Download pdf - Characteristics and outcomes of haematology patients admitted to the intensive care unit

AUDIT

doi: 10.1111/nicc.12005

Characteristics and outcomesof haematology patients admittedto the intensive care unitCaroline McCaughey, Bronagh Blackwood, Marie Glackin, Michele Bradyand Mary Frances McMullin

ABSTRACTAim: To profile the characteristics and outcomes of adult haematology patients admitted to the intensive care unit (ICU).Background: The role of intensive care support for haematology patients is contentious due to high mortality rates thus generating debateregarding the inappropriate use of limited resources versus denial of effective care.Methods: Medical notes, laboratory records and Intensive Care National Audit and Research Centre (ICNARC) data for all adult haematologypatients admitted to Belfast City Hospital ICU in 2009 were analysed.Results: Twenty one patients were admitted to the ICU; mean age was 56 years (SD 12·5), 52% were male and 82% (n=19) had amalignant diagnosis. The main indication for admission was neutropenic sepsis with associated organ impairment (n=18, 85%). ICU mortalitywas 43%. Three-month and six-month mortality rates were 62% and 67%, respectively. ICU survivors had lower acute physiology and chronichealth evaluation (APACHE II) scores, and decreased requirements for invasive ventilation and inotropic support. Of the post-six-month survivors,one had a relapse, one had responding disease and five remained in remission. Two patients have subsequently undergone a reduced intensityconditioning transplant.Conclusion: One third of patients survived for >6 months indicating that critically ill haematology patients can benefit from ICU admission,allowing progression to potentially curative therapies.Relevance to clinical practice: This study highlights the necessity of individualized assessment regarding patient suitability for admissionto a critical care facility, incorporating the perspective of both the haematologist and the intensivist.

Key words: Critical care • Haematology • Intensive care • Long-term outcome • Short-term outcome • Survival

INTRODUCTIONOverall survival for patients with haematologicalmalignancies is improving (Cuthbertson et al., 2008),with survival rates for some malignancies nowexceeding 80% (Gates et al., 2009). The progressin survival rates has been achieved by the useof intensified treatment protocols and advances in

Authors: C McCaughey, Practice Educator, Belfast Health and SocialCare Trust and School of Nursing and Midwifery, Queen’s UniversityBelfast, Belfast, UK; B Blackwood, Lecturer, School of Nursing andMidwifery, Queen’s University Belfast, Belfast, UK; M Glackin, NurseLecturer, School of Nursing and Midwifery, Queen’s University Belfast,Belfast, UK; MF McMullin, Professor of Haematology, Belfast Health andSocial Care Trust, Belfast, UK, Centre for Cancer Research and Cell Biology,Queen’s University Belfast, Belfast, UK; M Brady, Consultant Anaesthetist,Belfast Health and Social Care Trust, Belfast, UKAddress for correspondence: C McCaughey, Practice Educator,Belfast Health and Social Care Trust and School of Nursing and Midwifery,Queen’s University Belfast, Belfast, UK.E-mail: [email protected]

supportive care. However, these improvements haveled to an increased occurrence of therapy-relatedcomplications in the immune-compromised patients,necessitating emergency admission to an intensive careunit (ICU) (Staudinger et al., 2000; Benoit et al., 2003;Cuthbertson et al., 2008).

In the past two decades, mortality rates forhaematology patients admitted to ICU in an emergencyranged from 54% to 98%; mortality is higher ifpatients require mechanical ventilation or renalreplacement therapy or if they develop multi-organ failure (Maschmeyer et al., 2003; McGrathet al., 2010; Soares et al., 2010). Furthermore, post-allogeneic transplantation patients are particularly atrisk (Maschmeyer et al., 2003). However, over the pastfew years several centres have reported improvingsurvival rates in haematology patients admitted toICU despite increased intensity of treatments (Soareset al., 2005, Cuthbertson et al., 2008, McGrath et al.,2010). Indeed, the largest data set study in this field

© 2013 The Authors. Nursing in Critical Care © 2013 British Association of Critical Care Nurses • Vol 18 No 4 193

Characteristics and outcomes of haematology patients

(Hampshire et al., 2009) report an ICU mortality rateof 43·1%. Nevertheless, controversy still surrounds thebenefit versus futility of providing critical care forthis patient group, with an ongoing preconception thatICU resources for cancer patients may be less justifiable(Hill, 2010; Soares et al., 2010).

This study was conducted to profile patientcharacteristics and assess outcomes of haematologypatients admitted to the Belfast City Hospital ICUin order to establish a baseline, to identify areas forimprovement so that resources can be utilised mosteffectively and to compare outcomes against nationalstatistics.

METHODSSettingThe Belfast City Hospital haematology service isa regional tertiary referral centre for NorthernIreland and acts as a haematology unit to its localpopulation. It has a 29 bedded in-patient unit,incorporating six transplant beds designated for theprovision of autografts and sibling allografts. Inaddition to the management of patients with ahaematological malignancy, the service covers non-malignant haematological disorders and incorporatesa haemostasis and thrombosis comprehensive carecentre. In 2009, there were 829 admissions to the in-patient haematology unit.

The ICU had 12 beds up until September 2009,which was then reduced to 9 as part of the Belfast,Health and Social Care Trust (HSCT) reconfiguration.The ICU is staffed by a full-time team of intensivists(n = 8) with 24-h multidisciplinary support, offeringlevel 3 critical care, including invasive mechanicalventilation. In 2009, there were 660 admissionsto the ICU.

DesignRetrospective data were collected from medical notes,laboratory records and the Intensive Care NationalAudit and Research Centre (ICNARC) database for alladult haematology patients admitted to the ICU from1 January to 31 December 2009.

Data collectionData were collected on the following availablevariables: patient characteristics; APACHE II scores;requirement for invasive mechanical ventilation, renalreplacement therapy and inotropic support; ICU lengthof stay; mortality (ICU, 3 and 6 months); neutropaenia;pancytopaenia and liver toxicity.

Data analysisData were analysed using the SPSS software (SPSS,version 17.0). Descriptive statistics were used andresults were reported as mean ± SD or median andIQR.

ETHICAL APPROVALA sub-panel of the Northern Ireland Research EthicsCommittee reviewed the protocol. Approval was notrequired, and the study was registered with the BelfastHSCT Audit Department.

RESULTSPatient characteristicsFrom 1 January 2009 to 31 December 2009, therewere 829 admissions to the in-patient haematologyunit. Twenty-one adult patients were admitted tothe ICU. This constituted 2·5% of all haematologyadmissions in that year and 3·2% of all ICU admissions.Demographics, baseline data and ICU mortality arepresented in Table 1.

Eighty two percent (n = 19) had a malignantdiagnosis, with the majority having an acuteleukaemia (n = 8, 38%) or an aggressive lymphoma(n = 6, 28·5%). Table 2 categorizes the haematologydiagnoses.

Treatment historyBefore the haematology admission during which theywere transferred to ICU, 38% (n = 8) of patients hadnot been previously treated with chemotherapy: twohad a non-malignant diagnosis and six were newlydiagnosed with a haematological malignancy. Sevenpatients had been pre-treated with chemotherapy,four patients had a previous autograft, one hada previous sibling myeloablative allograft and onehad a previous sibling reduced intensity conditioning(RIC) allograft. At the point of ICU admission, 71%(n = 15) were post-chemotherapy patients (median9·9 days, range 1–22 days) and one was at day 4 postautograft.

At the time of ICU admission, nine patients (42·9%)were in remission or had stable disease, six (28·6%)had newly diagnosed malignant disease, two (9·5%)had newly diagnosed non-malignant disease and one(4·8%) had active, relapsed (n = 2, 9·5%) or refractorydisease (n = 1, 4·8%).

Reason for ICU admissionThe most prevalent reason for emergency admissionto the ICU was neutropenic sepsis and associated

194 © 2013 The Authors. Nursing in Critical Care © 2013 British Association of Critical Care Nurses


Table 1 Demographics and outcomes of patients admitted to ICU in 2009

Total ICU admissionsin 2009

Haematology admissionsin 2009

Admissions 660 21Male 395 (35·8%) 11 (52%)Age (mean) 60 56APACHE II 12 23ICU length of stay 4 days 4 daysICU mortality 11·5% 43%

Table 2 ICU mortality according to haematology diagnoses

Number admittedto ICU ICU mortality (%)

Acute leukaemia N = 8 N = 6 (75)Chronic leukaemia N = 1 N = 0 (0)High-risk myleodysplastic syndrome N = 2 N = 0 (0)Lymphoma N = 6 N = 3 (50)Myeloma N = 2 N = 0 (0)Non-malignant N = 2 N = 2 (100)Total 21

organ impairment (85%, n = 18), with respiratoryfailure being the most common complication of sepsis.Two patients had multi-organ failure at the point ofadmission, with neither surviving. In patients with anon-malignant diagnosis (n = 2), reasons for admissionwere acute haemolysis (Hb 2·8 on presentation) leadingto loss of consciousness, and seizures secondary tothrombotic thrombocytopaenia purpura.

Characteristics and outcomes of survivorsand non-survivorsICU mortality was 43% (n = 9), and 89% (n = 8) of thesedeaths were sepsis related. To contextualize this figure,throughout the same time period the crude mortalityrate for the Belfast HSCT was 2%. Non-survivors hadhigher admission APACHE II scores, and were moreinclined to have grade 3 or more bone marrow andliver toxicity (using the NCI Common TerminologyCriteria for Adverse Events (CTCAE) Version 4.02). Agreater percentage of non-survivors required invasiveventilation and inotropic support. Both survivors andnon-survivors had similar requirements for continuousrenal replacement therapy and had similar lengths ofstay in the ICU (see Table 3)

The 3-month mortality rate was 62%. Three ofthe four deaths within this time period occurredwithin 2 days of ICU discharge. The 6-month mortalityrate was 67%. Seven patients were alive more thansix months: one had relapsed disease; one had

Table 3 Characteristics and support requirements of survivors andnon-survivors

Survivors(n= 12)

Non-survivors(n= 9)

APACHE II score, (mean, SD) 20 (5) 28 (6·6)Neutropaenia≥ grade 3 (n%) 6 (50%) 6 (66·7%)Pancytopaenia≥ grade 3 (n %) 6 (50%) 6 (66·7%)Liver toxicity≥ grade 3 (n %) 3 (25%) 7 (78%)Invasive ventilation (n %) 6 (50%) 8 (90%)Inotropic support (n %) 3 (25%) 7 (78%)Continuous renal replacement therapy 4 (33%) 3 (33%)Maximum no. of organs supported (mean, SD) 2·25 (1·1) 3 (0·7)Days in ICU (median, IQR) 4·3 (2, 8) 4 (1, 7)

Table 4 Disease status of patients admitted to ICU

Haematological status Frequency Percentage

Newly diagnosed malignant disease n= 6 28.6Newly diagnosed non-malignant disease n= 2 9·5Remission/stable disease n= 9 42·9Active disease n= 1 4·8Relapsed disease n= 2 9·5Refractory disease n= 1 4·8Total n= 21 100

responding disease; and five were in remission, twoof whom have progressed to a potentially curativeRIC transplant.

Haematological status and outcomesOwing to the controversies associated with access tocritical care facilities for patients with malignant dis-ease, disease status and outcome were recorded. Whilstthe patient with active disease survived >6 months,those with relapsed or refractory disease had a maxi-mum survival of 112 days (Table 4).

DISCUSSIONMany studies have examined the characteristicsand outcomes of patients with solid tumours andhaematological malignancies in the ICU context(Staudinger et al., 2000; Benoit et al., 2003; Maschmeyeret al., 2003; Gruson et al., 2004; Darmon et al., 2005;Thiery et al., 2005; Ferra et al., 2007; McGrath et al.,2010; Soares et al., 2010; Schellongowski et al., 2011).This study is unique in that it addresses the specialityof haematology, incorporating both malignant andnon-malignant conditions; however, as there wereonly two patients with a non-malignant diagnosisadmitted to ICU in the audit timeframe, a review

© 2013 The Authors. Nursing in Critical Care © 2013 British Association of Critical Care Nurses 195


spanning a longer time period would have been moreinsightful. Our study showed a trend towards betteroutcomes for patients with active malignant diseasethan non-malignant disease and thus tentativelysupports the view that the severity of the acute illnessrather than malignancy-specific parameters determinepatient outcome from ICU (Staudinger et al., 2000;Massion et al., 2002; Maschmeyer et al., 2003; Darmonet al., 2005; Hill, 2010; Soares et al., 2010; Schellongowskiet al., 2011).

Our most common reason for ICU admission wasacute respiratory failure (secondary to sepsis), whichwas similarly reported by Ferra et al. (2007), Limet al. (2006), Soubani (2006), Azoulay et al. (2010) andSchellongowski et al. (2011). Additionally, 24% of ourpopulation were chemotherapy naive, emphasizingthe impact of disease-related as well as treatment-related acute complications in the haemato-oncologycontext, also reflected in Thiery et al. (2005) andSchellongowski et al.’s (2011) experience. Molina et al.(2012) conducted a large prospective, multicentreobservational study to clarify the role of differentventilatory strategies, given the high mortality ratesassociated with intubation. They found that whilst non-invasive mechanical ventilation (NIMV) can improveoutcomes in haematology patients admitted to the ICU,patients with NIMV failure experience a more severerespiratory impairment than those electively intubated.Their results suggest that diseases that may have a fastresponse to therapy, such as diuretics and inotropes forcardiogenic pulmonary oedema, or directed antibiotictherapy for a documented infection may benefit fromNIMV. However, in other causes of lung injury, NIMVmay not be effective for a prolonged time, thusincreasing the risk of failure and intubation. Molinaet al. therefore concluded that it appears critical tocarefully select candidates for NIMV, identifying thosewith rapidly reversible causes of respiratory failure, asin other cases, the delay in optimal respiratory supportwith invasive mechanical ventilation may increasethe risk of death. Furthermore, a randomized trialconducted by Wermke et al. (2012), recruiting a cohortof allogeneic transplant recipients, found an earlyinterventional strategy using NIMV (versus oxygenonly) was not associated with improved prognosis.However, it is acknowledged that the limited influenceof NIMV may be related to their modest samplesize and cross-over study design. They would concurwith Molina et al. that whilst early application ofNIMV may be of help, it must be combined withother modern treatment modalities for respiratoryfailure, as it is a complex process involving toxic,infectious and inflammatory mechanisms, potentiallyexacerbated by individual susceptibility factors; thus

early risk assessment and identification of respiratorydysfunction is vital.

This study demonstrates comparable outcomes tothose reported in Hill’s (2010) review of 27–57%hospital mortality in patients with haematologicalmalignancy and 20–61% in haematopoietic transplantpatients. Our ICU mortality rate of 43% is identical toan analysis of ICNARC data for patients with haema-tological malignancy admitted to UK critical care units1995–2007 (Hampshire et al., 2009). However, thisshould be interpreted in the context that ICNARCdata will incorporate haematology units associatedwith district general hospitals, which may not reflectthe same profile of complexity of haematology casesmanaged in a regional tertiary referral centre such asthe setting for our analysis.

A recent retrospective and prospective study ofn = 115 haematology patients admitted for critical carereported a mortality of 30·4%, which is significantlylower than previously reported in similar studies(McGrath et al., 2010). It is noted, however, when 6-month mortality rates are examined, these are 73%as compared with 67% found in our review. Theparticipants in the study also had fewer failed organs(median number of failed organs for survivors was1 and for non-survivors 2), whereas our analysisreported 2·25 and 3, respectively, perhaps indicatinga more critically ill cohort of patients; however,it is recognized that McGrath et al. used a largersample.

Several poor prognostic indicators for oncol-ogy/haematology patients have been suggested, whichare consistent with the finding of this review. A mul-ticentre prospective study of cancer patients (7% ofwhom had a haematological malignancy) by Soareset al. (2010) also concluded that the severity of organfailures and the need for mechanical ventilation werepredictors of mortality. An earlier observational studyby Massion et al. (2002) over a 10-year period alsoidentified these factors in a population of exclusivelyhaemato-oncology patients, furthermore identifyingbone marrow transplant and fungal infections asadverse predictors of outcome. The risk factor offungal infection was replicated in a case review of189 oncology and haematology cases requiring ICUtreatment, which concluded that given the high mor-tality associated with invasive fungal infection, itmay be justified to decide against ICU treatmentin patients with this condition who also requirevasopressor support and mechanical ventilation(Maschmeyer et al., 2003). Despite collecting data oncultured organisms in our study, no pattern emergedwith regard to a particular organism and patientoutcome.



The poor prognostic indictor of bone marrowtransplantation (BMT) is also reported by Staudingeret al. (2000) in a retrospective single centre study.Conversely, in a more recent study Lim et al. (2006)found no significant difference in overall survival inpatients receiving intensive chemotherapy and thosewho underwent allogeneic transplant. Both patients inour analysis who had undergone allogeneic transplantdied in the ICU, whereas the post autograft patientsurvived the episode.

Our analysis adds credence to the findings by Soareset al. (2010) and McGrath et al. (2010) that patientperformance status on admission to ICU influencesmortality rates, given the higher median APACHEII scores of non-survivors. There is some debate asto the value of prognostic models, with Afessa et al.(2003) finding the APACHE II tool to have moderatediscrimination for predicting hospital mortality inhaematopoietic stem cell transplant recipients, whilstBenoit et al. (2003) would caution that the APACHEII score lacks sensitivity in the haematologicalcontext. Indeed, systematic review suggests that ICUphysicians can discriminate more effectively betweensurvivors and non-survivors in the first 24 h thanscoring systems (Sinuff et al., 2006). Our experiencewould be in accordance with Thiery et al. (2005)who stress the importance of individualized patientselection and flexible admission criteria, advocatingcase by case decisions given the excess mortality in‘too well for ICU’ patients and the relatively goodsurvival in the ‘too sick for ICU’ cohort of their study.

Both Benoit et al. (2003) and, more recently,Schellongowski et al. (2011) found the need forvasopressor therapy as an adverse predictor ofoutcome, a finding suggested in our study as agreater percentage of non-survivors required inatropicsupport. Surprisingly, the later researchers did notidentify mechanical ventilation as a poor predictor ofoutcome in their population of patients with de novoacute myeloid leukaemia (AML).

The impact of neutropaenia in critical care cancerpatients remains controversial (Hill, 2010), and whilstnon-survivors in our study were inclined to have moreprofound bone marrow suppression, robust studiesby Staudinger et al. (2000), Ferra et al. (2007) andSchellongowski et al. (2011) conclude that this factoris not necessarily correlated with an adverse outcome.Our study found that the renal replacement therapyrequirements between survivors and non-survivorswere the same, concurring with Maschmeyer et al.’s(2003) conclusion that the need for renal replacementin itself is not an independent risk factor formortality in patients with haematological malignancy.Furthermore, the poor prognosis associated with

hyperbilirubinaemia (Lim et al., 2006; Pene et al.,2006) is mirrored in our analysis, with a greaterproportion of non-survivors than survivors havinggrade 3 liver toxicity (25% of survivors, 78% ofnon-survivors).

As reported by other authors (Maschmeyer et al.,2003; Darmon et al., 2005; Thiery et al., 2005; Limet al., 2006) our study adds weight to the dismaloutcome associated with multi-organ failure at thetime of ICU admission and underscores the necessityof individualized assessment of suitability for criticalcare intervention.

CONCLUSIONThe most common reason for admission to ICUwas neutropaenic sepsis-associated respiratory failureand median length of stay was 4 days (SD 4·7).One third of patients survived >6 months, in somecases allowing progression to potentially curativetherapies, thus adding to the growing body of evidencewhich challenges the preconception that patients withmalignancy are unsuitable for ICU admission. Thebenefit of ICU for patients with relapsed or refractorydisease however remains ambiguous.

The study has established that our outcomes arealigned with national statistics. It mirrors the findingsof previous studies that factors of poor performancestatus, liver impairment, invasive mechanical venti-lation requirement, need for inotropic support andmulti-organ failure contribute to increased risk of mor-tality for haematology/oncology patients requiringemergency critical care support. Although the sam-ple size is relatively small, analysis is retrospectiveand descriptive, and findings are subject to singlecentre bias, they are strengthened as the review wasconducted in a regional tertiary referral haematologycentre, which is supported by a level 3 multidisci-plinary ICU dealing with much greater acuity thanhigh-dependency units associated with some other cen-tres. Additionally, it encompassed haematology as awhole rather than solely haemato-oncology. This studyhighlights the necessity of individualized assessmentregarding patient suitability for admission to a criticalcare facility, incorporating the perspective of both thehaematologist and intensivist. Indeed Ferra et al. (2007)attribute improving outcomes in their institution to bein part due to closer collaboration between haematolo-gists and intensive care specialists. It is established thatcharacteristics of the underlying disease have consis-tently failed to predict long-term survival Hill’s (2010);thus the reversibility of the acute illness should be theprimary consideration, with the obvious recognitionthat in some cases escalating care is inappropriate.



WHAT IS KNOWN ABOUT THE TOPIC

• Intensified treatment protocols and advances in supportive care have led to increased overall survival for patients with haematologicalmalignancies.

• These improvements in survival have led to an increased occurrence of therapy-related complications in the immune-compromised patientsnecessitating emergency admission to an ICU.

• Mortality is higher if patients require mechanical ventilation or renal replacement therapy, or develop multi-organ failure.• Reversibility of the acute illness should be the primary consideration, with the obvious recognition that in some cases escalating care is

inappropriate.

WHAT THIS PAPER ADDS

• This paper adds to the growing body of evidence which challenges the preconception that patients with malignancy are unsuitable for ICUadmission.

• The benefit of ICU for patients with relapsed or refractory disease however remains ambiguous.• This study encompassed haematology as a whole rather than solely haemato-oncology.• This study highlights the necessity of individualized assessment regarding patient suitability for admission to a critical care facility,

REFERENCESAfessa B, Tefferi A, Dunn WF, Litzow MR, Peters SG. (2003).

Intensive care unit support and acute physiology and chronichealth evaluation 111 performance in hematopoietic stem celltransplant recipients. Critical Care Medicine; 31: 1715–1721.

Azoulay E, Mokart D, Lambert J, Lemiale V, Rabbat A, KouatchetA, Vincent F, Gruson D, Bruneel F, Epinette-Branche G,Lafabrie A, Hamidfar-Roy R, Cracco C, Renard B, TonnelierJ-M, Blot F, Chevret S, Schlemmer B. (2010). Diagnosticstrategy for hematology and oncology patients with acuterespiratory failure. Randomised controlled trial. AmericanJournal of Respiratory and Critical Care Medicine; 182: 1038–1046.

Benoit DD, Vandewoude KH, Decruyenaere JM, Hoste EA,Colardyn FA. (2003). Outcomes and early prognosticindicators in patients with a hematologic malignancy admittedto the intensive care unit for a life-threatening complication.Critical Care Medicine; 31: 104–112.

Cuthbertson BH, Rajalingam Y, Harrison S, McKirdy F, on behalfof the Scottish Intensive Care Society. (2008). The outcomeof haematological malignancy in Scottish intensive care units.Journal of the Intensive Care Society; 9: 135–140.

Darmon M, Thiery G, Ciroldi M, De Miranda S, Galicier L, RaffouxE, LeGall J-R, Schlemmer B, Azoulay E. (2005). Intensive carein patients with newly diagnosed malignancies and a need forcancer chemotherapy. Critical Care Medicine; 33: 2488–2493.

Ferra C, Marcos P, Misis M, Morgades M, Bordeje M-L, OriolA, Lloveras N, Sancho J-M, Xicoy B, Batlle M, Klamburg J,Feliu E, Ribera J-M. (2007). Outcome and prognostic factorsin patients with hematologiv malignancies admitted to theintensive care unit: A single centre experience. InternationalJournal of Hematology; 85: 195–202.

Gates P, Krishnasamy M, Wheeler G, Seymour JF, Schembri S.(2009). The development of a haematology late effects nurseconsultant role with a national scope. European Group forBlood and Marrow Transplantation – Nurses Group Newslet-ter; 8: www.ebmt.org/6nursesgroup/october09/ebmt-ng-lateeffects.html (accessed on 14/07/11).

Gruson, D., Vargas, F., Hilbert, G., Bui, N., Maillot. T., Mayet, T.,Pillet, O., Chene, G. Gbikpi-Benissan, G. (2004) Predictivefactors of intensive care unit admission in patients with

haematological malignancies and pneumonia. Intensive CareMedicine, 30, 965971

Hampshire PA, Welch CA, McCrossan LA, Francis K, HarrisonDA. (2009). Admission factors associated with hospital mor-tality in patients with haematological malignancy admitted toUK adult, general critical care units: a secondary analysis ofthe ICNARC case mix programme database. Critical Care; 13:R137.

Hill QA. (2010). Intensify, resuscitate or palliate: decision makingin the critically ill patient with haematological malignancy.Blood Reviews; 24: 17–25.

Lim Z, Pagliuca A, Simpson S, Cottam S, Ervine M, Ho AYL,Devereux S, Mufti G. (2006). Outcomes of patients withhaematological malignancies admitted to intensive care unit.A comparative review of allogeneic haematopoietic stemcell transplantation data. British Journal of Haematology; 136:448–450.

Maschmeyer G, Bertschat FL, Moesta KT, Hausler E, HeldTK, Nolte M, Osterziel KJ, Papstein V, Peters M, ReichG, Scmutzler M, Sezer O, Stula M, Wauer H, Wortz T,Wischnewsky M, Hohenberger P. (2003). Outcome analysisof 189 consecutive cancer patients referred to the intensivecare unit as emergencies during a two year period. EuropeanJournal of Cancer; 39: 783–792.

Massion PB, Dive AM, Doyen C, Bulpa P, Jamart J, Bosly A,Installe E. (2002). Prognosis of hematologic malignancies doesnot predict intensive care mortality. Critical Care Medicine; 30:2260–2270.

McGrath A, Chatterjee F, Whiteley C, Ostermann M. (2010). ICUand 6-month outcome of oncology patients in the intensivecare unit. Quarterly Journal of Medicine; 103: 397–403.

Molina R, Bernal T, Borges M, Zaragoza R, Bonastre J, GranadaRM, Rodriguez-Borregan JC, Nunez K, Seijas I, AyestaranI, Albaiceta GM. (2012). Ventilatory support in critically illhaematological patients with respiratory failure. Critical Care;16: 1–28.

Pene F, Aubron C, Azoulay E, Blot F, Thiery G, Raynard B,Schlemmer B, Nitenberg G, Buzyn A, Arnaud P, Socie G, MiraJP. (2006). Outcome of critically ill allogeneic hematopoieticstem cell transplant recipients: a reappraisal of indications



for organ failure supports. Journal of Clinical Oncology; 24:643–649.

Schellongowski P, Staudinger T, Kundi M, Laczika K, LockerGJ, Bojic A, Robak O, Fuhrmann V, Jager U, Valent P, SperrWR. (2011). Prognostic factors for intensive care admission,intensive care outcome, and post-intensive care survival inpatients with de novo acute myeloid leukemia: a single centerexperience. Haematologica; 96: 231–237.

Sinuff T, Adhikari NKJ, Cook DJ, Schunemann HJ, GriffithLE, Rocker LE, Walter SD. (2006). Mortality predictions inthe intensive care unit: comparing physicians with scoringsystems. Critical Care Medicine; 34: 878–885.

Soubani AO. (2006). Critical care considerations of hematopoieticstem cell transplantation. Critical Care Medicine; 34: S251–S267.

Soares M, Caruso P, Silva E, Teles JMM, Lobo SMA, Friedman G,Dal Pizzol F, Mello PVC, Bozza FA, Silva UVA, Torelly AP,Knibel MF, Rezende E, Netto JJ, Piras C, Castro A, FerreiraBS, Rea-Neto A, Olmedo PB, Salluh JIF. (2010). Characteristicsand outcomes of patients with cancer requiring admission tointensive care units: A prospective multicentre study. CriticalCare Medicine; 38: 9–14.

Soares M, Salluh JIF, Spector N, Rocco JR. (2005). Characteristicsand outcomes of cancer patients requiring mechanicalventilation support > 24 hours. Critical Care Medicine; 33:520–526.

Staudinger T, Stoiser B, Mullner M, Locker GJ, Laczika K, KnappS, Burgmann H, Wilfing A, Kofler J, Thalhammer F, FrassM. (2000). Outcome and prognostic factors in critically illcancer patient admitted to the intensive care unit. Critical CareMedicine; 28: 1322–1328.

Thiery G, Azoulay E, Darmon M, Ciroldi M, De Miranda S, Levy V,Fieux F, Moreau D, Le Gall JR, Schlemmer B. (2005). Outcomeof cancer patients considered for intensive care unit admission:a hospital wide prospective study. Journal of Clinical Oncology;23: 4406–4413.

Wermke M, Schiemanck S, Hoffken G, Ehninger G, BornhauserM, Illmer T. (2012). Respiratory failure in patients undergoingallogeneic hematopoietic SCT – a randomised trial on earlynon-invasive ventilation based on standard care hematologywards. Bone Marrow Transplantation; 47: 574–580.
