Alarid-Escudero Fernando, Andrews Jason R, Goldhaber-Fiebert Jeremy D
Department of Health Policy, School of Medicine, Stanford University, Stanford, CA, USA.
Center for Health Policy, Freeman Spogli Institute, Stanford University, Stanford, CA, USA.
Med Decis Making. 2024 Jan;44(1):5-17. doi: 10.1177/0272989X231205565. Epub 2023 Nov 13.
Compartmental infectious disease (ID) models are often used to evaluate nonpharmaceutical interventions (NPIs) and vaccines. Such models rarely separate within-household and community transmission, potentially introducing biases in situations in which multiple transmission routes exist. We formulated an approach that incorporates household structure into ID models, extending the work of House and Keeling.
We developed a multicompartment susceptible-exposed-infectious-recovered-susceptible-vaccinated (MC-SEIRSV) modeling framework, allowing nonexponentially distributed duration in exposed and infectious compartments, that tracks within-household and community transmission. We simulated epidemics that varied by community and household transmission rates, waning immunity rate, household size (3 or 5 members), and numbers of exposed and infectious compartments (1-3 each). We calibrated otherwise identical models without household structure to the early phase of each parameter combination's epidemic curve. We compared each model pair in terms of epidemic forecasts and predicted NPI and vaccine impacts on the timing and magnitude of the epidemic peak and its total size. Meta-analytic regressions characterized the relationship between household structure inclusion and the size and direction of biases.
Otherwise similar models with and without household structure produced equivalent early epidemic curves. However, forecasts from models without household structure were biased. Without intervention, they were upward biased on peak size and total epidemic size, with biases also depending on the number of exposed and infectious compartments. Model-estimated NPI effects of a 60% reduction in community contacts on peak time and size were systematically overestimated without household structure. Biases were smaller with a 20% reduction NPI. Because vaccination affected both community and household transmission, their biases were smaller.
ID models without household structure can produce biased outcomes in settings in which within-household and community transmission differ.
Infectious disease models rarely separate household transmission from community transmission. The pace of household transmission may differ from community transmission, depends on household size, and can accelerate epidemic growth.Many infectious disease models assume exponential duration distributions for infected states. However, the duration of most infections is not exponentially distributed, and distributional choice alters modeled epidemic dynamics and intervention effectiveness.We propose a mathematical framework for household and community transmission that allows for nonexponential duration times and a suite of interventions and quantified the effect of accounting for household transmission by varying household size and duration distributions of infected states on modeled epidemic dynamics.Failure to include household structure induces biases in the modeled overall course of an epidemic and the effects of interventions delivered differentially in community settings. Epidemic dynamics are faster and more intense in populations with larger household sizes and for diseases with nonexponentially distributed infectious durations. Modelers should consider explicitly incorporating household structure to quantify the effects of non-pharmaceutical interventions (e.g., shelter-in-place).
分区传染病（ID）模型常用于评估非药物干预措施（NPIs）和疫苗。此类模型很少区分家庭内部传播和社区传播，在存在多种传播途径的情况下可能会引入偏差。我们制定了一种将家庭结构纳入ID模型的方法，扩展了豪斯和基林的工作。
我们开发了一个多分区易感-暴露-感染-康复-易感-接种（MC-SEIRSV）建模框架，允许暴露和感染分区的持续时间呈非指数分布，该框架可追踪家庭内部和社区传播。我们模拟了因社区和家庭传播率、免疫力下降率、家庭规模（3人或5人）以及暴露和感染分区数量（各1 - 3个）而有所不同的疫情。我们将没有家庭结构的其他相同模型校准到每个参数组合的疫情曲线早期阶段。我们比较了每对模型在疫情预测以及预测的NPIs和疫苗对疫情峰值的时间和幅度及其总体规模的影响方面的差异。荟萃分析回归表征了纳入家庭结构与偏差大小和方向之间的关系。
其他方面相似但有无家庭结构的模型产生了等效的早期疫情曲线。然而，没有家庭结构的模型预测存在偏差。在没有干预的情况下，它们在峰值规模和疫情总体规模上存在向上偏差，偏差还取决于暴露和感染分区的数量。在没有家庭结构的情况下，模型估计的社区接触减少60%对峰值时间和规模的NPIs效应被系统地高估。NPIs减少20%时偏差较小。由于疫苗接种影响社区和家庭传播，其偏差较小。
在家庭内部传播和社区传播不同的情况下，没有家庭结构的ID模型可能会产生有偏差的结果。
传染病模型很少将家庭传播与社区传播区分开来。家庭传播的速度可能与社区传播不同，取决于家庭规模，并且可以加速疫情增长。许多传染病模型假设感染状态的持续时间呈指数分布。然而，大多数感染的持续时间并非呈指数分布，分布选择会改变模拟的疫情动态和干预效果。我们提出了一个用于家庭和社区传播的数学框架，该框架允许持续时间呈非指数分布，并包含一系列干预措施，并通过改变家庭规模和感染状态的持续时间分布来量化考虑家庭传播对模拟疫情动态的影响。未纳入家庭结构会在模拟的疫情总体过程以及在社区环境中差异实施的干预措施的效果方面产生偏差。在家庭规模较大的人群以及感染持续时间呈非指数分布的疾病中，疫情动态更快且更强烈。建模者应考虑明确纳入家庭结构，以量化非药物干预措施（例如就地避难）的效果。
