| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2005;79:383
© 2005 The Society of Thoracic Surgeons
London Health Sciences Centre, Department of Anesthesia and Perioperative Medicine, University of Western Ontario, 339 Windermere Rd, London, ON N6A 5A5, Canada
To the Editor:
We read with interest the report by Reston and associates [1] concerning a meta-analysis of short-term and midterm results after off-pump coronary artery bypass grafting (OPCABG). However, we believe their statement that "the present study suggests that OPCABG reduces length of hospital stay, operative morbidity, and operative mortality as compared with on-pump CABG" is not convincingly supported by the information presented. Our reasons are as follows:
No baseline data on patient characteristics were given. Therefore, it is difficult to delineate comparability between the intervention and control groups, not to mention between the randomized and prospective trials. Of greatest concern is whether the two operations were truly comparable. In other words, were the numbers of grafts in the treatment groups similar within each trial as well as across the different trial designs?
Reston and associates concluded that "no statistically significant differences were observed between the results of RCTs [randomized controlled trials] and nonrandomized studies for nine of 10 outcomes evaluated." However, this is entirely to be expected because the small number of randomized trials available for comparison was unlikely to provide sufficient power to show important differences ten RCTs versus five prospective controlled studies and 38 retrospective controlled studies). As the results were not reported separately for randomized and nonrandomized trials, the reader is given no opportunity to evaluate the power and the justification for this statement (ie, whether the confidence intervals for the differences across trial designs excluded clinically important differences). Failure to show a difference does not suggest equivalence [2, 3].
Reston and coauthors identified ten RCTs, but in fact, trials referenced as 9, 14, and 17 in the bibliography are duplicates and comprise the same patient population as reference 1 (Gianni D. Angelini, personal communication, 2003). This limits the number of original randomized trails to only seven.
Given that there is no assurance that the patient groups had a similar distribution of risk factors at baseline (ie, New York Heart Association status, age, ejection fraction) and that surgical performance was similar (ie, number of grafts performed), it is difficult to conclude from the results presented by Reston and colleagues that OPCABG is superior to conventional CABG for short-term and midterm results. Furthermore, there was inadequate power to draw valid conclusions about similarities or differences across trial designs, even when duplicate trials were included.
This analysis prematurely concludes that RCTs are on par with retrospective trials despite insufficient data to substantiate this statement. This is particularly discouraging in the area of surgery where randomized trials need to be encouraged rather than demoted without sufficient proof [4].
References
Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
