HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): John M. Murkin Permission Requests Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Murkin, J. M. Search for Related Content PubMed Articles by Murkin, J. M.

J Thorac Cardiovasc Surg 1996;111:900-901
© 1996 Mosby, Inc.

LETTERS TO THE EDITOR

Influence of perfusion technique and pH management strategy during coronary artery bypass surgery

Professor of Anaesthesia
University of Western Ontario
London, Ontario, Canada

Reply to the Editor:

On behalf of my coauthors I would like to thank Dr. Taggart for his insightful analysis of our paper "A Randomized Study of the Influence of Perfusion Technique and pH Management Strategy in 316 Patients Undergoing Coronary Bypass Surgery. I. Mortality and Cardiovascular Morbidity." He has identified a number of contentious issues regarding the study design, and he raises concern regarding overall morbidity and mortality rates in the nonpulsatile group.

With regard to study design, we agree with Dr. Taggart that the very real differences in composition of cardioplegic solutions, route of administration (one surgeon used blood cardioplegia administered retrogradely, n = 21; the other three administered the cardioplegic solution anterogradely), number of grafts, and duration of crossclamp between cardioplegia types (blood versus crystalloid), if controlled for, might be expected to result in differences in outcome across cardioplegia groups. It is precisely because of these many uncontrolled variables, however, that cardioplegic type was not significant in outcome (page 343: death, myocardial infarction, arrhythmia, and insertion of an intraaortic balloon pump vs cardioplegia type, univariate analysis, p = 0.11). More grafts and shorter crossclamp times were associated with use of crystalloid cardioplegia. Usage of blood cardioplegia was associated with the converse, but presumably enabled a longer ischemic time (crossclamp time) to be tolerated without increasing myocardial morbidity, thus resulting in no difference in outcome between cardioplegia groups.

It is also important to recognize, however, that surgical technique, encompassing all the variables of cardioplegia type, route of administration, number of grafts, and ischemic time, was controlled for by stratifying by surgeon (block design) the randomization of patients into pulsatile or nonpulsatile perfusion groups. As shown in Table II in the article, this powerful technique ensured equal distribution of these important parameters by the surgeon, thus enabling the crucial role of pulsatile perfusion to be unmasked.

I must disagree with the characterization of the term "double-blind" as being "totally misleading." Although the type of perfusion may have been apparent to the personnel in the operating room, the nurse specialist assigned to identify adverse outcomes was blinded to treatment group. Such blinding of observers is not irrelevant, even when absolutely objective end points such as mortality are used. Blinding becomes even more relevant when potentially contentious issues such as diagnosis of postoperative myocardial infarction are involved.

The specific concern that the decision regarding number of coronary grafts to be performed or that the need to use an intraaortic balloon pump could be influenced, either subconsciously or overtly, by knowledge of cardiopulmonary bypass perfusion type (pulsatile or nonpulsatile) is simply implausible. Before unblinding of these results, the four operating surgeons indicated no preference, either overt or covert, regarding cardiopulmonary bypass perfusion type. The claim that such knowledge did influence "the number of grafts they performed (as indeed was the case for surgeons using crystalloid cardioplegia)" is completely misleading and entirely incorrect.

The concern that mortality (5.1%) and morbidity (insertion of intraaortic balloon pump in 7%) rates in the nonpulsatile group appear "particularly excessive" in such "good-risk elective patients" appears reasonable. In fact, however, of the eight deaths in the nonpulsatile group, two were due to massive cerebrovascular accident caused by cerebral emboli and unrelated to primary myocardial dysfunction. Furthermore, of the other six patients dying, two had very low ejection fractions (<15%) before the operation (one of whom was having a reoperation) and two other patients were in coronary care before the operation, requiring hemodynamic support and anticoagulation. The subset of patients dying of myocardial ischemia was thus at much higher risk than would be predicted from looking at mean data for the group as a whole. Again, the main strength of this study is that patients were randomized and assigned to treatment groups in a blinded fashion. Moreover, there were no intraoperative crossovers.

A very real limitation of this study remains: patients were not stratified before the operation by cardiac risk into pulsatile or nonpulsatile groups. Despite this, as shown inTables I and II in the article, post-hoc analysis assured as much as possible the equivalency of risk factors and that uncontrolled variables (ischemic times, number of grafts etc.) did not differ between groups. We are thus left with the striking result that in a randomized trial, block stratified by surgeon (and surgical technique), pulsatile perfusion was associated with a significantly lower risk of morbidity and mortality than was nonpulsatile perfusion. We completely agree that these "dramatic conclusions" warrant "major reservations" and fully support the call for further objective clinical outcome data to confirm or refute these results.

This Article Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): John M. Murkin Permission Requests Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Murkin, J. M. Search for Related Content PubMed Articles by Murkin, J. M.