Legal forms

PCT as a diagnostic and prognostic tool in burn patients

Description
PCT as a diagnostic and prognostic tool in burn patients. Whether time course has a role in monitoring sepsis treatment
Categories
Published
of 8
48
Categories
Published
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Transcript
  PCTasadiagnosticandprognostictoolinburnpatients.Whethertimecoursehasarolein   monitoringsepsistreatment   A.Lavrentieva a, * ,S.Papadopoulou b ,J.Kioumis c  ,E.Kaimakamis c  ,M.Bitzani a a Papanikolaou   General   Hospital,   Burn   ICU,   Thessaloniki,   Greece b Papanikolaou   General   Hospital,   Burn   Surgery   Department,   Thessaloniki,   Greece c Papanikolaou   General   Hospital,   Pulmonary   Department,   Thessaloniki,   Greece 1.   Introduction Despite   major   advances   in   burn   care   and   improved   opportu-nities   for   therapeutic   intervention,   the   mortality   rate   of patients   with   severe   burn   due   to   septic   complications   is   stillconsiderable   [1,2].   Early   diagnosis   of    septic   complicationsfollowing    by   early   and   adequate   antibiotic   therapy   mayimprove   the   survival   rate   of    critically   ill   patients   [3,4].Since   burn   patients   are   in   a   state   of    chronic   systemicinflammatory   stimulation;   and   present   non-specific   symp-toms   of    systemic   inflammation,   traditional   laboratory   testsand   diagnostic   criteria   of    SIRS   and   sepsis   lack   diagnosticaccuracy   and   are   sometimes   misleading.   A   consensus   panelfor   the   American   Burn   Association   (ABA)   has   developedspecific   guidelines   for   the   diagnosis   of    sepsis   in   burn   patientthat   include   higher   thresholds   for   certain   parameters   such   astemperature,   heart   rate   and   respiratory   rate   [5].These   guide-lines   also   suggest   the   presence   of    thrombocytopenia,   insulinresistance,   feeding    intolerance,   increased   fluid   requirementsand   other   clinical   indicators   as   markers   of    inflammation   andinfection.   In   addition   to   these   clinical   signs,   documentedpresence   of    infection   and   a   clinical   response   to   antimicrobialsare   required.   The   consensus   conference   of    the   ABA,   in   anattempt   to   improve   the   current   definitions   of    sepsis,   suggeststhe   use   of    procalcitonin   (PTC)   and   the   other   inflammatory burns38(2012)356–363 a   r   t   i   c   l   e   i   n   f   o  Articlehistory: Accepted29August2011 Keywords: SepsisLocalizedinfectionProcalcitoninDiagnosticaccuracy a   b   s   t   r   a   c   t Objective: Toevaluatethediagnosticandprognosticperformanceofinflammatorymarkersforsepticandnonseptic(localized)bacterialinfectionsinpatientswithsevereburn. Methodsandresults: Dataof145patientswereprospectivelyincludedinthisstudy.Serumprocalcitoninandotherinflammatorymarkersweremeasuredwithin24hafterburnanddailythereafter.Maximumprocalcitonin(    p =0.004)wasindependentpredictorsofoutcomeinlogisticregressionanalysis.PCTthresholdsof1.5ng/ml,0.52ng/mland0.56ng/mlhadadequatesensitivityandspecificitytodiagnosesepsis,respiratorytractandwoundinfec-tionsrespectively.Athresholdvalueof7.8ng/mlinPCTconcentrationonday3wasassociatedwiththeeffectivenessofthesepsistreatmentwithanAUCof0.86(95%CI0.69–1.03,  p =0.002).C-reactiveproteinlevelsandWBCsshowednosignificantchangeoverthefirst3daysinthepatientswithsuccessfullytreatedsepsis(    p =0.93). Conclusion: Themaximumprocalcitoninlevelhasprognosticvalueinburnpatients.PCTcanbeusedasadiagnostictoolinpatientswithinfectiouscomplicationswithorwithoutbacteremiaduringICUstay.DailyconsecutivePCTmeasurementsmaybeavaluabletoolinmonitoringtheeffectivenessofantibiotictherapyinburnICUpatients. # 2011ElsevierLtdandISBI.Allrightsreserved.* Correspondingauthorat :   Hadzipanagiotidi2,Panorama,55236Thessaloniki,Greece.Tel.:+306949121458.E-mailaddress:alavrenti@gmail.com(A.Lavrentieva).  Available   online   at   www.sciencedirect.com journal   homepage:www.elsevier.com/locate/burns 0305-4179/$36.00 # 2011ElsevierLtdandISBI.Allrightsreserved.doi:10.1016/j.burns.2011.08.021  markers   in   order   to   more   comprehensively   define   thevariations   in   individual   response   to   burn   and   infection   [5].Procalcitonin   measurement   is   now   routinely   used   toconfirm   bacterial   infection   in   critically   ill   patients.   Procalci-tonin   also   seems   to   be   a   useful   marker   for   the   detection   of septic   complications   in   burn   patients   [6–10].   Besides   its   role   asa   marker   of    infection,   procalcitonin   has   been   shown   to   behelpful   in   determining    the   effectiveness   and   appropriateduration   of    antibiotic   therapy   in   critically   ill   patients   [11,12].Evidence   from   clinical   trials   shows   that   the   use   of    algorithmsbased   on   PCT   levels   leads   to   an   important   reduction   inantibiotic   use   in   critically   ill   patients   [13,14].   However   the   roleof    PCT   in   monitoring    the   effectiveness   of    antibiotic   therapy   inburn   ICU   patients   has   not   been   reported,   to   our   knowledge.The   diagnostic   performance   of    PCT   has   also   not   beenevaluated   for   a   variety   of    localized   bacterial   infections   inburns,   such   as   lower   respiratory   tract   infection,   burn   woundand   urinary   tract   infections.   No   reports   have   described   thePCT   levels   in   sepsis   caused   by   different   types   of    microorgan-isms   (Gram   negative   and   Gram   positive).The   objectives   of    this   study   were:    To   examine   the   accuracy   of    PCT   to   diagnose   sepsis   andlocalized   infectious   complications   in   burn   ICU   patients   andto   compare   the   diagnostic   value   of    PTC   with   traditionalmarkers   of    inflammation   such   as   C-reactive   protein   (CRP)and   white   blood   cells   count   WBC).    To   estimate   the   differences   in   PCT   level   according    to   thetype   of    microbial   agent   responsible   for   causing    the   infection.    To   evaluate   the   ability   of    PCT   to   assess   the   effectiveness   of antibiotic   therapy   in   these   patients.    To   evaluate   the   relationship   of    PCT   levels   to   patients   outcome. 2.   Materials   and   methods This   prospective   study   was   conducted   between   2005   and   2010and   performed   at   the   Burn   Unit   of    ‘‘G.   Papanikolaou’’   GeneralHospital   in   Thessaloniki.   The   study   protocol   was   approved   bythe   local   ethic   committee. Patients :   All   consecutive   patients   admitted   to   our   ICU   wereincluded   in   this   study.   Exclusion   criteria   were:   nonsurvivableburn   (decision   for   comfort   care   on   admission)   [ n :15],   admis-sion   for   non-burn   diagnosis   (TEN,   reconstructive   surgery)   [ n :8],admission   for   less   than   72   h   [ n :19],   age   less   than   18   years   [ n :2],and   burn   size   less   than   20%   of    total   burn   surface   area   (TBSA)[ n :35].We   prospectively   investigated   data   of    145   remaining patients.   Demographic   and   clinical   data   were   recorded   foreach   patient.   The   APACHE   II   score   (Acute   Physiology   andChronic   Health   Evaluation   II)   and   SAPS   II   score   (Simplified   AcutePhysiological   Score   II)   were   used   to   grade   illness   severity.   For   thediagnosis   of    multiple   organ   dysfunction   or   failure,   the   Sequen-tial   Organ   Failure   Assessment   score   (SOFA)   was   calculated   dailyuntil   discharge   from   ICU,   maximumSOFA   scores   and   ICUmortality   were   also   recorded.   In   all   patients   PCT   and   CRPplasma   concentrations   and   white   blood   cell   count   (WBC)   weremeasured   daily   during    ICU   stay   (firstsample   was   performedwithin   the   first   24   h   from   admission).   All   patients   had   a   urinarycatheter   in   placeto   monitor   urine   output. 2.1.   Infection   criteria Each   patient   was   examined   for   signs   and   symptoms   of infection   at   the   time   of    admission   and   daily   thereafter   untiltheir   discharge   from   the   ICU   or   their   death.   Diagnosis   of    sepsisand   septic   shock   were   performed   according    to   the   currentrecommendations   [5,15].Diagnosis   of    localized   infections   (respiratory   tract,   burnwound,   urinary   tract)   were   defined   according    to   the   AmericanBurn   Association   consensus   conference   recommendations   aswell   as   by   using    additional   criteria   recommended   in   bibliog-raphy   [5,16,17].Pneumonia   was   defined   by   the   presence   of    a   new   orprogressive   infiltrate   in   addition   to   at   least   two   out   of    threeclinical   features   (fever   greater   than   38   8 C,   leukocytosis   orleukopenia,   and   purulent   secretions)   plus   a   positive   quantita-tive   culture   of    samples   obtained   either   by   bronchoalveolarlavage,   or   by   protected   specimen   brush,   using    diagnosticthresholds   of    10 4 colony   forming    units   (CFU)/ml   and   10 3 CFU/ml   respectively.Catheter-associated   urinary   tract   infection   (CA-UTI)   wasdefined   by   the   presence   of    significant   bacteriuria   (  10 5 CFU/ml)   with   no   more   than   2   species   of    microorganisms   in   apatient   with   signs   and   symptoms   relative   to   the   urinary   tract.Burn   wound   infections   were   characterized   by   the   presenceof    the   following    criteria:   (a)   clinical   signs   of    wound   infection(purulent   secretions,   color   changes,   pain,   erythema,   unex-pected   change   in   the   appearance   and   the   depth   of    the   wound,conversion   of    partial   sickness   injury   to   full   sickness   necrosis,non-viable   grafts),   and   (b)   burn   wound   biopsy   with    10 5 colo-ony-forming    units/g    tissue,   or   quantitative   swab   with   countsof    10 6 bacteria   obtained   from   surface   swab   samples   or   ahistological   diagnosis   of    burn   wound   infection   based   on   theobservation   of    microorganisms   invading    viable   tissue   beneaththe   eschar   surface. 2.2.   Laboratory   measurement Blood   samples   from   patients   were   drawn   from   indwelling arterial   lines   for   measurement   of    CRP,   PCT   and   routinelaboratory   tests   including    WCC.   Serum   PCT   levels   weredetermined   by   immunoluminometric   assay   (Lumitest   PCT,Brahms   Diagnostica,   Berlin,   Germany).   The   lower   detectionlimit   was   0.08   ng/ml.   The   CRP   concentration   was   measured   bya   Boehringer   Mannheim   (BM)/Hitachi   automated   immuno-turbidimetric   (Tuna-quant,   BM,   Germany)   technique.   Thelower   detection   limit   was   set   at   0.2   mg/dl.   White   blood   cell(WBC)   and   platelet   counts   were   determined   by   using    anautomatic   counter   (Gene-s;   Coulter,   Paris,   France).   The   normalrange   for   the   WBCcount   was   4300–10,800   cells/mm 3 ,thenormal   range   for   the   platelet   count   was   150,000–450,000   cells/mm 3 .   The   blood   gas   analyzer   (Radiometer   ABL   725   analyzer;Radiometer   A/S,   Copenhagen,   Denmark)   directly   measuredlactate   levels.   The   range   of    normal   serum   lactate   levels   was0.5–2.2   mmol/l. 2.3.   Statistical   analysis Values   are   presented   as   median   and   interquartile   range   (IQR)(25th   to   75th   percentiles),   absolute   value   and   percentage,   or burns38(2012)356–363 357  mean      standard   deviation.   The   nonparametric   data   werecompared   with   the   Mann–Whitney   U   test,   and   categoricalvariables   were   compared   with   the   chi-square   test.   PCT   kineticis   expressed   as   delta   PCT   ( D PCT)   concentrations.   D PCT   wascalculated   as   the   difference   between   concentrations   on   day   0and   72   h   (day   0   to   72   h).   The   sensitivity,   specificity,   andpositive   likelihood   ratio   for   PCT   cutoff    levels   were   calculated.To   determine   the   predictive   ability   of    PCT   receiver   operating characteristic   (ROC)   curves   were   constructed   and   the   areasunder   the   curve   (AUCs)   were   calculated   with   95%   confidenceintervals   (CIs).   Analysis   of    the   coordinate   points   of    the   ROCcurves   revealed   the   best   threshold   values   for   prediction   of outcomes   in   each   case   based   on   the   supreme   combination   of sensitivity   and   specificity   levels   of    each   proposed   threshold.Correlations   between   inflammation   markers   and   TBSA   werestudied   using    Pearson   bivariate   correlation   coefficient.   Logis-tic   regression   analysis   was   used   to   examine   the   impact   of parameters   on   mortality.   A    p   value   of    less   than   0.05   wasconsidered   to   be   statistically   significant   in   all   tests.   Theanalyses   were   performed   using    SPSS   17.0   software   (SPSS   Inc.,Chicago,   IL,   USA). 3.   Results Data   of    145   patients   were   evaluated   (48.2      18.3   years,   55%male,   38.8      18%   TBSA,   11.5      4   APACHE   II,   29      10   SAPS   II,   4.4(2–6)   SOFA).   The   mortality   rate   of    patients   was   16.5%.   Multipleorgan   failure   due   to   septic   complications   was   the   main   causeof    death.   Concentrations   of    PCT   at   admission   and   maximumPCT   levels   were   0.69   (0.3–1.4)   ng/ml,   and   7.8   (0.3–139)   ng/ml,respectively.   No   correlations   between   TBSA   and   admissionPCT   (    p   =   0.299)   or   CRP   (    p   =   0.3)   levels   were   observed.   Acorrelation   between   the   maximum   PCT   and   TBSA   was   found( r   =   0.87,    p   =   0.004).   Maximum   procalcitonin   (    p   =   0.004)   andmaximum   SOFA   score   (    p   =   0.011)   were   independent   predictorsof    outcome   in   logistic   regression   analysis.Sepsis   of    bacterial   etiology   was   diagnosed   in   86   out   of    145(59.3%)   patients,   43   of    whom   had   septic   shock.   The   distribu-tion   of    septic   patients   who   metthe   clinical   and   laboratorycriteria   of    inflammation   and   infection   as   they   are   specified   byAmerican   Burn   Association   consensus   conference   recom-mendations   was   the   following:   temperature   ( > 39   8 C   or < 36.5   8 C)   –   58   (67.4%)   patients;   progressive   tachycardia   –   82(95.3%)   patients;   progressive   tachypnea   –   69   (80.2%)   patients;thrombocytopenia   –   44   (51%)   patients;   hyperglycemia   –   68(79%)   patients,   inability   to   continue   enteral   feedings   for   morethan   24   h   –   55   (64%)   patients.   Positive   cultures   were   identifiedin   75   patients   with   sepsis,   62   septic   patients   had   a   clinicalresponse   to   antibiotic   therapy   and   in   3   patients   the   diagnosiswas   confirmed   with   histological   tissue   analysis.Patients   with   septic   shock   had   significantly   higher   SOFAscores   [8.9   (IQR,   3–13)]   and   PCT   concentrations   [23.9   ng/ml(IQR,   1.6–34)]   on   day   1   of    sepsis   compared   with   patientswithout   septic   shock   [SOFA   score-5.6   (IQR,   3–6),   PCT-5.6   ng/ml(IQR,   0.4–8)],    p   =   0.04,    p   =   0.001,   respectively.   No   significantdifferences   were   found   in   CRP   and   WBCbetween   patients   withand   without   septic   shock. Table   1   –   Inflammatory   markers   and   SOFA   score   in   sepsis,   respiratory   tract    infection   (with   and   without    sepsis),    burnwound   infection   (with   and   without    sepsis)   and   urinary   tract    infection. PCT,   median(IQR),   ng/mlCRP,   median(IQR),   mg/dlWBC,   median(IQR),   (10 9    l  1 )Temp,   median(IQR),   ( 8 C)SOFA,   median(IQR) Sepsis(allsepticpatients, n :86)7.2(0.4–33)17.65(2–39)14.6(2–33)37.9(33–41)7(2–15)Pre   infection0.44(0.1–1.8)12.0(4.5–30.7)12.7(4.6–13.7)37.7(36–39)3(1–9)  p * < 0.001 < 0.0010.0040.198 < 0.001Sepsis(unknownetiology, n :11)8.9(0.6–21)19.5(15–26)15.4(3.3–24)38.5(35–39)7.9(4–14)Preinfection0.48(0.2–1.2)10.5(6–26.8)10.3(5–16.5)38.1(35.5–39)2.7(2–7)  p * < 0.001 < 0.001 < 0.010.07 < 0.001Bloodstreaminfection( n :28)10.4(4.1–25)24.3(11.8–38)18(8.4–23)38.6(38–40)6.7(5–8)Preinfection0.46(0.2–1.8)9.3(7–12.8)15.6(8.9–13.5)37.7(35–39)3(2–6)  p * < 0.001 < 0.0010.070.060.01Respir.tractinfection(withsepsis, n :33)6.3(1.6–19)23.6(17.7–27)12.2(5.8–13)37(33–38.5)6(5–9)Pre   infection0.25(0.1–0.4)13.7(11.5–15.4)7.9(5.8–11.4)36.2(35–37.8)4(3–7)  p * < 0.001 < 0.001 < 0.010.070.055Respir.tractinfection(withoutsepsis, n :24)0.73(0.1–4)17.0(10.5–38)10.9(24–26.7)38.1(34–40)5(2–9)Pre   infection0.38(0.1–3.6)16.150(10.5–28.7)9.9(4.8–18.5)37.1(35.5–39)3(1–9)  p * 0.0110.7430.480.013 < 0.001Woundinfection(withsepsis, n :14)4.5(0.8–8)13.5(3.8–15.7)20.5(11–24)37.6(35–39)6.3(5–9)Preinfection0.4(0.1–0.8)10.8(9–11)10(4.5–13)38.6(38=39)2.6(2–3)  p * < 0.0010.060.010.070.01Woundinfection(withoutsepsis, n :20)0.87(0.2–5.4)18.5(1.9–38.5013,400(5000–31,000)38.3(35–40)3(1–8)Pre   infection0.35(0.1–3.4)11.0(6.5–28)10,640(3800–22,300)37.3(36–40)3(1–7)  p * < 0.0010.120.050.3290.177Urinarytractinfection(withoutsepsis, n :9)0.36(0.1–0.4)14.2(7.7–14.5)11,360(7700–23,500)38.6(37–38.9)2.5(1–4)Preinfection0.2(0.1–0.34)12.9(7.3–21.2)10,500(8380–13,500)38.3(37–39)2.4(2–4)  p * 0.5450.2370.4110.3590.9PCT,procalcitonin;CRP,C-reactiveprotein;WBC,whitebloodcellcount;Temp,temperature;SOFA,SequentialOrganFailureAssessmentscore. * Differencesbetweenpreinfectionandinfectionlevels. burns38(2012)356–363 358  Out   of    86   septic   patients   11   had   sepsis   of    unknown   etiology,33   patients   had   sepsis   of    respiratory   tract   etiology,   28   patientshad   blood   stream   infection   and   14   patients   had   burn   woundinfection   with   sepsis.   All   patients   with   blood   stream   infectionmetthe   ABA   criteria   of    sepsis   and   were   considered   as   septicpatients.   No   septic   patients   with   urinary   tract   infection   werediagnosed.Septic   complications   were   caused   by   Gram   negativebacteria   (  Acinetobacter   baumannii ,   Pseudomonas   aeruginosa , Klebsiella    pneumoniae ,   Proteus   mirabilis )   in   74.4%   of    patientsand   were   caused   by   Gram   positive   bacteria   ( Staphylococcusaureus ,   Enterococcus    faecium , Enterococcus    faecalis ),in   25.6%   of patients.   Procalcitonin   concentrations   were   lower   in   patientswith   Gram   positive   sepsis   (2.9   (0.4–4.7)   ng/ml)   compared   to   thepatients   with   Gram   negative   sepsis   (8.2   (0.5–43)   ng/ml),  p   =   0.01.Localized   infections   were   diagnosed   in   53   patients.   Thetypes   of    localized   infection   were:   respiratory   tract   infection   in24   patients   (45.2%),   burn   wound   infection   in   20   patients   (37.7%)and   urinary   tract   infection   in   9   patients   (17%).   The   mainbacterial   strains   responsible   for   infection   were   Gram   negativebacteria:   P.   aeruginosa ,  A.   baumannii   and   K.    pneumoniae .Diagnosis   of    respiratory   tract   infection   (with   and   withoutsepsis)   was   confirmed   by   positive   quantitative   culture   of samples   obtained   by   bronchoalveolar   lavage   (21   patients)   andby   protected   specimen   brush   (36   patients).Diagnosis   of    wound   infectionwas   relied   on   clinical   signsof    infection   and   culture   data[burn   wound   biopsy(22patients),quantitative   swab   (9   patients)andhistologicalfindings   (3   patients)].The   levels   of    PCT   were   higher   in   patients   with   the   diagnosisof    sepsis,   respiratory   tract   infection   and   wound   infectioncompared   to   their   preinfection   concentrations   (Table   1).The   performance   of    PCT   in   the   diagnosis   of    septic   and   nonseptic   infectious   complications   is   shown   in   Table   2,   while   theROC   curve   for   PCT   as   a   predictor   for   sepsis   is   shown   in   Fig.   1.The   selected   value   of    1.5   for   PCT   was   based   on   the   ROCanalysis   since   different   PCT   measurements   led   to   slightly   lessdesirable   sensitivity   and   specificity   levels   (values   of    1.3   or   1.6had   sensitivity   and   specificity   of    88–90.4%   or   82.5–95.2%,respectively).In   62   patients   the   sepsis   was   successfully   treated,   theremaining    24   patients   died   due   to   septic   complications   andmultiple   organ   failure.   Patients   with   treatment   failure   hadhigher   PCT   levels   on   day   3   of    treatment   (19   ng/ml,   IQR,   5.4–33)as   compared   to   the   PCT   values   at   the   onset   of    sepsis   (10.56   ng/ml,   IQR,   2.6–29),    p   =   0.001.   C-reactive   protein   levels   were   alsoincreased   on   day   3   in   patients   with   treatment   failure   [22   mg/dl(IQR,   15–33)   vs.   28.2   mg/dl   (IQR,   18–49),    p   =   0.053].   The   PCTvalues   were   decreased   in   patients   with   successfully   treatedsepsis,   but   the   C-reactive   protein   levels   showed   no   significantchange   over   the   first   3   days   in   these   patients   (    p   =   0.93)   andwere   decreased   only   after   day   5   of    the   treatment   (Table   3).A   substantial   decrease   in   PCT   concentration   between   thefirst   and   the   second   time   points   (day   1   vs.   day   3,   D PCT)   was Fig.   1   –   Receiver   operating   characteristic   curve   displayingthe   diagnostic   performance   of    PCT   in   sepsis.Table   2   –   Diagnostic   performance   of    PCT,   in   sepsis,   and   in   localized   infection. Type   of    infection   Sepsis, n :86Respiratory   tract   infection(without   sepsis),   n :24Wound   infection(without   sepsis),   n :20 Cutoffvalue(ng/ml)1.50.520.56Sensitivity(%)88.377.075.6Specificity(%)92.388.580.5LR+11.477.674LR  0.130.20.29PPV(%)91.888.580.0NPV(%)88.883.377.3Diagnosticaccuracy(%)90.385.778.6AUC0.9660.8590.828ROC   significance0.00010.00010.000195%CILower0.9440.7410.735Upper0.9870.9780.921LR+,positivelikelihoodratio;LR  ,negativelikelihoodratio;PPV,positivepredictivevalue;NPV,negativepredictivevalue;ROC,receiveroperatingcharacteristicscurve;AUC,areaunderROC. burns38(2012)356–363 359
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x