颈源性头痛针灸治疗性文章的文献学质量分析

目的]对颈源性头痛的中医药临床治疗性文献进行科研设计质量的评价。【方法】运用循证医学和临床流行病学评价文献质量的原则和方法全面检索的颈源性头痛中医药临床治疗性文献并进行评价，评价的项目包括对照组的设置、随机化的应用、盲法、退出病例及随访、种族、样本量估计、诊断，纳入/排除／疗效评价标准、中医证型、不良反应、统计方法的应用。并使用Jadad评分量表评价其中的RCT文献。结果】共搜索到相关文献64篇，在37篇有对照组的文献中，73．0％的文献采用了随机对照方法，但随机的质量和可信度较差；盲法很少被采用；无一篇文章提及样本含量估计及患者种族；有20．3％的文献没有明确的诊断标准；有96．9％的文献提及疗效标准，但是其中有28篇（43．8％）是自拟的疗效标准；19篇文献提出明确的统计方法，但是其中有些方法的应用欠准确。27篇RCT文献Jadad评分结果：1篇为3分，3篇为2分，22篇为1分。结论】目前颈源性头痛中医治疗性文献的质量普遍不高，试验设计和统计学方法的运用均有待提高。

Quality analysis on articles in domestic journals about Acupunctue-moxibustion treatment of Cervicogenic headache

Objective] To carry out a qualitative analysis on research design about the clinical treatment of Cervicogenic headache in domestic journals. Methods] In accordance with the principle of clinical epidemiology, the scale in the literatures from 1979 to 2012 about the clinical evaluation on the treatment of Cervicogenic headache in the domestic journals was evaluated. Application settings, randomized control group application, blind method, withdraw from the case and follow-up, racial, estimation of sample size, diagnosis/inclusion/exclusion/effect evaluation criterion, TCM syndrome, adverse reactions, statistical method. And use the Jadad scale evaluation of the RCT document.Results] The 73.0% of the literature used a randomized controlled trial, but the quality and reliability of randomdesign were poor the blind method was rarely used no any article mentioned the quantity of samples and the race of the patients. The 20.3% had no clear diagnostic criteria. The 96.9% of the literatures mentioned the therapeutic standard, but in 28 of them(43.8%) the standard was the self-worked. The 19 literatures used a clear statistical methods, but some of themmethod were less accurate. Other problems, requiring attention were the comparison of baseline between groups and the choice of indicators of the outcome, and so on. Conclusions] Generally speaking, the quality in Acupunctue-moxibustion treatment of Cervicogenic headache was not so high, and the experimental design and the statistical methods should to be enhanced.