逍遥散治疗慢性疲劳综合征研究的方法学及报告质量评价

Quality Assessment of Methodology and Reporting of Clinical Trials Involving Xiaoyao San for Chronic Fatigue Syndrome

目的评价逍遥散治疗慢性疲劳综合征(CFS)随机对照试验(RCT)和病例系列研究的方法学及报告质量。方法检索Pub Med、CBM、CNKI、VIP和Wan Fang Data数据库,检索时间均从建库截至2013年11月,收集逍遥散治疗CFS的RCT和病例系列研究,分别参考Cochrane手册5.1.0版的偏倚风险评估工具和CONSORT声明2010版,并结合9项中医药特色指标评价纳入RCT的方法学和报告质量;分别参考NICE病例系列研究质量评价标准和STROBE声明第四版评价纳入病例系列研究的方法学和报告质量。结果初检出相关文献117篇,最终纳入27篇文献,包括11个RCT和16个病例系列研究。Cochrane偏倚风险工具评价结果显示:54.5%的RCT采用了较为合理的随机分配方法,仅9.1%的RCT采用了分配隐藏和盲法;72.7%的RCT采用了意向治疗(ITT)分析却不清楚是否存在失访;研究均无选择性报告。中医药特色指标补充评价结果显示:54.5%的RCT未进行中医证候诊断,研究中采用的中医证候疗效标准不统一;干预措施的质控措施质量参差不齐;没有RCT严格符合多中心研究。CONSORT声明2010版评价结果显示:文献均不能从文题识别是随机临床试验;样本含量计算、随机化实施细节、流程图的使用、结果分析类型、试验注册及方案等的报告率均为0。NICE病例系列研究质量评价结果显示:纳入病例系列研究中仅6.25%符合多中心研究;81.25%的研究纳入排除标准不明确;均未做到连续招募患者和使收集的数据达到预期目标,也均未做到分层分析结局。STROBE声明评价结果显示:纳入的病例系列研究均未在题目中明确注明是病例系列研究;研究设计、样本大小、受试者流程、偏倚、局限性、可推广性的报告率均为0。结论逍遥散治疗CFS RCT和病例系列研究的报告质量较差,研究存在不同程度的方法学问题,建议今后在设计和报告RCT和病例系列研究时分别参考CONSORT声明和STROBE声明,以提高研究质量,增强逍遥散对CFS疗效证据的强度,推进中药的现代化和国际化。

Objective To investigate the methodological and reporting quality of clinical trials involving Xiaoyao San for chronic fatigue syndrome. Methods We searched Pub Med, CBM, CNKI, VIP and Wan Fang Data to identify randomized controlled trials（RCTs） about Xiaoyao San for chronic fatigue syndrome. The methodological and reporting quality of included RCTs was respectively evaluated according to the assessment tool of risk of bias of the Cochrane Handbook 5.1.0 and the CONSORT 2010 statement, combined with complementary assessment by the characteristic indicators of traditional Chinese medicine（TCM）. The methodological and reporting quality of included case series study was respectively assessed by the methods recommended by the Britain＇s National Institute for Clinical Excellence（NICE） and the STROBE statement. Results A total of 27 clinical trials were included, involving 11 RCTs and 16 case series studies. According to the assessment tool of risk of bias of the Cochrane Handbook, 54.5% of the RCTs performed proper random method, 9.1% conducted allocation concealment and blinding, 72.7% selected intention-to-treat（ITT） analysis without the report of loss to follow-up, and no RCT existed selective reports. Corresponding to the characteristic indicators of TCM, 54.5% of the RCTs did not conduct TCM syndrome diagnosis, the curative effect standard of TCM syndrome was discrepant, and no RCT was multi-center study. The CONSORT 2010 statement indicated that no RCT explained sample size estimation, implementation details of randomization, flow diagram of participant, use of ITT and clinical trial registration. According to the items recommended by Britain＇s NICE, 6.25% of the case series studies were multi-center, 81.25% did not report clear inclusion and exclusion criteria, and no case series study performed continuous patient recruitment and stratification analysis of outcome. The STROBE statement indicated that no case series study reported research design, sample size, flow chart, bias, limitations and generalizability. Conclusion The quality of clinical trials about Xiaoyao San for chronic fatigue syndrome is still low in methodological and reporting aspects. It is suggested that the future clinical trials should be conducted with references of CONSORT statement and STROBE statement, to propel the modernization and internationalization of TCM.