基于神经网络模型的我国儿科用药上市后安全性再评价模型与风险控制机制研究

There is a lack of enough clinical trials before the listing of pediatric medication because of the particularity of pediatric patients，which leads to a general phenomenon about experiential drug use and high risk in hospitals after the listing. The adverse reaction reports after drug marketing belong to passive monitoring，which are unable to calculate the incidence of adverse drug reactions, so it’s difficult to measure the actual risk of pediatric medication. This study will take the initiative to monitor the usage of pediatric representative drugs in the medical institutions, for getting the safety re-evaluation index, including the basic data and the medication situation of pediatric patients, the occurrence of adverse reactions, the laboratory indicators before and after treatment and so on. This study will construct the safety reevaluation model of pediatric medication after listing in China by using neural network model, to predict the adverse reactions of patients and find out the potential risk points of pediatric drugs. Based on Natural Science Foundation of China “Research of risk-control mechanism of high-risk medication in hospitals in China” in which we have studied the risk-control mechanism of high-risk medication in hospitals, with the early research results, the risk-control mechanism based on the safety reevaluation of pediatric drugs will be established by using probabilistic risk assessment（PRA）. Corresponding risk control measures are taken to reduce the risk of pediatric medicine and promote the safety of pediatric medicine according to different risk levels， improve the safety management system of pediatric medication, and support a powerful umbrella for the healthy growth of children in China.

由于儿童患者的特殊性，儿科用药上市前缺乏足够的临床试验研究，导致上市后医院经验用药现象普遍，风险高发。药品上市后的不良反应报告属于被动监测，无法计算发生率，因此当前儿科用药的实际风险难以评估。本课题拟通过对儿科代表性药物在医疗机构使用情况进行主动监测，获取患者的基本资料、用药情况、不良反应发生情况、治疗前/后的实验室指标等安全性再评价指标；利用神经网络模型构建我国儿科用药上市后安全性再评价模型，对患者不良反应情况进行预测，发现儿科用药潜在的风险点。在此研究基础上，结合课题组前期完成的风险控制机制研究，利用风险概率评估技术（PRA）构建基于儿科用药安全性再评价的风险控制机制，针对不同的风险等级采取相对应的风险控制措施，来降低儿科用药风险，促进儿科用药安全，完善儿科用药安全管理体系，为我国儿童健康成长撑起一把强大的保护伞。

由于儿童患者的特殊性，儿科用药上市前缺乏足够的临床试验研究，导致上市后医院经验用药现象普遍，风险高发。药品上市后的不良反应报告属于被动监测，无法计算发生率，因此当前儿科用药的实际风险难以评估。本课题拟通过对儿科代表性药物在医疗机构使用情况进行主动监测，获取患者的基本资料、用药情况、不良反应发生情况、治疗前/后的实验室指标等安全性再评价指标；利用神经网络模型构建我国儿科用药上市后安全性再评价模型，对患者不良反应情况进行预测，发现儿科用药潜在的风险点。在此研究基础上，结合课题组前期完成的风险控制机制研究，利用风险概率评估技术（PRA）构建基于儿科用药安全性再评价的风险控制机制，针对不同的风险等级采取相对应的风险控制措施，来降低儿科用药风险，促进儿科用药安全，完善儿科用药安全管理体系，为我国儿童健康成长撑起一把强大的保护伞。