Published online before print September 13, 2007, doi: 10.1161/STROKEAHA.107.498006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2007;38:e143.)
© 2007 American Heart Association, Inc.
Letters to the Editor |
Response to Letter by Bereczki
Stroke Clinical Research Unit, Department of Neurology, West China Hospital, Sichuan University, Chengdu, China
Response:
We thank Dr Bereczki for his comments on and interest in our study. We agree with Dr Bereczki, for visual-type readers, Figures 2 and 3 might suggest that TCPMs improve neurological deficit and save lives. However, we rigorously discussed these figures in the corresponding text (see Discussion section of the original article).1 The reasons that we retained the Figures were: (1) we hoped to present honestly the real situations of trials in the area of TCPMs for stroke in China although the uniform positive results in these Figures were so unbelievable. In fact, these phenomena were very common in Chinese trials2,3; (2) to avoid misleading readers, we critically assessed quality of trials and explained these results cautiously in the Abstract and Discussion sections.1 We assumed that readers would read not only Figures and Tables, but also the text; (3) we clearly stated that the results of primary outcome, "death or dependency", were negative and were not conclusive regarding the effectiveness because of limited data. The positive results were only from secondary outcomes, which we thought could justify further studies but could not be used as indicators for effectiveness.
So, we think there are at least 2 important lessons from this review. First, when reading a systematic review we should pay more attention to critical appraisal for trial quality rather than just read the attractive forest plot of meta-analysis. Second, the most important thing is that the design of clinical trials needs to be improved in the future, not only in the area of TCPMs but also in other areas including western medicine in China. Therefore, we stressed that the recently developed CONSORT guideline4 should be used for reporting clinical trials. We must admit, before properly designed randomized trials prove their value, the results in Figures 2 and 3 may be beautiful traps rather than wonders!
Acknowledgments
Disclosures
None.
References
1. Wu B, Liu M, Liu H, Li W, Tan S, Zhang S, Fang Y. Meta-analysis of traditional Chinese patent medicine for ischemic stroke. Stroke. 2007; 38: 1973–1979.
2. Tang JL, Zhan SY, Ernst E. Review of randomized controlled trials of traditional Chinese medicine. BMJ. 1999; 319: 160–161.
3. Vickers A, Goyal N, Harland R, Rees R. Do certain countries produce only positive results? A systematic review of controlled trials. Control Clin Trials. 1998; 19: 159–166.[CrossRef][Medline] [Order article via Infotrieve]
4. Gagnier JJ, Boon H, Rochon P, Moher D, Barnes J, Bombardier C. Reporting randomized, controlled trials of herbal interventions: an elaborated CONSORT statement. Ann Intern Med. 2006; 144: 364–367.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||