基于临床试验注册信息调查影响我国临床试验质量的因素

目的：调查影响我国临床试验质量的主要因素。 

方法：随机选择从 2016 年 3 月 15 日至 12 月 31 日在中国临床试验注册中心注册的临床试验，对其中随机对照试验的研究计划书按照“临床随机对照试验计划书制订指南（SPIRIT）”，包括对数据管理制度、原始数据共享声明进行评估。 

结果：从该时间段注册的2205个试验中按 1∶4 的比例随机抽取了 457 个试验，其中 393 个为治疗性的随机对照试验。国家药理基地实施的上市前临床试验研究计划书质量较好，其中 4 个试验为最高质量（1.02%），21 个较好（5.34%）。129 个试验（28.23%）描述了正确的数据管理制度，包括使用纸质病例记录表和电子数据收集与管理系统（electronic data capture，EDC）；392 个试验（85.77%）声明愿共享原始数据。 结论  临床研究计划书质量低、缺乏专业的数据管理制度是我国很多临床试验的共同问题。我们相信我国大多数临床研究者有很好的学习能力和愿望，这可从多数研究者愿意共享原始数据反映出来。我们建议从三方面为我国临床研究者提供知识教育和技术支撑：(一) 推广普及 SPIRIT；(二) 提供标准化数据管理平台；(三) 推广临床试验透明化包括共享原始数据是临床研究伦理要求的理念。

Objectives: To explore potential important factors that impacts the quality of Chinese trials. 

Methods: We randomly selected clinical studies registered in the Chinese Clinical Trial Registry during March 15th, 2016 to December 31st, 2016. The randomized controlled trials protocols were retrieved to assess the quality based on the SPIRIT guideline, their data management plan and statement of sharing individual participant data were also investigated. Results 457 studies were randomly selected from 2 205 studies by a rate of 1:4. Among them, 393 were randomized controlled trials. Pre-market trials of new medicines conducted by the State Clinical Study Bases had better quality of protocols. In total, 4 protocols were rated as high quality (1.02%) and 21 as higher quality (5.34%). 129 studies in 457 (28.23%) described a correct data management system including a paper case record form and an electronic data capture. 392 studies (85.77%) stated that public sharing IPD will be available.

Conclusions: Poorly developed protocol and lack of professional data management system are common issues in some Chinese clinical studies. We feel confident that most Chinese investigators are good in learning considering that they are willing to share the IPD of their studies. Providing education and technical support focus on three technical aspects are crucial: introducing SPIRIT for developing protocol, providing standardized data management system, and introducing the concept of transparency include sharing IPD which is an essential requirement of clinical study ethics.