ORIGINALARTICLEContinuouspositiveairwaypressureventilationversusBi-levelpositiveairwaypressureventilationinpatientswithbluntchesttraumaR.EmanShebla,*,SaadRabieSamraa,MagidM.Abderaboha,MohammadS.MousabaDepartmentofChestDiseases,FacultyofMedicine,ZagazigUniversity,EgyptbDepartmentofAnesthiology,FacultyofMedicine,ZagazigUniversity,EgyptReceived17July2014;accepted17November2014Availableonline29December2014KEYWORDSChesttrauma;Noninvasivepositivepressureventilation;CPAP;BiPAPAbstract Introduction: The use of positive pressure ventilation has decreased the overall morbid-ityand mortalityassociatedwithblunt chesttrauma,but invasivemechanicalventilation(IMV)isassociated with many complications. The role of noninvasive ventilation (NIV) for the managementof patients with blunt chest trauma has not been well established. The aim of this study was to com-paretheefciencyofCPAPversusBiPAPinavoidingIMV.Patientsandmethod: Thisstudywascarriedout intheperiodbetweenApril 2011andApril2103,on40patientsadmittedtoICUwithbluntchesttraumawithacuterespiratorydistressthathaddeteriorated despite aggressive medical management. Patients were randomly assigned toreceiveeithercontinuouspositiveairwaypressureventilation(CPAP) (group1) n = 15, Bi-levelpositiveairwaypressureventilation(BiPAP)(group2)n = 15orIMV(group3)n = 10.Results: Improvement in gas exchange and relieve of respiratory distress was noticed in the threestudiedgroupsafterthestartofassistedventilation. Fourpatientsingroup1(26.7%)andthreepatientsingroup2(20%) requiredendotracheal intubation. Therewasnosignicant differenceinthelengthofstayinICUbetweenthethreegroups(10 5 daysingroup1, 11 4ingroup2 and 10 6 in group 3.Pneumoniadeveloped in onepatientin group 1 (6.6%) and in2 patientsin group 2 (13.3%) and in 3 patients in group 3 (30.3%). Pneumothorax developed in one patient ingroup 1 (6.6%) and in no patients in group 2 (0%) and in one patient in group 3 (10%). As regardsmortalitynomortalitieswereobservedingroups1and2butonepatientingroup3(10%)died.Conclusion: BothCPAPandBiPAParesafeandefcient techniquesinmanagingrespiratoryfailureandreducingtheincidenceofintubationinpatientswithbluntchesttrauma.2014TheEgyptianSocietyofChestDiseasesandTuberculosis.ProductionandhostingbyElsevierB.V.Allrightsreserved.*Correspondingauthor.Tel.:+2001125520503.E-mailaddress:emanshebl3000@yahoo.com(R.EmanShebl).Peer reviewunder responsibility of The EgyptianSocietyof ChestDiseasesandTuberculosis.EgyptianJournalofChestDiseasesandTuberculosis(2015)64,203208HOSTEDBYTheEgyptianSocietyofChestDiseasesandTuberculosisEgyptianJournalofChestDiseasesandTuberculosiswww.elsevier.com/locate/ejcdtwww.sciencedirect.comhttp://dx.doi.org/10.1016/j.ejcdt.2014.11.0160422-7638 2014TheEgyptianSocietyofChestDiseasesandTuberculosis.ProductionandhostingbyElsevierB.V.Allrightsreserved.IntroductionChesttraumaisoneimportantfactorfortotalmorbidityandmortalityintraumatizedemergencypatients. Thelethalityofisolated chest traumas is about 58%. Up to 25% of all deathscaused by trauma are related to chest injuries [1], and mortalitydramaticallyincreasesasafunctionofincreasedchesttraumaforce[2].Chest injuries often occur in combination with other severeinjuries, such as extremity, head, brain and abdominal injuries[1]. Theimpact of ablunt traumais typicallyconductedtomany different intrathoracic structures; hence nearly all organsof thethoraciccavitycanbeinvolvedinchest trauma. Themostcommontypesofdamagethatresultfromchesttraumaincludeinjuriestotheribs, lungcontusion, hematomaofthechestwall, pleural effusion, pneumothoraxandhaemothorax[3].PathophysiologicalaspectsRespiratoryimpairment: damagetotheosseousstructureofthe thoraxbyribandsternumfractures destabilizes the ribcageandimpairs spontaneous breathingmechanics substan-tially; this condition is amplied by pain, which further reducesbreathing function. Direct traumatic damage to the lung leadsto an extravasation of protein-rich uid with an altered surfac-tant composition [4]. Disturbance of diffusion, the reduction ofcomplianceandfunctional residual capacity, ventilationper-fusionmismatchandintrapulmonaryshuntdevelopwithsub-sequentreducedoxygenationandelevatedPaCO2levels[5,6].Afterseverechesttrauma, intrapulmonaryshuntingcanalsobe caused by a disruption of pulmonary capillaries and extrav-asationintothealveolarspaces. Aspirationof bloodand/orgastriccontents, fat embolismtothelungduetolongbonefracturesandsystemicinammatoryresponsesyndromemayadditionallyexacerbate respiratorydecits andmayleadtoacuterespiratorydistresssyndrome(ARDS)[7].Cardiovascular impairment: a reduction in normalintraventricular llingbytensionpneumothorax, pericardialtamponade or massive hemorrhage may result in alife-threatening reduction in cardiac output. Moreover,intracardiac structural damage or heart contusions withconcomitant arrhythmias are additional contributors toreducedcardiacoutput[6].Management of patients with blunt chest trauma focuses oninterventionssuchasthestabilizationoffractures,pulmonarytoilet, effective physiotherapy, andearlyandadequate paincontrol [8,9]. These patients are at highriskfor developingrespiratoryfailure[10] withreportsofupto20%ofpatientswithblunt chest traumadevelopingacutelunginjury(ALI)or acute respiratory distress syndrome (ARDS) [8]. Intubationratesrangefrom23%to75%anddependontheseverityofthe trauma, the degree of the underlying lungdisease, andthe intensity of initial management and monitoring [8,11].The use of positive pressure ventilation has decreased the over-all morbidity and mortality associated with blunt chesttrauma, butendotracheal intubationandmechanical ventila-tionareassociatedwithahighriskofnosocomialpneumoniaandprolongedmechanical ventilation[12].Theroleofnonin-vasiveventilation(NIV)forthemanagementofpatientswithblunt chest trauma has not beenwell established[13]. Theaimof this studywas tocompareNIV(CPAPandBiPAP)withinvasive mechanical ventilation(IMV) inmanagementofpatientswithbluntchesttraumaandtocompareefciencyofCPAPversusBiPAPinavoidingintubationandIMV.PatientsandmethodThisclinicalstudywascarriedouton40patientsadmittedtointensive care unit withblunt chest trauma (either isolatedchest trauma or as a part of polytrauma) in the periodbetween April 2011 and April 2013. The inclusion criteriawereacute respiratorydistress that haddeteriorateddespiteaggressivemedical management, includingseveredyspneaatrest, arespiratoryrate greater than35breaths per minute;and the partial pressure of arterial oxygen (PaO2)less60 mmHgwhilethepatient wasbreathingoxygenthroughaVenturi maskwithFiO2upto60%; andactivecontractionof the accessory muscles of respiration or paradoxicalabdominal motion.Patientswithanyofthefollowingwereexcluded: trachealintubationindicatedfor any other reason, contraindicationfor non-invasive ventilation (active gastro-intestinal hemor-rhage, lowlevel of consciousness, multiorganfailure, airwaycontrol problems, hemodynamicinstability), traumaticbraininjury, facial trauma, skull basefracture, orbitbasefracture,cervical injury with specic treatment contraindicating a facialmask[10].Allpatientsweresubjectedto-Completemedicalhistory.-Clinicalexamination.-Laboratory investigations (renal and hepatic function tests,serumelectrolytes, blood sugar ,complete blood count,arterial bloodgasanalysis, andmicrobiological investiga-tionswhenpneumoniawassuspected).-Radiological investigations (plain Xray and computedtomography on the chest for all patients and for other bodypartsasindicated).-TheInjurySeverityScale(ISS):wasevaluatedasthemea-sureofanatomicinjuryforsixbodyregions:(1)thehead-neck, (2) theface, (3) thethorax, (4) theabdomen-pelvis,(5)theextremitiesand(6)theexternal.TheISSwascalcu-latedasthesumofthesquaresofthehighestabbreviatedinjury scale grade in each of the three most severely injuredbodyregions[14].-Simpliedacutephysiologicscore(SAPS) wascalculated,thisscoretakesintoaccount14variables(age, heartrate,systolicbloodpressure,bodytemperature,respiratoryrateorneedforventilatorysupport, urinaryoutput, white-cellcount, hematocrit, Glasgowcomascore, andserumglu-cose, potassium, sodium, bicarbonate, andureanitrogenconcentrations). Arangeof04isassignedforeachvari-able (range of possible scores, 056). Higher scores indicateahigherriskofdeath[15].Patients were randomized to receive CPAP (group 1)n = 15 (11 males, 4 females with mean age 31.8 13.8),BiPAP(group2)n = 15(10males, 5femaleswithmeanage31.8 13.1), andpatientswhometinclusioncriteriabutdidnot show cooperation received IMV (group 3) n = 10 (7 males,3femaleswithmeanage30.6 12.7).204 R.EmanShebletal.MethodsofmechanicalventilationBiPAPThepatients wereconnectedtoaventilator (BiPAPVision,Respironics Inc., Murrysville, Pa., USA). Initial ventilator set-tings were: inspiratory positive airway pressure (IPAP) was setat 8 cmH2O, positive end-expiratorypressure at 5 cmH2OandFiO2at 100%. Then, IPAPandEPAPweretitratedtoreachthe clinical targets of respiratoryrate (RR) less than25 breaths/min and tidal volume greater than or equal to8 mL/kgfor IPAPandSpo2greater thanor equal to90%on FiO2 < 60% for EPAP titration, while minimizingpatientsintoleranceandleaksaroundthemask[10].NIVintheCPAPmodeUsing Tranquility, Healthdyne, USA, pressure was initially setto 3 cm H2O for 5 min then titrated according to patients tol-eranceandcomfortandclinicalmonitoring[10].ConventionalinvasiveventilationPatientswereintubatedandconnectedtoventilators(PuritanBennett 7200 (PuritanBennett, OverlandPark, Kans.) andtheServo900 C(SiemensElema,Uppsala, Sweden).Strategyof protectivemechanical ventilationwas appliedbylimitingpeaklungdistensionandpreventingend-expiratorycollapsewith low tidal volumes (Tidal volumes of 6 mL/kg of predictedbody weight), limited plateau pressure 6consecu-tive hours, withdrawal fromNIVwas attempted daily in30-minspontaneous breathing trials. Predenedcriteria forfailureofthespontaneousbreathingtrial were: Spo2 < 90%or PaO2 < 60 mmHg with FiO2 > 0.6, RR > 30 breaths/min,oractivationoftheaccessoryrespiratorymuscles[10].StatisticalanalysisResultsaregivenasmeans SD. Demographicandphysio-logic characteristics of the twogroups were comparedwithuseofStudentst-testforcontinuousdataandchi-squaretestforcategorical data. TheSPSSpackage(SPSS, Chicago)wasusedforall analyses. P < 0.05wasconsideredasstatisticallysignicant.ResultsThis study was carried out between April 2011 and April 2013 ,on 40 patients admitted to the intensive care unit (ICU) due toblunt chest trauma (either isolated chest trauma or as a part ofpolytrauma).ThepatientswererandomlyassignedtoreceiveeitherCPAP(group1)n = 15(11male,4femaleswithmeanage 31.8 13.8), BiPAP (group 2) n = 15 (10 male, 5 femaleswithmeanage31.8 13.1), andpatientswhomet inclusioncriteria but didnot showcooperationreceivedIPPV(group3)n = 10(7male,3femaleswithmeanage30.6 12.7).Table1shows thebase-linecharacteristics of thestudiedthree patient groups: the ratio of the partial pressure of arterialoxygen to the fraction of inspired oxygen (PaO2:FiO2), arterialPh., partial pressureof carbondioxide(PaCO2), respiratoryrate(RR), heart rate(HR), injury severity score (ISS) andsimplied acute physiological score (SAPS), all showednon-signicantdifferences.Fig. 1showsimprovementoftheratioofthepartialpres-sure of arterial oxygen to the fraction of inspired oxygen(PaO2:FiO2) after 1 hof mechanical ventilationinthethreegroups.After one hour of mechanical ventilation there was adecreaseinPaCO2andtheRRandHR(Figs. 24)fromthebaselineinthethreegroups.Table 2 shows a signicant increase in mean PaO2:FiO2 anddecreaseinRR,andHRafterMVinthethreegroups.Table 3 shows non-signicant difference in mean change ofPaO2:FiO2amongthethreegroups.As regards the patient outcomes in the three studied groupsasshownin(Table4): fourpatientsingroup1(26.7%)andthreepatients ingroup 2(20%)requiredendotrachealintuba-tionwithnosignicant difference P > .05. Thereasons forintubationinthesepatient groupsareshowninTable5, thefailureofnoninvasiveventilationtomaintainthePaO2aboveTable1 Baselinecharacteristicsofthestudiedpatientgroups.Group1(N = 15) Group2(N = 15) Group3(N = 10) PAge(years) 31.8 13.8 31.8 13.1 30.6 12.7 0.97Male/female 11:4 10:5 7:3 0.92ArterialPh. 7.4 0.066 7.4 0.065 7.4 0.07 0.93PaCO2(mmHg) 39.6 8.09 39.6 7.7 38.8 7.9 0.95PaO2:FiO2216.9 32.9 217.1 38.2 221.7 33.2 0.93RR(breath/min) 34.2 2.37 34.3 2.34 34.2 2.1 0.99HR(beat/min) 119 2.4 118.8 2.4 118.1 2 0.63Meanbloodpressure(mmHg) 75 6.3 76.2 5.2 74.8 4.7 0.9ISS 42.9 2.1 42.3 1.4 43.2 2.2 0.51SAPS 15.6 1.3 15.4 1.1 15.2 1.4 0.82ISS:injuryseverityscore.SAPS:simpliedacutephysiologicalscore.ContinuousversusBi-levelpositiveairwaypressureventilationinbluntchesttrauma 20560 mmHg(2patients ingroup1and2patients ingroup2,inabilitytocorrect dyspnea(onepatient ingroup1andonepatientingroup2),andhemodynamicinstability(onepatientingroup1).Therewasnosignicantdifferenceinthelengthofstayintheintensivecareunitbetweenthethreegroups(10 5 daysin group 1, 11 4 in group 2 and 10 6 in group 3) (p = .8).Pneumonia developed in one patient in group 1 (6.6%) andin2patientsingroup2(13.3%)andin3patientsingroup3(30.3%).Pneumothoraxdevelopedinonepatientingroup1(6.6%)and in no patients in group 2 (0%) and in one patient in group3 (10%), with no signicant difference between the threegroupsP = .08. Asregardsmortalityonepatientingroup3died due to septic shock while there was no mortality in groups1and2.DiscussionAgeneraloptimalventilatorystrategythatisapplicabletoallpatientsafterchesttraumadoesnotexist.Understandingthepathophysiology of individual patients, with their specickinds of lungdamage after trauma, andaccordinglyimple-menting ventilation strategies may support the respiratory sys-temand prevent further ventilator-associated lung injury(VALI). VALIhasthepotential toinduceacutelunginjury(ALI)orARDS,aswellasmultipleorganfailure[16,17].Cli-nicians canselect betweentwodifferent strategies toapplymechanical ventilation: noninvasive ventilationandinvasivemechanical ventilation. The advantages of noninvasive ventila-tion are the avoidance of complications related to endotrachealintubation, avoidanceofsedationandparalysisandtheeasyremoval and reinstitution of NIV, if needed. However, trachealintubationshouldnever be delayedif the respiratorystatusworsensundernoninvasiveventilation[12].Theroleofnoninvasiveventilationforthemanagementofpatients with blunt chest trauma has not been well established[12]. Although the safety of NIV has been assessed in a numberof observational studies in patients with blunt thoracic injuries[8,13],theevidenceregardingtheuseofNIVinthissettingisinconsistent [12]. The aim of this study was to compare nonin-vasiveventilation(CPAPandBiPAP)withinvasivemechani-cal ventilationinmanagement of patients withblunt chesttraumaandcomparetheefciencyofCPAPversusBiPAPinavoidingintubationandIMV.This study included 40 patients: 15 patients received CPAP,15 patients received BiPAP and 10 patients were intubated andreceived conventional invasive mechanical ventilation .Thepatients who received invasive mechanical ventilation, metthecriteriaofNIVbuttheywereuncooperative,thistogetherwith the similarity of ISS and SAPS between the present studygroupsmadecomparisonpossibleasinsomepreviousstudies[18,19] whichcomparedinvasivewithnoninvasiveventilationandthereweresignicantdifferencesbetweenpatientgroupsFigure 1 Ratio of partial pressure of arterial oxygen to thefractionofinspiredoxygen(PaO2:FiO2)atthebaselineandafter1 hofmechanicalventilationinthestudiedpatientgroups.3737.53838.53939.540PRE PCO2 POST PCO2group1group2group3Figure2 Partial pressureofarterial carbondioxideatthebaselineandafter1 hofmechanicalventilationinthestudiedpatientgroups.Figure3 Heartrateatthebaselineandafter1 hofmechanicalventilationinthestudiedpatientgroups.0510152025303540base line one hour aergroup1group2group3Figure 4 Respiratory rate at the base line and after 1 h ofmechanicalventilationinthestudiedpatientgroups.206 R.EmanShebletal.asregardstheISSandtheseverityofunderlyingconditionastheinvasivegroupinvolvedpatientswithabsoluteindicationofIMVlikecomaandhemodynamicinstability.InthisstudyCPAPandBiPAPlikeconventional invasivemechanical ventilation were efcient in improving gasexchangeandrelievingrespiratorydistress(Figs.14).FourpatientsintheCPAP-ventilationgroup(26.7%)andthree patients inthe BiPAPgroup(20%) requiredendotra-cheal intubation, sothesuccessratetoavoidintubationwas73.3%in the CPAPgroup and 80%in the BiPAPgroup(Table4), thisagrees withthendingof Duggal et al. 2013[20]. Thereasonsfor intubation(Table5)were failureofnon-invasive ventilation tomaintain the PaO2above 60 mmHg(fourpatients), itsinabilitytocorrectdyspnea(twopatients)andhemo-dynamicinstability(onepatient).Nosocomial pneumonia and pneumo-thorax were the mostcommonly reported adverse events associated with NIV use inprevious studies, and the rate ranged from8%to 13.8%[20,21]. The rate of pneumothorax reportedintwostudiesrangedfrom5.5%to24%[21,22]. Inthis studypneumoniadevelopedinonepatientingroup1(6.6%)andin2patientsingroup2(13.3%)andin3patientsingroup3(30.3%)andpneumothorax developedinone patient ingroup1 (6.6%)and in no patients in group 2 (0%) and in one patient in group3 (10%), with no signicant difference between the threegroupsP = .08.This study showed non-signicant differences as regards thelengthof stayintheintensivecareunit betweentheCPAP,BiPAP and the conventional invasive ventilation groups(10 5, 11 4 daysand. 10 6)(Table4), thisndingdidnot agree withthat of aprevious studywhichshowedthatthelengthofstayinICUwaslowerinpatientswithNIVusecomparedtoinvasive mechanical ventilation[12]. This canbeexplainedbythat inthispreviousstudytheconventionalinvasive patient group included patients with more severeunderlying condition who were excluded fromthe presentstudypatient grouponinvasive ventilation(like comatosedpatients).In this study no patients in the noninvasive-ventilationgroupandonepatient intheconventional-ventilationgroupdiedintheintensivecareunit.Theobservedlowmortalityinthesestudypatients maybeexplainedbythat patients withsevere central nervous systemdamage and a lowGlasgowcomascalescorewereexcluded;studieshaveshownthatcen-tral nervous systemdamage is anindependent predictor ofmortalityinpolytraumapatients[20].Table2 ChangeinmeanPaO2:FiO2,RR,HRbeforeandafterMVineachgroup.Before After Pairedt PGroup1 MeanPaO2:FiO2216.9 32.9 299 50.2 5.9 0.00RR 34.2 2.3 23.4 1.8 17.2 0.00HR 119 2.4 104.6 4.7 14.4 0.00Group2 MeanPaO2:FiO2217.1 38.2 292 58.9 4.5 0.00RR 34.3 2.3 23.3 2.5 21.7 0.00HR 118.8 2.4 104 4.3 20.2 0.00Group3 MeanPaO2:FiO2221.7 33.2 322 23.4 11.7 0.00RR 34.2 2.1 24.4 2.4 25.2 0.00HR 118.1 2 102.1 3.1 21 0.00Table3 MeanPaO2:FiO2changeamongthestudiedgroups.MeanPaO2:FiO2change F PGroup1 86.87 42.8 0.58 0.5Group2 82.0 48.1Group3 100.3 27.07Table4 OutcomeofMVinthestudiedpatientgroups.Outcome Group1 Group2 Group3 Sig.Nosocomialpneumonia 1(6.6%) 2(13.3%) 3(30.3%) 0.000008Developmentofpneumothorax 1(6.6% 0(0%) 1(10%) 0.08Needforinvasiveventilation(failureofNIV) 4(26.7%) 3(20%) 0.3Lengthofhospitalstay(dayes) 10 5 11 4 10 6 0.8Mortality 0(0.0%) 0(0.0%) 1(10%) 0.00004Table5 ReasonsfortheNIVfailure.Reasonoffailure Group1(4/15) Group2(3/15) PFailuretomaintainthePaO2above60 mmHg 2Patients 2Patients 0.11Inabilitytocorrectdyspnea 1Patient 1Patient 0.2Hemodynamicinstability. 1Patient 0 0.02Totalfailure 4 3 0.28ContinuousversusBi-levelpositiveairwaypressureventilationinbluntchesttrauma 207ConclusionBothCPAPandBiPAPare safeandefcient techniques inmanaging respiratory failure andreducing the incidence ofintubationinpatients withblunt chest traumaandtheyareassociatedwithfewseriouscomplications, ashortstayattheintensivecareunit, soearlyidenticationof at-riskpatientswith early institution of CPAP or BiPAP in appropriatepatientsmaybeofagreatbenet.ConictofinterestWehavenoconictofinteresttodeclare.References[1] J.H. Devitt, R.F. McLean, J.P. Koch, Anaesthetic managementof acute blunt thoracic trauma, Can. J. Anaesth. 38 (1991) 506510.[2] H.C. Pape, D. Remmers, J. Rice, M. Ebisch, C. Krettek, H.Tscherne, Appraisal ofearlyevaluationofbluntchesttrauma:Development of a standardized scoring system for initial clinicaldecisionmaking,J.Trauma.49(2000)496504.[3] A. Wustner, O. Gehmacher, S. Hammerle, C. Schenkenbach, H.Hafele, G. Mathis, Ultrasound diagnosis in blunt thoracictrauma,UltraschallMed.26(2005)285290.[4] M. Aufmkolk, R. Fischer, G. Voggenreiter, C. Kleinschmidt,K.P. Schmit-Neuerburg, U. Obertacke, Local effect of lungcontusiononlungsurfactant compositioninmultipletraumapatients,Crit.CareMed.27(1999)14411446.[5] S. Wanek, J.C. Mayberry, Blunt thoracic trauma: ail chest,pulmonary contusion, and blast injury, Crit. Care Clin. 20(2004)7181.[6] E.A. Ullman, L.P. Donley, W.J. Brady, Pulmonary traumaemergency department evaluation and management, Emerg.Med.Clin.NorthAm.21(2003)291313.[7] K.Raghavendran,R.H.Notter,B.A.Davidson,J.D.Helinski,S.L. Kunkel, P.R. Knight, Lung contusion: inammatorymechanisms and interaction with other injuries, Shock 32(2009)122130.[8] Y. Klein, S.M. Cohn, K.G. Proctor, Lung contusion:pathophysiologyandmanagement, Curr. Opin. Anaesthesiol.17(2002)6568.[9] S.M. Cohn, J.J. Dubose, Pulmonarycontusion: anupdateonrecent advances in clinical management, World J. Surg. 17(2010)19591970.[10] L.D. Nelson, Ventilatorysupport of the traumapatient withpulmonary contusion, Respir. Care Clin. N. Am. 17 (1996) 425447.[11] H. Tanaka, K. Tajimi, Y. Endoh, K. Kobayashi, Pneumaticstabilization for ail chest injury: an 11-year study, Surg. Today17(2001)1217.[12] M. Walz, G. Mo llenhoff, G. Muhr, CPAP-augmentedspontaneous respirationinthoracic trauma: analternative tointubationinGerman,Unfallchirurg17(1998)527536.[13] N. Xirouchaki, E. Kondoudaki, M. Anastasaki, C.Alexopoulou, S. Koumiotaki, D. Georgopoulos, Noninvasivebilevel positive pressure ventilation in patients with bluntthoracictrauma,Respiration17(2005)517522.[14] S.P. Baker, B. ONeill, W. Haddon, W.B. Long, The injuryseverityscore: amethodfor describingpatients withmultipleinjuries and evaluating emergency care, J. Trauma 14 (1974) 87196.[15] J.R.LeGall,S.Lemeshow,F.Saulnier, Anewsimplied acutephysiology score (SAPS II) based on a European/NorthAmericanmulticenterstudy,JAMA27(1993)29572963.[16] O. Gajic, F. Frutos-Vivar, A. Esteban, R.D. Hubmayr, A.Anzueto, Ventilator settings as a risk factor for acute respiratorydistresssyndromeinmechanicallyventilatedpatients,IntensiveCareMed.31(2005)922926.[17] A.S. Slutsky, Ventilator-inducedlunginjury:frombarotraumatobiotrauma,Respir.Care50(2005)646659.[18] C.B. Hurst, R.D. DeHaven, Branson, Use of CPAP mask as thesole mode of ventilatory support in trauma patients with mild tomoderate respiratory insufciency, J. Trauma 17 (17) (1985)10651068.[19] G. Hernandez, R. Fernandez, P. Lopez-Reina, R. Cuena, A.Pedrosa,R.Ortiz,P.Hiradier,Noninvasiveventilationreducesintubationinchest trauma-relatedhypoxemia: a randomizedclinicaltrial,Chest17(2010)7480.[20] A. Duggal, P. Perez, E. Golan, L. Tremblayand, T. Sinuff,Safetyandefcacyof noninvasiveventilationinpatientswithblunt chest trauma: a systematic review, Crit. Care 17 (4) (2013)R142.[21] M. Gunduz, H. Unlugenc, M. Ozalevli, K. Inanoglu, H. Akman,Acomparative study of continuous positive airway pressure(CPAP) and intermittent positive pressure ventilation (IPPV) inpatientswithailchest,Emerg.Med.J.17(2005)325329.[22] C.T. Bolliger, S.F. Van Eeden, Treatment of multiple ribfractures: randomized controlled trial comparing ventilatorywithnonventilatorymanagement,Chest17(1990)943948.208 R.EmanShebletal.