Медицинские услуги
© 2017-2021 Сайт «БЕЛМЕД.БАЙ», info@belmed.by
Robert H. Habib, Anoar Zacharias, Thomas A. Schwann, Christopher J. Riordan
Eur J Cardiothorac Surg 2005;28:512-513
We read with care and interest the recent article by Kuduvalli and colleagues [1] which investigated the role of perioperative transfusions on coronary artery bypass surgery (CABG) outcomes. We largely agree with the authors' conclusion that perioperative transfusions are associated with worse operative and 1-year mortality. Indeed, it conforms with the findings reported by our group in 2002 [2]. Yet, study design limitations (partly recognized by the authors) have lead them to underestimate the true adverse effects of excessive hemodilution on cardiopulmonary bypass (CPB). This was exacerbated by the authors' apparent unfamiliarity with findings of another report [3] that were not discussed in the article. There, in a systematic analysis of 3800 consecutive CABG patients with CPB and after adjusting for intraoperative and post-operative RBC transfusions, an independent association between increasing CPB hemodilutional anemia levels and worse operative outcomes and 6-year mortality were documented [3].
We wish to make (pose) the following comments (questions) to the authors that are primarily related to their handling of the two primary independent variables (hemoglobin levels and RBC transfusions):
(1) It is unclear why the authors chose the lowest laboratory hemoglobin (LL Hb; gdl–1) after arrival in the intensive care unit as opposed to the lowest Hb including during surgery. This is especially puzzling given their handling of the RBC transfusion data, which included units provided during surgery. Such inconsistency is a source of substantial confounding. The authors should elucidate why the intraoperative Hb was ignored while intraoperative transfusions were not. It is also noteworthy that their propensity model did not distinguish between transfusions during or after surgery. Why not?
(2) The pre- and intra-operative Hb values were not provided—nor were the CPB and ischemic times. Importantly, the reported median (interquartiles) LL Hb value of 14.1 (13.0–15.0)gdl–1 for the No RBC Transfusion cohort (Table 1) is surprisingly high and suggests phenomenally high preoperative Hb levels. Even the 30% off-pump patients cannot explain such values. In fact, such LL Hb values are higher than what is reported for preoperative Hb in many published CABG series. Please explain these data as they are critical to the interpretation of the study as a whole.
(4) The number of patients in each of the patient subgroups (A–D) in Figs. 4 and 5—which are based on the four possible combinations of RBC transfusion (Yes/No) and LL Hb >10gdl–1 (Yes/No)—and the corresponding P-values for these comparisons were not provided in the paper. Taking into account the above points, we contend that (a) the apparent lack of significance is a consequence of the number of patients in some of the groups, and (b) the systematically worse survival trends in the LL Hb <10gdl–1 (irrespective of transfusion status) is indicative of the importance of CPB hemodilutional anemia on outcomes.
In summary, and given the importance of the topic, we hope the authors would consider reporting to the readers their outcomes analysis results after incorporating the above outlined study design changes. We suggest that in doing so they might find significant independent adverse effects of CPB hemodilutional anemia and RBC transfusions on outcomes. Such findings would advance the position that changes to current CPB practice guidelines are needed such that both ‘RBC transfusions’ as well as their primary cause ‘CPB hemodilutional anemia’ are avoided.
References
1. Kuduvalli M, Oo AY, Newall N, Grayson AD, Jackson M, Desmond MJ, Fabri BM, Rashid A. Effect of peri-operative red blood cell transfusion on 30-day and 1-year mortality following coronary artery bypass surgery. Eur J Cardiothorac Surg 2005;27(4):592-598.[Abstract/Free Full Text]
2. Engoren M, Habib RH, Zacharias A, Schwann TA, Riordan CJ, Durham SJ. Effect of blood transfusion on long-term survival after cardiac surgery. Ann Thorac Surg 2002;74:1180-1186.[Abstract/Free Full Text]
3. Habib RH, Zacharias A, Schwann TA, Riordan CJ, Durham SJ, Shah A. Adverse Effects of low hematocrit during adult cardiopulmonary bypass: should current practice be changed?. J Thorac Cardiovasc Surg 2003;125(6):1438-1450.[Abstract/Free Full Text]
LETTER TO THE EDITOR
REPLY TO HABIB ET AL.
Manoj Kuduvalli, Antony D. Grayson , Michael J. Desmond , Brian M. Fabri
Eur J Cardiothorac Surg 2005;28:513-514
We thank Dr Habib and his colleagues for their interest in our paper [1]. They have highlighted a few issues with the paper, and we would like to take this opportunity to clarify them in the same order the questions have been raised.
1. We did not have the data for the intra-operative haemoglobin values. Hence, they were not used. This limitation has been discussed at length on page 596 of the paper. We have also put forward our rationale for using LL Hb, as this value was still the strongest predictor of the need for transfusion. Our records of transfusion are for 24-h periods, hence we were not in a position to distinguish between intra-operative and postoperative transfusions for the day of the operation. The issue of haemoglobin levels and triggers for RBC transfusions is a difficult one to deal with, and is probably the reason why Engoren and colleagues elected altogether to ignore data regarding haemoglobin levels in their paper [2].
2. The median haemoglobin values in the non-transfused cohort were admittedly higher than expected. One probable reason is that a significant proportion (30%) of patients were done off-pump using cell salvage in theatre and postoperatively. Secondly, the residual pump blood in all the patients done on bypass in our centre is routinely processed through the cell saver before transfusion, thus significantly reducing the effects of haemodilution.
3. Inclusion of patients transfused only after postoperative day 3 in the ‘No transfusion’ group is arguable. One of the effects of no transfusion in the first 72h is that it may be made inevitable thereafter. Delaying transfusion that was subsequently needed might have led to complications that affected outcome adversely. We therefore felt that including these patients in the transfused group might have biased the results in favour of the ‘No transfusion’ group. We have, however, rerun the analysis including these 28 patients in the ‘Transfusion’ group and also excluding them from the study entirely and the overall conclusions of our study remain unchanged.
4. After excluding 30-day mortality, the number of patients transfused with a LL Hb>10gmdL–1 was 520 compared to 380 patients transfused with LL Hb<10gmdL–1. Habib and colleagues are correct to point out that the apparent lack of significance in survival between these two groups is more likely a consequence of sample size, with the difference in freedom from death approaching significance in these two groups (96.7 versus 95.4%; P=0.082). We take note of their previously published paper [3] that deals with the importance of CPB haemodilutional anaemia on outcomes. We entirely agree that it is best to avoid excessive haemodilution on bypass, but what is excessive? We do not have the data to comment on that. Habib and colleagues were right in their paper [3] to suggest that a causal link between increasing haemodilution and morbidity/mortality after bypass can be demonstrated only by a randomised controlled trial.
References
1. Habib RH, Zacharias A, Schwann TA, Riordan CJ. The independent effects of cardiopulmonary bypass haemodilutional anaemia and transfusions on CABG outcomes. Eur J Cardiothorac Surg 2005;28:512-513.[Free Full Text]
2. Engoren M, Habib RH, Zacharias A, Schwann TA, Riordan CJ, Durham SJ. Effect of blood transfusion on long-term survival after cardiac surgery. Ann Thorac Surg 2002;74:1180-1186.[Abstract/Free Full Text]
3. Habib RH, Zacharias A, Schwann TA, Riordan CJ, Durham SJ, Shah A. Adverse effects of low haematocrit during adult cardiopulmonary bypass: should current practice be changed?. J Thorac Cardiovasc Surg 2003;125(6):1438-1450.[Abstract/Free Full Text]