Nitrous oxide and postoperative cardiovascular morbidity
Portsmouth, UK
* E-mail: danieljconnor{at}porthosp.nhs.uk
Editor—I read with interest Professor Myles' article on the effect of nitrous oxide on plasma homocysteine and folate in patients undergoing major surgery.1 I would like to query the conclusion with respect to cardiovascular morbidity in this cohort. In the patient characteristic data in Table 1, the patients are well matched until the medical conditions are considered. However, smokers, patients with coronary artery disease, patients with heart failure, and patients with previous stroke seem to be over-represented in the nitrous oxide group. There is no comment as to whether this reached statistical significance, or could have affected the cardiovascular morbidity experienced.
I look forward to the results coming through from ENIGMA-II to determine the impact of nitrous oxide on patients with pre-existing cardiovascular morbidity, though this will not affect my practice as I ceased to use N2O many years ago.
Melbourne, Australia
* E-mail: p.myles{at}alfred.org.au
Editor—We thank Dr Connor for his letter. Random allocation to groups should result in comparable baseline characteristics, but it remains possible that small differences can still exist. In our study, there were slightly more patients with some risk factors for coronary artery disease in the nitrous oxide group (as suggested by Dr Connors), yet there were slightly more patients with other risk factors in the control groups (older age, diabetes). These random differences are inevitable, and are unlikely to explain the differences in outcome that we observed.
The prime purpose of presenting baseline characteristics—typically Table 1 of a clinical trial—is to demonstrate comparable groups in all respects except that of the intervention of interest (in our case, exposure to nitrous oxide). Although some choose to compare these factors with hypotheses testing, it is wrong and potentially misleading.2 If groups are randomly allocated, then significance tests only test the success of randomization. A significance level of 0.05 merely suggests that one in 20 comparisons will be significant purely by chance. Importantly, there may be a clinically significant difference between the groups which is not detected by significance testing, yet such an imbalance may have an important effect on the outcome of interest. We, also, await the results of ENIGMA-II with interest.
References
1 Myles PS, Chan MTV, Leslie K, Peyton P, Paech M, Forbes A. Effect of nitrous oxide on plasma homocysteine and folate in patients undergoing major surgery. Br J Anaesth (2008) 100:780–6.
2 Myles PS, Gin T. Statistical Methods for Anaesthesia and Intensive Care (2000) London: Butterworth Heinemann. 126–7.
Read all E-letters
CiteULike 
Connotea 
Del.icio.us What's this?
E-letters:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||