To comprehend the dynamics of infectious disease transmission, it is imperative to incorporate human protective behavior into models of disease spreading. While models exist for both infectious disease and behavior dynamics independently, the integration of these aspects has yet to yield a cohesive body of literature. Such an integration is crucial for gaining insights into phenomena like the rise of infodemics, the polarization of opinions regarding vaccines, and the dissemination of conspiracy theories during a pandemic. We make a threefold contribution. First, we introduce a framework to describe models coupling infectious disease and behavior dynamics, delineating four distinct update functions. Reviewing existing literature, we highlight a substantial diversity in the implementation of each update function. This variation, coupled with a dearth of model comparisons, renders the literature hardly informative for researchers seeking to develop models tailored to specific populations, infectious diseases, and forms of protection. Second, we advocate an approach to comparing models' assumptions about human behavior, the model aspect characterized by the strongest disagreement. Rather than representing the psychological complexity of decision-making, we show that "influence-response functions'' allow one to identify which model differences generate different disease dynamics and which do not, guiding both model development and empirical research testing model assumptions. Third, we propose recommendations for future modeling endeavors and empirical research aimed at selecting models of coupled infectious disease and behavior dynamics. We underscore the importance of incorporating empirical approaches from the social sciences to propel the literature forward.

中文翻译：

---

耦合传染病和行为动力学。模型假设回顾。


为了理解传染病传播的动态，必须将人类保护行为纳入疾病传播模型。虽然传染病和行为动力学的模型都独立存在，但这些方面的整合尚未产生一个有凝聚力的文献体系。这种整合对于深入了解信息流行病的兴起、关于疫苗的看法两极分化以及大流行期间阴谋论的传播等现象至关重要。我们做出三重贡献。首先，我们引入了一个框架来描述耦合传染病和行为动力学的模型，描述了四个不同的更新函数。回顾现有文献，我们强调了每个更新函数的实现存在很大差异。这种变化，再加上缺乏模型比较，使得文献对于寻求开发针对特定人群、传染病和保护形式量身定制的模型的研究人员几乎没有信息。其次，我们提倡一种方法来比较模型对人类行为的假设，模型方面的特征是最强烈的分歧。我们不是代表决策的心理复杂性，而是表明“影响-反应函数”允许人们识别哪些模型差异会产生不同的疾病动态，哪些不会，从而指导模型开发和实证研究测试模型假设。第三，我们为未来的建模工作和实证研究提出了建议，旨在选择传染病和行为动力学耦合的模型。我们强调结合社会科学的实证方法来推动文献向前发展的重要性。