DTIC ADA629607: Comparison of Mortality Associated with Sepsis in the Burn, Trauma, and General Intensive Care Unit Patient: A Systematic Review of the Literature

0.76 MB·English

by  Defense Technical Information Center

Report Documentation Page Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302 Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it

does not display a currently valid OMB control number
1. REPORT DATE 2. REPORT TYPE 3. DATES COVERED
01 JAN 2015 N/A -
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Comparison of mortality associated with sepsis in the burn, trauma, and
5b. GRANT NUMBER
general intensive care unit patient: a systematic review of the literature
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
Mann E. A., Baun M. M., Meininger J. C., Wade C. E.,
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
United States Army Institute of Surgical Research, JBSA Fort Sam REPORT NUMBER
Houston, TX
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release, distribution unlimited
13. SUPPLEMENTARY NOTES
14. ABSTRACT
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF
ABSTRACT OF PAGES RESPONSIBLE PERSON

a REPORT b ABSTRACT c THIS PAGE UU 13
unclassified unclassified unclassified
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
SHOCKJANUARY2012 SEPSISINBURN,TRAUMA,ANDICUPATIENTS 5
injury associated with burn or trauma, these patients are rou- first international consensus conference on sepsis (12); thus, the date range
searched was limited to 1990 to 2010. Articles were excluded from review
tinely excluded from large multicenter trials that strive for
if theprimaryoutcomemeasure was limitedto infection,bacteremia,organ
maximum homogeneity in the population studied. Further- failure,oranyotheroutcomenotdirectlyrelatedtosepsis(infectiousprocesses
more, the burn patient has been proposed as representative of coupled with organ failure). Perinatal, non critically ill, emergency depart
ment, and oncology populations were also excluded. To improve generaliz
the universal model of trauma (10). Oftentimes, outcomes for
abilityofICUpopulationsincludedintheanalysis,furtherexclusioncriteria
burn and trauma patients are compared, yet no data exist to eliminated singlecenter studies, or a secondary analysis of the same pop
suggest whether prevalence or survival associated with sepsis ulation in a published study; studies of general ICU populations (primarily
medical, surgical, or combination) with fewer than 1,000 patients enrolled
is similar. The purpose of this systematic review of the liter-
wereexcludedtopromoteequityamongnumberofstudiesincludedineach
aturewastodeterminetheassociationofsepsiswithoutcomes group.Thisstepwasdeemednecessarybecauseofthelargenumberofsepsis
by means of a comparative analysis of patients with sepsis in reportsintheliteratureanddifficultyaggregatingtheoverwhelmingnumber
ofavailablestudiestoserveasacomparisongroup.
burnandtraumainjurywithageneralcriticallyillpopulation.
Sepsisrelated definitions used in this analysis include (a) septicemia or
Outcomes were mortality during ICU stay, during hospital bacteremia:positiveinfectioninthebloodstream;(b)sepsis:twoormoreof
stay, or at 28 days after hospital admission. As a secondary thecriteriaforSIRS,pluspositivecultureorclinicalsuspicionofinfection;(c)
severe/complicated sepsis:sepsis criteria andpresence ofat least onefailed
purpose, when reported, the prevalence of sepsis was also
organsystem;and(d)septicshock:severesepsisinthepresenceofhemody
compared. namicfailureunresponsivetofluidtherapy,andrequiringvasopressorsup
port (6). The ACCP/SIRS criteria for sepsis include presence of infection
METHODS with at least two of the following: temperature greater than 38-C or less
than36-C;heartrategreaterthan90beats/min;respiratoryrategreaterthan
A systematic review of the literature (11) was conducted using the 20 breaths/min or arterial carbon dioxide tension less than 32 mmHg; or
MEDLINE (PubMed), Cochrane Library, and ProQuest (Dissertations and whitebloodcellcountlessthan4,000orgreaterthan12,000cells/2L(6).
Theses) scientific databases. The following keywords and MeSH headings Adult populations selected for this study include (a) burn: thermal or
were used: Bsepsis,[ septicemia,[ Bseptic shock,[ Bepidemiology,[ Bburns,[ chemicalinjuryincivilianormilitarypatients;(b)trauma:mechanicalinjury,
Bthermalinjury,[Btrauma,[Bwoundsandinjuries,[Bcriticalcare,[Bintensive includingblunt,penetrating,ormotorvehicleaccidentincivilianormilitary
care,[Boutcomes,[andBmortality.[Additionalarticleswereidentifiedfrom patients;and(c)generalICU:patientsrequiringmedicalorsurgicalintensive
referencelistsduringfulltextreview. care management (such as mechanical ventilator support or cardiovascular
Studies were considered for inclusion based on review of abstracts that support)andnotprimarilycomposedofburn ortraumainjuredpatients.The
reportedclinicalstudies(retrospectiveorprospectivedesign)publishedinthe primaryoutcomeofinterestforthisreviewwasmortality,variouslydefinedas
English language, for primarily adult populations (918 years of age), with deathduringICUstay,duringhospitalstay,orwithin28daysafterhospital
information on survival of sepsis in a critically ill population. Significant admission. Prevalence of sepsis was reported when this information was
changesinclinicalpracticeinthetreatmentofsepsishaveoccurredsincethe availablefordescriptiveratherthananalyticalpurposes.
FIG.1. Resultsofsearchstrategy.
Copyright © 2011 by the Shock Society. Unauthorized reproduction of this article is prohibited.
6 SHOCKVOL.37,NO.1 MANNETAL.
Methodologicalquality 2005(20).Mortalityfornonsepticpatientswas12%compared
For inclusion, studies needed an evidence rating of level II (evidence with33%forsepticpatients(P=0.06)inasingle-centerstudy
obtainedfromatleastonewelldesignedrandomizedcontrolledtrial)tolevel
led by Cumming et al. (18).
IV (evidence from welldesigned, casecontrol or cohort studies) (13). The
quality of included studies was required to be either Bhigh[ (A grade) or
Bgood[(Bgrade)(14).ThestudiesthatwereconsideredBgood[didnotpro Traumastudies
videspecificcriteriaforsepsisdiagnosis.Riskofbias,suchasselectiveout
Traumastudieswithreportedoutcomesofpatientswithsep-
come reporting or outcome concealment, was considered during quality
assessment,andnosystematicbiaswasnoted. sis(n = 11) covered the time period from 1990 to 2009, with
two prospective observational studies and nine retrospective
RESULTS
reviews(Table2).Themechanismofinjuryvaried,withseven
A total of 38 articles that met inclusion criteria were re- that included victims of polytrauma; three included motor
viewed(burn=9,trauma=11,generalICU=18)(Fig.1).Of vehicleaccidentsorbluntormechanicalinjury;andonecom-
the nine burn studies, two studies consisted of patients inclu- prised combat injuries. Most studies reported results from
dedinpreviouslyreportedanalyses(15,16),sothesesubjects multiple centers or regional databases (n = 6) compared with
werecountedasingletimeinthetotalnumberofburnpatients reports from a single center (n= 5).
reported.AllstudieshadanevidenceratingoflevelsIIthrough The population for this analysis includes 3,719 septic pa-
IV; quality ofincluded studies was judged Bgood[or higher. tients from a pool of more than 70,000 trauma patients. Un-
surprisingly,relativelyyoungmalespredominate;frequencyof
Burnstudies male sex ranged from 68% to 100% with mean age range of
The nine studies reporting sepsis outcomes for the burn 34 to 49 years. The injury severity scores (ISSs) ranged from
population, representing the time period from 1991 to 2005, 19.3 to 47. One study reported the ISS for deaths associ-
include three prospective observational studies and six ret- ated with sepsis versus deaths without sepsis as 28 T 14 vs.
rospective record reviews (Table 1). The majority were con- 13T12(PG0.001),respectively(30).However,anotherstudy
ducted at a single center (n = 7), and one study is a summary reported no difference between ISS for the combination of
reportfromtheABAthatrepresentsdatafrom70burncenters sepsis and trauma compared with trauma alone (29 T 10 vs.
in 30 US states (20). A total of 2,106 burn patients with the 32T13,notstatisticallysignificant)(34).Sourcesofinfection
diagnosis ofsepsisofa poolof134,159 burnadmissionscom- or infecting organisms were not provided in any of the inclu-
prise this review. The reported mean or median age for all ded studies. The definition of sepsis, when noted (n = 7),
studieswaslessthan45years,withmalebeingthepredominant varied among reports with the ACCP/SCCM definitions (24)
sex(56%Y75%)inallbutonereport(22).Burnedpatientswith used in five studies (30Y32, 34, 35) and International Classi-
sepsis represent a relatively severely injured population with fication of Diseases, Ninth Revision (ICD-9) codes coupled
30% to 76% total body surface area (TBSA) burn. Reported withdeathcertificateinformationusedintwostudies(26,29).
sepsisprevalencerangedfrom8%(15)oftheburnpopulationto Mortality associated with trauma complicated by sepsis
greater than 42% in four studies (42%Y50%) (16, 18, 21, 23), ranged from less than 7% in four studies (25, 28, 29, 33) to
andtwostudiesreportedprevalenceof50%and65%(22,23). 10%to23%insixotherstudies(26,27,30,31,35),withone
Thecriteriaforthediagnosisofsepsisvariedgreatlyamong study reporting 46% mortality among combat-related trauma
the studies, with two referring to the ACCP/SCCM sepsis patients with sepsis (34). One study found the difference be-
definitions(24),threethatreliedonclinicalcriteriainaddition tween death from trauma alone to be significant from trauma
toapositivecultureresult,twowithcultureresultsalone,and coupled with sepsis (7.6% vs. 23%, P G 0.001) (30). Another
onewithclinicalcriteriaalone.Onlytworeportsdescribedthe team also found mortality to differ between trauma patients
primary site of infection associated with sepsis of blood (19) without sepsis and those with sepsis (9.3% vs. 36.9%, P = 0.01)
and wound (21) for bacterial and fungal causes, respectively. (31). Wafaisade et al. (35) reported no significant decrease in
Primary organisms associated with sepsis were identified as mortality associated with sepsis in the German trauma pop-
gram-negativeinthreestudies(65%Y72%)(19,22,23),gram- ulation from 1993 to 2008, with reported mortality during
positive in two studies (61%Y62%) (15, 17), and fungal from consecutive 4-year periods of 16.2%, 21.5%, 22%, and
a single study (21). 18.2%,respectively(n=28,829;P=0.054).Duringthesame
Mortality associated with sepsis varied with degree of ill- period,theauthorsreportadecreaseintheprevalenceofsepsis
ness reported. The study that differentiated between uncom- of 14.8%, 12.5%, 9.4%, and 9.7%, respectively (P G 0.0001)
plicatedsepsisandsepticshockreportedmortalityratesof6% (35). These findings suggest a reduction in the absolute mor-
to 11% to 27% to 63%, respectively (16). The ABA National tality associated with sepsis, with an increase in the propor-
Burn Repository 10-year review (n = 3,488) reported pulmo- tion of sepsis-related deaths over time in the German trauma
nary failure/sepsis as the primary cause of death for 11.3%, population.
multiple organ failure for 27.5%, and burn wound sepsis for
4% of patients (20). Variables associated with increased mor- Generalcriticalcarestudies
tality from sepsis in the burn population were identified as Studiesofcriticallyillpatientswithsepsis(n=18)covered
multiple organ failure (15Y17, 20), TBSA burn (15), and thetimeperiodfrom1979to2008andincludednineprospec-
presenceofinhalationinjury(15,16).Theoverallmortalityin tive randomized trials or observational studies and nine ret-
the National Burn Repository burned population without sep- rospective studies (Table 3). One country was represented
sis is reported as decreasing from 6.2% in 1995 to 4.7% in in 10 studies, with eight studies including up to 37 different
Copyright © 2011 by the Shock Society. Unauthorized reproduction of this article is prohibited.
SHOCKJANUARY2012 SEPSISINBURN,TRAUMA,ANDICUPATIENTS 7
TABLE1. Burnstudiesincludedinanalysis(n=9)
Design/evidence
level(13)/quality
Burnstudies Studyperiod grade(14) Center/location Purpose Subjects Sepsissubjects
Bangetal.1998(17) 1992 1996 Retroreview/level Singlecenter; Retrospectivereview 943admits,280 79/280(28.2%)
IV/gradeA Kuwait ofsepticemicburn (30%)ICUadmits, ICUpatients;
patients(data 79/280(28.2%) 118episodes
includedinBang2004) septic
Bangetal.2004(15) 1992 2001 Retroreview/level Singlecenter; Studydemographic/ 2,082hospitaladmits, 166/2,082(8%);
IV/gradeA Kuwait clinicalfactors 166(8%)sepsis 253episodes
associatedwithburn
sepsisinKuwait
Cummingetal.(18) 1998 1999 Prospobs/level Singlecenter; Quantifycomplications 85ICUadmits Uncomplicated
2001(dataincluded IV/gradeA Parkland oforgandysfunction (920%TBSA); sepsis
inFitzwater2003) (Dallas,Tex) andsepsisafter SSorseptic 43/85(50.6%);
burninjury shock=12(14%) severesepsis
12/85(14.1%)
D’Avignonetal.(19)2010 1991 2003 Retroreview/level Singlecenter;USAISR Retrospectivereview 97ICUpatient NR
IV/gradeA (SanAntonio,Tex) ofautopsyreportsto autopsies,
determineincidence 27=bacterialsepsis,
ofdeathattributableto 5=viralsepsis
bacterialorviralcause
Fitzwateretal.(16)2003 1998 2000 Prospobs/level Singlecenter; Definerelationships n=175ICUadmits Allsepsis=79/175
IV/gradeA Parkland betweensepsis, (920%TBSA, (45%);
(Dallas,Tex) MOD,anddeath 916y);sepsis:79 uncomplicated
afterburntrauma (45%);complicated sepsis49/175
sepsis30(17%) (28%),49/79
(62%);severe
sepsis14/175
(8%),14/79(18%);
septicshock16/175
(9%),16/79(20%)
Milleretal.(20)2006 1995 2005 Retroreview/level 70Center,30US Tenyearreviewof 126,642burnhospital Septicemia
IV/gradeB states nationalburndata admits(peds/adults); complication
repository 18,964with 1,554/18,964;
complications; 1.2%ofall
6,797/126,642deaths patientcases
(5.6%) (1,554/126,642)
Murrayetal.(21)2008 1991 2003 Retroreview/level Singlecenter; Twelveyearreviewof 3,751ICUadmits, 43/97(44%)
IV/gradeA USAISR(San fungalinfectionand 228(6.1%)deaths, fungalelements
Antonio,Tex) relatedmortalityin 97autopsies identified
burnautopsy
Sharmaetal.(22)2006 2000 2004 Retroreview/ Singlecenter;India Fiveyearreviewof 334autopsycases; NR
levelIV/gradeA autopsycasesto 216/334(65%)
determinerateof ‘‘septicemiadue
infection/sepsisin toburns’’
burnpatients
Sjobergetal.(23)2003 1997 1999 Prospobs/level 2centers;Zimbabwe Evaluationofpredicting 50ICUsubjects;sepsis 21/50(42%)
IV/gradeA septicemiainburn 21/50(42%) sepsis;16/21
patientsbyusing (76%)positive
woundsurface, tissueCX
tissueculture
techniques,and
bloodcultures
BP indicates blood pressure; CX, culture; gmj, Gram negative; gm+, grampositive; HR, heart rate; INH, inhalation injury; IQR, interquartile range; LOC, level of
consciousness; MOD, multiple organ dysfunction; NR, not reported; Obs, observational; peds, pediatric; prosp, prospective; retro, retrospective; temp, temperature;
USAISR,USArmyInstituteofSurgicalResearch.
countries; between 12 and 847 centers participated in the in- this analysis among 2.08 billion studied. Various estimates of
cluded reports. The populations studied comprised hospital- prevalence of sepsis in general medical and surgical ICU pop-
ized or ICU patients who subsequently developed sepsis as ulations reviewed in this analysis were reported, ranging from
an inpatient (7 studies) and patients admitted to the ICU with 8%(38)or1.6%to3.2%(41,42)ofhospitaladmissions,upto
diagnosis of sepsis, severe sepsis, or septic shock (two, eight, 12% to 21% (45, 46) or 19% to 37.4% (7, 48) of ICU admis-
and one study, respectively). Ultimately, more than 31.6 mil- sions. The mean age of patients ranged from 57 to 64 years;
lion septic patients were included in the studies used for three studies reported that 60% to 82% of patients with sepsis
Copyright © 2011 by the Shock Society. Unauthorized reproduction of this article is prohibited.
8 SHOCKVOL.37,NO.1 MANNETAL.
Demographics Sepsismortality Sepsisdefinition Site Organism
Agemean(range)26y 29%(23/79)TBSA72% PositivebloodCXbased NR 118episodes:62%gm+,
(45d 75y);male56%; (38% 90%) onclinicalsuspicion 25%gmj,13%mixed
TBSA46%(10 90);
INH14/79(18%)
Agemean(range)26y, 23.5%(39/166);mean PositivebloodCXbased NR 61.3%gm+;12%gmj;
5T1.4(1 70);60%male; age31;MOFcause onclinicalsuspicion 12.7%mixed
TBSA42%(2 95);INH death71.8%
39/166(23.5%),26/39
(67%)INHdiedofsepsis
Uncomplicatedsepsis (n=85)Agemedian Severesepsis:4/12(33.3%); ACCP/SCCM NR
43/85(50.6%); (IQR)35(24 48); nosepsis:9/73(12.3%),
severesepsis12/85(14.1%) male75.3%;TBSA P=0.06
median30(23 40);
INH1/15(7%)
n=27:Agemedian(range) 27/97(27.8%) Autopsyconcurwith Blood,pulmonary, 70.4%gmj(Pseudomonas
45(2 95);male74%; clinicalstatus;positive wound aeruginosa50%);
TBSA43%(2 81); bloodculture;pneumonia; 18.5gm+;11.1mixed
INHbacteremic33%, histologicaltissueinvasion
nonbacteremic24%
(P=0.38)
Sepsis:agemedian38 Sepsis3/49(6%);severe ACCP/SCCM;severe NR NR
(IQR26 49);male86%; sepsis2/14(14%); sepsis:MODscoreQ3;
TBSAmedian37% septicshock10/16(63%) sepsisshock:
(IQR29 52).IHN22%: pressororacidosis
sepsis11/49(22%),
severesepsis3/14(21%),
septicshock7/16(44%)
Agemean33y;male70%; MOF27.5%(958/3,488); NR NR NR
TBSA%920=17%; pulmonaryfail/sepsis
INH6.5%:lived5%, 11.3%(395/3,488);
died30.7% burnwoundsepsis
4.1%(142/3,488)
Fungusattributable Mortalityattributableto Fungalelementspresent Wound,pulmonary, Aspergillus13/14cases
mortality=14/97, fungalinfection14/43 inautopsyreportandcause abdominal withfatalfungalinfection
agemedian(range) (33%),14/97(14.4%) ofdeathbypathologist (92.8%);woundprimary
42(24 67),male73%, sourceofinfection
TBSA76%(8 92)
Age21 25=30%; 216/334(65%) Splenicbloodculturefor NR 65%gmj,11%mixed
male32%;TBSA patientswithpremorbid
mean51%(range, cultures
G30to980)
Septic:agemedian(range) 8/16(50%)positive Temp,BP,HR,LOC NR 72%gmj;23%gm+
23(12 56),TBSAmedian tissueculturedied (woundtissue)
30%(12 70):survived
22(12 30)died40
(30 70);maleNR
were older than 65 years (42, 51, 52), and two reported 40% The definitions for sepsis varied among reports, with the
to46%ofsepticpatientswereolderthan75years(42,52).One majorityusingtheACCP/SCCMdefinitions(24)(n=9)(2,7,
study reported a significant age difference between septic and 37,40Y42,45,47,48)andtheremainderusingICD-9codesor
nonsepticpatientsof61yearsversus54years(PG0.001)(38), medicalrecorddiagnosesandpresenceofinfectionwithorgan
respectively.However,anotherstudyreportednosignificantdif- dysfunction[n=5(7,38,43,49,51)andn=5(36,39,44,46,
ference between groups (septic 65 years vs. nonseptic 64 years, 50), respectively]. The primary source of infection was iden-
P 9 0.05) (7). Males comprised between 47% and 64% of the tifiedasthelungorrespiratorysysteminall15studieswhere
patientsinthisanalysis. source was reported. The second most common source of
Copyright © 2011 by the Shock Society. Unauthorized reproduction of this article is prohibited.
SHOCKJANUARY2012 SEPSISINBURN,TRAUMA,ANDICUPATIENTS 9
TABLE2. Traumastudiesincludedinanalysis(n=11)
Design/evidence
level(13)/quality Location
Traumastudies Studyperiod grade(14) (no.centers) Purpose Primarypopulation
Espositoetal.(25)1995 1990 1991 Retroreview/level Montana(NR) Determinerateofpreventable Mechanicaltrauma
IV/gradeB mortalityandinappropriate
carefromtraumaticdeathin
aruralstate
Hodgsonetal.(26)2000 1991 1997 Retroreview/level Ontario(1) Determinemissedinjuriesinblunt Blunttrauma
IV/gradeB traumaandaccuracyof
recordedcauseofdeath
Maioetal.(27)1996 1994 ProsObserv/level Michigan(NR) Determinepreventabledeathrate, Trauma
IV/gradeB inappropriatecareinruralstate
MarsonandThomson(28) 1995 1997 Retroreview/level Brazil(NR) Impactofpre hospitalcare MVAtrauma
2001 IV/gradeB systemonMVAmortality,
autopsyevaluation
Meislinetal.(29)1997 1991 1993 Retroreview/level Arizona(NR) Examinetraumaticdeathina Blunt/penetratingtrauma
IV/gradeB UScounty
Osborneetal.(30)2004 1996 1997 Retroreview/level Pennsylvania(28) Characterizeepidemiologyof Blunt/penetratingtrauma
IV/gradeA sepsisintrauma
Pluradetal.(31)2010 2000 2009 Retroreview/level CA(1) Associationbetweenraceand Trauma
IV/gradeA incidenceandsurvival
posttraumaticsepsis
Probstetal.(32)2009 1973 1990 Retroreview/level Germany(1) Longtermmortalityandcauseof Polytrauma
IV/gradeA deathaftermultipleinjuries
Stewartetal.(33)2003 1995 2001 Retroreview/level TX(1) Identifypreventablecausesof Trauma
IV/gradeB traumaticdeath
Surbatovicetal.(34)2007 1999 Retroreview/level Serbia(1) Evaluateprognosticvalueof Trauma;combat
IV/gradeA immuneresponseincombat
casualties
Wafaisadeetal.(35)2011 1993 2008 Retroreview/level Germany(149); Assesschangeinincidence, Trauma
IV/gradeA CentralEurope(17) outcome,riskfactorsof
sepsisintrauma
hospindicateshospital;mech,mechanical;micro,microbiology;MODS,multipleorgandysfunctionsyndrome;MVA,motorvehicleaccident;NR,notreported;ns,non
significant;observ,observational;pros,prospective;retro,retrospective;SIRS,systemicinflammatoryresponsesyndrome.
infection resulting in sepsis was the abdomen or gastrointes- (n = 16) provided hospital mortality outcomes for patients
tinal system (n = 8) (36Y40, 45, 46, 50), the genitourinary withsepsisandseveresepsisthatrangedfrom18.5%to53.6%
tract(n=3)(43,44,47),andblood(n=3)(2,7,48).Ofthe12 (2, 7, 37, 39, 41Y49, 51) and up to 87% for sepsis associated
studies in which a primary infecting organism was identified, with failure ofmore than five organ systems(41).
ninereportedgram-positivebacteriaaspredominant(25%Y56%) Further analysis by Annane et al. (38) described the differ-
(7, 36, 40, 43, 45, 46, 48Y50) compared with two reports of ence in septic shock mortality compared with nonseptic
gram-negativebacteria (41%Y49%) as cause ofsepsis(37,39). shock patients of 61.2% vs. 13.2%, respectively, for the gen-
Of note, 9 studies showed no identified organism associated eral ICU population; matched septic shock patients with con-
withclinicaldiagnosisofsepsiswithinarangeof15%to50% trols (n = 5,473 per group) revealed mortality of 53.8% vs.
ofthe time (36, 38Y40, 43,45,46, 49, 53). 28.2%(PG0.001),respectively.InanotherstudyledbyAlberti
Mortality associated with sepsis during ICU stay was re- et al. (37), mortality was reported as 17% for noninfected
ported as ranging from 26.5% to 61% (38, 39, 45), with four patients compared with 53% of patients who presented to the
studiesreporting28-daymortalityof17%to33%(36,40,45, ICUwithongoinginfection.Numberofinvolvedorgansystems
50). The two randomized controlled trials testing drotrecogin hasbeenassociatedwithincreasedmortality;onefailedsystem
alfa (activated) (DrotAA) therapy for septic shock that re- versus two is associated with increased mortality from 11%
ported 28-day mortality for both intervention and control to 49% (P = 0.001) (46). Over time, mortality associated with
groups reached different conclusions. The majority of studies sepsis has declined from 45% in 1993 to 38% in 2003 in one
Copyright © 2011 by the Shock Society. Unauthorized reproduction of this article is prohibited.
10 SHOCKVOL.37,NO.1 MANNETAL.
Sepsissubjects Demographics ISS Sepsismortality Sepsisdefinition
Mechanicaltrauma: Agemean(range) 37(range,1 75) 2%(5/324) NR
324/629(52%of 42(2 95),age n=227
traumarelateddeaths) median36;male74%
108traumadeaths Agemedian(range) NR 17%(18/108) Deathcertificate/autopsy
39(2 90);male72% report;SIRScriteria
65hospitaladmits Agemean37.4T25; 46.8(range,5 75) 10%(2/20)preventable; NR
(25/65diedin male71.6% 3%(2/65)hospitalized
hospital38%)
243hospitaldeaths Agemean34preintervention, NR Preintervention3.1%(4/128), NR
(preinterventionn=128, 35postintervention; postintervention5.2%
postinterventionn=115 male81.3%preintervention, (6/115);overall4.1%
82.8%postintervention (10/243)
340hospitalized Agemean49.3; 25.6 SurviveG60min=3%; ICD9code,death
male67.9% 4 24h=5.9% 7.6%; certificate,autopsy
93wk=7.1% report
30,303hospitalized;2% Sepsis:agemean48.8T21; Sepsis:28.1T14; 23%(nonseptic7.6%, SIRSandinfection
sepsis(606/30,303) maleNR;primarysource: no sepsis:12.9T11 PG0.001) (ACCP/SCCM)
pulmonary (PG0.001)
3,998ICUadmits; Agemean36.7T19; 19.3T12.7 Septic:36.9%(250/677); SIRS(ACCP/SCCM)
16.9%(677/3,998) male79%;Hispanic62.4% nonseptic:9.3% andinfectioussource
(310/3,321)(P=0.01);
total:14%(560/3,998)
408(inhospitaldeaths); Agemean29.4T15.8; Hospdeaths Inhospital:11%sepsis SIRSand‘‘clinically
103(postdischargedeaths) male73% 29.2T10.2 (45/408) manifestinfection’’
(ACCP/SCCM)
753hospitaldeaths Agemean42.5T25.4, 41T20.6 CombinedwithMODS/other NR
agemedian39; 3%(23/753)
maleNR
76ICUadmits:sepsis=56; Agemean(range)26.8 Sepsis/traumamean Nonsurvivors(n=36), SIRSandpositiveblood
nonsepsis=20 (11 72);male100% 29T10.4;trauma sepsis/trauma32/56 culture(ACCP/SCCM)
31.7T12.5(P=ns) (56%),trauma4/20(20%)
3,042/28,829admits; Agemean44T19; 33T13 1993 1996=16.2%, SIRS(ACCP/SCCM),no
septictotal=10.2%; male81% 1997 2000=21.5%, microdatainregistry
1993 1996=14.8%, 2001 2004=22%,
1997 2000=12.5%, 2005 8=18.2(P=0.054);
2001 2004=9.4%, overallseptic19.5%vs.
2005 2008=9.7% nonseptic12.5%
(PG0.0001) (PG0.0001)
report (42) and decreased during the period of 1979 to 1984 care studies are especially representative of a worldwidepop-
from28%toonly18%in1995to2005inanotherstudy,despite ulation in which multiple countries were included in many of
anincreaseintheoverallincidenceofsepsis(49). thereportsandthusprovideahomogenouscomparisonforthis
analysis.Mostoftheincludedstudiesarepredominatelychart
reviews or retrospective in nature, 25 (66%) of 38, which isa
DISCUSSION
limitationto the completeness of reported data.
Sepsis is associated with poor outcomes in all patient pop- The age of included patients appears to be different when
ulations. This review is the first to compare sepsis outcomes theburn(G45years)andtrauma(34Y49years)groupsarecon-
inthreedistinctpatientpopulations:burn,trauma,andgeneral trasted with the older general ICU population (57Y64 years).
medical/surgicalcriticalcarepatients.Studiesidentifiedthrough Association of increased age with worse outcomes for patients
a systematic review of the literature represent the available re- with sepsis would seem to favor better outcomes for the burn
portsdescribingmortalityassociatedwithsepsisforthesespeci- population(relativelyyoungergroupinthisreview)wheninfact
ficgroupsofpatients,overthepasttwodecades. the mortality rateamong burnpatientswas moresimilartothe
An international population is represented in this review; older general ICU population than to the younger trauma pa-
the majority of burn studies originate from a single center in tients. Perhaps the effect of age is overcome by the degree of
contrast to the fact that most trauma and all general critical burn injury, presence of inhalation injury, and multiple organ
care studies were conducted in multiple centers. The critical failure(16).Theratioofmalestofemalesisgreaterintheburn
Copyright © 2011 by the Shock Society. Unauthorized reproduction of this article is prohibited.
SHOCKJANUARY2012 SEPSISINBURN,TRAUMA,ANDICUPATIENTS 11
TABLE3. Generalcriticalcarestudiesincludedinanalysis(n=18)
Design/evidence
General Study level(13)/quality
ICUstudies period grade(14) Countries/centers Purpose Subjects
Abrahametal.(36)2005 2002 2004 PRCT/level 34/516 APCforsepsisw/low Severesepsis2,613
ADDRESS II/gradeA riskofdeath
Albertietal.(37)2002 1997 1998 Cohortobserv/level 8/28 Incidenceofinfectionand Sepsis3,239/14,364(22.5%)ICU
IV/gradeA ICUoutcome admits;1,115/3,239(34%)septic;
944/3,239(29%)severesepsis;
1,180/3,239(36%)septicshock)
Angusetal.(2)2001 1995 Cohortobserv/level 7USstates/847 Incidence,cost,outcomeof Septic192,980/6,621,559(3%)
IV/gradeA severesepsisinUS hospitaladmits
Annaneetal.(38)2003 1993 2000 Retroreview/level France/22 Updateepidemiologyof 100,554ICUadmits;rate8.2%
IV/gradeA septicshock (8,251/100,554);1993:7%;
2000:9.7%
Bealeetal.(39)2009 2002 2005 Cohortobserv/level 37/276 Internationalsepsisregistry Severesepsis12,570(2burn)
IV/gradeA
Bernardetal.(40)2001 1998 2000 PRCT/level 11/164 APCphase3trialformortality Severesepsis1,690
PROWESS II/gradeA reductionforseveresepsis
Dombrovskiyetal.(41) 1995 2002 Retroreview/level US(NewJersey)/NR Trendseveresepsis Hospitaladmit7,364,550;sepsis:
2005 IV/gradeA (dataincludedin hospitalization,mortality, 233,432(3.2%);severesepsis:
2007report) fatalityrate,andimpact 87,675(1.19%)
age,race,sex
Dombrovskiyetal.(42) 1993 2003 Retroreview/level US/NR Trendseveresepsis Hospitaladmits391,571,824;sepsis
2007 IV/gradeA hospitalization,mortality, 8,403,766(2.15%);severesep
casefatalityrate 2,857,476(G1%)
Esperetal.(43)2006 1979 2003 Retroreview/level US/NR Factorsthatmayinfluence Hosp930million;sepsis12,505,082
IV/gradeA healthcaredisparitieson (1.3%)
incidenceofsepsis
Ferreretal.(44)2008 2005 2006 Before/afterprosp/level Spain/59ICU Determineifeducation Sepsis:2,319;septicshock
IV/gradeA programonSSCimproves 1,842/2,319(79.4%)
careandsepsismortality (pre:854;post:1,465)
Finferetal.(45)2004 1999(3mo) Prospobserv/level Australia,NZ/21 Documentincidenceand 691/3,543ICUadmits;11.8%
IV/gradeA (23ICU) outcomeofseveresepsisin (95%CI10.9 12.6)
AustraliaandNewZealand
Guidetetal.(46)2005 1997 2001 Retroreview/level France/12(35ICU) Studyincidenceandseverity ICUadmits96,193;severe
IV/gradeA oforgandysfunction sepsis=20,963/96,193(21.4%);
associatedwithsepsis 924hICU=18,273/65,910(27.7%)
Levyetal.(47)2010 2005 2008 Observ/PIproject/level Europe,NorthAmerica, Determinecompliancewith Septic15,022(52%ED,13%ICU,
III/gradeA SouthAmerica/165 severesepsisbundles 35%ward
andmortality
Martinetal.(48)2009 2003 2004 Prospobserv/level Canada/11ICU Determineacquisition,timing, ICUadmits6,298;severesepsis
IV/gradeA andoutcomesofsepsis 1,198/6,298(19%)
Martinetal.(49)2003 1979 2000 Retroreview/level US/NR Determineindependenteffect 750millionhospitaladmits;severe
IV/gradeA ofageonsepsis sepsis10,319,418(1.4%)
Vincentetal.(50)2005 2001 2003 Prospobservlevel 25/361 OpenlabeltrialofAPC ICUsepsis2,375(84%hadQ2OD)
ENHANCE III/gradeA treatmentforseveresepsis
Vincentetal.(7)2006 2002 Prospobserv/level 24/19ICU Defineincidenceofsepsisand ICUadmits3,174;sepsis1,177
SOAP IV/gradeA characteristicsofpatientsin (37.4%);severesepsis=930
EuropeanICUs (30%);sepshock=462(15%)
Weyckeretal.(51)2003 1991 2000 Retroreview/level US/NR Estimatemortalityandmedical Severesepsis16,019
IV/gradeA chargesamongsevere
sepsispatients
Abdindicatesabdominal;abx,antibiotics;APC,activatedproteinC;Am,America;a/w,associatedwith;CI,confidenceinterval;GI,gastrointestinal;gmj,gramnegative;
gm+,grampositive;GU,genitourinary;hosp,hospital;intervene,interventional;mech,mechanical;MODS,multipleorgandysfunctionsyndrome;MVA,motorvehicle
accident;no.,number;NR,notreported;ns,nonsignificant;observ,observational;OD,organdysfunction;PRCT,prospectiverandomizedcontrolledtrial;pros,prospective;
resp,respiratory;retro,retrospective;SIRS,systemicinflammatoryresponsesyndrome;SSC,SurvivingSepsisCampaign;Tx,treatment;UTI,urinarytractinfection.
Copyright © 2011 by the Shock Society. Unauthorized reproduction of this article is prohibited.
12 SHOCKVOL.37,NO.1 MANNETAL.
Demographics 28Daymortality Hospitalmortality ICUmortality Organism/source Sepsisdefinition
Intervention:agemean Intervention:18.5%; 29%gm+;24%gmj;11%mix; Infection and
58.6T17;male58.5%. Control:17% 30%none/lung,ABD,UTI sepsisinducedOD
control:58.8T17;male56.3%
924hICUagemedian64 Noninfected16.9%; 37%gm+;49%gmj;10% ACCP/SCCM
(27 83);61.1%male infectICUadmit fungus/resp,GI
53.6%
Agemean63.8;49.6%male 28.60% NR/resp,blood,GU,ABD ACCP/SCCM
Sepsis:agemean61.4T16.6; 61.2%(n=8,251)septic NR;severesepsisunidentified Diagnosisinmedical
nonsepsis53.9T19, shockpatients,13.2% pathogen~20%/lung, record
(PG0.001);maleseptic (n=92,293)nonsepsis ABD,UTI
63.3%,nonseptic58% shockpatients
Age60.4T17.5;male59.3% 49.60% 39.20% 32.4%gm+;41.4%gmj;8.7% InfectandOD(cid:1)1a/w
fungus;34%undetermined/lung, sepsis
ABD,UTI,blood
Intervention:age60.5T17.2; Intervention:24.7% 25% 26%gm+;22% 23%gmj; ACCP/SCCM:infection
male56.1%;control: 210/850;control: 13% 15%mixed;9%fungus; andSIRSandOD(cid:1)1
age60.6T16.5;male58% 30.8%259/840 33%negativeCX/lung,
(P=0.01) ABD,UTI
Severesepsisage960y75.3%, 38%with1OD;87% NR/NR ACCP/SCCM,sepsis
975y46.5%;male49.1% with5 6OD andOD,ICD9codes
Severesepsisage965y60%; 1993:45%;2003: NR/NR ACCP/SCCM,sepsis
975y40%;male50.7% 37.7% withOD,
ICD9codes
Age60.5y(95%CI60.4 60.7): 20.3%(CI19.9 20.6) 49% 56%gm+;microbedoc ICD9codes
malesNR infection52%/resp,GU,GI infection/sepsis
Pre:age67.4T16;male61.9%; Pre:44%;Post: NR/resp,GU,UTI Specificsepsiscriteria,
post:62.1T16;male60.2% 39.7%,P=0.04 shock,andOD
(overall41.2%)
Agemedian60.7T17.2; 32.4%(224/691) 37.5%(259/691) 26.5%(183/691) 48%gm+;38.5%gmj;13.2% Severesepsis=infection,
male57% other;CX+57.8%episodes/pulm, SIRS(ACCP/SCCM),
ABD,blood,skinUTI OD:PROWESS
0OD:age55.7T19,male57%; 0OD=14.5%; 28% 42%gm+;22.7 33.5% InfectwithOD(cid:1)1and2
1OD:age58.2T18,male 1OD=11.3%; gmj;2% 4%fungus;nodoc
63.1%;2OD:age62.1T16, 2OD=49%, infectSS150%,SS2
male64.2%,PG0.001 PG0.001 40%/pulm,ABD,CV
NR Intervention30.80%; NR/lung,UTI,ABD Suspectedinfection,
control:37% Q2SIRS,Q1OD
(ACCP/SCCM)
Age61.2T16.5;male58.8% 38.1%(CI35.4 40.8) 35.9%gm+;27.8%gmj;6.4% ACCP/SCCMand
yeast;14%other;15% PROWESS
missing/lung,blood,UTI
1979 1984:age57.4T29,male 1979 1984=27.8%; 52.1%gm+;37.6%gmj;4.6% Medicalrecorddiagnosis
49.6%;1995 2000: 1995 2000=17.9% fungus;specificorganism codes,ICD9
age60.8T14,male48% 51%/NR
Age59.1T17;male58.2% 25.3%(early22.9%, 26.6%gm+;43.4%gmj;3.7% Infection,3of4SIRS,
924h27.4%) fungi/lung,ABD,UTI Q1OD
Sepsis:age65(range,51 74), Allpatients=24%(747); 40%gm+;38%gmj;17%fungi; Infect,abx,
male63%;Nonseptic:age64 septic=36%(413); 18%mixed:clinicalsignsonly ACCP/SCCM;
(49 74),male61% nonseptic=17%(334) 40%/resp,blood,ABD,UTI severesepsisQ1OD
(PG0.05)
Age965y81.2%;male53.4% 21.2%T0.3% NR/NR ICD9codes,
infect,OD
Copyright © 2011 by the Shock Society. Unauthorized reproduction of this article is prohibited.