Better Strategies for Containing COVID-19 Epidemics–A Study of 25 Countries via an Extended SEIR Model

Jia Gu, Han Yan, Yaxuan Huang, Yuru Zhu, Haoxuan Sun, Xinyu Zhang, Yuqing Wang, Yumou Qiu, Song Xi ChenmedRxiv 2020 · DOI

Abstract

We evaluate the effectiveness of COVID-19 control strategies of 25 countries which have endured more than four weeks of community infections. With an extended SEIR model that allows infections in both the exposed and infected states, the key epidemic parameters are estimated from each country’s data, which facilitate the evaluation and cross-country comparison. It is found quicker control measures significantly reduce the average reproduction numbers and shorten the time length to infection peaks. If the swift control measures of Korea and China were implemented, average reductions of 88% in the confirmed cases and 80% in deaths would had been attained for the other 23 countries from start to April 10. Effects of earlier or delayed interventions in the US and the UK are experimented which show at least 75% (29%) less infections and deaths can be attained for the US (the UK) under a Five-Day Earlier experiment. The impacts of two removal regimes (Korea and Italy) on the total infection and death tolls on the other countries are compared with the natural forecast ones, which suggest there are still ample opportunity for countries to reduce the final death numbers by improving the removal process.

Introduction

In the early months of the COVID-19 pandemic, the global community faced a critical challenge: understanding the efficacy of various containment strategies in real-time. Traditional Susceptible-Exposed-Infected-Recovered (SEIR) models often failed to account for a defining characteristic of SARS-CoV-2—pre-symptomatic transmission. This study, published in May 2020, addresses this gap by introducing an extended SEIR model to evaluate the responses of 25 countries and simulate the impact of alternative policy timelines and clinical efficiencies.

Methodology: The vSEIdR Model

To capture the nuances of the pandemic, the researchers developed the vSEIdR (varying coefficient Susceptible-Exposed-Infected-Recovered) model. This framework introduced two pivotal adjustments to the standard epidemiological model:

  1. Varying Coefficients: Unlike static models, vSEIdR uses time-varying rates for infection, recovery, and death. This allows the model to reflect real-time changes in government policy, social distancing compliance, and healthcare improvements.
  2. Exposed Infection State: Crucially, the model allows for infections to occur in both the Exposed (pre-diagnosis) and Infected (diagnosed) states. This better reflects the clinical reality that many transmissions happen before a case is officially recorded or symptoms appear.

The study utilized data from Johns Hopkins University, the WHO, and various national health ministries. Key parameters, such as the diagnosis rate ($\alpha$) and the time-varying reproduction number ($R_t$), were estimated using kernel-weighted local regression and frequentist estimation techniques.

Key Insights

The Advantage of Early Intervention

The study categorized countries into "Quick Action" (intervention within 13 days of community spread) and "Slow Action" groups. The results were stark: the Quick Action group reached the turning point ($R_t < 1$) in an average of 28.7 days, nearly 10 days faster than the Slow Action group, which took 37.8 days on average.

Higher Transmission Potential

By accounting for pre-symptomatic transmission, the model estimated a basic reproduction number ($R_0$) averaging 5.70 across the studied countries. This is significantly higher than many early pandemic estimates (which often cited 2.0–3.0), suggesting that the virus was inherently more contagious than initially believed when accounting for the "Exposed" pool.

The "Five-Day" Counterfactual

One of the study's most compelling contributions is its "what-if" analysis regarding the timing of government intervention. For the United States, the model suggests that implementing aggressive measures just five days earlier would have reduced total cases by 80% and deaths by 75%. Conversely, the UK was found to be less sensitive to intervention timing due to its lower diagnosis rate, which allowed the "Exposed" pool to amplify the epidemic for a longer duration.

Diagnosis and Removal Regimes

The efficiency of a country's healthcare infrastructure—specifically its "diagnosis rate" and "removal regime" (the process of recovery or death)—played a decisive role in the final outcome:

  • Diagnosis Rate: The US maintained a higher diagnosis rate (0.17) than the UK (0.10). A higher diagnosis rate acts as a force multiplier for interventions, bringing the infection peak earlier and reducing the total size of the outbreak.
  • Removal Regime: By comparing South Korea’s high-recovery/low-death regime to Italy’s overwhelmed system, the authors demonstrated that improving medical care and recovery processes can substantially reduce final death tolls even after the infection peak has passed.

Conclusion

The study concludes that the global trajectory of COVID-19 was highly sensitive to the timing of intervention and the speed of clinical diagnosis. If the 23 other countries studied had adopted the swift strategies seen in South Korea or China, the researchers estimate an 88% reduction in confirmed cases and an 80% reduction in deaths by April 10, 2020. These findings underscore that while social distancing is vital, the speed of the diagnostic process and the quality of medical removal regimes are equally critical in managing a pandemic's final toll.

Summary updated

Abstract

We evaluate the effectiveness of COVID-19 control strategies of 25 countries which have endured more than four weeks of community infections. With an extended SEIR model that allows infections in both the exposed and infected states, the key epidemic parameters are estimated from each country’s data, which facilitate the evaluation and cross-country comparison. It is found quicker control measures significantly reduce the average reproduction numbers and shorten the time length to infection peaks. If the swift control measures of Korea and China were implemented, average reductions of 88% in the confirmed cases and 80% in deaths would had been attained for the other 23 countries from start to April 10. Effects of earlier or delayed interventions in the US and the UK are experimented which show at least 75% (29%) less infections and deaths can be attained for the US (the UK) under a Five-Day Earlier experiment. The impacts of two removal regimes (Korea and Italy) on the total infection and death tolls on the other countries are compared with the natural forecast ones, which suggest there are still ample opportunity for countries to reduce the final death numbers by improving the removal process.
BibTeX
@misc{gu2020better,
  author = {Jia Gu and Han Yan and Yaxuan Huang and Yuru Zhu and Haoxuan Sun and Xinyu Zhang and Yuqing Wang and Yumou Qiu and Song Xi Chen},
  title = {{Better Strategies for Containing COVID-19 Epidemics–A Study of 25 Countries via an Extended SEIR Model}},
  year = {2020},
  doi = {10.1101/2020.04.27.20081232},
  howpublished = {medRxiv},
}
Figure