Objective: To construct a predictive model to assess the risk of severe respiratory syncytial virus infection among children under five years in China, conduct preliminary validation of this model by using external data, and develop an individual risk assessment tool available for their parents. Methods: The admission after RSV infection was used as a marker of severe infection. Based on the evidence of RSV hospitalization-related risk factors and real-world data, such as the prevalence of various risk factors in children under five years old in China, a Monte Carlo-based individual RSV hospitalization risk prediction model for children under five years old was constructed. Taking Suzhou City as an example, the model was externally validated, and an interactive risk prediction tool (RSV HeaRT) was developed on the WeChat mini-program platform. Results: The estimation model showed that in children under five years old in China if the population did not have any risk factors for severe RSV infection, the RSV annual hospitalization rate was 2.2/1 000 (95%CI: 0.9/1 000-7.5/1 000). Based on this baseline hospitalization rate and the prevalence of related risk factors in Suzhou, the model predicted an RSV hospitalization rate of 8.0/1 000 (95%CI: 4.6/1 000-24.4/1 000) for children under five years old annually in Suzhou, which was close to the reported RSV hospitalization rate in literature (10/1 000-20/1 000). In the developed RSV HeaRT WeChat mini-program, target users (such as parents of children) could input basic information, disease history, and social environmental factors of the child into the mini-program, and the tool could provide real-time feedback on the following predicted results: First, the relative risk of hospitalization due to RSV infection in current children compared to general children; Second, the probability of hospitalization due to RSV infection within the next year; Third, the relative risk of adverse outcomes during hospitalization in the event of RSV infection. Conclusion: This study is based on real-world evidence related to RSV hospitalization risk and constructs an RSV hospitalization risk prediction model suitable for Chinese children based on the combination of the current prevalence of risk factors in children under five years old in China. The accuracy of the prediction model results has been preliminarily demonstrated. Based on this design, the RSV HeaRT developed can facilitate parents to evaluate the hospitalization risk of children.
目的： 构建适用于中国五岁以下儿童的呼吸道合胞病毒（RSV）重症感染风险预测模型，应用外部数据对模型进行初步验证，开发可供儿童家长使用的个体化风险预测工具。 方法： 以RSV感染后入院作为重症感染标志，应用RSV住院相关危险因素证据和中国5岁以下儿童中各危险因素流行率等真实世界数据，构建基于蒙特卡罗方法的5岁以下儿童个体RSV住院风险预测模型，以苏州市为例，对模型进行外部验证，并在微信小程序平台开发交互式风险预测工具（RSV HeaRT）。 结果： 模型估算结果显示，在中国5岁以下儿童中，如人群不具有任何RSV重症感染风险因素，则RSV 年住院率为2.2/1 000（95%CI：0.9/1 000~7.5/1 000），基于该基线住院率，代入苏州市相关危险因素流行率，模型预测苏州市每年5岁以下儿童RSV住院率为8.0/1 000（95%CI：4.6/1 000~24.4/1 000），与文献报道的苏州RSV住院率（10/1 000~20/1 000）接近。在开发的RSV HeaRT微信小程序中，目标用户（如儿童家长）在小程序中输入儿童的基本信息、疾病史和社会环境因素等信息后，工具可实时反馈以下预测结果：与一般儿童相比，当前儿童因感染RSV而住院的相对风险；未来一年内因感染RSV而住院的概率；如发生RSV感染住院，住院期间出现不良结局的相对风险。 结论： 本研究基于RSV住院风险相关的真实世界证据，结合中国5岁以下儿童的危险因素流行现状，构建了适用于中国儿童的RSV住院风险预测模型，初步证明了该预测模型结果的准确性，基于此设计开发的RSV HeaRT可便于家长对儿童住院风险进行评估。.