Department of Applied Mathematics, Hong Kong Polytechnic University, Hong Kong, China.
Research Institute for Future Food, The Hong Kong Polytechnic University, Hong Kong, China.
Proc Natl Acad Sci U S A. 2023 Mar 7;120(10):e2211422120. doi: 10.1073/pnas.2211422120. Epub 2023 Feb 27.
The two nearby Amazonian cities of Iquitos and Manaus endured explosive COVID-19 epidemics and may well have suffered the world's highest infection and death rates over 2020, the first year of the pandemic. State-of-the-art epidemiological and modeling studies estimated that the populations of both cities came close to attaining herd immunity (>70% infected) at the termination of the first wave and were thus protected. This makes it difficult to explain the more deadly second wave of COVID-19 that struck again in Manaus just months later, simultaneous with the appearance of a new P.1 variant of concern, creating a catastrophe for the unprepared population. It was suggested that the second wave was driven by reinfections, but the episode has become controversial and an enigma in the history of the pandemic. We present a data-driven model of epidemic dynamics in Iquitos, which we also use to explain and model events in Manaus. By reverse engineering the multiple epidemic waves over 2 y in these two cities, the partially observed Markov process model inferred that the first wave left Manaus with a highly susceptible and vulnerable population (≈40% infected) open to invasion by P.1, in contrast to Iquitos (≈72% infected). The model reconstructed the full epidemic outbreak dynamics from mortality data by fitting a flexible time-varying reproductive number [Formula: see text] while estimating reinfection and impulsive immune evasion. The approach is currently highly relevant given the lack of tools available to assess these factors as new SARS-CoV-2 virus variants appear with different degrees of immune evasion.
附近的亚马逊城市伊基托斯和马瑙斯经历了 COVID-19 的爆发，并且在疫情的第一年（2020 年），很可能是世界上感染率和死亡率最高的地区。最先进的流行病学和建模研究估计，这两个城市的人口在第一波疫情结束时接近获得群体免疫（超过 70%的人感染），因此得到了保护。这使得很难解释为何仅仅几个月后，马瑙斯又出现了更致命的第二波 COVID-19，同时还出现了一种新的令人关注的 P.1 变体，这对毫无准备的人口造成了灾难。有人认为第二波疫情是由再感染驱动的，但这一事件在疫情历史上引起了争议和谜团。我们提出了一个伊基托斯传染病动力学的数据分析模型，我们也用它来解释和模拟马瑙斯的事件。通过对这两个城市 2 年多的多次疫情进行反向工程，部分观察到的马尔可夫过程模型推断，第一波疫情使马瑙斯留下了一个高度易感和脆弱的人群（约 40%的人感染），容易受到 P.1 的侵袭，而伊基托斯（约 72%的人感染）则不然。该模型通过拟合灵活的时变繁殖数 [Formula: see text]，同时估计再感染和脉冲免疫逃避，从死亡率数据中重建了完整的疫情爆发动力学。考虑到新出现的 SARS-CoV-2 病毒变体具有不同程度的免疫逃避，目前缺乏评估这些因素的工具，因此这种方法非常重要。
