Impact of International Nosocomial Infection Control Consortium (INICC) strategy on central line-associated bloodstream infection rates in the intensive care units of 15 developing countries

Please download to get full document.

View again

of 7
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.
Document Description
Impact of International Nosocomial Infection Control Consortium (INICC) strategy on central line-associated bloodstream infection rates in the intensive care units of 15 developing countries
Document Share
Document Tags
Document Transcript
  Impact   of    an   International   Nosocomial   Infection   Control   Consortiummultidimensional   approach   oncentral   line-associated   bloodstreaminfection   rates   in   adult   intensive   care   unitsineightcitiesin   India NamitaJaggi a ,CamillaRodrigues b ,VictorDanielRosenthal c, *,SubhashKumarTodi d ,SwetaShah e ,   NarinderSaini f  ,ArpitaDwivedy g ,F.E.Udwadia h ,Preeti   Mehta i ,MuraliChakravarthy  j ,SanjeevSingh k ,SamirSahu l ,   DeepakGovil k ,AshitHegd b ,FarahadKapadia b ,ArpitaBhakta d ,   MahuyaBhattacharyya d ,Tanu   Singhal e ,ReshmaNaik e ,VatsalKothari e ,AmitGupta f  ,SuvinShetty g ,SheenaBinu g ,PreethiPinto g ,ArunaPoojary h ,GeetaKoppikar h ,LataBhandarkar h ,ShitalJadhav h ,NeerajChavan h ,ShwetaBahirune h ,ShilpaDurgad h ,GitaNataraj i ,   PallaviSurase i ,B.N.Gokul  j ,R.Sukanya  j ,LeemaPushparaj  j ,KavithaRadhakrishnan k a  Artemis   Health   Institute,   New   Delhi,   India b PD   Hinduja   National   Hospital   &    Medical   Research   Centre,   Mumbai,   India c International   Nosocomial   Infection   Control   Consortium,   Corrientes    Ave   #4580,   Floor    12,    Apt    D,Buenos    Aires,   1195,    Argentina d  AMRI    Hospitals,   Kolkata,   India e Kokilaben   Dhirubhai    Ambani   Hospital,   Mumbai,   India f  Pushpanjali   Crosslay   Hospital,   Ghaziabad,   India g Dr    L.H.   Hiranandani   Hospital,   Mumbai,   India h Breach   Candy   Hospital   Trust,   Mumbai,   India i Seth   GS    Medical   College,   Mumbai,   India  j Fortis   Hospitals,   Bangalore,   India k  Amrita   Institute   of    Medical   Sciences   &    Research   Center,   Kochi,   India l Kalinga   Hospital,   Bhubaneswar,   India International    Journal   of    Infectious   Diseases   17   (2013)   e1218–e1224 A   R    T   I   C   L    E   I   N   F   O  Articlehistory: Received   18   March   2013 Receivedinrevisedform14June2013 Accepted   11    July   2013 Corresponding    Editor:   Eskild   Petersen,Aarhus,   Denmark Keywords: Catheter-related   infectionsBundleInternational   Nosocomial   Infection   ControlConsortiumMultidimensional   approachHand   hygieneDeveloping   countries SU   MMA   R    Y  Objective: Toevaluatetheimpactof    theInternationalNosocomialInfection   Control   Consortium(INICC)multidimensionalinfectioncontrolapproachoncentralline-associatedbloodstreaminfection(CLABSI)ratesineightcitiesofIndia. Methods: Thiswas   aprospective,before-and-aftercohortstudyof    35650patientshospitalizedin16adultintensivecareunitsof11hospitals.Duringthebaselineperiod,outcomesurveillanceof    CLABSI   wasperformed,applyingthedefinitionsoftheCDC/NHSN(USCenters   forDiseaseControland   Prevention/NationalHealthcareSafety   Network).Duringtheintervention,theINICCapproachwasimplemented,whichincludedabundleofinterventions,education,outcome   surveillance,processsurveillance,feedbackonCLABSIrates   andconsequences,andperformancefeedback.RandomeffectsPoissonregressionwasusedforclusteringof    CLABSIratesacrosstimeperiods. Results: Duringthebaselineperiod,9472centralline   (CL)-daysand61CLABSIswererecorded;duringtheinterventionperiod,80898CL-daysand404CLABSIswererecorded.Thebaselineratewas   6.4CLABSIsper   1000CL-days,whichwasreducedto   3.9CLABSIsper   1000CL-daysinthesecond   yearandmaintainedfor36monthsoffollow-up,accountingfora53%   CLABSIratereduction(incidencerateratio0.47,95%confidenceinterval0.31–0.70;  p   =   0.0001). Conclusions: Implementing   thesixcomponentsoftheINICCapproachsimultaneouslywasassociatedwithasignificantreductionintheCLABSI   rateinIndia,which   remainedstableduring36monthsof follow-up.  2013InternationalSocietyforInfectiousDiseases.PublishedbyElsevierLtd.Allrightsreserved. * Corresponding   author.   Tel.:/fax:   +54   11   4861   5826. E-mail   address:   (V.D.   Rosenthal). URL: Contents   lists   available   at   ScienceDirect InternationalJournalofInfectiousDiseases jou   rnal   h   o   mepag   e:www.elsevier   .co   m    /locate/ijid 1201-9712/$36.00   –   see   front   matter      2013   International   Society   for   Infectious   Diseases.   Published   by   Elsevier   Ltd.   All   rights   reserved.  1.   Introduction Central   line-associated   bloodstream   infections   (CLABSIs)   areresponsible   for   increased   lengths   of    hospital   stay   and   increasedattributable   mortality   in   high-income   countries 1 and   in   limited-resource   countries, 2 including   India. 3 CLABSIs   are   also   responsiblefor   increased   health   care   costs,   as   reported   in   studies   from   high-income   countries 1 and   from   some   limited-resource   countries   inLatin   America; 2,4 however,   no   data   on   costs   of    CLABSIs   areavailable   from   India.The   incidence   ofCLABSI   ismany   times   underestimated   inlimited-resource   countries,   as   basic   infection   control   programs   arenot   systematically   implemented. 5 Device-associated   healthcare-acquired   infection   (DA-HAI)   ratesin   the   intensive   care   units   (ICUs)of    limited-resource   countries   are   three   to   five   times   higher   than   inhigh-income   countries,   as   reported   by   the   International   Nosoco-mial   Infection   Control   Consortium   (INICC)   in   pooled   studies, 5 andparticularly   inIndia. 3 In   developing   countries,   the   socioeconomic   level   of    the   countryhas   an   impact   on   DA-HAI   rates   in   the   pediatric 6 and   neonatal 7 ICUsettings.   However,   far   too   little   attention   has   been   paid   to   this,   asonly   two   studies   addressing   this   issue   havebeen   published.   Theresults   of    one   ofthe   studies   showed   that   lower-middle-incomecountries   had   higher   CLABSI   rates   than   upper-middle-incomecountries   in   pediatric   ICUs(12.2   vs.   5.5   per   1000   central   line   (CL)-days). 6 Similarly,   inthe   other   study,   CLABSI   rates   were   significantlyhigher   in   neonatal   ICU   patients   from   low-income   countries   than   inthose   from   lower-middle-income   countries   or   upper-middle-income   countries   (37.0   vs.   11.9   (  p   <   0.02),   and   vs.   17.6(  p <   0.05)   CLABSIs   per   1000   catheter-days,   respectively). 7 Unfor-tunately,   no   studies   from   developing   countries   that   have   analyzedthis   issue   inadult   ICUs   (AICUs)   are   available.Indeveloped   countries,   ithas   been   demonstrated   thatsurveillance   is   fundamental   to   the   prevention   of    CLABSIs,   whichcan   be   reduced   by   more   than   30%. 8 Implementing   infection   controlbundles,   including   the   five   interventions   (1)   hand   hygiene,   (2)   skinantisepsis   with   chlorhexidine,   (3)   maximal   barriers,   (4)   insertionintothe   subclavian   vein,   and   (5)   timely   CL    removal,   has   beenassociated   with   areduction   in   the   incidence   density   of    CLABSI   indeveloped   countries. 9 Thepresentstudywasdesignedto   determinetheeffectofamultidimensionalprogram   forCLABSIpreventionin16   AICUsof 11   hospitals,in   eightcities   ofIndia.   Ourprogram   wasimple-mentedfromSeptember   2004toFebruary   2012and   includedsixsimultaneousinterventions:   (1)practice   bundle,   (2)educa-tion,   (3)outcomesurveillance,   (4)   process   surveillance,(5)feedbackof    CLABSI   rates,and(6)performancefeedback   oninfectioncontrolpractices.Thedesign   oftheINICCmultidimen-sionalapproachfollows   thebasicrecommendationspublished   intheguidelinesoftheSocietyforHealth   Care   Epidemiologyof America(SHEA)   and   the   InfectiousDiseasesSociety   ofAmerica(IDSA)in2008. 10 So   far   there   has   been   no   systematic   research   on   the   effect   of such   an   approach   inIndia,   the   second   most   populous   country   in   theworld,   with   a   population   of    around   1300   000   000   people.   Thisprovided   sufficient   ethical   and   theoretical    justification   forconducting   this   particular   study,   and   through   itspublication,increase   and   spread   awareness   of    thispublic   health   burden   inIndia. 11 2.   Methods  2.1.   Background   on   INICC  Founded   in   Argentina   in1998,   the   INICC   was   the   firstmultinational   research   network   established   to   control   and   reduceHAIsat   the   international   level   through   the   analysis   of    datacollected   on   avoluntary   basis   by   apoolof    hospitals   worldwide. 12 The   goals   of    the   INICC   include   the   development   ofadynamic   globalhospital   network   that   applies   systematic   surveillance   of    HAIs   withstandardized   definitions   and   methodologies,   the   promotion   of evidence-based   infection   control   practices,   and   the   performance   of applied   infection   control   research   toreduce   ratesof    HAIs,   theassociated   mortality,   excess   lengths   of    hospital   stay,   costs,   andbacterial   resistance. 13  2.2.   Setting    and   study   design The   study   was   conducted   in   16   AICUs   in   11   hospitals,   allmembersofthe   INICC,   in   eight   cities   of    India.   Each   hospital   hadbeen   activelyparticipating   inthe   INICC   surveillance   program   for   aminimum   of 6   months,withan   infection   control   team   (ICT)comprisedofinfectioncontrol   professionals   (ICPs)   and   amedical   doctor   witha   formaleducation   and   background   in   internal   medicine,   critical   care,infectious   diseases,   and/or   hospital   epidemiology.This   prospective,   before-and-after   study   was   performed   overtwo   time-periods:   the   baseline   period   and   the   intervention   period.The   institutional   review   board   at   each   hospital   approved   the   studyprotocol   (Figure   1).  2.3.Baseline    period The   baselineperiod   included   onlytheperformanceof    outcomesurveillanceandprocess   surveillance.Thelength   ofthebaselineperiodwas3monthsforthefollowing   threereasons:   (1)   This   is   thetime   neededto   conduct   the   following   activitiesatINICCheadquarters   inArgentina   on   a   monthlybasis:   receivingthosecasereport   forms(CRF)completed   atallparticipating   ICUsfromIndia;conductinga   validation   process   of    filledCRFs;sending 6.442.611.882.493.514.794. 1-3months 4-6months 7-9months 10-12months 13-18months 19-24third year Central Line-Associated Blood Stream Infecon Rates Figure   1.   Central   line-associated   blood   stream   infection   rates   by   period. N.Jaggi   etal.    /    International    Journal   of    Infectious   Diseases   17(2013)   e1218–e1224   e1219  queries   to   participating   ICUs;receiving   andanalyzingrepliestoqueries;   uploading   andanalyzingCRF   data;producingmonthlyreportscontaining   chartsandtables   with   theresultsof    outcomeandprocesssurveillance;   sendingmonthlyreportsto   each   ICU;presenting   the   monthly   report   to   healthcareworkers   (HCWs)at   theparticipatingICUsatmonthly   infectioncontrol   meetings,with   the   aimof    providingfeedbackonCLABSIratesand   theirconsequences   andperformancefeedback   to   increaseawarenessabout   CLABSIsandto   improve   compliancewithinfection   controlpractices. 11 (2)Thesample   size   of    patientsandnumberof    monthsofdatacollection   during   thebaselineperiodaresufficientforcomparison   with   the   samplesizeof    patientsandnumberof monthsof    data   collection   during   theintervention   period.From   astatisticalperspective,theissue   is   addressedbyconsideringthechangesin   ratesovertime.Therelativelyshort   baseline   periodmayhavehadan   impacton   thestandarderrors   ofourestimates.However,   we   found   that   this   didnotbiastheresults,   becausetherewere   no   systematicdifferencesbetween   thetwogroups.   (3)Ourpriority   wasto   start   theintervention   asearly   aspossibleinordertoachieve   thedesired   results—reducing   theCLABSIand   mortalityratesas   soon   aspossible.  2.4.Intervention    period The   intervention   period   was   initiatedafter   3months   of participation   in   the   INICC   program.   This   was   a   cohort   study,   andforthat   reason,   each   ICU    joined   the   INICC   program   atadifferent   time.Thus,by   theday   we   analyzedthe   impact   of    the   INICCintervention,the   ICUshad   undergone   different   lengths   ofintervention.   Theaveragelength   of    the   intervention   period   was   17.54   months(standard   deviation   8.9months,   range   5–36   months).  2.5.   INICC    multidimensional   infection   control   approach The   INICC   multidimensional   infection   control   approach   includ-ed   the   following   items:   (1)   bundle   of    infection   control   interven-tions,   (2)   education,   (3)   outcome   surveillance,   (4)   processsurveillance,   (5)   feedback   of    CLABSI   rates,   and   (6)   performancefeedback   of    infection   control   practices.  2.5.1.   Components   of    the   central   line-care   bundle    for    CLABSI  Components   of    the   CL-care   bundle   for   CLABSI   were   as   follows: 10 1.   Perform   hand   hygiene   before   CL    insertion   or   manipulation; 14 2.   Use   sterile   gauze   or   a   transparent   sterile   dressing   tocover   theinsertion   site; 10 3.   Maintain   a   good   condition   of    the   sterile   dressing.   Change   thegauze   every   48   hand   the   transparent   dressing   every   7   days; 10 4.   Remove   the   CL    asearly   as   possible,   when   notnecessary; 10 5.   Change   the   administration   set   every   96   h;   unless   used   for   fat,nutrition,   or   blood   precuts,   and   in   these   cases   change   every24h; 10 6.   Use   achlorhexidine-based   antiseptic   for   skin   preparation; 10 7.   Preferably   use   the   subclavian   vein; 10 8.   Use   an   all-inclusive   catheter   cartorkit; 15 9.   Use   maximal   sterile   barrier   precautions   during   CL    insertion; 10 10.   Avoid   using   several   times   those   vials   meant   to   be   used   onlyonce; 10 11.   Disinfect   line   hubs,   needleless   connectors,   and   infection   portsbefore   accessing   the   CL. 10 Other   effective   interventions   were   discussed   but   notappliedbecause   of    budget   limitations,   namely   the   following   five   practiceswere   partially   applied   or   not   applied: 1.   Use   of    a   split   septum   instead   ofmechanical   valves   orathree-way   stopcock; 10 2. Use   ofachlorhexidine   impregnated   sponge   at   the   insertionsite; 10 3. Daily   bath   with   chlorhexidine; 10 4. Use   of    antimicrobial   impregnated   catheters; 10 5. Use   ofclosed   collapsible   flexible   containers   instead   of    opensemi-rigid   vented   or   glass   vented   IV   containers. 16  2.5.2.   Education Monthly   sessions   ofeducation   for   CLABSI   prevention   wereprovided   by   the   ICP   tothe   HCWs   in   charge   ofthe   insertion,   care,and   maintenance   of    CLs,   based   on   SHEA   and   IDSA   guidelines. 10,11  2.5.3.   INICC    surveillance   methods TheINICC   surveillance   program   included   two   components:outcome   surveillance   (DA-HAI   rates   and   their   adverse   effects,including   mortality   rates)   and   process   surveillance   (adherence   tohand   hygiene   (HH)   and   other   basic   preventive   infection   controlpractices). 13 Investigators   were   required   to   complete   outcome   and   processsurveillance   forms   in   their   ICUs,   which   were   then   sent   for   monthlyanalysis   to   the   INICC   headquarters   office   in   Buenos   Aires.  2.5.4.Outcome   surveillance Outcome   surveillance   included   rates   of    CLABSI   per   1000   CL-days,   use   of    invasive   devices   (CL,   mechanical   ventilator,   andurinary   catheter),   severity   of    illness   score,   underlying   diseases,   useofantibiotics,   culture   taken,   microorganism   profile,   bacterialresistance,   length   ofstay,   and   mortality   in   the   ICU. 13 CLABSI   definitions   and   surveillance   methods   were   in   accor-dance   with   the   definitions   for   HAIdeveloped   by   the   US   Centers   forDisease   Control   and   Prevention   (CDC)   for   the   National   HealthcareSafety   Network   (NHSN)   program. 17 Additionally,   INICC   methodswereadapted   to   the   limited-resource   setting   of    developingcountries,   due   to   their   different   socioeconomic   status. 13 The   ASIS(Average   Severity   of    Illness   Score)   was   used   instead   of    the   APACHEII   score   (Acute   Physiology   and   Chronic   Health   Evaluation)   due   tobudget   limitations   of    participating   ICUsin   thislimited-resourcecountry.   Thus,   we   used   the   ASIS   score,   as   historically   used   by   theCDC   National   Nosocomial   Infections   Surveillance. 18  2.5.5.   Definitions When   CLABSI   was   suspected,   the   CL    was   removed   asepticallyand   the   distal   5cm   ofthe   catheter   cut   off    and   cultured   using   thestandardized   semi-quantitative   method. 17,18 Concomitant   bloodcultures   were   drawn   percutaneously   in   most   cases.   At   eachhospital,   standard   laboratory   methods   wereused   to   identifymicroorganisms,   and   standardized   susceptibility   testing   wasperformed. 17,18 A   laboratory   confirmed   CLABSI   was   defined   inthe   following   case:   A   patient   with   aCL    from   whom   a   recognizedpathogen   isisolated   from   one   or   more   percutaneous   blood   culturesafter   48   h   ofcatheterization;   the   pathogen   cultured   from   the   bloodisnot   related   to   an   infection   at   another   site;   and   the   patient   has   oneormore   of    the   following   signs   orsymptoms:   fever   (  38   8 C),   chills,or   hypotension.   Inthe   case   of    skin   commensals   being   isolated(diphtheroids,   Bacillus   spp , Propionibacterium   spp ,   coagulase-negative   staphylococci,   and   micrococci),   the   organism   had   to   havebeen   recovered   from   two   or   more   separate   blood   cultures. 17–19  2.5.6.   Process   surveillance Process   surveillance   was   designed   to   assess   compliance   witheasily   measurable   keyinfection   control   practices,   such   assurveillance   of    compliance   ratesfor   HH   practices   and   specificmeasures   for   the   prevention   of    CLABSI. 13 Due   to   budget   limitations,   only   five   outof11   components   of    thebundle   were   monitored: N.    Jaggi   et    al.    /    International    Journal   of    Infectious   Diseases   17    (2013)   e1218–e1224 e1220  1.   The   hand   hygiene   (HH)   compliance   rate   was   based   on   thefrequency   with   which   HH   was   performed   asindicated   in   HCWinfection   control   training.   Observing   ICPs   were   trained   to   recordHHopportunities   and   compliance   on   a   form,   during   randomlyselected   observation   periods   of    30   min   to   1   h,   3times   a   week.   Inparticular,   the   INICC   direct   observation   comprised   the   ‘‘FiveMoments   for   Hand   Hygiene’’   asrecommended   by   the   WorldHealth   Organization   (WHO).   The   ‘Five   Moments’   included   themonitoring   of    the   following   moments:   (1)   before   patientcontact,   (2)   before   an   aseptic   task,   (3)   after   body   fluid   exposurerisk,   (4)   after   patient   contact,   and   (5)   after   contact   with   patientsurroundings. 20 Although   HCWs   knew   that   hand   hygienepractices   were   regularly   monitored,   they   were   notinformedof    the   schedule   for   HH   observations. 2.   Data   on   compliance   with   CL    care   measures   were   recorded   5   daysaweek   on   a   form   that   evaluated   if    infection   control   procedureswere   correctly   carried   out   by   the   HCW.   The   ICP   observing   theactivity   in   the   AICU   completed   a   standardized   form   thatcontained   the   following   data:   total   number   of    inserted   CLs   foreach   patient   for   the   whole   ICU;   total   number   of    dressings   placedto   protect   the   puncture   site,   in   order   to   evaluate   the   number   of patients   with   asterile   dressing. 13 3.   The   ICP   completed   astandardized   form   that   contained   the   totalnumber   of    dressings   in   correct   condition   and   an   evaluation   of whether   the   dressing   was   clean,   dry,   and   correctly   adhered   tothepuncture   site,   so   as   to   evaluate   the   number   ofdressings   inthe   correct   condition. 13 4.   The   ICP   also   completed   a   standardized   form   that   contained   thetotal   number   of    cases   in   which   the   dates   in   the   administrationsetwere   written,   with   the   aim   of    measuring   the   number   of patients   with   the   number   ofdays   of    the   administration   set   inplace,   and   evaluating   ifthe   set   was   replaced   by   orbefore   96   h. 13 5.   Finally,   the   ICP   completed   a   standardized   form   including   thedateof    insertion   and   removal   ofeach   CL,   toevaluate   the   numberof    days   the   CL    was   inserted   and   the   earliest   possible   removal   of theCL    when   not   necessary.  2.5.7.   Feedback   on   DA-HAI    rates Every   month,   the   INICC   research   team   atINICC   headquarters   inBuenos   Aires,   prepared   and   sent   to   each   ICT   a   final   report   of    theresults   of    outcome   surveillance   data   sent   by   the   investigators   ateach   hospital,   i.e.,   monthly   DA-HAI   rates,   length   of    stay,   bacterialprofile   and   resistance,   and   mortality. 13 Feedback   on   DA-HAI   rates   was   provided   toHCWs   working   inthe   AICU   by   communicating   the   patient   outcomes.   The   resultingrates   were   reviewed   by   the   ICT   at   monthly   meetings,   where   chartswereanalyzed.   Statistical   graphs   and   visuals   were   displayed   inprominent   locations   inside   the   ICU   to   provide   an   overview   of    DA-HAIrates.   This   infection   control   tool   isimportant   to   increaseawareness   of    patient   outcomes   inthe   ICU   and   to   enable   ICT   and   ICUstaff    to   focus   on   the   necessary   issues   and   apply   specific   strategiestoimprove   high   DA-HAI   rates.  2.5.8.   Performance    feedback Upon   processing   the   hospital   process   surveillance   data   on   amonthly   basis,   the   INICC   research   team   at   INICC   headquarters   inBuenos   Aires   prepares   and   sends   to   each   ICT   afinal   report   of    theresults   of    process   surveillance   rates,   including   compliance   with   HHand   care   of    CLs. 13 Performancefeedback   is   providedto   HCWsworking   in   theAICUbycommunicatingtheassessment   ofpractices   routinelyperformed   bythem.   Theresulting   ratesarereviewed   bytheICTatmonthlymeetings,where   charts   areanalyzed;   statisticalgraphsandvisualsarepostedinside   theICUto   provide   anoverview   of ratesmeasuring   compliance   withinfectioncontrol   practices.   Thisinfectioncontrol   toolis   keytoenablingthe   ICTandICUstaff    tofocuson   thestrategiesnecessary   to   improvelow   compliancerates.  2.6.   Statistical   methods Patient   characteristics   during   the   baseline   and   interventionperiods   in   each   ICU   were   compared   using   Fisher’s   exact   test   fordichotomous   variables   and   an   unmatched   Student’s   t  -test   forcontinuous   variables.   95%   Confidence   intervals   (CI)   were   calculat-edusing   VCStat   (Castiglia).   Relative   risk   (RR)   ratios   with   95%confidence   intervals   (CI)   were   calculated   for   comparisons   ofratesof    CLABSI   using   Epi   Info   v.   6.  p -Values   of  < 0.05   by   two-sided   testswere   considered   significant.We   conducted   two   types   of    analysis   to   evaluate   the   impact   of ourintervention   on   CLABSI   rates.   First,   we   performed   an   analysis   tocompare   the   data   of    the   first   3   months   (baseline   period)   with   theremaining   pooled   months   (intervention   period),   using   the   RR,   95%CI,and    p -value.   Second,   in   order   to   analyze   a   progressive   CLABSIrate   reduction,   we   used   Poisson   regression.   We   divided   the   dataintothe   first   3   months   (baseline   period),   followed   by   a9-monthperiod   (intervention   period),   and   two   annual   follow-up   periods   forthe   second   and   third   years.   We   compared   the   CLABSI   rates   for   eachfollow-up   period   with   the   baseline   CLABSI   rate.   For   this   compari-son,   we   used   the   baseline   data   only   of    those   hospitals   thatcontributed   to   the   follow-up   during   that   period   (i.e.   excluding   fromthe   baseline   those   hospitals   with   long   lengths   offollow-up   thatonlycontributed   a   shorter   length   of    surveillance).   We   used   randomeffects   Poisson   regression   to   account   for   clustering   of    CLABSI   rateswithin   hospitals   across   time   periods.   These   models   were   estimatedusing   Stata   11.0.   For   thisanalysis   we   used   the   incidence   rate   ratio(IRR),   95%   CI,   and    p -value. 3.Results Duringthewhole   study,   atotalof35   650patients   werehospitalizedin   16   AICUsduring173056days,   amounting   to90370CL-days.Participatinghospitals   were   classified   accordingto   thetypeofhospitalandtypeofAICU.Thefirst   ICUstoparticipatewere   enrolledin   September2004,   andthemostupdateddata   includedouranalysisdatesfrom   February   2012(Table1).Patient   characteristics,   such   as   thoracic   surgery,   immunecompromise,   mechanical   ventilator   use,   and   ASIS,   were   similarin   the   two   periods   (Table   2).   However,   the   mean   age   of    patientswas   higher   during   the   intervention   period   (Table   2).In   relation   to   compliance   rates,   during   the   baseline   period,   HHcompliance   improved   significantly   and   the   CL    use   ratio   wasreduced.   Compliance   with   other   measures,   such   as   date   on  Table   1 Characteristics   ofthe   participating   ICUs   and   hospitals   (from   September   2004   to January   2011)Data   AICUs,   n   AICU   patients,   n Type   of    ICU,   n   (%)Cardiac   surgical   1   (6%)   3024Cardiac   medical   1   (6%)   2348Surgical   3   (19%)   5053Medical   4   (25%)   5230Adult   step-down   1   (6%)   1353Medical   surgical   6   (38%)   17   743AllAICUs   16   (100%)   35   650Type   of    hospital,   n   (%)Private   community   9   (82%)   28   682Academic   teaching   1   (9%)   2348Public   1   (9%)   4620Allhospitals   11   35   650ICU,   intensive   care   unit;   AICU,   adult   ICU. N.    Jaggi   etal.    /    International    Journal   of    Infectious   Diseases   17(2013)   e1218–e1224   e1221  administration   set,   placed   dressing   and   correct   condition   of dressing,   were   high   and   similar   in   the   two   periods   (Table   2).During   the   baseline   period,   we   recorded   9472   CL-days,   for   a   CL usemean   of    0.54.   There   were   61   CLABSIs,   for   an   overall   baselinerate   of    6.4   CLABSIs   per   1000   CL-days   (Table   2).Merging   all   datafrom   the   intervention   period,   during   the   implementation   of    themultidimensional   infection   control   program,   we   recorded   80   898CL-days,   for   aCL    use   mean   of    0.52.   There   were   404   CLABSIs   for   anincidence   density   of    3.9   per   1000   CL-days.   These   results   showed   anoverall   reduction   in   the   CLABSI   rate   from   baseline   of    39%(from   6.4to3.9   CLABSIs   per   1000   CL-days;   RR    0.61,   95%   CI0.5–0.8;  p =   0.0001)   (Table   2).Onthe   other   hand,   using   Poisson   regression,   we   found   aprogressive   CLABSI   reduction.   The   baseline   CLABSI   rate   (duringfirst   3   months   of    the   study)   was   6.4   CLABSIs   per   1000   CL-days,which   reduced   progressively   during   the   intervention   period   to   4.1CLABSIs   per   1000   CL-days   in   the   third   year,   amounting   to   a53%CLABSI   rate   reduction   (IRR    0.47,   95%   CI   0.31–0.70;    p   =0.0001)(Table   3).The   microorganism   profile   isshown   in   Table   4.The   predomi-nant   microorganism   in   both   phases   was   Klebsiella   spp . 4.   Discussion This   study   set   out   with   the   aim   of    assessing   the   effect   ofamultidimensional   infection   control   approach   inthe   ICU   setting   inIndia,   a   limited-resource   country.   If    compared   with   rates   indeveloped   countries,   the   baseline   rate   ofCLABSI   found   in   this   study(9.0   per   1000   CL-days)   isaround   eight-fold   higher   than   in   the   USA(1.1   per   1000   CL-days;   as   determined   by   the   CDC/NSHN 21 )andhigher   than   the   CLABSI   rate   determined   by   KISS   (1.4   per   1000   CL-days). 22  Table   2 Characteristics   of    patients,   hand   hygiene   compliance,   central   linecare   compliance,   central   line   use,   and   CLABSI   rates   inphase   1   (baseline   period)   and   phase   2   (interventionperiod)Patient   characteristics   Phase   1—baseline   Phase   2—intervention   RR    95%   CI    p -ValueStudy   period   by   hospital   in   months,   mean    SD   (range) 3   17.54    8.9(5–36) --   -Number   of    patients   3747   31   903--   -Bed-days, a n   17   395   155   661Number   of    CL    days, b n   9472   80   898CLduration,   mean    SD 2.53  4.4 2.54    5.8 --   0.9Number   of    MVdays, b n   4088   35   636MVuse, c mean   0.240.23   0.970.94–1.0   0.13Age,   years,   mean    SD 55.0  18.0 56.3    18.0 --   0.0001ASIS   score,   mean    SD 2.62  1.14 2.6    1.1 --   0.7Male   2590   (69%)   20   602   (65%)   0.940.9–0.97   0.0013Female   1156   (31%)   11   259(35%)   --   -Thoracic   surgery,   n   (%)11   (1%)   91   (1%)   0.970.52–1.82   0.94Immune   compromise,   n   (%)   11   (1%)   142   (1%)   1.520.82–2.81   0.18Hand   hygiene   compliance,   n   (%)73%   (1467/2013)   84%   (17   914/21   327)   1.151.1–1.22   0.0001Compliance   with   date   on   administration   set   %   ( n/n )97%   (4405/4532)   96%   (39   581/41   247)   0.990.96–1.02   0.42Compliance   with   placed   dressing   %   ( n/n )   99%   (4493/4532)   98%   (40   248/41   247)   0.980.95–1.02   0.313Compliance   with   correct   condition   ofdressing   %( n/n )   94.4%   (4277/4532)   92%   (37   904/41   247)   0.970.94–1.01   0.1CL    use, c mean   0.540.52   0.950.93–0.97   0.0001No.   of    CLABSI,   n   61   404CLABSI   rate   per   1000   CL-days   (95%   CI)6.4(4.9–8.3)   3.9   (3.5–4.3)   0.610.46–0.81   0.0007ASIS,Average   Severity   of    Illness   Score;   CI,confidence   interval;   CL,   central   line;   CLABSI,   central   line-associated   bloodstream   infection;   MV,mechanical   ventilation;   RR,   relativerisk;SD,   standard   deviation. a Bed-days:   the   total   number   of    days   thatpatients   are   inthe   ICU   during   the   selected   time-period. b CL-days:   the   total   number   ofdays   of    exposure   to   a   central   line   for   allof    the   patients   in   the   selected   population   during   the   selected   time-period;   MV-days:   the   totalnumberof    days   of    mechanical   ventilation   for   allof    the   patients   inthe   selected   population   during   the   selected   time-period. c CL    and   MVuse   ratios   were   calculated   bydividing   the   totalnumber   of    CL-days   or   MV-days   by   the   totalnumber   of    bed-days.  Table   3 CLABSI   stratified   by   length   ofparticipation   ofeach   ICU   in   the   INICC;   Poisson   regression   analysisTime   since    joining   INICC   Number   of ICUsCL-days,   n   CLABSI,   n   Crude   LCBIrate/1000   CL-daysIRR    accounting   forclustering   by   ICU  p -Value1–3months   (baseline)   16   9472   61   6.44(95%   CI4.9–8.3)   1.0   -4–6months   16   8814   23   2.61(95%   CI1.7–3.9)   0.42   (95%   CI   0.26–0.68)   0.00017–9   months   14   12   210   23   1.88(95%   CI1.2–2.8)   0.31   (95%   CI   0.19–0.51)   0.000110–12   months   13   8429   21   2.49(95%   CI1.5–3.8)   0.30   (95%   CI   0.18–50.2)   0.000113–18   months   15   18   220   64   3.51(95%   CI2.7–4.5)   0.41   (95%   CI   0.28–0.59)   0.000119–24   months   12   14   412   69   4.79(95%   CI3.7–6.1)   0.48   (95%   CI   0.34–0.70)   0.0001Third   year   9   18   813   78   4.15(95%   CI3.3–5.2)   0.47   (95%   CI   0.31–0.70)   0.0001CI,confidence   interval;   CL,   central   line;   CLABSI,   central   line-associated   bloodstream   infection;   ICU,   intensive   care   unit;   INICC,   international   nosocomial   infection   controlconsortium;LCBI,   laboratory-confirmed   bloodstream   infection;   IRR,   incidence   rate   ratio.  Table   4 Microorganism   related   to   CLABSI   in   AICUs   in   phase   1   (baseline   period)   and   phase   2(intervention   period)Microorganisms   isolated   Baseline,   %( n )Intervention,   %( n ) Klebsiella   spp   32%   (11)   25%   (72)  Acinetobacter    spp   18%   (6)   10%   (30) Candida   spp   12%   (4)   15%   (42) Staphylococcus   aureus   12%   (4)   2%   (7) Enterococcus   spp   6%   (2)   4%   (11) Pseudomonas   spp   6%   (2)   16%   (45) Escherichia   coli   3%   (1)   7%   (19) Enterobacter    spp   0%   (0)   7%   (20) Flavobacterium   spp   0%   (0)   5%   (15)Coagulase-negative   staphylococci   6%   (2)   4%   (11) Stenotrophomonas   spp   0%   (0)   2%   (6) Other    microorganisms   6%   (2)   3%   (9)Total   100%   (34)   100%   (287)AICU,   adult   intensive   care   unit;   CLABSI,   central   line-associated   bloodstreaminfection. N.    Jaggi   et    al.    /    International    Journal   of    Infectious   Diseases   17    (2013)   e1218–e1224 e1222
Similar documents
View more...
Search Related
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