| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anaesthesia, Critical Care and Pain Medicine, Royal Infirmary of Edinburgh, Edinburgh, UK
To the Editor:
The Kadoi et al. (1) study of the influence of the rate of rewarming from hypothermic cardiopulmonary bypass and diabetes mellitus on jugular bulb oxygen saturation (SjO2) and cognitive outcome from coronary artery bypass grafting has several important methodological limitations. Their measurement of SjO2 is based on fiberoptic oximetry, which is unreliable during CABG surgery (2,3) Although they may have found a strong correlation in previous studies between SjO2 measured by fiberoptic and bench oximetry, they have not confirmed the accuracy of their method in the present study. Moreover, correlation is an inappropriate way of comparing methods (4). Also, they have not corrected for arterial carbon dioxide tension as we have advocated, and such correction might have profoundly influenced the interpretation of their within- and between-patient variances in SjO2 during CABG surgery (5,6). We have argued against the use of dichotomization of cognitive data in the setting of cardiac surgery, as Kadoi et al. have done, because they are continuous variables (3). Our position is supported by a recent and extensive review of the subject by MacCallum et al., which outlines the many drawbacks to this approach and concludes that "dichotomization of quantitative measures has substantial negative consequences" (7).
References
Department of Anesthesiology and Reanimatology, Gunma University School of Medicine, Maebashi, Gunma, Japan
In Response:
I appreciate the comments of Dr. Alston. In this study, my colleagues and I chose to perform a dichotomous outcome analysis. In a recent report by researchers at Duke University, Grigore et al. (1) reported that continuous measures of analysis was more sensitive than dichotomous measures of analysis for detecting the postoperative cognitive dysfunction. However, statistical analysis does not translate to clinical importance (2). There is no acceptable agreement regarding which type of analysis is clinically more useful (3). We know that PaCO2 is one of the determinants for SjVO2 value. However, we found that PaCO2 was not a significant predictor in this study. The SjVO2 value expressed in Fig. 1 was measured with the use of a GO-oximeter and jugular venous blood. Our measurement of SjVO2 value was not based on fiberoptic oximetry. Bland-Altman analysis between oxymetry catheter values for SjVO2 and simultaneous SjVO2 values obtained from samples of jugular venous blood measured in a CO-oximeter results in a bias of 0.045%, with precision of 0.657% (Fig. 1). Therefore, we do not think that our measurements have the methodological limitations that Dr. Alston describes.
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
