Share this article on:

Failing to Reject the Null Hypothesis Does Not Mean that the Null Hypothesis Is True

McIlroy, David R. MBBS, FANZCA

doi: 10.1213/01.ANE.0000156679.52332.42
Letters to the Editor: Letters & Announcements

Department of Anaesthesia and Pain Management, Alfred Hospital, Melbourne, Australia, d.mcilroy@alfred.org.au

Back to Top | Article Outline

To the Editor:

Schwann et al.’s (1) conclusion that “avoidance of CPB does not reduce renal morbidity for patients with normal preoperative renal function or for those with mild preoperative renal insufficiency” must be questioned in light of a power analysis of their data.

They are to be commended for choosing a clinically relevant endpoint by which to measure renal dysfunction. However, serum creatinine is a crude marker that, when starting with near-normal renal function, will only begin to increase when glomerular filtration has been reduced by at least 50% (2). Consequently, the patients in this study were at relatively low risk for developing clinically detectable renal dysfunction; their observed incidence of around 8% in both groups is in keeping with this. Unfortunately, with such a low background incidence of renal dysfunction the power of the study is unacceptably low. Given the multifactorial etiology of the renal insult in cardiac surgery, it would seem reasonable to consider even a 25% reduction in the incidence of renal dysfunction with off-pump coronary artery bypass (i.e., from 8% to 6%) to be a clinically relevant outcome. However, to have an 80% power to detect such a difference in this study would have required more than 5700 patients to be enrolled. As it was, with a total of 339 patients, this study had a power of <10% to detect a 25% difference between groups in the incidence of renal dysfunction.

The possibility of a Type II error is so large that this negative result must be viewed with caution. In fact, the ethics behind the use of funding for such a trial must also be questioned when a negative result was likely. The best potential for these data now is as part of a subsequent meta-analysis.

David R. McIlroy, MBBS, FANZCA

Department of Anaesthesia and Pain Management

Alfred Hospital

Melbourne, Australia

d.mcilroy@alfred.org.au

Back to Top | Article Outline

References

1. Schwann NM, Horrow JC, Strong MD III, et al. Does off-pump coronary artery bypass reduce the incidence of clinically evident renal dysfunction after multivessel myocardial revascularization? Anesth Analg 2004;99:959–64.
2. Sladen RN. Renal physiology. In: Miller RD, ed. Anesthesia, 4th ed. New York: Churchill Livingstone, 1994:663–88
© 2005 International Anesthesia Research Society