In a retrospective study by Venkatesan et al.1 assessing the effects of five preoperative cardiovascular drugs on short-term mortality after coronary artery bypass grafting in 16 192 patients, they showed that statins significantly reduced perioperative mortality, but angiotensin-converting enzyme inhibitors, α-2 adrenergic agonists, calcium channel blockers, and β-blockers did not. Many aspects of this study were done correctly. It included a large sample of patients and used five multivariable logistic regression analysis models and a further Cox regression analysis model to adjust and control the effects of potential confounders on study outcomes. In a retrospective study, the multivariable logistic regression analysis is useful for adjusting the patients’ baseline characteristic and controlling selection biases. However, a limitation of multivariable logistic regression analysis is the assumption of a particular mathematical relation between intervention and measured outcome. To obtain the true inferences of multivariable logistic regression analysis for adjusted hazard ratios of measured outcomes, all of the known risk factors affecting measured outcome must be taken into account within the model. If an important risk factor is missed, the multivariate adjustment for hazard ratios of measured outcomes can be biased and a spurious association between intervention and outcome of interest may be obtained. In our view, several important issues of this study were not well addressed.
First, perioperative blood management was not provided and included in multivariate adjustments. Actually, preoperative anaemia is highly prevalent among coronary artery bypass grafting patients and can significantly increase postoperative short-term mortality risk by 3.4.2 Similarly, both haemodilution anaemia (a haematocrit of <24%) during cardiopulmonary bypass and intraoperative transfusion of as little as 1 or 2 units of red blood cells are common among coronary artery bypass grafting patients and they have been independently associated with increased short-term morbidity and mortality after on-pump coronary artery bypass grafting.3 Furthermore, postoperative anaemia is sustained for more than 50 days in 44% of patients undergoing coronary artery bypass grafting and has been significantly associated with impaired postoperative short-term outcomes. When the haemoglobin level is considered as a continuous variable, every 1 mg dl−1 decrease in haemoglobin is associated with a 13% increase in cardiovascular events and a 22% increase in all-cause mortality.4
Second, the readers were not provided with detail of coronary artery bypass grafting surgery and intraoperative managements. It has been shown that emergency coronary artery bypass grafting is an independent predictor of postoperative 30-day mortality.5 In a meta-analysis including data from a total of more than 100 000 patients from 22 studies, Takagi et al.6 concluded that off-pump coronary artery bypass grafting was associated with worse postoperative survival compared with on-pump coronary artery bypass grafting. Moreover, in patients undergoing on-pump coronary artery bypass grafting, the number of anastomoses, excessive duration of the operation per graft, nonuse of a left internal thoracic artery graft, bleeding, and aortic cross-clamp time were all shown to be risk factors of postoperative short-term mortality.7,8
Finally, the authors did not provide data on postoperative complications. In fact, surgical complications and organ injury in the early postoperative period have an impact on short-term mortality following coronary artery bypass grafting. The available literature provides compelling evidence that postoperative low cardiac output syndrome or use of intra-aortic balloon pump, arrhythmia, haemorrhage requiring re-exploration, acute myocardial and kidney injuries, pulmonary and gastrointestinal complications, dysfunction or loss of bypass grafts, sepsis, stroke and prolonged stay in the ICU are significant risk factors for short-term mortality after coronary artery bypass grafting.7–10 Especially, the influence of stroke on short-term mortality is obvious and devastating.10
Thus, we argue that in this study the nonconsideration of the above mentioned risk factors in the multivariable logistic regression model may have biased the true effect of preoperative cardiovascular drugs on short-term outcomes after coronary artery bypass grafting surgery.
Acknowledgements related to this article
Assistance with the letter: none.
Financial support and sponsorship: none.
Conflicts of interest: none.
1. Venkatesan S, Okoli GN, Mozid AM, et al. Effects of five preoperative cardiovascular drugs on mortality after coronary artery bypass surgery: a retrospective analysis of an observational study of 16192 patients. Eur J Anaesthesiol
2. Boening A, Boedeker RH, Scheibelhut C, et al. Anemia before coronary artery bypass surgery as additional risk factor increases the perioperative risk. Ann Thorac Surg
3. Paone G, Likosky DS, Brewer R, et al. Membership of the Michigan Society of Thoracic, Cardiovascular Surgeons. Transfusion of 1 and 2 units of red blood cells is associated with increased morbidity and mortality. Ann Thorac Surg
4. Westenbrink BD, Kleijn L, de Boer RA, et al. IMAGINE Investigators. Sustained postoperative anaemia is associated with an impaired outcome after coronary artery bypass graft surgery: insights from the IMAGINE trial. Heart
5. Rosato S, Biancari F, Maraschini A, et al. Identification of very high risk octogenarians undergoing coronary artery bypass surgery: results of a multicenter study. Heart Vessels
6. Takagi H, Umemoto T. All-Literature Investigation of Cardiovascular Evidence (ALICE) Group. Worse long-term survival after off-pump than on-pump coronary artery bypass grafting. J Thorac Cardiovasc Surg
7. Puddu PE, Brancaccio G, Leacche M, et al. OPerative-RISK Italian Study Group. Prediction of early and delayed postoperative deaths after coronary artery bypass surgery alone in Italy. Multivariate predictions based on Cox and logistic models and a chart based on the accelerated failure time model. Ital Heart J
8. Simchen E, Galai N, Zitser-Gurevich Y, et al. Sequential logistic models for 30 days mortality after CABG: preoperative, intra-operative and postoperative experience: the Israeli CABG study (ISCAB). Three models for early mortality after CABG. Eur J Epidemiol
9. Kamiya H, Tanzeem N, Akhyari P, et al. Impact of severe postoperative complications after cardiac surgery on mortality in patients aged over 80 years. Ann Thorac Cardiovasc Surg
10. Hedberg M, Boivie P, Engström KG. Early and delayed stroke after coronary surgery: an analysis of risk factors and the impact on short- and long-term survival. Eur J Cardiothorac Surg