3种肿瘤学期刊2000-2003年刊登的治疗类文献分析

目的 评价已刊登的肿瘤治疗类文献所提供的证据的正确性及其参考价值。方法 选择被引频次在我国排名前三位的肿瘤学期刊2000年1月至2003年5月刊登的化学治疗类论文。从论文类型、试验设计、例数、疗效及不良反应等结果的描述和分析等方面对文献进行评价。结果 3种杂志中刊登的治疗类论文共93篇，其中18篇(19．3％)为随机试验，37篇(39．8％)为非随机试验，38篇(40．9％)为回顾性分析；28篇(30．1％)研究或总结分析的病例数偏少(不足30例)，37篇(39．8％)仅评价远期疗效，15篇(16．1％)未描述或仅介绍部分病例的不良反应情况。56篇研究论文中，有9篇(16．4％)未对病例流失的原因进行说明。非随机临床疗效试验研究中，有54．0％(20／37)没有设置对照组。结论 国内期刊刊登的肿瘤治疗类论文尚有很大一部分研究设计不够完善，其论证强度不高；在引用或参考这些文献时，应非常慎重。

Analysis of articles on therapy published in 3 oncology journals during 2000-2003

Objective The aim of this overview is to analyze the accuracy and the clinical reference significance of the proofs provided by the published articles on cancer therapy. Methods The articles on cancer chemotherapy published from January 2000 to May 2003 in 3 Chinese oncology journals, which are 3 top journals in invitation frequency list, are chosen, and analyzed in the aspects of articles type, trials design (mainly for control setting), cases number, effectiveness result and side effects evaluation, elc. Results Of 93 articles in 3 journals, 18 (19.3 %) were on randomized trials, 37 (39.8 %) on trials, 38 (40.9 %) on retrospective study; the case number was less than 30 cases in 28 (30.1 %), and only long-term effectiveness was evaluated in 37 (39.8 %), no or not all side effects of all cases effects were described in. 15 (16.1 %). Of 56 articles on clinical trials, the reasons for lost cases were not explained in 9(16.4 %). In the nonrandomized clinical effectiveness trials, 20 (54.0 %) were not controlled trials. Conclusion The trials design is incomplete and the proof is poorly convincing in most articles on cancer chemotherapy in Chinese journals. The author suggests that attention should be paid when one is citing or referring to these articles.