# A retrospective assessment of COVID-19 vulnerability index indicators and mortality rates pre-COVID-19 (2018–2020) and during COVID-19 (2020–2022) in a health and demographic surveillance site, Soweto, South Africa

**Authors:** Takwanisa Machemedze, Chodziwadziwa Whiteson Kabudula, Jean Juste Harrisson Bashingwa, Beth A. Tippett Barr, Nellie Myburgh, Sana Mahtab, Cleopas Hwinya, Stephen Tollman, Ziyaad Dangor, Shabir A. Madhi

PMC · DOI: 10.1186/s12963-025-00387-9 · Population Health Metrics · 2025-06-20

## TL;DR

This study in Soweto, South Africa, found that mortality rates doubled during the COVID-19 pandemic compared to before, with vulnerable groups being at higher risk.

## Contribution

The study evaluates a pre-developed vulnerability index to assess mortality risk before and during the pandemic in a specific South African population.

## Key findings

- Mortality rates during the COVID-19 period were more than double those before the pandemic.
- Individuals in the highest vulnerability tertile had a significantly lower risk of death compared to the lowest tertile.
- The risk of dying during the pandemic was at least double for each vulnerability indicator compared to pre-COVID-19.

## Abstract

Before COVID-19, knowledge on pandemic vulnerability and mortality in South Africa was largely limited to the context of HIV/AIDS. We evaluated mortality rates and risk of death, prior to and during the COVID-19 pandemic, in relation to an individual’s COVID-19 vulnerability, based on a scoring algorithm developed in South Africa.

The analysis was undertaken using data from a health and demographic surveillance system (HDSS) in Soweto and Thembelihle, Gauteng, South Africa. Health and demographic population-based data have been collected from the HDSS area since 2018. Using indicators included in a COVID-19 Vulnerability Index, previously developed in South Africa, the current study established a composite COVID-19 vulnerability index, stratified into tertiles. The risk of death pre-COVID-19 (1 January 2018–28 February 2020) and during the COVID-19 period (1 March 2020–31 December 2021) was analysed. A Cox proportional hazard model was used to compare the risk of death between the two time periods. Statistical analyses were conducted using Stata software version 17.

Before COVID-19, overall mortality rates were 8.1 (95% CI 7.6–8.8), 7.0 (95% CI 6.4–7.7) and 6.1 (95% CI 5.5–6.7) per 1000 person-years in the lowest, middle, and highest tertile of vulnerability index, respectively. All cause-mortality across all tertiles more than doubled during the COVID-19 period compared to pre-COVID-19 (15.5 against 7.2). The mortality rates during the COVID-19 era were 17.1 (95% CI 16.3–18.0), 14.5 (95% CI 13.4–15. 7) and 13.7 (95% CI 12.8–14.7) per 1000 person-years in the lowest, middle, and highest tertiles, respectively. Overall, individuals in the highest tertile of COVID-19 vulnerability were at a significantly lower risk of death relative to those in the lowest tertile (aHR 0.9, 95% CI 0.8–1.0, p < 0.05). The risk of dying during the COVID-19 period for vulnerable individuals was at least double compared to the pre-COVID-19 period for each of the individual vulnerability indicators.

All-cause mortality during the COVID-19 era was significantly higher than the pre- COVID-19 period, with the increase observed across all vulnerability tertiles. It is important to identify vulnerable individuals and communities during the early stages of a pandemic to inform prioritisation of public health intervention.

The online version contains supplementary material available at 10.1186/s12963-025-00387-9.

## Linked entities

- **Diseases:** COVID-19 (MONDO:0100096)

## Full-text entities

- **Diseases:** death (MESH:D003643), HIV/AIDS (MESH:D015658), COVID-19 (MESH:D000086382)

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC12181823/full.md

## Figures

1 figure with captions in the complete paper: https://tomesphere.com/paper/PMC12181823/full.md

## References

11 references — full list in the complete paper: https://tomesphere.com/paper/PMC12181823/full.md

---
Source: https://tomesphere.com/paper/PMC12181823