| Objective: The blood transfusion rates are varied widely,more than 10 times,from 7.8% to 92.8% in adult patients undergoing cardiac surgery with cardiopulmonary bypass(CPB).Controversy of transfusion threshold contributes substantially to this variation,as the thresholds recommended by different transfusion guidelines are different.Therefore,to investigate the “optimal” transfusion threshold promotes appropriate transfusion and avoids the risk for transfusion exposure.Transfusion trigger means a critical hemoglobin(Hb)level,at which when Hb level lower than that,a transfusion needs to initiate.Therefore,the aim of transfusion is to increase hemoglobin level to avoid anemia related injury.The nadir Hb level may be more related to anemia injury,through which we try to determine the “optimal” transfusion threshold.Additionally,there are two factors closely related to transfusion trigger: 1)patients’ condition of disease.The degree of anemia that patients could tolerant may vary according to their different compensatory ability.2)clinicians’ subjective transfusion decision.As the practitioners of clinical transfusion,the transfusion decision made by clinicians will also influence the overall effect of transfusion practice,and the two factors influence on each other.In this study,on the basis of “Perioperative information of adult cardiac surgery under CPB” database,the risk factors of blood transfusion,clinical decision of blood transfusion in cardiac surgery and what factors influenced the decision are investigated.Finally,the "optimal" blood transfusion threshold is further explored and provides evidences for future studies.Materials and Methods:This study included three parts as follows:(1)The first part was a retrospective study.Collected perioperative data of patients undergoing cardiac surgery with CPB at West China Hospital of Sichuan University from 2011.1.1 to 2017.6.30 and the Second Affiliated Hospital in the School of Medicine of Zhejiang University from 2013.9.1 to 2017.6.30 retrospectively.Especially Hb level and blood transfusion in different time points.A database of “Perioperative information of adult cardiac surgery under CPB” through Epi Data 3.1 was established,and the characters of patients were analyzed.After excluding patients died on table and with missing transfusion information,patients were grouped according to receiving transfusion or not during perioperative period,the risk factors of transfusion and the relationship between transfusion and in-hospital mortality,infection and composite adverse event(including new on-set myocardial infarction,stroke and acute kidney injury)after surgery were analyzed by uni-and multivariate logistic regression.(2)The second part was cross-sectional survey.In the view of the results in “part 1”and the current transfusion guideline in cardiac surgery,a questionnaire of “transfusion threshold and related information in adult cardiac surgery” is designed and the online survey was completed by anesthesiologists,cardiac surgeons and critical care physicians.The clinicians’ choice of transfusion threshold in cardiac surgery were analyzed.We divided clinicians into three groups: picking lower threshold than guideline group(chose threshold lower than Hb<8.0g/d L as guideline recommended),guideline threshold group(chose threshold of Hb<8.0g/d L as guideline recommended),and higher threshold than guideline group(chose threshold higher than Hb<8.0g/d L as guideline recommended),and analyzed the factors that influence the recognition of appropriate blood transfusion and further investigated the implementation of transfusion guideline in clinicians including the reasons.(3)The third part was a retrospective study.Based on this database used in “Part 1”,patients with valve surgery or coronary artery bypass graft(CABG)or combination were included andpatients were excluded if the following scenarios is present: emergency operation,died on table,having difficulty in wean from CPB,combination with aortic surgery,missed postoperative Hb value,having blood disease or coagulopathy.Then patients were grouped according to postoperative nadir Hb level,and the relative risk(RR)of composite adverse event and postoperative in-hospital death were evaluated by uni-and multivariate logistic regression and propensity score matching(PSM)analysis,and to determine the critical Hb values that is associated with lowest risk of adverse outcomes.According to this critical Hb value,patients were further divided into two groups,either lower than the critical Hb group or higher than the critical Hb group;the odds ratio of composite adverse event and death were calculated.Subgroup analysis was performed on the basis of “Part 1”and “Part 2” results to assess the “optimal” transfusion trigger in different clinical conditions.Results:(1)Totally,8521 patients with cardiac surgery under CPB were enrolled in this database,of which,85.6% was valvular surgery.After excluding 10 patients underwent emergency surgery,19 patients died on table and 32 patients missed transfusion data,8460 patients were included into subsequent analyses.The perioperative transfusion rate was 47.6% and the risk factors in perioperative blood transfusion as follows: elderly(for every 10 years older,the a OR increased to 1.04 times(95%CI 1.03-1.05);the a OR in female patients increased to 2.35 times(95%CI 2.05-2.70),compared to male patients;lower BMI(every 1kg/M2,a OR was 1.09(95%CI 1.08-1.12)),increased in the grade of pulmonary hypertension(a OR was 1.14(95%CI 1.06-1.22)in each increase),patients with hyperlipidemia(a OR increased to 1.17 times(95%CI 1.01-1.34)),patient with myocardial infarction(a OR increased to 2.13 times(95%CI 1.42-3.21));Hb decreased in perioperative period(for every 1g/d L decreased in Hb level in preoperative,during operation,and postoperative period,the a OR increased to 1.59-(95%CI 1.53-1.67),1.17-(95%CI 1.08-1.27)and 1.76-(95%CI 1.08-1.27)times,respectively);The duration of CPB(for every 10 min increased in CPB time,the a OR increased to 1.01 times(95%CI 1.00-1.01)),and the increased in postoperative drainage(every 100 ml increased,a OR increased to 1.00 times(95%CI 1.00-1.01)).Compared to patients without transfusion RBCs(Red blood cells),when RBCs≤3U,the odds ratio of death,infection and composite adverse event were 3.76(95%CI 1.59-8.91),1.88(95% CI 1.14-3.11),and 1.42(95%CI 1.19-1.69);as transfusion RBCs 3-6U,the odds ratio of death,infection and composite adverse event were 3.85(95%CI 1.54-9.63),2.11(95% CI 1.24-3.59),and 2.37(95%CI 1.98-2.85).When transfusion RBCs>6U,the odds ratio of death,infection and composite adverse event were 41.85(95%CI19.32-90.67),6.08(95% CI 3.55-10.41),and 5.16(95%CI 4.07-6.55).(2)In the second part of cross-sectional survey,a total of 1755 valid questionnaires were collected.The distribution of recognized appropriate transfusion threshold varied widely as follows: Hb<12.0g/d L(0.2%),Hb<11.0g/d L(0.5%),Hb<10.0g/d L(17.5%),Hb<9.0g/d L(16.8%),Hb<8.5g/d L(2.8%),Hb<8.0g/d L(34.8%),Hb<7.5g/d L(4.7%),Hb<7.0g/d L(16.0%)and Hb<6.0g/d L(6.8%).Among them,481 clinicians chose a threshold lower than guideline recommended,611 chose guideline recommended threshold and 663 chose a threshold higher than guideline recommended.The different professional experience of clinicians(as “the levels of care”,“working years” et al)or the choice of transfusion related information(including “discretion of transfusion threshold of cardiac and non-cardiac surgery”,“the awareness of risk of Hb decrease” and “the risk of transfusion”)did not influence the recognition of appropriate blood transfusion(P>0.05).However,compared with clinicians who self-identified as “knowing transfusion guideline well”,clinicians who self-identified as “knowing little about transfusion guideline” were more prone to choose a higher transfusion trigger than guideline recommended(OR1.51,95%CI 1.14-1.98).Among1755 clinicians,only 23.2%(408)clinicians chose to “follow transfusion guideline”,42.5%(746)chose to “not follow” and 34.2%(601)chose “unclear”.Clinicians who self-identified “know transfusion guideline” preferred “not to perform guideline” when compared to the ones who self-identified “know little about guideline”(OR 1.42,95%CI 1.08-1.86).The main reason for “not follow guideline” was subjective(57.3%),including “special clinical condition of patients”,“disagreement on the transfusion threshold between clinicians” and “disagreement on guideline”,the other reason was objective(42.7%),as “the medical cost” and “lack of blood products”.Among the clinicians who chose “not follow the guideline”,clinicians who chose “subjective reasons” preferred to higher transfusion threshold than the clinician who chose “objective reasons”(OR 1.90,95%CI 1.02-3.51).(3)In the third part of retrospective cohort study,8206 patients were finally analyzed.The incidence of composite adverse event was 19.8%(1628)and the mortality was 1.3%(109).Patients with nadir Hb of 9.0-9.9 g/d L showed the lowest incidence of composite adverse event(12.3%).Compared with nadir Hb at 9.0-9.9 g/d L,the relative risk of composite adverse event increased stepwise as nadir Hb fell below 9.0 g/d L: for 8.5-8.9 g/d L,a RR was 1.44(95%CI 1.14-1.83),PSMRR was 1.30(95%CI 1.00-1.69);for 8.0-8.4 g/d L,a RR was 1.56(95%CI 1.23-1.99),PSMRR was 1.38(95%CI 1.06-1.79);for 7.5-7.9 g/d L,a RR was 1.66(95%CI 1.31-2.11),PSM RR was 1.55(95%CI 1.17-2.06);for 7.0-7.4 g/d L,a RR was 2.22(95%CI 1.75-2.83),PSMRR was 1.73(95%CI 1.34-2.24);for <7.0 g/d L,a RR was 4.00(95%CI 3.18-5.04),PSMRR was 3.37(95%CI 2.59-4.37).When the nadir Hb above 9.0-9.9 g/d L,the relative risk of composite adverse event was as follows: for 10.0-10.9 g/d L,a RR was 0.91(95%CI 0.65-1.27),PSMRR was 0.92(95%CI 0.62-1.37);for 11.0-11.9 g/d L,a RR was 1.03(95%CI 0.61-1.66),PSMRR was 1.14(95%CI 0.57-2.28);for ≥12.0 g/d L,a RR was 1.69(95%CI 0.61-3.99),PSMRR was 3.22(95%CI 0.77-13.50).As postoperative nadir Hb decreased from ≥ 10.0 g/d L to <7.0 g/d L,the mortality was increased from 0.0% to 4.1%.Compared with nadir Hb at 9.0-9.9 g/d L,the relative risk of mortality when nadir Hb fell below 9.0 g/d L was as follows: for 8.5-8.9 g/d L,a RR was1.48(95%CI 0.52-4.81);for 8.0-8.4 g/d L,a RR was 1.28(95%CI 0.38-4.50);for 7.5-7.9 g/d L,a RR was 1.89(95%CI 0.68-6.05);for 7.0-7.4 g/d L,a RR was 2.84(95%CI 1.09-8.86);for <7.0 g/d L,a RR was 5.36(95%CI 2.20-16.12).Compared to nadir Hb level ≥9.0g/d L,when nadir Hb level <9.0g/d L,the relative risks of composite adverse event and mortality were 1.92(95%CI 1.64-2.24)and 3.78 times(95%CI 1.51-9.45),respectively.Further subgroup analyses showed: no matter the age,sex,comorbities as diabetes mellitus,hyperlipidemia,preoperative anemia or not or left ventricular function,pulmonary hypertension,renal function,valve surgery or CABG or combination,CPB duration less or more than 2h,when nadir Hb <9.0 g/d L,the risk of composite adverse event was higher than nadir Hb≥9.0g/d L.Especially in patient with age>70 years,or with diabetes mellitus,hyperlipidemia,preoperative anemia,renal function injury,CABG alone or combine with valve surgery,CPB duration>2h,the risk of low nadir Hb was increased compared to patients with age≤70 years,or not with diabetes mellitus,hyperlipidemia,preoperative anemia,or with normal renal function,valve surgery,CPB duration ≤2h,the a RR was 2.03 vs.1.58,2.06 vs.1.85,2.08 vs.1.78,3.03 vs.1.66,2.16 vs.1.66,2.37 vs.1.78,2.16 vs.1.91 and 1.68 vs.1.50,respectively.Conclusion:On the basis of established a database of “perioperative information of adult cardiac surgery under CPB”,we found that perioperative blood transfusion rate was high and dose-dependently increased the risk of postoperative in-hospital mortality,infection and composite adverse event.The reasons were not only related to perioperative transfusion risk factors as age,sex,BMI,and co-morbidities et al,but also related to the transfusion decision of clinicians,including the choice of threshold and not follow the guideline.We found that compared to nadir Hb 9.0-9.9 g/d L,the relative risk of composite adverse event increased when nadir Hb<9.0 g/d L,and the relative risk of mortality increased when nadir Hb<7.5 g/d L in valve surgery or CABG or combination.The risk of composite adverse event and mortality increased in nadir Hb<9.0g/d L when compared to nadir Hb ≥9.0g/d L.Therefore,the “optimal” transfusion threshold may be set at Hb<9.0g/d L in such operations. |
PDF Full Text Request