- Email: [email protected]
Age of blood and survival after massive transfusion
+Model TRACLI-2900; No. of Pages 5
ARTICLE IN PRESS Disponible en ligne sur
ScienceDirect www.sciencedirect.com Transfusion Clinique et Biologique xxx (2017) xxx–xxx
Original article
Age of blood and survival after massive transfusion Importance de l’âge dans une transfusion massive C.C. Sanz ∗ , A. Pereira Transfusion Service, Hospital Clínic, Villarroel 170, 08036 Barcelona, Spain
Abstract Background. – Massive transfusion is the clinical scenario where the presumed adverse effects of stored blood are expected to be more evident because the whole patient’s blood volume is replaced by stored blood. Objective. – To analyse the association between age of transfused red blood cells (RBC) and survival in massively transfused patients. Methods. – In this retrospective study, clinical and transfusion data of all consecutive patients massively transfused between 2008 and 2014 in a large, tertiary-care hospital were electronically extracted from the Transfusion Service database and the patients’ electronic medical records. Prognostic factors for in-hospital mortality were investigated by multivariate logistic regression. Results. – A total of 689 consecutive patients were analysed (median age: 61 years; 65% males) and 272 died in-hospital. Projected mortality at 2, 30, and 90 days was 21%, 35% and 45%, respectively. The odds ratio (OR) for in-hospital mortality among patients who survived after the 2nd day increased with patient age (OR: 1.037, 95% CI: 1.021–1.054; per year P < 0.001), with the number of RBC unit transfused in the first 48 hours (OR: 1.060; 95% CI: 1.038–1.020 per unit; P < 0.001), and the percentage of such RBC stored for more than 28 days (1.010, 95% CI: 1.005–1.018 per percent point; P = 0.01). Conclusion. – Mortality after massive transfusion was associated with a higher proportion of old RBCs transfused in the first 48 hours. Other factors associated with poor prognosis were older patient’s age and larger volumes of transfused RBCs. © 2017 Elsevier Masson SAS. All rights reserved. Keywords: Blood transfusion; Red blood cells; Survival
Résumé Contexte. – La transfusion massive est le scénario clinique où les effets néfastes présumés du sang stocké devraient être plus évidents car l’ensemble du volume sanguin du patient est remplacé par du sang stocké. Objectif. – Analyser l’association entre l’âge des globules rouges transfusés (RBC) et la survie chez des patients massivement transfusés. Méthodes. – Dans cette étude rétrospective, les données cliniques et transfusionnelles de tous les patients consécutifs massivement transfusés entre 2008 et 2014 dans un grand hôpital de soins tertiaires étaient extraites électroniquement de la base de données du service Transfusion et des dossiers médicaux électroniques des patients. Les facteurs pronostiques pour la mortalité hospitalière ont été étudiés par une régression logistique multiple. Résultats. – Au total, 689 patients ont été analysés (âge médian : 61 ans, 65 % d’hommes) et 272 sont morts à l’hôpital. La mortalité prévue à 2, 30 et 90 jours était respectivement de 21 %, 35 % et 45 %. Le rapport de cote (OR) pour la mortalité hospitalière chez les patients qui ont survécu après le 2e jour augmenta avec l’âge du patient (OR : 1.037, IC 95 % : 1.021–1.054, par an p < 0,001), avec le nombre d’unités RBC transfusées dans les premières 48 heures (OR : 1,060 ; IC 95 % : 1,038–1,020 par unité ; p < 0,001) et le pourcentage de ces RBC stockés pendant plus de 28 jours (1,010, IC 95 % : 1,005–1,018 par point de pourcentage, p = 0,01). Conclusion. – La mortalité après une transfusion massive a été associée à une proportion plus élevée de vieux globules rouges transfusés au cours des premières 48 heures. D’autres facteurs associés à un mauvais pronostic ont été l’âge des patients et de plus grands volumes de globules rouges transfusés. © 2017 Elsevier Masson SAS. Tous droits r´eserv´es. Mots clés : Transfusion sanguine ; Globules rouges ; Survie ∗
Corresponding author. E-mail address: [email protected] (C.C. Sanz).
http://dx.doi.org/10.1016/j.tracli.2017.04.005 1246-7820/© 2017 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005
+Model TRACLI-2900; No. of Pages 5 2
ARTICLE IN PRESS C.C. Sanz, A. Pereira / Transfusion Clinique et Biologique xxx (2017) xxx–xxx
1. Introduction Critical haemorrhage requiring massive transfusion is a leading cause of death in trauma patients and a relevant contributor to surgery-related mortality [1]. In addition, massive transfusion accounts for a large share of all the blood components used in many trauma centres and tertiary-care hospitals [2,3]. Clinical research on the haemotherapy of critical haemorrhage has traditionally been focused on the prevention and treatment of the coagulopathy associated with trauma and massive transfusion. As a result, long established paradigms have changed in recent years, with emphasis now being placed on controlled hypotensive resuscitation and early, proactive transfusion of plasma and platelets [4]. In contrast, little attention has been paid to the attributes of transfused RBCs. Preclinical studies have shown that transfusion of stored RBCs impairs the microcirculation dynamics and the delivery of oxygen to tissues [5]. However, the clinical relevance of such pathophysiologic features remains a matter of discussion because retrospective and prospective observational studies have yielded conflicting results [5–8], and the few available clinical trials may have been unable to detect infrequent but clinically relevant effects [9–11]. Massive transfusion is a unique scenario where the presumed adverse clinical effects of stored blood are expected to be more evident. First, patients usually present in a critical condition in which any disequilibrium of the already precarious homeostasis can be life threatening. Second, the whole patient’s blood volume may end up replaced by stored blood. Nevertheless, clinical guidelines on the management of critical haemorrhage and massive transfusion do not make any recommendation about the age of RBCs [12–16]. The present study was aimed at investigating the association between age of transfused RBCs and in-hospital, all-cause mortality in a large series of massively transfused patients. Additionally, we provide a clinical description of an unselected population of massive transfusion recipients as it was seen in a large, tertiary-care hospital.
We also recorded whether the haemorrhage triggering the massive transfusion begun out-of-hospital or while the patient was already admitted in the hospital. In order to avoid selection and recall biases, all the variables were selected automatically from the electronic databases and only the cause of haemorrhage was assigned subjectively from a pre-established list after reviewing the medical records. Over the period under study, all the RBC units were stored in SAG-M and leukoreduced before storage. Platelet units consisted in a 5-donor pooled concentrate or an equivalent aphaeresis unit. All the plasma was from male donors and most of the units were photo-inactivated with methylene blood. 2.1. Statistics The study’s main outcome was in-hospital mortality (up to the 90th admission day) in patients surviving more than 48 hours after starting the massive transfusion. Patients whose length of stay was longer than 90 days were censored at that follow-up time point (see Fig. S1). Continuous variables were summarized as median and interquartile range (IQR). Statistical comparisons of variables were done by the Mann-Whitney U-test for continuous or ordered data and the Chi-square test for categorical variables. The overall survival curve was drawn by the method of Kaplan and Meier. Patient- and transfusion-related factors associated with in-hospital mortality were investigated by multivariate binary logistic regression. Factors analysed for a possible prognostic significance included sex, age, ABO/Rh blood group, whether the haemorrhage begun in-hospital or out-of-hospital, number of RBC units transfused within the first 48 hours, proportion of RBCs transfused within the first 48 hours that had been stored for more than 28 days, plasma- and platelet-to-RBC ratios, and calendar year (from 2008 to 2014). For the purpose of prognostic analysis only RBC units transfused within the first 48 hours were taken into account. All the statistical analyses were performed by using Stata, version 11 (www.satata.com). The study was approved by the Hospital Clinic’s Committee for Ethics in Research (reference: HCB/2016/0533).
2. Material and methods 3. Results For the purpose of this retrospective study, massive transfusion was defined as the infusion of 8 or more RBC units in less than 24 hours. The Transfusion Service database was queried for all consecutive adult patients (> 16 years) who met the above criterion from January 2008 to December 2014. We also retrieved the date of delivery of every unit of RBC, plasma, platelets, and cryoprecipitate received by the patients within the first 15 days after the massive transfusion as well as the storage time of transfused RBCs. In patients who had more than one episode of massive transfusion, only the first one was taken into account. Surviving patients were followed up until hospital discharge or the study closing date on July 1st, 2015. Information on the medical condition triggering the massive transfusion, the patient’s sex and age, clinical course until death or hospital discharge, and use of manufactured coagulation products was obtained from the patients’ electronic medical records.
A total of 689 patients met the inclusion criteria and constituted the subject of this study (four patients were excluded because RBCs were actually used in exchange transfusion for haemoglobinopathy or paludism). Median age was 61 (IQR: 45–71) years and 65% patients were males. The main patient characteristics at the time of the massive transfusion are shown in Table 1. As can be seen, 56% of patients were already admitted in the hospital at the time of massive bleeding and the most frequent underlying conditions were cardiac surgery (36% of cases), abdominal surgery (17%), and liver transplant (16%). In patient in whom the haemorrhage started out-of-hospital, blunt or penetrating trauma (33% of cases), gastrointestinal bleeding (21%), and ruptured aneurisms of large vessels (16%) were the most frequent conditions. Patients already in-hospital at the start of massive haemorrhage were older than those in whom the
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005
+Model
ARTICLE IN PRESS
TRACLI-2900; No. of Pages 5
C.C. Sanz, A. Pereira / Transfusion Clinique et Biologique xxx (2017) xxx–xxx
3
Table 1 Characteristics of patients undergoing massive transfusion.
Age in years, median (IQR)a Sex (n = 683)a Males Females Blood groupa A O B AB Rh positive Rh negative Underlying conditiona Haemorrhage in hospital Cardiovascular surgery Abdominal surgery Liver transplant Urologic/gynaecologic procedures Gastrointestinal bleeding Other Total Haemorrhage out-of-hospital Traumatic injuries Gastrointestinal bleeding Ruptured aortic aneurysm of great vessels Intra-abdominal bleeding Stab or firearm wounds Other Total Transfusion data, median (IQR) First 48 hours pRBC Plasma Platelets First 15 daysb pRBC Plasma Platelets a b
Whole series
Patients surviving > 48 h
61 (45–71)
61 (47–71)
451 (66%) 232 (34%)
339 (65%) 186 (35%)
305 (44%) 301 (44%) 62 (9%) 20 (3%) 566 (82%) 124 (18%)
240 (45%) 229 (43%) 48 (9%) 15 (3%) 434 (82%) 98 (18%) Fig. 1. Distribution of the length of storage in 10,989 RBC units transfused in the first 48 hours to 689 massively transfused patients.
141 (36%)
112 (34%)
64 (17%) 61 (16%) 44 (11%)
57 (17%) 58 (18%) 42 (13%)
22 (6%)
16 (5%)
52 (14%) 384 (100%)
45 (14%) 330 (100%)
98 (33%) 64 (21%)
59 (30%) 51 (26%)
48 (16%)
26 (13%)
22 (7%)
15 (8%)
15 (5%)
9 (5%)
53 (18%) 300 (100%)
40 (20%) 200 (100%)
12 (10–18) 12 (8–18) 2 (1–3)
12 (10–16) 10 (8–16) 2 (0–3)
15 (11–22) 12 (8–20) 2 (1–4)
15 (11–21) 12 (8–18) (0–4)
Not all patients have all data. Includes blood component units transfused in the first 48 hours.
haemorrhage started out-of-hospital (median [IQR] 63 [51–71] years and 55 [39–71] years, respectively; P = 0.0003). The plasma- and platelet-to-RBC ratio by the first 24 hours were 0.9 (IQR: 0.5–1.3) and 0.1 (IQR: 0.0–0.2), respectively. There was no difference in both ratios between the in-hospital and the out-of-hospital groups. The quantity of blood components transfused within the first 48 hours and by the 15th day
are summarised in Table 1. Cryoprecipitate and/or manufactured fibrinogen concentrate was used within the first 48 hours in 187 patients; prothrombin complex concentrate was used in 95 patients, and rFVIIa in 24. The length of storage of the 10,989 RBC units transfused within the first 48 hours is shown in Fig. 1. Among these units, 8568 (78.0%) were older than 14 days, 1411 (12.8%) had been stored for more than 28 days, and only 496 (4.5%) were younger than 7 days. Patients of B or AB blood group received significantly older RBCs than A/O blood group patients (Supplementary Fig. S2A; P = 0.0001). Two hundred and sixty-two patients received at least one RBC unit older than 28 days. When measured as the proportion of RBCs older than 28 days received within the first 48 hours, the figures ranged from 9.0% in group O patients to 13.0% in group A, and 27.1% in group B and AB patients (P < 0.001). Age of transfused RBCs was relatively invariable for most of the study period but increased in calendar years 2013 and 2014 as compared with previous years (Supplementary Fig. S2B). There was no correlation between the number of RBC units transfused in the firsts 48 hours and the proportion of such units that had been stored for more than 28 days (Spearman rank correlation: −0.002, P = 0.95; Supplementary Fig. S3). Fig. 2 shows the projected survival after massive transfusion for the whole series. As can be seen, estimated survival at 2, 30, and 90 days from massive transfusion was 79% (95% CI: 75%–81%), 65% (95% CI: 61%–69%), and 54% (95% CI: 49%–59%), respectively. Since early mortality is strongly determined by the cause of the critical haemorrhage and the patient’s condition at the start of massive transfusion rather than by the characteristics of the transfused blood, we investigated factors associated with in-hospital mortality in the selected population of 532 patients who survived more than two days. By the 90th follow-up day, 106 (20%) out of these 532 patients have died and 426 were censored (see follow-up flow-card in Supplementary Fig. S1). Factors independently associated with higher mortality between the 2nd and 90th postransfusion day are shown in Table 2. As can be seen, after adjustment for other potential prognostic and
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005
+Model TRACLI-2900; No. of Pages 5 4
ARTICLE IN PRESS C.C. Sanz, A. Pereira / Transfusion Clinique et Biologique xxx (2017) xxx–xxx
Fig. 2. In-hospital projected survival of 689 massively transfused patients followed up until death or hospital discharge. Table 2 Factors predicting in-hospital mortality in massively transfused patients who survived more than 2 days and adjustment covariates included in the logistic regression model. Covariate
Odds ratio (95% CI)
P
Age in years RBC units in the first 48 h Proportion of RBCs used in the first 48 h stored > 28 days (%) Male sex Ratio plasma-to-RBC (first 24 h) Ratio platelets-to-RBC (first 24 h) Haemorrhage starting out-of-hospital Blood group A/O (as opposed to B/AB) Blood group Rh+ (as opposed to Rh−) Calendar year (2008 to 2014)
1.03 (1.02–1.04) 1.07 (1.04–1.10) 1.01 (1.005–1.018)
< 0.001 < 0.001 0.01
1.12 (0.68–1.82) 1.22 (0.97–1.53) 2.99 (0.61–3.45) 1.37 (0.46–1.83) 0.92 (0.88–1.14)
0.65 0.08 0.17 0.18 0.81
1.27 (0.68–2.35)
0.44
1.00 (0.88–1.01)
0.89
confounder factors listed in the table, the odds for dying increased by 3.7% for every additional year in the patient’s age, by 6.9% for every additional RBC unit transfused in the first 48 hours, and by 1.0% for every percent point increase in the fraction of such RBC units that had been stored for more than 28 days. 4. Discussion In this study of an unselected population of massively transfused adult patients seen in a large tertiary-care, academic hospital over a 7-year period, we found that nearly half the patients were projected to have died in hospital after the massive transfusion. Among patients who survived beyond two days from the start of massive transfusion, factors independently associated with mortality were older age, a greater number of RBC units transfused in the first two days, and a large proportion of such units stored for more than 28 days. The mortality rate observed in our study was comparable to that found in other unselected populations of massively
transfused patients [1,3]. We found a high mortality rate in the first 24–48 hours, as previously reported from other series [1,3], but also that nearly half the patients who eventually died, did so days or weeks after the massive transfusion, not rarely after a long hospital stay. This observation suggests that many patients did not actually die because of haemorrhagic shock or as an immediate consequence of the condition triggering the massive transfusion, but of complications arising after the initial critical period. Despite the general agreement on the morphological and physiological changes suffered by stored RBCs, the so called “storage lesion”, and the adverse biological effects of transfusing stored blood found in preclinical studies, there is no agreement on the clinical consequences of such findings [5–8]. Observational studies, both retrospective and prospective, have produced a wide variety of results, sometime conflicting to one another (revised by Wang et al. [6] and Alexander et al. [7]). Recent clinical trials have failed to confirm that transfusion of young RBCs improves survival as compared with transfusion of older blood [9–11]. Although results from these trials can rule out a large clinical effect mediated by old blood, concern remains about the trials’ ability to detect small but clinically relevant effects [8,17], and about extrapolation of results to patient populations and clinical conditions different from those analysed in the trials. For instance, the INFORM investigators excluded massively transfused patients and the median volume transfused per patient was 2 RBC units [11]. Similarly, in the ABLE [9] and the RECESS [10] trials, patients received on average 4 RBC units over a period of several days or even weeks. This is a quite different situation from that of the massively transfused patient in whom stored blood may account for nearly all the circulating RBCs. In this regard, previous authors have noted that age of blood is associated with mortality only in patients transfused with large volumes of RBCs, while the association disappears in patients who receive smaller transfusion volumes [18]. Moreover, it is worth noting that a recent subanalysis of the ABLE trial disclosed a survival advantage in the younger RBC arm for those patients who had received more than 5–7 RBC units [19]. As it has already been pointed out by previous authors [20], our study shows that patients of B and AB blood group are more likely to end up receiving significantly older RBCs than patients of O or A blood groups. In the present study, the ABO blood group was not an independent predictor for shorter survival, perhaps because of low number of group B/AB patients and the fact that a fraction of the transfusion support they received were given with younger, blood group O RBCs. Nevertheless, this observation warrants for exploring specific RBC inventory management and transfusion policies for these patients. The metrics to measure the age of transfused RBCs has varied widely among studies (e.g. age of the oldest transfused RBC unit, average age of all transfused RBCs, number of units older than a given age). In the present study, we employed the proportion of all transfused RBCs that were older than a given threshold as the most appropriate yardstick. Indeed, in massive haemorrhage, many of the transfused RBCs are lost through the bleeding sites, so that it can be entertained that old RBC units transfused after the massive bleeding was controlled would have a higher
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005
+Model TRACLI-2900; No. of Pages 5
ARTICLE IN PRESS C.C. Sanz, A. Pereira / Transfusion Clinique et Biologique xxx (2017) xxx–xxx
potential for negative effects than those transfused earlier. This is also the reason why we took into account the volume transfused in the first 48 hours instead of a shorter period of time. The traditional definition of massive transfusion entails the replacement of the patient’s blood volume in less than 24 hours, what usually is measured by transfusing 10 or more RBC units (and the corresponding colloids, crystalloids or plasma) for the adult patient. However, it has been shown that this 10-unit criterion excludes a large proportion of patients with critical bleeding [21], so that we relied on an 8-unit threshold as a more inclusive one. The 28-day storage threshold used in this study was the result of balancing statistical power (i.e. enough patients in the old blood group) with the capability to capture the prognosis of those patients who received a large proportion of the oldest RBCs. It should not be taken, therefore, as a threshold with a definite biological meaning. In fact, the timing of the potentially adverse effects associated with RBC storage has not yet been defined, and the most discriminating storage threshold based on a biological rational remains unknown [17]. The present study has several limitations. First, the retrospective design does not allow to control for unknown confounding factors. Second, the data we used came from a single institution, so that the results might not be generalized to other institutions with different transfusion policies in massive haemorrhage. Third, clinical characteristics were aggregated into large diagnostic categories, so that granular clinical data that may have influenced on survival were not analysed. Four, the only outcome evaluated was survival, and other information of the clinical curse (e.g. infection, organ failure, etc.) was omitted from the analysis. On the positive side, it is worth mentioning that all the variables analysed were objective and extracted electronically from the databases, so that biases related to selection and interpretation of clinical data and outcomes can reasonably be ruled out. Moreover, although we did not analyse clinical variables known to influence on survival in some specific causes of massive transfusion (i.e. severity scores in trauma patients), it seems quite unlikely that such clinical variables might have confounded the prognostic effect of the age of transfused RBCs. In this regard, it is worth mentioning that length of storage was unrelated to the volume of transfused RBCs. In conclusion, this study shows that delayed mortality after massive transfusion was higher in patients who received a larger proportion of old RBCs in the first two days. Other factors associated with poor prognosis were older patient’s age and larger volumes of transfused RBCs. Authorship CS and AP designed research, collaborate in the acquisition, analysis and interpretation of data, drafted the paper and approved the final version of the manuscript. Disclosure of interest The authors declare that they have no competing interest.
5
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.tracli.2017.04.005. References [1] Halmin M, Chiesa F, Vasan SK, et al. Epidemiology of massive transfusion: a binational study from Sweden and Denmark. Crit Care Med 2016;44:468–77. [2] Como JJ, Dutton RP, Scalea TM, et al. Blood transfusion rates in the care of acute trauma. Transfusion 2004;44:809–13. [3] Dzik WS, Ziman A, Cohen C, et al. Survival after ultramassive transfusion: a review of 1360 cases. Transfusion 2016;56:558–63. [4] Sihler KC, Napolitano LM. Massive transfusion: new insights. Chest 2009;136:1654–67. [5] Flegel WA, Natanson C, Klein HG. Does prolonged storage of red blood cells cause harm? Br J Haematol 2014;165:3–16. [6] Wang S, Sun J, Solomon SB, et al. Transfusion of older stored blood and risk of death: a meta-analysis. Transfusion 2012;52:1184–95. [7] Alexander PE, Barty R, Fei Y, et al. Transfusion of fresher vs older red blood cells in hospitalized patients: a systematic review and meta-analysis. Blood 2016;127:400–10. [8] Glynn SA, Klein HG, Ness PM. The red blood cell storage lesion: the end of the beginning. Transfusion 2016;56:1462–8. [9] Lacroix J, Hébert PC, Fergusson DA, et al. Age of transfused blood in critically ill adults. N Engl J Med 2015;372:1410–8. [10] Steiner ME, Ness PM, Assmann SF, et al. Effects of red-cell storage duration on patients undergoing cardiac surgery. N Engl J Med 2015;372:1419–29. [11] Heddle NM, Cook RJ, Arnold DM, et al. Effect of short-term vs long-term blood storage on mortality after transfusion. N Engl J Med 2016;375:1937–45. [12] Stainsby D, MacLennan S, Thomas D, et al. Guidelines on the management of massive blood loss by the British Committee for standards in haematology. Br J Haematol 2006;135:634–41. [13] Gaarder C, Naess PA, Frischknecht Christensen E, et al. Scandinavian guidelines – “The massively bleeding patient”. Scand J Surg 2008;97:15–36. [14] Dzik WH, Blajchman MA, Fergusson D, et al. Clinical review: Canadian national advisory committee on blood and blood products–massive transfusion consensus conference 2011: report of the panel. Crit Care 2011;15:242. [15] Patient blood management guidelines: module 1. Critical bleeding/massive transfusion. Australian National Blood Authority; 2011 [available from URL: http://www.blood.gov.au/system/files/documents/ pbm-module-1.pdf; accessed on February 20th, 2016]. [16] Pham HP, Shaz BH. Update on massive transfusion. Br J Anaest 2013;111:i71–82. [17] Pereira A. Will clinical studies elucidate the connection between the length of storage of transfused red blood cells and clinical outcomes? An analysis based on the simulation of randomized controlled trials. Transfusion 2013;53:34–40. [18] Weinberg JA, McGwing G, Griffin RL, et al. Age of transfused blood: an independent predictor of mortality despite universal leukoreduction. J Trauma 2008;65:279–84. [19] Mack JP, Kahn SR, Tinmouth A, Fergusson D, Hébert PC, Lacroix J. Dosedependent effect of stored red blood: results of a sub-group analysis of the age of blood evaluation (ABLE) trial. 58th Congress of the American Hematology Association, San Diego (USA); 2016 [available from URL: https://ash.confex.com/ash/2016/webprogram/Paper97411.html]. [20] McDaniel LM, Triulzi DJ, Cramen J, et al. Massive transfusion protocols does not result in preferential use of older red blood cells. J Blood Transfus 2014;2014:328967. [21] Zatta AJ, McQuilten ZK, Mitra B, et al. Elucidating the clinical characteristics of patients captured using different definitions of massive transfusion. Vox Sang 2014;107:60–70.
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005
ARTICLE IN PRESS Disponible en ligne sur
ScienceDirect www.sciencedirect.com Transfusion Clinique et Biologique xxx (2017) xxx–xxx
Original article
Age of blood and survival after massive transfusion Importance de l’âge dans une transfusion massive C.C. Sanz ∗ , A. Pereira Transfusion Service, Hospital Clínic, Villarroel 170, 08036 Barcelona, Spain
Abstract Background. – Massive transfusion is the clinical scenario where the presumed adverse effects of stored blood are expected to be more evident because the whole patient’s blood volume is replaced by stored blood. Objective. – To analyse the association between age of transfused red blood cells (RBC) and survival in massively transfused patients. Methods. – In this retrospective study, clinical and transfusion data of all consecutive patients massively transfused between 2008 and 2014 in a large, tertiary-care hospital were electronically extracted from the Transfusion Service database and the patients’ electronic medical records. Prognostic factors for in-hospital mortality were investigated by multivariate logistic regression. Results. – A total of 689 consecutive patients were analysed (median age: 61 years; 65% males) and 272 died in-hospital. Projected mortality at 2, 30, and 90 days was 21%, 35% and 45%, respectively. The odds ratio (OR) for in-hospital mortality among patients who survived after the 2nd day increased with patient age (OR: 1.037, 95% CI: 1.021–1.054; per year P < 0.001), with the number of RBC unit transfused in the first 48 hours (OR: 1.060; 95% CI: 1.038–1.020 per unit; P < 0.001), and the percentage of such RBC stored for more than 28 days (1.010, 95% CI: 1.005–1.018 per percent point; P = 0.01). Conclusion. – Mortality after massive transfusion was associated with a higher proportion of old RBCs transfused in the first 48 hours. Other factors associated with poor prognosis were older patient’s age and larger volumes of transfused RBCs. © 2017 Elsevier Masson SAS. All rights reserved. Keywords: Blood transfusion; Red blood cells; Survival
Résumé Contexte. – La transfusion massive est le scénario clinique où les effets néfastes présumés du sang stocké devraient être plus évidents car l’ensemble du volume sanguin du patient est remplacé par du sang stocké. Objectif. – Analyser l’association entre l’âge des globules rouges transfusés (RBC) et la survie chez des patients massivement transfusés. Méthodes. – Dans cette étude rétrospective, les données cliniques et transfusionnelles de tous les patients consécutifs massivement transfusés entre 2008 et 2014 dans un grand hôpital de soins tertiaires étaient extraites électroniquement de la base de données du service Transfusion et des dossiers médicaux électroniques des patients. Les facteurs pronostiques pour la mortalité hospitalière ont été étudiés par une régression logistique multiple. Résultats. – Au total, 689 patients ont été analysés (âge médian : 61 ans, 65 % d’hommes) et 272 sont morts à l’hôpital. La mortalité prévue à 2, 30 et 90 jours était respectivement de 21 %, 35 % et 45 %. Le rapport de cote (OR) pour la mortalité hospitalière chez les patients qui ont survécu après le 2e jour augmenta avec l’âge du patient (OR : 1.037, IC 95 % : 1.021–1.054, par an p < 0,001), avec le nombre d’unités RBC transfusées dans les premières 48 heures (OR : 1,060 ; IC 95 % : 1,038–1,020 par unité ; p < 0,001) et le pourcentage de ces RBC stockés pendant plus de 28 jours (1,010, IC 95 % : 1,005–1,018 par point de pourcentage, p = 0,01). Conclusion. – La mortalité après une transfusion massive a été associée à une proportion plus élevée de vieux globules rouges transfusés au cours des premières 48 heures. D’autres facteurs associés à un mauvais pronostic ont été l’âge des patients et de plus grands volumes de globules rouges transfusés. © 2017 Elsevier Masson SAS. Tous droits r´eserv´es. Mots clés : Transfusion sanguine ; Globules rouges ; Survie ∗
Corresponding author. E-mail address: [email protected] (C.C. Sanz).
http://dx.doi.org/10.1016/j.tracli.2017.04.005 1246-7820/© 2017 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005
+Model TRACLI-2900; No. of Pages 5 2
ARTICLE IN PRESS C.C. Sanz, A. Pereira / Transfusion Clinique et Biologique xxx (2017) xxx–xxx
1. Introduction Critical haemorrhage requiring massive transfusion is a leading cause of death in trauma patients and a relevant contributor to surgery-related mortality [1]. In addition, massive transfusion accounts for a large share of all the blood components used in many trauma centres and tertiary-care hospitals [2,3]. Clinical research on the haemotherapy of critical haemorrhage has traditionally been focused on the prevention and treatment of the coagulopathy associated with trauma and massive transfusion. As a result, long established paradigms have changed in recent years, with emphasis now being placed on controlled hypotensive resuscitation and early, proactive transfusion of plasma and platelets [4]. In contrast, little attention has been paid to the attributes of transfused RBCs. Preclinical studies have shown that transfusion of stored RBCs impairs the microcirculation dynamics and the delivery of oxygen to tissues [5]. However, the clinical relevance of such pathophysiologic features remains a matter of discussion because retrospective and prospective observational studies have yielded conflicting results [5–8], and the few available clinical trials may have been unable to detect infrequent but clinically relevant effects [9–11]. Massive transfusion is a unique scenario where the presumed adverse clinical effects of stored blood are expected to be more evident. First, patients usually present in a critical condition in which any disequilibrium of the already precarious homeostasis can be life threatening. Second, the whole patient’s blood volume may end up replaced by stored blood. Nevertheless, clinical guidelines on the management of critical haemorrhage and massive transfusion do not make any recommendation about the age of RBCs [12–16]. The present study was aimed at investigating the association between age of transfused RBCs and in-hospital, all-cause mortality in a large series of massively transfused patients. Additionally, we provide a clinical description of an unselected population of massive transfusion recipients as it was seen in a large, tertiary-care hospital.
We also recorded whether the haemorrhage triggering the massive transfusion begun out-of-hospital or while the patient was already admitted in the hospital. In order to avoid selection and recall biases, all the variables were selected automatically from the electronic databases and only the cause of haemorrhage was assigned subjectively from a pre-established list after reviewing the medical records. Over the period under study, all the RBC units were stored in SAG-M and leukoreduced before storage. Platelet units consisted in a 5-donor pooled concentrate or an equivalent aphaeresis unit. All the plasma was from male donors and most of the units were photo-inactivated with methylene blood. 2.1. Statistics The study’s main outcome was in-hospital mortality (up to the 90th admission day) in patients surviving more than 48 hours after starting the massive transfusion. Patients whose length of stay was longer than 90 days were censored at that follow-up time point (see Fig. S1). Continuous variables were summarized as median and interquartile range (IQR). Statistical comparisons of variables were done by the Mann-Whitney U-test for continuous or ordered data and the Chi-square test for categorical variables. The overall survival curve was drawn by the method of Kaplan and Meier. Patient- and transfusion-related factors associated with in-hospital mortality were investigated by multivariate binary logistic regression. Factors analysed for a possible prognostic significance included sex, age, ABO/Rh blood group, whether the haemorrhage begun in-hospital or out-of-hospital, number of RBC units transfused within the first 48 hours, proportion of RBCs transfused within the first 48 hours that had been stored for more than 28 days, plasma- and platelet-to-RBC ratios, and calendar year (from 2008 to 2014). For the purpose of prognostic analysis only RBC units transfused within the first 48 hours were taken into account. All the statistical analyses were performed by using Stata, version 11 (www.satata.com). The study was approved by the Hospital Clinic’s Committee for Ethics in Research (reference: HCB/2016/0533).
2. Material and methods 3. Results For the purpose of this retrospective study, massive transfusion was defined as the infusion of 8 or more RBC units in less than 24 hours. The Transfusion Service database was queried for all consecutive adult patients (> 16 years) who met the above criterion from January 2008 to December 2014. We also retrieved the date of delivery of every unit of RBC, plasma, platelets, and cryoprecipitate received by the patients within the first 15 days after the massive transfusion as well as the storage time of transfused RBCs. In patients who had more than one episode of massive transfusion, only the first one was taken into account. Surviving patients were followed up until hospital discharge or the study closing date on July 1st, 2015. Information on the medical condition triggering the massive transfusion, the patient’s sex and age, clinical course until death or hospital discharge, and use of manufactured coagulation products was obtained from the patients’ electronic medical records.
A total of 689 patients met the inclusion criteria and constituted the subject of this study (four patients were excluded because RBCs were actually used in exchange transfusion for haemoglobinopathy or paludism). Median age was 61 (IQR: 45–71) years and 65% patients were males. The main patient characteristics at the time of the massive transfusion are shown in Table 1. As can be seen, 56% of patients were already admitted in the hospital at the time of massive bleeding and the most frequent underlying conditions were cardiac surgery (36% of cases), abdominal surgery (17%), and liver transplant (16%). In patient in whom the haemorrhage started out-of-hospital, blunt or penetrating trauma (33% of cases), gastrointestinal bleeding (21%), and ruptured aneurisms of large vessels (16%) were the most frequent conditions. Patients already in-hospital at the start of massive haemorrhage were older than those in whom the
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005
+Model
ARTICLE IN PRESS
TRACLI-2900; No. of Pages 5
C.C. Sanz, A. Pereira / Transfusion Clinique et Biologique xxx (2017) xxx–xxx
3
Table 1 Characteristics of patients undergoing massive transfusion.
Age in years, median (IQR)a Sex (n = 683)a Males Females Blood groupa A O B AB Rh positive Rh negative Underlying conditiona Haemorrhage in hospital Cardiovascular surgery Abdominal surgery Liver transplant Urologic/gynaecologic procedures Gastrointestinal bleeding Other Total Haemorrhage out-of-hospital Traumatic injuries Gastrointestinal bleeding Ruptured aortic aneurysm of great vessels Intra-abdominal bleeding Stab or firearm wounds Other Total Transfusion data, median (IQR) First 48 hours pRBC Plasma Platelets First 15 daysb pRBC Plasma Platelets a b
Whole series
Patients surviving > 48 h
61 (45–71)
61 (47–71)
451 (66%) 232 (34%)
339 (65%) 186 (35%)
305 (44%) 301 (44%) 62 (9%) 20 (3%) 566 (82%) 124 (18%)
240 (45%) 229 (43%) 48 (9%) 15 (3%) 434 (82%) 98 (18%) Fig. 1. Distribution of the length of storage in 10,989 RBC units transfused in the first 48 hours to 689 massively transfused patients.
141 (36%)
112 (34%)
64 (17%) 61 (16%) 44 (11%)
57 (17%) 58 (18%) 42 (13%)
22 (6%)
16 (5%)
52 (14%) 384 (100%)
45 (14%) 330 (100%)
98 (33%) 64 (21%)
59 (30%) 51 (26%)
48 (16%)
26 (13%)
22 (7%)
15 (8%)
15 (5%)
9 (5%)
53 (18%) 300 (100%)
40 (20%) 200 (100%)
12 (10–18) 12 (8–18) 2 (1–3)
12 (10–16) 10 (8–16) 2 (0–3)
15 (11–22) 12 (8–20) 2 (1–4)
15 (11–21) 12 (8–18) (0–4)
Not all patients have all data. Includes blood component units transfused in the first 48 hours.
haemorrhage started out-of-hospital (median [IQR] 63 [51–71] years and 55 [39–71] years, respectively; P = 0.0003). The plasma- and platelet-to-RBC ratio by the first 24 hours were 0.9 (IQR: 0.5–1.3) and 0.1 (IQR: 0.0–0.2), respectively. There was no difference in both ratios between the in-hospital and the out-of-hospital groups. The quantity of blood components transfused within the first 48 hours and by the 15th day
are summarised in Table 1. Cryoprecipitate and/or manufactured fibrinogen concentrate was used within the first 48 hours in 187 patients; prothrombin complex concentrate was used in 95 patients, and rFVIIa in 24. The length of storage of the 10,989 RBC units transfused within the first 48 hours is shown in Fig. 1. Among these units, 8568 (78.0%) were older than 14 days, 1411 (12.8%) had been stored for more than 28 days, and only 496 (4.5%) were younger than 7 days. Patients of B or AB blood group received significantly older RBCs than A/O blood group patients (Supplementary Fig. S2A; P = 0.0001). Two hundred and sixty-two patients received at least one RBC unit older than 28 days. When measured as the proportion of RBCs older than 28 days received within the first 48 hours, the figures ranged from 9.0% in group O patients to 13.0% in group A, and 27.1% in group B and AB patients (P < 0.001). Age of transfused RBCs was relatively invariable for most of the study period but increased in calendar years 2013 and 2014 as compared with previous years (Supplementary Fig. S2B). There was no correlation between the number of RBC units transfused in the firsts 48 hours and the proportion of such units that had been stored for more than 28 days (Spearman rank correlation: −0.002, P = 0.95; Supplementary Fig. S3). Fig. 2 shows the projected survival after massive transfusion for the whole series. As can be seen, estimated survival at 2, 30, and 90 days from massive transfusion was 79% (95% CI: 75%–81%), 65% (95% CI: 61%–69%), and 54% (95% CI: 49%–59%), respectively. Since early mortality is strongly determined by the cause of the critical haemorrhage and the patient’s condition at the start of massive transfusion rather than by the characteristics of the transfused blood, we investigated factors associated with in-hospital mortality in the selected population of 532 patients who survived more than two days. By the 90th follow-up day, 106 (20%) out of these 532 patients have died and 426 were censored (see follow-up flow-card in Supplementary Fig. S1). Factors independently associated with higher mortality between the 2nd and 90th postransfusion day are shown in Table 2. As can be seen, after adjustment for other potential prognostic and
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005
+Model TRACLI-2900; No. of Pages 5 4
ARTICLE IN PRESS C.C. Sanz, A. Pereira / Transfusion Clinique et Biologique xxx (2017) xxx–xxx
Fig. 2. In-hospital projected survival of 689 massively transfused patients followed up until death or hospital discharge. Table 2 Factors predicting in-hospital mortality in massively transfused patients who survived more than 2 days and adjustment covariates included in the logistic regression model. Covariate
Odds ratio (95% CI)
P
Age in years RBC units in the first 48 h Proportion of RBCs used in the first 48 h stored > 28 days (%) Male sex Ratio plasma-to-RBC (first 24 h) Ratio platelets-to-RBC (first 24 h) Haemorrhage starting out-of-hospital Blood group A/O (as opposed to B/AB) Blood group Rh+ (as opposed to Rh−) Calendar year (2008 to 2014)
1.03 (1.02–1.04) 1.07 (1.04–1.10) 1.01 (1.005–1.018)
< 0.001 < 0.001 0.01
1.12 (0.68–1.82) 1.22 (0.97–1.53) 2.99 (0.61–3.45) 1.37 (0.46–1.83) 0.92 (0.88–1.14)
0.65 0.08 0.17 0.18 0.81
1.27 (0.68–2.35)
0.44
1.00 (0.88–1.01)
0.89
confounder factors listed in the table, the odds for dying increased by 3.7% for every additional year in the patient’s age, by 6.9% for every additional RBC unit transfused in the first 48 hours, and by 1.0% for every percent point increase in the fraction of such RBC units that had been stored for more than 28 days. 4. Discussion In this study of an unselected population of massively transfused adult patients seen in a large tertiary-care, academic hospital over a 7-year period, we found that nearly half the patients were projected to have died in hospital after the massive transfusion. Among patients who survived beyond two days from the start of massive transfusion, factors independently associated with mortality were older age, a greater number of RBC units transfused in the first two days, and a large proportion of such units stored for more than 28 days. The mortality rate observed in our study was comparable to that found in other unselected populations of massively
transfused patients [1,3]. We found a high mortality rate in the first 24–48 hours, as previously reported from other series [1,3], but also that nearly half the patients who eventually died, did so days or weeks after the massive transfusion, not rarely after a long hospital stay. This observation suggests that many patients did not actually die because of haemorrhagic shock or as an immediate consequence of the condition triggering the massive transfusion, but of complications arising after the initial critical period. Despite the general agreement on the morphological and physiological changes suffered by stored RBCs, the so called “storage lesion”, and the adverse biological effects of transfusing stored blood found in preclinical studies, there is no agreement on the clinical consequences of such findings [5–8]. Observational studies, both retrospective and prospective, have produced a wide variety of results, sometime conflicting to one another (revised by Wang et al. [6] and Alexander et al. [7]). Recent clinical trials have failed to confirm that transfusion of young RBCs improves survival as compared with transfusion of older blood [9–11]. Although results from these trials can rule out a large clinical effect mediated by old blood, concern remains about the trials’ ability to detect small but clinically relevant effects [8,17], and about extrapolation of results to patient populations and clinical conditions different from those analysed in the trials. For instance, the INFORM investigators excluded massively transfused patients and the median volume transfused per patient was 2 RBC units [11]. Similarly, in the ABLE [9] and the RECESS [10] trials, patients received on average 4 RBC units over a period of several days or even weeks. This is a quite different situation from that of the massively transfused patient in whom stored blood may account for nearly all the circulating RBCs. In this regard, previous authors have noted that age of blood is associated with mortality only in patients transfused with large volumes of RBCs, while the association disappears in patients who receive smaller transfusion volumes [18]. Moreover, it is worth noting that a recent subanalysis of the ABLE trial disclosed a survival advantage in the younger RBC arm for those patients who had received more than 5–7 RBC units [19]. As it has already been pointed out by previous authors [20], our study shows that patients of B and AB blood group are more likely to end up receiving significantly older RBCs than patients of O or A blood groups. In the present study, the ABO blood group was not an independent predictor for shorter survival, perhaps because of low number of group B/AB patients and the fact that a fraction of the transfusion support they received were given with younger, blood group O RBCs. Nevertheless, this observation warrants for exploring specific RBC inventory management and transfusion policies for these patients. The metrics to measure the age of transfused RBCs has varied widely among studies (e.g. age of the oldest transfused RBC unit, average age of all transfused RBCs, number of units older than a given age). In the present study, we employed the proportion of all transfused RBCs that were older than a given threshold as the most appropriate yardstick. Indeed, in massive haemorrhage, many of the transfused RBCs are lost through the bleeding sites, so that it can be entertained that old RBC units transfused after the massive bleeding was controlled would have a higher
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005
+Model TRACLI-2900; No. of Pages 5
ARTICLE IN PRESS C.C. Sanz, A. Pereira / Transfusion Clinique et Biologique xxx (2017) xxx–xxx
potential for negative effects than those transfused earlier. This is also the reason why we took into account the volume transfused in the first 48 hours instead of a shorter period of time. The traditional definition of massive transfusion entails the replacement of the patient’s blood volume in less than 24 hours, what usually is measured by transfusing 10 or more RBC units (and the corresponding colloids, crystalloids or plasma) for the adult patient. However, it has been shown that this 10-unit criterion excludes a large proportion of patients with critical bleeding [21], so that we relied on an 8-unit threshold as a more inclusive one. The 28-day storage threshold used in this study was the result of balancing statistical power (i.e. enough patients in the old blood group) with the capability to capture the prognosis of those patients who received a large proportion of the oldest RBCs. It should not be taken, therefore, as a threshold with a definite biological meaning. In fact, the timing of the potentially adverse effects associated with RBC storage has not yet been defined, and the most discriminating storage threshold based on a biological rational remains unknown [17]. The present study has several limitations. First, the retrospective design does not allow to control for unknown confounding factors. Second, the data we used came from a single institution, so that the results might not be generalized to other institutions with different transfusion policies in massive haemorrhage. Third, clinical characteristics were aggregated into large diagnostic categories, so that granular clinical data that may have influenced on survival were not analysed. Four, the only outcome evaluated was survival, and other information of the clinical curse (e.g. infection, organ failure, etc.) was omitted from the analysis. On the positive side, it is worth mentioning that all the variables analysed were objective and extracted electronically from the databases, so that biases related to selection and interpretation of clinical data and outcomes can reasonably be ruled out. Moreover, although we did not analyse clinical variables known to influence on survival in some specific causes of massive transfusion (i.e. severity scores in trauma patients), it seems quite unlikely that such clinical variables might have confounded the prognostic effect of the age of transfused RBCs. In this regard, it is worth mentioning that length of storage was unrelated to the volume of transfused RBCs. In conclusion, this study shows that delayed mortality after massive transfusion was higher in patients who received a larger proportion of old RBCs in the first two days. Other factors associated with poor prognosis were older patient’s age and larger volumes of transfused RBCs. Authorship CS and AP designed research, collaborate in the acquisition, analysis and interpretation of data, drafted the paper and approved the final version of the manuscript. Disclosure of interest The authors declare that they have no competing interest.
5
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.tracli.2017.04.005. References [1] Halmin M, Chiesa F, Vasan SK, et al. Epidemiology of massive transfusion: a binational study from Sweden and Denmark. Crit Care Med 2016;44:468–77. [2] Como JJ, Dutton RP, Scalea TM, et al. Blood transfusion rates in the care of acute trauma. Transfusion 2004;44:809–13. [3] Dzik WS, Ziman A, Cohen C, et al. Survival after ultramassive transfusion: a review of 1360 cases. Transfusion 2016;56:558–63. [4] Sihler KC, Napolitano LM. Massive transfusion: new insights. Chest 2009;136:1654–67. [5] Flegel WA, Natanson C, Klein HG. Does prolonged storage of red blood cells cause harm? Br J Haematol 2014;165:3–16. [6] Wang S, Sun J, Solomon SB, et al. Transfusion of older stored blood and risk of death: a meta-analysis. Transfusion 2012;52:1184–95. [7] Alexander PE, Barty R, Fei Y, et al. Transfusion of fresher vs older red blood cells in hospitalized patients: a systematic review and meta-analysis. Blood 2016;127:400–10. [8] Glynn SA, Klein HG, Ness PM. The red blood cell storage lesion: the end of the beginning. Transfusion 2016;56:1462–8. [9] Lacroix J, Hébert PC, Fergusson DA, et al. Age of transfused blood in critically ill adults. N Engl J Med 2015;372:1410–8. [10] Steiner ME, Ness PM, Assmann SF, et al. Effects of red-cell storage duration on patients undergoing cardiac surgery. N Engl J Med 2015;372:1419–29. [11] Heddle NM, Cook RJ, Arnold DM, et al. Effect of short-term vs long-term blood storage on mortality after transfusion. N Engl J Med 2016;375:1937–45. [12] Stainsby D, MacLennan S, Thomas D, et al. Guidelines on the management of massive blood loss by the British Committee for standards in haematology. Br J Haematol 2006;135:634–41. [13] Gaarder C, Naess PA, Frischknecht Christensen E, et al. Scandinavian guidelines – “The massively bleeding patient”. Scand J Surg 2008;97:15–36. [14] Dzik WH, Blajchman MA, Fergusson D, et al. Clinical review: Canadian national advisory committee on blood and blood products–massive transfusion consensus conference 2011: report of the panel. Crit Care 2011;15:242. [15] Patient blood management guidelines: module 1. Critical bleeding/massive transfusion. Australian National Blood Authority; 2011 [available from URL: http://www.blood.gov.au/system/files/documents/ pbm-module-1.pdf; accessed on February 20th, 2016]. [16] Pham HP, Shaz BH. Update on massive transfusion. Br J Anaest 2013;111:i71–82. [17] Pereira A. Will clinical studies elucidate the connection between the length of storage of transfused red blood cells and clinical outcomes? An analysis based on the simulation of randomized controlled trials. Transfusion 2013;53:34–40. [18] Weinberg JA, McGwing G, Griffin RL, et al. Age of transfused blood: an independent predictor of mortality despite universal leukoreduction. J Trauma 2008;65:279–84. [19] Mack JP, Kahn SR, Tinmouth A, Fergusson D, Hébert PC, Lacroix J. Dosedependent effect of stored red blood: results of a sub-group analysis of the age of blood evaluation (ABLE) trial. 58th Congress of the American Hematology Association, San Diego (USA); 2016 [available from URL: https://ash.confex.com/ash/2016/webprogram/Paper97411.html]. [20] McDaniel LM, Triulzi DJ, Cramen J, et al. Massive transfusion protocols does not result in preferential use of older red blood cells. J Blood Transfus 2014;2014:328967. [21] Zatta AJ, McQuilten ZK, Mitra B, et al. Elucidating the clinical characteristics of patients captured using different definitions of massive transfusion. Vox Sang 2014;107:60–70.
Please cite this article in press as: Sanz CC, Pereira A. Age of blood and survival after massive transfusion. Transfusion Clinique et Biologique (2017), http://dx.doi.org/10.1016/j.tracli.2017.04.005