Year : 2008  |  Volume : 11  |  Issue : 2  |  Page : 137--138

Authors' reply


Bharathi H Scott 
 Department of Anesthesiology, SUNY at Stony Brook, Health Sciences Center, L4-060, Stony Brook, NY 11794-8480, USA

Correspondence Address:
Bharathi H Scott
Department of Anesthesiology, SUNY at Stony Brook, Health Sciences Center, L4-060, Stony Brook, NY 11794-8480
USA




How to cite this article:
Scott BH. Authors' reply.Ann Card Anaesth 2008;11:137-138


How to cite this URL:
Scott BH. Authors' reply. Ann Card Anaesth [serial online] 2008 [cited 2022 May 20 ];11:137-138
Available from: https://www.annals.in/text.asp?2008/11/2/137/41592


Full Text

The Editor,

We thank the authors [1] for their interest in our paper. [2] There is a lot of uncertainty as to the level of haematocrit at which blood transfusion should be triggered during CPB. Only thing that is clear is that while on CPB, the decision to transfuse should be based on the patient's clinical situation and risk of organ ischaemia and potential for, or presence of, active bleeding. We want to emphasise that transfusion trigger in patients undergoing cardiac surgery and CPB remains contentious, and the impact of transfusion on outcomes has shown conflicting and confusing results. We believe that the lowest haematocrit that can be safely tolerated in patients with coronary disease is not clearly known. We did not delve into further discussion on causes of higher transfusion rates as our goal was to examine the impact of transfusion on resource utilisation, morbidity, and mortality. The authors of the letter state that we included only myocardial infarction, diabetes mellitus, hypertension, congestive heart failure, and renal failure in our definition of "no preoperative morbidity" and we did not include preoperative haematocrit, chronic obstructive pulmonary disease, cerebro vacular accident, and peripheral vascular disease. To the contrary, we do use all of these pre-op morbidity variables in our definition of preoperative morbidity. Although this is discussed near the end of the "Discussion" section, we do apologise for not making this explicit, upfront, in the paper. They also mention sex in this context, but sex is not a premorbid condition. It is a demographic variable that, like age and BSA, we account for in the analyses.

Next, the authors exemplify five surgical variables that are associated with resource utilisation and question our conclusions because we did not give "reference to these variables". In fact, there are 27 clinical/surgery variables measured at around the time of surgery. All are correlated to some extent with both resource utilisation and transfusion; and generally, these variables tend to be correlated with each other. Overall, there is much multicollinearity among these variables. This is not unexpected, but it creates serious complications and limitations in the analysis of predictors of resource utilisation; but here, the following observations clarify the matter and offer an explanation to the authors' concern. Of the 27 variables that are correlated with utilisation, transfusion and red blood cell count are by far the most highly correlated; and these, on a binary scale are almost interchangeable. Therefore, transfusion is the best single predictor of postoperative resource utilisation.

We appreciate these authors' insights and welcome the opportunity to clarify our results.

References

1Juneja R, Mehta Y. Authors' reply on Blood transfusion is associated with increased resource utilization, morbidity and mortality in cardiac surgery. Ann Cardiac Anaesth 2008;11:136.
2Scott BH, Seifert FC, Grimson R. Blood transfusion is associated with increased resource utilization, morbidity and mortality in cardiac surgery. Ann Card Anaesth 2008;11:15-9.