中药治疗复发性流产随机对照试验存在的方法学问题及改进建议

Methodological issues and suggestions for improvement in randomized controlled trials of Chinese herbal medicine for recurrent miscarriage

目的:复发性流产是妇产科临床常见疾病,现代医学尚缺乏有效治疗措施。中医对复发性流产的认识和治疗有丰富的古今文献记载,但其疗效和安全性仍缺乏高质量临床研究的验证。本研究对所有已发表或未发表的中药治疗复发性流产的随机对照试验存在的方法学问题进行评价,提出相应的改进建议,为提高临床研究的证据质量提供依据。方法:从中国知网、重庆维普中文科技期刊数据库、中国生物医学文献数据库、万方数据、Pub Med和Cochrane Library6个数据库中,全面系统地检索中药治疗复发性流产的随机对照试验文献,检索截止日期为2011年12月15日。参照Cochrane协作网系统评价员手册、CONSORT声明和CONSORT for TCM声明,评价纳入的研究在设计与报告方面的方法学质量。结果:共检索获得49篇中药治疗复发性流产的随机对照试验,均在中国发表,但这些研究的方法学质量普遍低下。存在的主要问题有未报告随机方法、样本量的估算、对象招募流程;未说明是否采用盲法;对受试对象的诊断标准以及病因、年龄、流产次数、中医辨证等基线资料未报告或报告不全面;治疗措施过于复杂;对照措施选择不当;随访报告存在偏倚;结局评价采用症状改善、实验室检查指标等中间结局,少有采用终点结局(如活产率)。结论:中药治疗复发性流产的随机对照试验存在诸多方法学问题,建议严格按照临床研究的基本原则加以实施和报告,还应报告明确的诊断标准和详细的基线资料,根据研究目的制定合理的中药干预措施和对照措施,并报告终点结局。

OBJECTIVE： Recurrent miscarriage is a common disease in clinical obstetrics and gynecology. There is no curative treatment for recurrent miscarriage in conventional medicine. Traditional Chinese medicine （TCM） has been widely used in the treatment of recurrent miscarriage in China for thousands of years. However, randomized controlled trials （RCTs） with high quality evaluating the effectiveness and safety of Chinese herbal medicine （CHM） still lack. This study was conducted in order to identify methodological problems in published or unpublished RCTs of CHM in the treatment of recurrent miscarriage, and provide suggestions for TCM researchers to conduct high-quality clinical research. METHODS： Literature searching was conducted to collect published and unpublished RCTs from six electronic literature databases, including China National Knowledge Infrastructure, Chongqing VIP Chinese Scientific Journals Database, Chinese Biomedical Literature Database, Wanfang Data, PubMed, and Cochrane Library, till December 15, 2011. The Cochrane Handbook, the CONSORT Statement and the CONSORT for TCM statement were used to assess the methodological quality in the designing and reporting of included trials. RESULTS： A total of 49 RCTs were included in this critical appraisal. Most of the trials were of high risk of bias with flawed study design and poor methodological quality. The main problems included： no report on random methods or sequence concealment; no calculation of sample size; no flow chart of enrollment and outcomes; no report on blinding methods; no or incomplete report on baseline data such as diagnosis criteria and causes of miscarriage recurrent, age, times of spontaneous abortion, TCM syndrome differentiation; complex CHM treatments and inappropriate control interventions; no report on follow-up and withdrawals; inappropriate outcome measurements such as middle-outcome or laboratory test results; no report on final outcome （live-birth rate）. CONCLUSION： There are many methodological problems in RCTs of CHM for recurrent miscarriage. It is suggested that four issues should be taken into consideration for TCM research： basic principles of clinical research are needed in design and reporting of RCTs; definite diagnosis criteria and baseline data should be reported in detail; appropriate treatment and control interventions should be selected according to the assumption and objective of research; final outcome （live-birth rate） should be reported in outcome evaluation.