engaged in testing between 1st September and 27th November 2021—Hazards Ratio (HR) = 0.48, 95% Confidence Intervals (CIs) = 0.47–0.50. Individuals with a previous registered positive test were also less likely to have a registered positive test (e.g., individuals engaged in testing between 1st September and 27th November 2021—HR = 0.16, 95% CIs = 0.15–0.18. However, the Omicron period saw smaller effect sizes for both vaccination status and previous registered positive test.
Conclusions
Changing patterns of SARS-CoV-2 infections during the Delta and Omicron waves reveals a dynamic pandemic that continues to affect diverse communities in sometimes unexpected ways.
Extended Summary
This research examined how COVID-19 infection patterns varied by vaccination status, previous infection, and neighbourhood deprivation across the Delta and Omicron waves in England. Using electronic health records from 2.7 million residents in Cheshire and Merseyside between June 2021 and March 2022, the study analysed registered positive SARS-CoV-2 tests through three distinct time periods covering the Delta and Omicron variants. Cox regression models were employed to examine associations between infection risk and key explanatory variables including vaccination status, previous positive tests, and neighbourhood socioeconomic deprivation measured by the Index of Multiple Deprivation. The research revealed striking changes in infection patterns over time, challenging assumptions about which communities were most affected by COVID-19. Initially, infection rates were highest in the most socioeconomically deprived neighbourhoods, consistent with earlier pandemic patterns. However, from September 2021 onwards, this relationship reversed, with higher rates occurring in the least deprived areas during much of the Delta wave. During the Omicron period, patterns became more complex and inconsistent between different community groups. Vaccination provided significant protection during the Delta waves, with fully vaccinated individuals showing 52-60% lower likelihood of testing positive compared to unvaccinated people. However, this protective effect diminished considerably during the Omicron period, particularly for those with one or two vaccine doses. Booster vaccinations (third doses) maintained better protection against Omicron than initial vaccination courses. Previous infection offered even stronger protection than vaccination during Delta waves, with those who had previously tested positive being 84-86% less likely to test positive again. This natural immunity effect also weakened during Omicron, though it remained substantial after statistical adjustment. The study identified 11% of Omicron-period positive tests as reinfections, with rates notably higher in deprived areas. These findings illuminate the dynamic nature of pandemic spread, showing how socially patterned immunity from vaccination and prior infection created complex geographical and social patterns of risk. The research demonstrates that COVID-19’s impact evolved beyond simple deprivation-based inequalities to encompass more nuanced community-level factors including vaccination coverage, previous exposure, and behavioural changes. This work has important implications for public health planning, suggesting that pandemic responses must account for changing community vulnerabilities rather than assuming consistent risk patterns. Understanding these shifting dynamics is crucial for targeted interventions and communicating evolving risks to different population groups as societies adapt to living with endemic COVID-19.
Key Findings
- COVID-19 infection patterns reversed from affecting deprived areas most to affecting affluent neighbourhoods from September 2021
- Vaccination protection weakened significantly during Omicron, with booster doses maintaining better effectiveness than initial courses
- Previous infection provided stronger protection than vaccination during Delta waves but diminished considerably against Omicron variant
- Reinfection rates during Omicron reached 11% of positive tests, with higher rates observed in socioeconomically deprived areas
- Complex community-level factors beyond deprivation became increasingly important in determining infection risk patterns over time
Citation
@article{green2022changing,
author = {Mark A. Green; Daniel J. Hungerford; David M. Hughes; Marta Garcia-Fiñana; Lance Turtle; Christopher Cheyne; Matthew Ashton; Gary Leeming; Malcolm G. Semple; Alex Singleton; Iain Buchan},
title = {Changing patterns of SARS-CoV-2 infection through Delta and Omicron waves by vaccination status, previous infection and neighbourhood deprivation: a cohort analysis of 2.7 M people},
journal = {BMC Infectious Diseases},
year = {2022},
volume = {22(1)},
doi = {10.1186/s12879-022-07878-2}
}