UBC Faculty Research and Publications

Anemia prevalence and incidence and red blood cell transfusion practices in aneurysmal subarachnoid hemorrhage:… English, Shane W; Chassé, Michaël; Turgeon, Alexis F; Lauzier, François; Griesdale, Donald; Garland, Allan; Fergusson, Dean; Zarychanski, Ryan; van Walraven, Carl; Montroy, Kaitlyn; Ziegler, Jennifer; Dupont-Chouinard, Raphael; Carignan, Raphaëlle; Dhaliwal, Andy; Mallick, Ranjeeta; Sinclair, John; Boutin, Amélie; Pagliarello, Giuseppe; Tinmouth, Alan; McIntyre, Lauralyn Jul 4, 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-13054_2018_Article_2089.pdf [ 685.21kB ]
JSON: 52383-1.0368793.json
JSON-LD: 52383-1.0368793-ld.json
RDF/XML (Pretty): 52383-1.0368793-rdf.xml
RDF/JSON: 52383-1.0368793-rdf.json
Turtle: 52383-1.0368793-turtle.txt
N-Triples: 52383-1.0368793-rdf-ntriples.txt
Original Record: 52383-1.0368793-source.json
Full Text

Full Text

RESEARCH Open AccessAnemia prevalence and incidence andred blood cell transfusion practices inaneurysmal subarachnoid hemorrhage:results of a multicenter cohort studyShane W. English1,2* , Michaël Chassé3, Alexis F. Turgeon4, François Lauzier4,5, Donald Griesdale6, Allan Garland7,Dean Fergusson2, Ryan Zarychanski7, Carl van Walraven8, Kaitlyn Montroy2, Jennifer Ziegler7,Raphael Dupont-Chouinard4, Raphaëlle Carignan4, Andy Dhaliwal9, Ranjeeta Mallick2, John Sinclair10,Amélie Boutin11, Giuseppe Pagliarello12, Alan Tinmouth2, Lauralyn McIntyre1,2 and on behalf of the CanadianCritical Care Trials GroupAbstractBackground: Whether a restrictive strategy for red blood cell (RBC) transfusion is applied to patients with aneurysmalsubarachnoid hemorrhage (aSAH) is unclear. To inform the design and conduct of a future clinical trial, we sought todescribe transfusion practices, hemoglobin (Hb) triggers, and predictors of RBC transfusion in patients with aSAH.Methods: This is a retrospective cohort study of all consecutively admitted adult patients with aSAH at four tertiarycare centers from January 1, 2012, to December 31, 2013. Patients were identified from hospital administrativedischarge records and existing local aSAH databases. Data collection by trained abstractors included demographic data,aSAH characteristics, Hb and transfusion data, other major aSAH cointerventions, and outcomes using a pretested casereport form with standardized procedures. Descriptive statistics were used to summarize data, and regression modelswere used to identify associations between anemia, transfusion, and other relevant predictors and outcome.Results: A total of 527 patients met inclusion eligibility. Mean (±SD) age was 57 ± 13 years, and 357 patients (67.7%)were female. The median modified Fisher grade was 4 (IQR 3–4). Mean nadir Hb was 98 ± 20 g/L and occurred onmedian admission day 4 (IQR 2–11). RBC transfusion occurred in 100 patients (19.0%). Transfusion rates varied acrosscenters (12.1–27.4%, p = 0.02). Patients received a median of 1 RBC unit (IQR 1–2) per transfusion episode and a mediantotal of 2 units (IQR 1–4). Median pretransfusion Hb for first transfusion was 79 g/L (IQR 74–93) and did not varysubstantially across centers (78–82 g/L, p = 0.37). Of patients with nadir Hb < 80 g/L, 66.3% received a transfusioncompared with 2.0% with Hb nadir ≥ 100 g/L (p < 0.0001). Predictors of transfusion were history of oral anticoagulantuse, anterior circulation aneurysm, neurosurgical clipping, and lower Hb. Controlling for numerous potential confounders,transfusion was not independently associated with poor outcome.Conclusions: We observed that moderate anemia remains very common early in admission following SAH. Only one-fifthof patients with SAH received RBC transfusions, mostly in cases of significant anemia (Hb < 80 g/L), and this did notappear to be associated with outcome.Keywords: Subarachnoid hemorrhage, Cerebral aneurysm, Anemia, Red blood cell transfusion, Cohort study* Correspondence: senglish@ohri.ca1Department of Medicine (Critical Care), The Ottawa Hospital, Civic CampusRoom F202, 1053 Carling Avenue, Ottawa, ON K1Y 4E9, Canada2Clinical Epidemiology Program (Centre for Transfusion Research), OttawaHospital Research Institute, Ottawa, ON, CanadaFull list of author information is available at the end of the article© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.English et al. Critical Care  (2018) 22:169 https://doi.org/10.1186/s13054-018-2089-7BackgroundMost atraumatic subarachnoid hemorrhages are caused bya ruptured cerebral artery aneurysm (aneurysmal sub-arachnoid hemorrhage [aSAH]). Patients who survive theprimary event are at high risk of complications, includingdelayed cerebral ischemia (DCI) such as with vasospasm,which may result in further neurologic deficits andincreased likelihood of death [1]. Initial managementfocuses on patient stabilization and support; securing ofthe aneurysm; and monitoring for, preventing, and treat-ing potential complications [2, 3]. Medical complications,including anemia, are common and affect up to 50% ofpatients with aSAH [4]. Among them, both anemia andtransfusion of red blood cells (RBC) have been associatedwith complications and poor outcomes [4]. Anemicpatients with aSAH are not only more likely to receiveRBC transfusions but also at an increased risk of ischemiccomplications [4]. Those who receive RBC transfusionshave been shown to have less favorable hospital outcomes,including severe disability [5].Preclinical studies in brain injury suggest that RBC trans-fusion to treat anemia improves oxygen delivery [6]. Onlyone small trial (N = 44) has compared two transfusion tar-gets (100 g/L and 115 g/L) in aSAH, but it was underpow-ered to examine clinically important outcomes [7]. Anothersmall mixed brain injury population trial (N = 102) com-pared a hemoglobin (Hb) target with a transcranial oxygen-ation target, but it, too, was underpowered to examineclinically important outcomes [8]. Despite this absence ofevidence, current aSAH management guidelines recom-mend considering RBC transfusion in anemic patients atrisk for cerebral ischemia, but they do not suggest specifictransfusion thresholds to guide clinicians [1, 3]. Theserecommendations are in contrast with current stated aSAHmanagement practice derived from surveys [9] andevidence from clinical trials in other critically ill adult andpediatric populations, which support a more restrictiveRBC transfusion approach [10, 11].Canadian epidemiologic data on aSAH is sparse andmore than 20 years old [12]. These data are limited tocharacterizing hospitalizations and case fatality rates withno published data on transfusion practices, which havechanged significantly in critically ill patients over the lastdecade. In collaboration with the Canadian Critical CareTrials Group, we conducted a retrospective, multicentercohort study of consecutive patients with aSAH to betterunderstand current RBC transfusion practices.MethodsStudy design and objectivesWe conducted a multicenter retrospective observationalstudy of patients with aSAH admitted to one of fourCanadian academic tertiary care hospitals in four differentprovinces. This study was approved by the local researchethics board of each participating center, which waivedthe need for informed consent.Our a priori primary objective was to describe RBCtransfusion practices, including the distribution of Hbvalues at which patients are transfused, the proportion ofpatients transfused, the median number of RBC transfu-sions per patient, the predictors of anemia and transfu-sion, and practice variations between participating centers.Our secondary objective was to examine the association ofRBC transfusion and clinical outcome.Patient selectionWe included all consecutive admissions for aSAH betweenJanuary 1, 2012, and December 31, 2013. We identifiedpotentially eligible patients by screening two sources: (1)hospital discharge abstracts, including all dischargeabstracts that listed the diagnosis of primary SAH usingInternational Classification of Diseases, Tenth Revision(ICD-10), codes (I60.0 to I60.9), and, where available, (2)local database or patient repositories, including all patientswith a diagnosis of aSAH obtained from an existing localhospital, intensive care unit (ICU), or neurosurgicaldatabase. The local database had to include consecutivecases and be searchable by principal diagnoses that werenot ICD code-generated.Each patient underwent chart review by a trained dataabstractor. For final inclusion in the cohort, patients had tobe ≥ 18 years of age and to have sustained an aSAH. Thediagnosis of aSAH required both evidence of SAH (by atleast one of the following: blood in subarachnoid space asdemonstrated by neuroimaging [e.g., computed tomog-raphy {CT}, magnetic resonance imaging, lumbar puncturedemonstrating xanthochromia and/or > 5 × 106 RBCs/L, orblood in subarachnoid space as observed at postmortemautopsy report) and (2) evidence of aneurysm rupture asthe cause (by at least one of the following: aneurysmobserved on angiography, aneurysm demonstrated onneuroimaging report [computed tomographic angiographyor magnetic resonance angiography], or aneurysm observedon postmortem examination report). No exclusion criteriawere used, because we wanted to describe all adult patientswith true aSAH as established by the inclusion criteria.Similar criteria have been used in previous cohort studiesexamining anemia and RBC transfusion in aSAH [5, 13].Patient managementAll four study centers are academic tertiary care referralhospitals with complete neurosurgical, neurointerventional,and intensive care services. Management of patients withaSAH was in accordance with local practice and proceduresand the most recent aSAH management guidelines [3].Specifically, acute care goals included early treatment ofhydrocephalus and securing of ruptured aneurysm, as wellas traditional management of intracranial hypertension withEnglish et al. Critical Care  (2018) 22:169 Page 2 of 9goal-directed osmotic therapy. All sites fostered early detec-tion of DCI with close clinical observation with or withoutserial surveillance imaging (e.g., transcranial Doppler).Medical complications of aSAH, including cardiac dysfunc-tion, electrolyte abnormalities, and temperature dysregula-tion, were managed according to local practice and inaccordance with accepted standards.At each site, acute care management of patients withaSAH occurred in either a level 3 ICU for mechanicallyventilated and/or unstable patients or in a level 2high-acuity unit until such time that the aneurysm wassecured, the patient was off all life supports and vasoactivemedications, and in the absence of DCI. During the studyperiod, no participating center had a specific RBC transfu-sion protocol for managing aSAH, nor did any of themhave specific transfusion guidelines, and the decision toinitiate as well as the timing of RBC transfusion was at thediscretion of the treating team.Data collectionData collection was completed by trained data abstractorsusing a previously prepared and piloted case report formand a standardized operations manual tested for com-pleteness, clarity, and ease of use. We used multiplesources, including electronic and paper medical records aswell as laboratory and imaging reports. An ICU day wasconsidered any amount of time in a single calendar dayadmitted to a level 2 or 3 unit (Additional file 1: Supple-mental Material). SAH severity was captured with Huntand Hess grade [14], World Federation of NeurologicalSurgeons grade [15], and/or the modified Fisher scale[16]. When not explicitly reported, modified Fisher scalegrade was calculated using the first available CT scan.The date and time of each RBC transfusion adminis-tered was recorded. Multiple units of RBCs transfused onthe same calendar day were considered to be part of asingle transfusion episode, unless separate transfusionsoccurred intraoperatively (this was considered a separatetransfusion episode). A pretransfusion Hb (defined as themost recent Hb [within 48 h] drawn prior to the initiationof RBC transfusion) was captured for each transfusion epi-sode. In the 6 h preceding an RBC transfusion, we identi-fied episodes of active bleeding (defined as more than250 ml of blood loss in 1 h or active blood loss associatedwith hemodynamic instability). We captured admissionHb, daily nadir Hb values for the first 21 days of admis-sion, and hospitalization and ICU stay nadir Hb. We de-fined anemia as a Hb value less than or equal to 100 g/Lbecause this value has clinical significance and representsan important threshold in multiple previously publishedtransfusion trials [11, 17–19]. Diagnostic criteria for vaso-spasm and cerebral infarction were established a priori(Additional file 1: Supplemental Material).We recorded vital status at discharge, discharge destin-ation (home, other hospital, rehabilitation, long-term care/nursing home, or other), and functional status at dischargeapplying the modified Rankin Scale (mRS) criteria (from 0to 6) (Additional file 1: Supplemental Material). Functionalstatus was ascertained using available documentation, in-cluding physiotherapy and occupational therapy dischargenotes and rehabilitation assessments. We dichotomizedstatus at discharge to good (mRS 0–3) or poor (mRS 4–6)outcome.Statistical analysesDescriptive statistics were used to summarize the studydata. Continuous variables are presented as mean with SDor median with IQR, depending on their distribution.Frequency and proportion estimates are presented aspoint estimates with 95% CIs. We used the chi-square orFisher’s exact test, as appropriate, to examine differencesin transfusion rates among different trigger thresholds.Predictors of anemia, RBC transfusion, and poor neuro-logic functional outcome were tested using a randomeffects generalized linear model to account for clusteringat the center level. Potential predictor variables for eachmodel were set a priori and included variables identifiedpreviously in the literature and those with clinical signifi-cance. For anemia, these included admission Hb, age, gen-der, history of oral anticoagulant, modified Fisher grade,aneurysm size, method of securing the aneurysm (clip vscoil), and presence of vasospasm (prior to onset ofanemia). For RBC transfusion, predictors considered in-cluded age, gender, history of oral anticoagulant, modifiedFisher grade, location of aneurysm (anterior vs posteriorcirculation), method of securing the aneurysm (clip vscoil), admission Hb, presence of anemia, and vasospasmor cerebral infarction (prior to the first RBC transfusion).Owing to the clinical correlation between the latter two,only vasospasm was considered in the final model. Finally,predictors of poor outcome entered in the model includedall of the previously mentioned variables and otherpreexisting comorbidities, presenting Glasgow Coma Scale(GCS) score, need for mechanical ventilation, hospitallength of stay, need for ICU admission, need for tracheos-tomy, or need for percutaneous gastrostomy tube.Similarly, given the clinical correlation between the needfor mechanical ventilation, need for ICU admission, andneed for tracheostomy, we did not include need for ICUadmission in the final model, because it was least signifi-cant in univariate analysis. Cerebral infarction wasincluded (instead of vasospasm) owing to its significantcorrelation during univariate analysis. We examined formulticollinearity among predictors. Statistical interactionbetween certain variables was considered and included inthe model only if statistically significant. Statistical signifi-cance was considered as a p value < 0.05. All analyses wereEnglish et al. Critical Care  (2018) 22:169 Page 3 of 9completed using SAS 9.3 software (SAS Institute, Cary,NC, USA).ResultsOf the 886 screened patients, 42 had no SAH, and 317had SAH resulting from causes other than aneurysmrupture, leaving a cohort of 527 patients. Baseline anddisease-specific characteristics, cointerventions, anddisease-related complications are presented in Table 1.AnemiaAnemia at presentation was found in 29 patients (5.5%).Among the other 498 patients, the hospitalizationincidence of acquired anemia was 52.0% (47.6–56.4%).The overall prevalence of anemia within the first 21 daysof admission was 51.8% (95% CI 47.5–56.1%). The meanhospitalization nadir Hb was 98.0 ± 19.5 g/L and occurredon median admission day 4 (IQR 2–11). The proportionof anemic patients varied across centers (p = 0.005)(Additional file 1: Figure S1). Female sex, history of oralanticoagulant use, modified Fisher grade 3 or 4, admissionHb, and aneurysm secured by neurosurgical clipping wereindependent predictors of anemia (see Table 2).Primary outcome: RBC transfusion practicesOverall, 100 patients (19.0%, 95% CI 15.6–22.3%) under-went transfusion of at least 1 unit of packed red bloodcells (PRBCs). A median of 1 unit of PRBCs (IQR 1–2)per transfusion episode was received, and a median of2 units of PRBCs (IQR 1–3) was received during thehospitalization. The median pretransfusion Hb of thefirst transfusion was 79 g/L (IQR 74–93), and themedian was 80 g/L (IQR 72–86) for all transfusions. Atotal of 35 patients (6.6%) had an intraoperative transfu-sion, and 24 patients (4.6%) received ≥1 unit of PRBCswithin 6 h of active bleeding.We grouped patients by their Hb nadir (in nontrans-fused patients) and pretransfusion Hb of first transfusion(in transfused patients) in Hb increments of 10 g/L. RBCtransfusion rates were highest in those patients with a Hb< 80 g/L (p < 0.0001) (Fig. 1). Among patients with a nadirHb of 100 g/L or higher, only 7.0% of patients received anRBC transfusion.Transfusion rates varied from 12.5% to 27.4% acrossparticipating sites (p = 0.016) (Additional file 1: Figure S2).However, median pretransfusion Hb did not vary signifi-cantly between sites, ranging from 78 to 82 g/L (p = 0.37)(Fig. 2).Predictors of RBC transfusionUnivariate predictors of at least 1 RBC transfusion includedhistory of home anticoagulant use, lower admission Hb,ruptured aneurysm in the anterior cerebral circulation,secured aneurysm by neurosurgical clipping, and anemia(Table 3). In our regression model, only lower admissionHb, ruptured aneurysm secured by neurosurgical clipping,and anemia remained independent risk factors for transfu-sion. Among patients with vasospasm, the only predictor oftransfusion was moderate anemia (OR 9.59, 95% CI 2.75–33.48, p = 0.0004).We stratified patients on the basis of their nadir Hb toexamine differences in transfusion predictors. Among themost severely anemic patients (Hb < 80 g/L, n = 98), 66.3%(65 patients; 95% CI 57.0–75.7%) of patients received atransfusion with no identifiable significant predictors inour model (Table 4). In the moderate anemia group (80 ≤Hb < 100 g/L, n = 175), 17.1% (30 patients; 95% CI 11.6–22.7%) received a transfusion. Independent predictors oftransfusion in this group were increasing age (per 10-yearTable 1 Patient characteristics, interventions, and complicationsData(N = 527)CharacteristicsAge, yr, mean ± SD 57 ± 13Female, no. (%) 357 (67.7)Comorbidities, no. (%)Hypertension 233 (44.2)Heart disease 41 (7.8)Active tobacco smoker 185 (35.1)Home medications, no. (%)Antiplatelet 67 (12.7)Oral anticoagulant 20 (3.8)Statin 100 (19.0)aSAH characteristicsPresenting GCS, mean ± SD 11 ± 5Modified Fisher scale score, median (IQR) 4 (3–4)AneurysmSize, mm, mean ± SD 6.9 ± 4.3Posterior circulation, no. (%) 182 (34.5)Anterior circulation, no. (%) 327 (62)Interventions and cointerventions, no. (%)Surgical clipping 209 (39.7)Endovascular coiling 275 (52.5)Postadmission day aneurysm secured, median (IQR)a 0 (0–1)EVD 236 (44.8)RBC transfusion 100 (19.0)Need for mechanical ventilation 216 (49.8)Complications, no. (%)Vasospasm 142 (26.9)New ischemic neurologic lesions 104 (19.7)Abbreviations: aSAH Aneurysmal subarachnoid hemorrhage, EVD Externalizedventricular drain, GCS Glasgow Coma Scale, RBC Red blood cellsaEleven dates missingEnglish et al. Critical Care  (2018) 22:169 Page 4 of 9increase OR 1.18, 95% CI 1.10–1.26), male sex (OR 1.42,95% CI 1.03–1.96), and neurosurgical clipping of theaneurysm (OR 4.60, 95% CI 1.89–11.17). Of the 254patients in the nonanemic group (nadir Hb ≥ 100 g/L), 5patients (2.0%, 95% CI 0.6–4.5%) received a transfusion.There were insufficient events to identify independentpredictors of transfusion in this group.RBC transfusion association with outcomeNeurologic functional outcome at discharge was missingfor 66 patients (12.5%). By hospital discharge, 93 (20.2%)patients had died and 220 (47.7%) had a poor neurologicoutcome (mRS 4–6). Although both anemia (Hb < 100 g/L)and RBC transfusion were predictors of poor outcome(ORs 2.44 and 3.72, 95% CIs 1.67–3.57 and 2.23–6.19,respectively), neither of them was a significant independentpredictor when controlling for these and other factors thatare likely to affect outcome (Table 5). Independentpredictors of poor neurologic outcome included increasingage; history of hypertension, stroke, or ever smoking; homeoral anticoagulant use, high-grade modified Fisher grade;anterior circulation aneurysm; poor GCS at presentation;need for mechanical ventilation; and need for percutaneousgastrostomy tube (Table 4).DiscussionIn this multicenter retrospective review of patients withaSAH, we found that although moderate anemia incidenceremains very high (51.8%), RBC transfusion is uncommon(19.0%). Significant RBC transfusion rates were onlyobserved with severe anemia (Hb ≤ 80 g/L), suggesting a“restrictive” RBC transfusion practice among the hospitalsTable 2 Predictors of anemia (Hb ≤ 100 g/L) during admission for subarachnoid hemorrhage, clustered by centerUnivariate analysis Multivariable model (n = 437; 238 events)Variable OR 95% CI p Value Variable OR 95% CI p ValueAge (increase by 10 yr) 1.02 0.97–1.08 0.472 Age (increase by 10 years) 0.99 0.91–1.09 0.913Sex (female vs other) 2.82 1.76–4.50 < 0.0001 Sex (female vs other) 1.91 1.08–3.40 0.027Hx of oral AC use 6.17 2.79–14.97 < 0.0001 Hx of oral AC use 5.98 2.11–16.95 0.001Admission hemoglobin(increase by 10 g/L)0.35 0.32–0.39 < 0.0001 Admission hemoglobin (increase by 10 g/L) 0.55 0.43–0.71 < 0.0001Modified Fisher grade 3–4 vs 0–2a 1.42 0.98–2.08 0.067 Modified Fisher grade 3–4 vs 0–2a 1.96 1.07–3.57 0.028Anterior circulation (vs posterior) 0.98 0.65–1.48 0.929 Anterior circulation (vs posterior) 0.58 0.30–1.13 0.107Aneurysm size (increase by 1 mm) 1.03 1.02–1.05 < 0.0001 Aneurysm size (increase by 1 mm) 1.03 0.98–1.07 0.226Clip (vs other) 3.19 2.09–4.85 < 0.0001 Clip (vs other) 4.62 2.46–8.69 < 0.0001Vasospasm (preanemia) 0.78 0.52–1.17 0.223 Vasospasm (preanemia) 1.49 0.79–2.80 0.214Abbreviations: AC Anticoagulant, antiplt Antiplatelet, Hb Hemoglobin, Hx History, RBCTx Red blood cell transfusion, SAH Subarachnoid hemorrhageaDichotomized as high (3–4) vs low (0–2) gradeFig. 1 Red blood cell transfusion rates across hemoglobin thresholds. Proportion of patients transfused according to hemoglobin threshold. Errorbar depicts proportion 95% CIEnglish et al. Critical Care  (2018) 22:169 Page 5 of 9in the study. Although the proportion of patients trans-fused differed across participating sites, the median pre-transfusion Hb did not differ, ranging from 78 to 82 g/L.In this large and robust dataset, controlling for other clin-ical factors that may influence outcome, RBC transfusionwas not correlated with negative outcomes. Further,although our data are not able to demonstrate if RBCtransfusion is in fact helpful in the improvement ofneurologic outcome, in the current cohort, it is a markerof disease severity.Our findings are important and add significantly to theaSAH and transfusion literature for several reasons. First,this is the first multi-institutional study examining RBCtransfusion practice in this patient population. Second, wehave described independent predictors of RBC transfusionusing a robust dataset. Further, we have attempted todescribe how these factors may change depending on thelevel of Hb of a given patient. Last, we have demonstratedwith a large, multicenter dataset populated with a largenumber of variables that, when controlling for otherimportant factors, RBC transfusion does not appear to bea predictor of poor neurologic functional outcome athospital discharge, but we cannot conclude that it is eitherharmful or helpful in this population.The common occurrence of anemia following aSAH hasbeen previously described [4, 5]. Despite changes incommon management strategies of patients with aSAHsince these earlier descriptions, including a movementaway from hypervolemia and hemodilution (two of theso-called triple-H therapies) in favor of hyperdynamictherapy for DCI management [20], our findings suggestthat this has had little impact on anemia incidence.Nonetheless, our findings also suggest that RBC transfu-sion is uncommon relative to the incidence of anemia andFig. 2 Red blood cell transfusion threshold site variation. Whisker box plot of pretransfusion hemoglobin according to site. The center line in thebox represents the median, and the outside lines of the box represent the first and third quartiles. The lines extending from the boxes demonstratethe minimum and maximum valuesTable 3 Predictors of one or more red blood cell transfusions during admission for subarachnoid hemorrhage, clustered by centerUnivariate analysis Multivariable model (n = 463; 95 events)Variable OR 95% CI p Value Variable OR 95% CI p ValueAge (increase by 10 years) 1.07 0.90–1.28 0.458 Age (increase by 10 years) 1.06 0.78–1.45 0.711Sex (male vs other) 0.45 0.19–1.10 0.080 Sex (male vs other) 0.84 0.45–1.57 0.593Hx of oral AC use 1.93 1.01–3.68 0.046 Hx of Oral AC use 1.00 0.60–1.67 0.993Admission Hb (increase by 10 g/L) 0.40 0.39–0.41 < 0.0001 Admission Hb (increase by 10 g/L) 0.81 0.69–0.96 0.013Modified Fisher grade 3–4 vs 0–2a 1.02 0.90–1.15 0.765 Modified Fisher Grade 3–4 vs 0–2a 1.02 0.77–1.36 0.876Anterior circulation (vs posterior) 1.46 1.17–1.83 0.001 Anterior circulation (vs Posterior) 1.18 0.90–1.55 0.231Clip (vs other) 4.39 3.25–5.92 < 0.0001 Clip (vs other) 2.44 1.21–4.94 0.013Anemia (Hb ≤100 g/L) 28.17 11.06–71.75 < 0.0001 Anemia (Hb ≤100 g/L) 17.38 5.11–59.13 < 0.0001Vasospasm (pre-RBCTx) 1.62 0.91–2.88 0.100 Vasospasm (pre-RBCTx) 1.11 0.40–3.09 0.845Cerebral infarct (pre-RBCTx) 1.08 0.96–1.23 0.206Abbreviations: AC Anticoagulant, antiplt Antiplatelet, Hb Hemoglobin, Hx History, RBCTx Red blood cell transfusionaDichotomized as high (3–4) vs low (0–2) gradeEnglish et al. Critical Care  (2018) 22:169 Page 6 of 9is largely restricted to those with severe anemia (medianHb ≤ 80 g/L). This “restrictive RBC transfusion” practiceis in keeping with a stated practice from a 2010 NorthAmerican survey [21]. Our observed practice is vastlydifferent from the two transfusion triggers (100 g/L vs115 g/L) employed in the only randomized controlled trialexamining RBC transfusion thresholds in patients withaSAH and from the current aSAH management guide-lines, which include a recommendation to consider RBCtransfusion in anemic patients at risk for cerebralischemia [3].Four prior studies [5, 22–24] and an abstract publica-tion [25] report adjusted analyses of the effect of RBCtransfusion on poor outcome, using a variety of controlvariables. Additionally, three prior studies [5, 26, 27](and two potentially related abstracts [28, 29]) report ad-justed analyses of effect of RBC transfusion on mortality.Two demonstrated a statistically significant association:1. RBC transfusion increased the odds of death bythreefold (OR 3.16, 95% CI 1.02–9.69) when control-ling for nadir Hb (and its interaction with RBC trans-fusion) in an RBC transfusion propensity scoreTable 4 Predictors (multivariable model) of red blood cell transfusion, stratified by nadir hemoglobinVariable Hb < 80 g/L(n = 83; 61 RBCTx events)80 ≤ Hb < 100 g/L(n = 168; 32 RBCTx events)OR 95% CI p Value OR 95% CI p ValueAge (increase by 10 years) 1.03 0.82–1.29 0.801 1.18 1.10–1.26 < 0.0001Male sex (vs other) 0.48 0.13–1.87 0.293 1.42 1.03–1.96 0.031Fisher grade 3–4 (vs 0–2) 1.29 0.76–2.18 0.347 0.64 0.29–1.42 0.268Anterior circulation (vs posterior) 1.77 0.65–4.79 0.262 0.85 0.60–1.22 0.385Clip (vs other) 1.00 0.16–6.18 1.000 4.60 1.89–11.17 0.001Vasospasm (pre-RBCTx) 0.81 0.23–2.91 0.751 1.57 0.68–3.62 0.294Hb Hemoglobin, RBCTx Red blood cell transfusionTable 5 Predictors of poor outcomea, clustered by centerUnivariate analysis Multivariable model (n = 404; 199 events)Variable OR 95% CI p Value Variable OR 95% CI p ValueAge (increase by 10 years) 1.65 1.28–2.12 0.0001 Age (increase by 10 year) 1.58 1.30–1.93 < 0.0001Sex (male vs other) 0.92 0.71–1.20 0.548 Sex (male vs other) 0.96 0.35–2.63 0.936History of hypertension 2.06 1.56–2.72 < 0.0001 History of Hypertension 1.47 1.06–2.03 0.022History of ischemic stroke 9.22 3.81–22.35 < 0.0001 History of Ischemic stroke 7.85 2.29–26.92 0.001History of heart disease 2.99 1.40–6.41 0.005 History of Heart disease 1.02 0.33–3.12 0.976Smoking history 0.54 0.38–0.78 0.0009 Smoking history 0.58 0.34–0.99 0.047Home oral anticoagulant 8.44 1.22–58.54 0.031 Home Oral anticoagulant 5.38 1.48–19.64 0.011Fisher grade 3–4 vs 0–2 2.83 2.31–3.45 < 0.0001 Fisher Grade 3–4 vs 0–2 1.60 1.45–1.77 < 0.0001Anterior circulation (vs posterior) 1.17 1.00–1.37 0.056 Anterior circulation (vs Posterior) 1.54 1.10–2.14 0.011GCS (for each increase by 1) 0.82 0.81–0.84 < 0.0001 GCS (for each increase by 1) 0.95 0.90–1.00 0.045Need for mechanical ventilation 9.58 6.89–13.31 < 0.0001 Need for Mechanical Ventilation 4.94 3.14–7.78 < 0.0001Clip (vs other) 0.86 0.48–1.54 0.602 Clip (vs other) 0.58 0.24–1.42 0.233RBCTx 3.72 2.23–6.19 < 0.0001 RBCTx 1.50 0.83–2.71 0.175Anemia (Hb ≤ 100 g/L) 2.44 1.67–3.57 < 0.0001 Anemia (Hb ≤100 g/L) 1.06 0.57–1.97 0.860Vasospasm 1.35 0.84–2.18 0.217 Cerebral Infarct 1.78 0.75–4.21 0.188Cerebral infarct 2.83 1.73–4.65 < 0.0001 Hospital length of stay 1.00 0.99–1.00 0.456Hospital length of stay 1.01 1.00–1.02 0.016 Need for tracheostomy 1.97 0.45–8.57 0.365Need for ICU/IMCU admission 1.38 0.48–3.92 0.549 Need for PEG tube 8.62 1.12–66.56 0.039Need for tracheostomy 7.96 2.89–21.89 < 0.0001Need for PEG tube 25.22 6.28–101.37 < 0.0001Abbreviations: GCS Glasgow Coma Scale, Hb Hemoglobin, ICU Intensive care unit, IMCU Intermediate care unit, PEG Percutaneous gastrostomy tube, RBCTx Redblood cell transfusiona67 missingEnglish et al. Critical Care  (2018) 22:169 Page 7 of 9analysis of predicted mortality using a cohort fromtwo hospitals [26]2. A fourfold increase in odds of death or poor neurologicoutcome was seen with RBC transfusion (OR 4.3, 95% CI1.9–9.3) in a single-center cohort when controlling for age,SAH severity, and anemia [5].The differences found in our study may be attributableto its larger, more current sample, higher event rates, andmulticenter contribution to the cohort, distinguishing itfrom all of these previous studies.The strengths of this study lie in the rigorous method-ology employed to minimize some of the biases inherentin retrospective cohort studies. Specifically, we set theprotocol a priori and undertook data collection by trainedabstractors using a piloted data collection tool and a set ofprocedures and definitions. We made a significant effortto ensure complete data abstraction with multiple datasources. These efforts led to our ability to gather robustdata that allowed for controlled analyses that have limitedother studies.Our study is not without limitation, however. We can-not eliminate all of the inherent potential biases. Althoughwe were successful in obtaining modified Fisher grades forthe majority of patients, the documenting of otherprognostic scales was poor, limiting our ability to use thesedata in our models. We applied very stringent and conser-vative diagnostic criteria for vasospasm, as a measure ofDCI, that included the need for imaging, clinical change,and treatment initiation. Owing to variances and inad-equacies in the documentation, we believe that our findingof a vasospasm incidence of 26.9% is a significant under-estimation of the true incidence. As such, this may haveprecluded our ability to demonstrate vasospasm as animportant predictor of RBC transfusion, both overall andin the anemia subgroups. Further, we were unable tocollect undocumented factors that may have led to thedecision to transfuse a patient to the extent that it isunclear if or to what degree a single Hb value influencedthe decision. Last, as we studied a retrospective cohort, wewere limited to measuring outcome at hospital dischargebecause 6- or 12-month follow-up was not ascertained. Itis clear that recovery from aSAH occurs over months andeven years, and this shorter follow-up may bias towardworse outcome [30, 31].ConclusionsOur study suggests that current practice includes a restrict-ive transfusion strategy and that RBC transfusion in this co-hort of patients was not associated with a worse outcomewhen controlling for other important factors. Rigorousrandomized controlled trials to better understand the roleand timing of RBC transfusion in this patient populationare still needed.Additional fileAdditional file 1: Supplemental methods: expanded definitions.Supplemental data: oral anticoagulant data. Figure S1. Anemia(hemoglobin, ≤100 g/L) rates across centers. Figure S2. RBC transfusionrates across centers. (DOCX 70 kb)AbbreviationsaSAH: Aneurysmal subarachnoid hemorrhage; CT: Computed tomography;DCI: Delayed cerebral ischemia; GCS: Glasgow Coma Scale; Hb: Hemoglobin;ICU: Intensive care unit; mRS: Modified Rankin Scale; PEG: Percutaneousgastrostomy tube; PRBCs: Packed red blood cells; RBC: Red blood cellAcknowledgementsWe acknowledge the significant contributions to this work by our latecolleague and friend Dr. Cheemun Lum. The authors also acknowledgethe tremendous efforts of ICU research coordinators Irene Watpool, DeniseFoster, Marie-Claude Tremblay and Caroline Leger. Thank-you to Drs. V.McCredie and D. Scales from the Canadian Critical Care Trials Group fortheir thorough and critical review of the manuscript.FundingThis work was completed with the support of a competitive peer-reviewedseed grant from the departments of medicine and critical care at the Universityof Ottawa (Ottawa, ON, Canada).Availability of data and materialsThe data that support the findings of this study are available from thecorresponding author upon reasonable request.Authors’ contributionsSWE, LM, and DF came up with the study concept. SWE, LM, DF, MC, AFT, FL,DG, AG, AT, GP, RM, and AB developed the study design and protocol. SWE,MC, AFT, FL, DG, AG, KM, RC, RDC, RZ, CvW, JS and AD collected data. SWE,LM, MC, and RM performed analysis. SWE prepared the first draft of themanuscript. All coauthors provided input and critical review of themanuscript leading to the final version. All authors read and approved thefinal manuscript.Ethics approval and consent to participateThis study was approved by the Ottawa Health Sciences NetworkResearch Ethics Board (no. 20140349-01H) as well as ethics committeeseach participating center, which waived the need for informed consent.Consent for publicationNot applicable.Competing interestsThe authors declare that they have no competing interests.Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Author details1Department of Medicine (Critical Care), The Ottawa Hospital, Civic CampusRoom F202, 1053 Carling Avenue, Ottawa, ON K1Y 4E9, Canada. 2ClinicalEpidemiology Program (Centre for Transfusion Research), Ottawa HospitalResearch Institute, Ottawa, ON, Canada. 3Department of Medicine, Universityof Montreal, Montreal, QC, Canada. 4Department of Anesthesiology & CriticalCare, Université Laval, Quebec City, QC, Canada. 5Department of Medicine,Université Laval, Quebec City, QC, Canada. 6Deparment of Anesthesiology,Pharmacology & Therapeutics, University of British Columbia, Vancouver, BC,Canada. 7Department of Medicine, University of Manitoba, Winnipeg, MB,Canada. 8Department of Medicine, University of Ottawa, Ottawa, ON, Canada.9Department of Medicine, University of Saskatchewan, Regina, SK, Canada.10Department of Surgery (Neurosurgery), University of Ottawa, Ottawa, ON,Canada. 11Department of Social and Preventive Medicine, Université Laval,English et al. Critical Care  (2018) 22:169 Page 8 of 9Quebec City, QC, Canada. 12Department of Surgery, University of Ottawa,Ottawa, ON, Canada.Received: 15 January 2018 Accepted: 7 June 2018References1. Diringer MN, Bleck TP, Claude Hemphill J, Menon D, Shutter L, Vespa P,et al. Critical care management of patients following aneurysmalsubarachnoid hemorrhage: recommendations from the NeurocriticalCare Society’s multidisciplinary consensus conference. Neurocrit Care.2011;15:211–40.2. Bederson JB, Connolly ES, Batjer HH, Dacey RG, Dion JE, Diringer MN, et al.Guidelines for the management of aneurysmal subarachnoid hemorrhage: astatement for healthcare professionals from a special writing group of theStroke Council, American Heart Association. Stroke. 2009;40:994–1025.3. Connolly ES, Rabinstein AA, Carhuapoma JR, Derdeyn CP, Dion J, HigashidaRT, et al. Guidelines for the management of aneurysmal subarachnoidhemorrhage: a guideline for healthcare professionals from the AmericanHeart Association/American Stroke Association. Stroke. 2012;43:1711–37.4. Sampson TR, Dhar R, Diringer MN. Factors associated with the developmentof anemia after subarachnoid hemorrhage. Neurocrit Care. 2010;12:4–9.5. Kramer AH, Gurka MJ, Nathan B, Dumont AS, Kassell NF, Bleck TP.Complications associated with anemia and blood transfusion in patientswith aneurysmal subarachnoid hemorrhage. Crit Care Med. 2008;36:2070–5.6. Le Roux PD. Anemia and transfusion after subarachnoid hemorrhage.Neurocrit Care. 2011;15:342–53.7. Naidech AM, Shaibani A, Garg RK, Duran IM, Liebling SM, Bassin SL, et al.Prospective, randomized trial of higher goal hemoglobin after subarachnoidhemorrhage. Neurocrit Care. 2010;13:313–20.8. Leal-Noval SR, Arellano-Orden V, Muñoz-Gómez M, Cayuela A, Marín-Caballos A, Rincón-Ferrari MD, et al. Red blood cell transfusion guidedby near infrared spectroscopy in neurocritically ill patients withmoderate or severe Anemia: a randomized, controlled trial. JNeurotrauma. 2017;34:2553–9.9. Kramer A. Hemoglobin thresholds for a clinical trial comparing liberaland restrictive transfusion strategies in patients with aneurysmalsubarachnoid hemorrhage: a multidisciplinary North American survey[abstract]. Neurocrit Care. 2010;13(Suppl 1):S207.10. Lacroix J, Hébert PC, Hutchison JS, Hume HA, Tucci M, Ducruet T, et al.Transfusion strategies for patients in pediatric intensive care units. NEngl J Med. 2007;356:1609–19.11. Hébert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, et al. Amulticenter, randomized, controlled clinical trial of transfusion requirements incritical care. Transfusion Requirements in Critical Care Investigators, CanadianCritical Care Trials Group. N Engl J Med. 1999;340:409–17.12. Østbye T, Levy AR, Mayo NE. Hospitalization and case-fatality rates forsubarachnoid hemorrhage in Canada from 1982 through 1991: theCanadian Collaborative Study Group of Stroke Hospitalizations. Stroke.1997;28:793–8.13. Kramer AH, DA Z, Bleck TP, Dumont AS, Kassell NF, Nathan B.Relationship between hemoglobin concentrations and outcomes acrosssubgroups of patients with aneurysmal subarachnoid hemorrhage.Neurocrit Care. 2009;10:157–65.14. Hunt WE, Hess RM. Surgical risk as related to time of intervention in therepair of intracranial aneurysms. J Neurosurg. 1968;28:14–20.15. World Federation of Neurological Surgeons. Report of World Federation ofNeurological Surgeons Committee on a Universal SubarachnoidHemorrhage Grading Scale. J Neurosurg 1988;68:985–986.16. Frontera JA, Claassen J, Schmidt JM, Wartenberg KE, Temes R, Connolly ESJ,et al. Prediction of symptomatic vasospasm after subarachnoid hemorrhage:the modified Fisher scale. Neurosurgery. 2006;58:21–7.17. Carson JL, Terrin ML, Noveck H, Sanders DW, Chaitman BR, Rhoads GG, et al.Liberal or restrictive transfusion in high-risk patients after hip surgery. NEngl J Med. 2011;365:2453–62.18. Murphy GJ, Pike K, Rogers CA, Wordsworth S, Stokes EA, Angelini GD,et al. Liberal or restrictive transfusion after cardiac surgery. N Engl JMed. 2015;372:997–1008.19. Hajjar LA, Vincent JL, Galas FRBG, Nakamura RE, Silva CMP, Santos MH, et al.Transfusion requirements after cardiac surgery: the TRACS randomizedcontrolled trial. JAMA. 2010;304:1559–67.20. Dankbaar JW, Slooter AJ, Rinkel GJ, Schaaf IC. Effect of different componentsof triple-H therapy on cerebral perfusion in patients with aneurysmalsubarachnoid haemorrhage: a systematic review. Crit Care. 2010;14:R23.21. Kramer AH, Diringer MN, Suarez JI, Naidech AM, Macdonald LR, Le Roux PD.Red blood cell transfusion in patients with subarachnoid hemorrhage: amultidisciplinary North American survey. Crit Care. 2011;15:R30.22. Kumar MA, Boland TA, Baiou M, Moussouttas M, Herman JH, Bell RD, et al.Red blood cell transfusion increases the risk of thrombotic events inpatients with subarachnoid hemorrhage. Neurocrit Care. 2014;20:84–90.23. Naidech AM, Drescher J, Ault ML, Shaibani A, Batjer HH, Alberts MJ. Higherhemoglobin is associated with less cerebral infarction, poor outcome, anddeath after subarachnoid hemorrhage. Neurosurgery. 2006;59:775–9.24. Smith MJ, Le Roux PD, Elliott JP, Winn HR. Blood transfusion and increasedrisk for vasospasm and poor outcome after subarachnoid hemorrhage. JNeurosurg. 2004;101:1–7.25. Matsushima K, Eastman A, Shafi S, Burris A, Tyner T, Frankel H. Transfusionincreases infection without affecting neurologic outcome in spontaneoussubarachnoid hemorrhage [abstract]. Crit Care. 2009;13(Suppl 1):P100.26. Festic E, Rabinstein AA, Freeman WD, Mauricio EA, Robinson MT,Mandrekar J, et al. Blood transfusion is an important predictor ofhospital mortality among patients with aneurysmal subarachnoidhemorrhage. Neurocrit Care. 2013;18:209–15.27. Broessner G, Lackner P, Hoefer C, Beer R, Helbok R, Grabmer C, et al.Influence of red blood cell transfusion on mortality and long-termfunctional outcome in 292 patients with spontaneous subarachnoidhemorrhage. Crit Care Med. 2009;37:1886–92.28. Festic E, William F, Rabinstein A, Gajic O. Blood transfusion is a predictor ofhospital mortality in aneurysmal subarachnoid hemorrhage [abstract]. CritCare Med. 2010;38(12 Suppl):A90.29. Festic E, Rabinstei AA, Freema WD, Dupon SA, Robinso MT, Diverti GD, et al.Association of anemia and blood transfusion with the outcome of patientswith aneurysmal subarachnoid hemorrhage [abstract 0390]. Intensive CareMed. 2010;36(Suppl 2):S183.30. Ogden J, Utley T, Mee E. Neurological and psychosocial outcome 4 to 7years after subarachnoid hemorrhage. Neurosurgery. 1997;41:25–34.31. Navi BB, Kamel H, Hemphill JC, Smith WS. Trajectory of functionalrecovery after hospital discharge for subarachnoid hemorrhage.Neurocrit Care. 2012;17:343–7.English et al. Critical Care  (2018) 22:169 Page 9 of 9


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items