The
accuracy
and
validity
of
a
weekly
point-prevalence
survey
for
evaluating
the
trend
of
hospital-acquired
infections
in
a
university
hospital
in
Turkey
§
Cemal
Ustun
a,*
,
Salih
Hosoglu
b,
Mehmet
Faruk
Geyik
c,
Zafer
Parlak
d,
Celal
Ayaz
ba
DepartmentofInfectiousDiseasesandClinicalMicrobiology,MinistryofHealthElazigTeachingHospital,Elazig,Turkey
bDepartmentofInfectiousDiseasesandClinicalMicrobiology,DicleUniversityHospital,Diyarbakir,Turkey c
DepartmentofInfectiousDiseasesandClinicalMicrobiology,DuzceUniversityHospital,Duzce,Turkey
d
DepartmentofInfectiousDiseasesandClinicMicrobiology,MinistryofHealthElbistanGeneralHospital,Kahramanmaras,Turkey
1. Introduction
Hospital-acquiredinfections(HAIs)areanimportantcauseof morbidityandmortality,aswellassignificantlyincreasedhospital stays,additionalantibioticutilization,andhealthcarecosts.1–6The surveillanceofHAIsisacrucialcomponentofaqualifiedinfection controlprogramandiswidelyacceptedasaprimarystepinthe controlofHAIs.1,6–13Thedifficultiesassociatedwithsurveillance ofHAIshaveledtoavarietyofmethodologicalapproaches,which many experimental studies have tested.9,13,14 For example, the incidencesurveyisregardedasthemostpowerfulmethod,anda gold standard for evaluating the burden of HAIs. However, incidencestudiesareexpensivebecausedatahavetobecollected over a long period and require more experienced investiga-tors.6,8,11 However, point-prevalence studies are less expensive
and time-consuming, and can be performed more easily than incidence studies.6–8,13–16 In addition, these studies increase awarenessof theproblemathospitalsand arewidelyaccepted andrecommendedbymanyinvestigators,particularlywhenthey canberepeatedatregularintervals.7,10,17
Therepeatedprevalencesurveyisusedtoevaluateaninfection controlprogram,followthetrendsofHAIs,measuretheadverse effects andcostsofHAIs,and determinetherateofdeviceand antibioticusage.8,16 In developing countries,because oflimited resources, therepeated point-prevalencesurveymaybea good alternativeforthesurveillanceofHAIs.
Theaimofthisstudywastodeterminethetrendandextentof HAIsbyweeklypoint-prevalencesurvey(WPS),andexaminethe accuracyand validityofWPSbycomparing thismethodwitha prospective-activeincidencesurvey(PIS).
2. Methods 2.1. Setting
This study was performed across all departments of Dicle UniversityHospital(DUH)betweenJanuaryandDecember2006.
InternationalJournalofInfectiousDiseases15(2011)e684–e687
ARTICLE INFO
Articlehistory:
Received18November2010 Receivedinrevisedform4May2011 Accepted16May2011
CorrespondingEditor:HubertWong, Vancouver,Canada. Keywords: Hospital-acquiredinfection Infectioncontrol Surveillance Prevalence Incidence SUMMARY
Objective:Toevaluatethevalidityofaweeklypoint-prevalencesurvey(WPS)bycomparingitwitha prospective-activeincidencesurvey(PIS).
Methods:WPSandPISwereconductedatatertiaryreferralhospitalbetweenJanuaryandDecember 2006.EachWednesday,aninfectioncontrolteamreviewedallclinicalrecordsofpatientswith hospital-acquiredinfections(HAIs)byWPS.RoutinePISwasconductedwithdailyvisitsbythesameteam.The RhameandSudderthformulawasusedforconvertingthedatabetweenWPSandPIS.
Results:Duringthestudyperiod,1287HAIs weredetected in37466patientsbyWPS. Themean observedprevalenceandcalculatedprevalencewere5.42%and5.45%,respectively.Thereanimation intensivecareunit(ICU)(49.4%)andburnsunit(27.6%)hadthehighestprevalencerates.Pneumonia (0.94%)andurinarytractinfections(0.37%)werethemostfrequentinfections.Overall602HAIswere detectedin545patientsbyPIS.Themeanobservedincidenceandcalculatedincidencewere 2.42/1000-admissionsand2.41/1000-admissions,respectively.TheCriticalcareICU(37.0/1000-admissions)and burns unit (24.8/1000-admissions) had the highest incidences of HAI. Pneumonia (0.64/1000-admissions)andurinarytractinfections(0.37/1000-admissions)werethemostfrequentinfections. Conclusions: Thisstudyconfirmsacloserelationshipbetweenprevalenceandincidencedata.WPSmay beausefulmethodforfollowingHAIswhenPIScannotbeperformed.
ß2011InternationalSocietyforInfectiousDiseases.PublishedbyElsevierLtd.Allrightsreserved.
§
Thisstudywaspresentedasa poster(P-30)attheEighthCongress ofthe InternationalFederationofInfectionControl,Budapest,Hungary,October18–21, 2007.
*Correspondingauthor.Tel.:+904242381000x1241;fax:+904242121461. E-mailaddress:drcustun@gmail.com(C.Ustun).
ContentslistsavailableatScienceDirect
International
Journal
of
Infectious
Diseases
j o urn a l hom e pa ge : ww w. e l s e v i e r. c om/ l o ca t e / i j i d1201-9712/$36.00–seefrontmatterß2011InternationalSocietyforInfectiousDiseases.PublishedbyElsevierLtd.Allrightsreserved.
DUHisan1150-bedtertiaryreferralcenter,andthelargesthospital inthesoutheastofTurkey.Thehospitalis25yearsoldandhas33 separateclinics,includingareanimationintensivecareunit(ICU) andaburnsunit.Annually,about40000patientsaretreatedatDUH, andin2006,theproportionofhospitalizedpatientswas77%.
During2006,WPSandPISwereconductedacrossall depart-mentsofthehospitalbythecentralinfectioncontrolcommittee, andallhospitalizedpatientswereincludedinthestudy.ForWPS data collection, the central infection control committee was composedofasurveillanceteam,includingaspecialistphysician, two resident physicians, and two infection control nurses.The teamwasexperiencedandtrainedinHAIs.Hospitalwardswere classifiedintotwogeneraltypes:surgicalandinternalclinics.The Criticalcare ICU and theburns unit wereclassified as surgical clinics. Subsequently, the team was divided into two groups includingaresidentphysicianandanurse,andemployedtorecord HAIdatainboththesurgicalandinternalclinics.Thisstudywas directedby thespecialistphysician,who wasa memberof the centralinfectioncontrolcommittee.
2.2. Definitionsanddatacollection
ThediagnosisofHAIswasmadeaccordingtotheCentersfor Disease Control and Prevention criteria18 and the National Nosocomial Infections Surveillance System methodology.19 AsymptomaticbacteriuriawasnotcategorizedasanHAI.
Each Wednesday during the prevalence study, WPS was performed by the team. On this day the team reviewed the clinicalandlaboratoryrecordsofallhospitalizedpatients.Patients were detected according to positive cultures, symptoms of infection,and antibiotic treatment for HAIs. Patient data were recorded on a standard form, including the total number of hospitalizedpatientsandthenumberandtypesofHAI.Theratesof HAIinallclinicswerethencalculated.PISwasperformedbasedon patientclinicalandlaboratoryrecordsbythesameteamwithdaily visitsto alldepartments ofthe hospital.Positive cultures from patientswereobtainedfromthecentralmicrobiologylaboratory bythe team.Subsequently, theteam visited allpatients at the bedsidewiththeirclinicphysicianandnurses.AllcaseswithHAI wererecordedonastandardform.Ifapatienthadsymptomsand signsof infection,themedical and nursingnotes, microbiology reports, temperature, and antibiotic treatment charts were reviewed. Urinary tract infections, pneumonia, surgical site infections,bacteremia,sepsis,burninfections,woundinfections, catheter-related infections, intraperitoneal infections, abscess, empyema,meningitis,andorthopedicprosthesisinfectionswere recordedbybothWPSandPIS.Theteamfilledoutaworksheetfor
each patient diagnosed with HAI. The data recorded on the standardformswerethentransferredtoaMicrosoftOfficeExcel 2003spreadsheet(MicrosoftCorp.,Redmond,WA,USA). 2.3. Interconversionofincidenceandprevalencedata
TheRhameandSudderthformula20wasusedforconvertingthe datafromincidencetoprevalence,andviceversa.Accordingtothis formula, the prevalence rate of HAIswas calculated asfollows: P=I[(LN INT)/LA],wherePisprevalence,Iisincidence,LNisthe lengthofhospitalizationofpatientshavingoneormoreHAI,INTis theaverageintervalbetweenadmissionandonsetofthefirstHAIfor patientshavingoneormoreHAI,andLAistheaveragelengthof hospitalization of all the hospitalized patients duringthe study period.
2.4. Statisticalanalysis
For each week duringthestudy period,HAI prevalencewas calculatedastheratioofthenumberofHAIstothetotalnumberof hospitalizedpatientsonthedayoftheWPS.Themeanprevalence fortheyearwascalculatedbyaveragingtheweeklyprevalences. Themeanprevalenceswerepresentedwitharange(minimum– maximum)ofobservedprevalences.Meanprevalencefortheyear wasalsocalculated as‘biweekly’and‘monthly’(byconsidering onlydatafromeverysecondoreveryfourthweek,respectively). TheincidenceofHAIswascalculatedastheratioofthenumberof HAIstothenumberofpatientadmissions(per1000-admissions)in 2006.Statistical analyseswerecarried outusingSPSS software, version13.0(SPSSInc.,Chicago,IL,USA).
3. Results
Duringthe studyperiod, 1287HAIswere detectedin 37466 patients by WPS. According to WPS results, the mean weekly observedprevalencerateofHAIswas5.42%(range1.9–8.4%)overthe studyperiod.Accordingtothebiweeklyandmonthlyresults,the meanobservedprevalenceratesofHAIswere5.5%(range3.2–8.4%) and5.4%(range3.2–7.1%),respectively.Figure1showsthetrendof weeklymeanprevalenceratesofHAIsforinternalclinicsandsurgical clinicsduringthestudyperiod.AccordingtoWPSresults,pneumonia (0.94%),urinarytractinfections(0.37%),andbacteremia(0.35%)were themostfrequentinfections(Table1).TheCriticalcareICUhadthe highestprevalencerate(49.4%),followedbytheburnsunit(27.6%), neurology(10.5%),andthegeneralsurgeryICU(8.4%)(Table2).
Duringthesamestudyperiod,atotalof40100patientswith 249000 admissionswereexamined byPIS.Atotalof602HAIs
0 2 4 6 8 10 12 14 16 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Prevalence rates (%) Weeks
Internal clinics Surgical clinics
Figure1.Thetrendofweeklymeanprevalenceratesofhospital-acquiredinfections(HAI)forinternalclinicsandsurgicalclinicsduringtheweeklypoint-prevalencesurvey study.
weredetectedin545patientsbythesurveymethod.Themean observedincidenceofHAIswas2.42/1000-admissions.Themost frequent infections were pneumonia (0.64/1000-admissions), urinary tract infections (0.37/1000-admissions), and wound infections (0.24/1000-admissions) (Table 1). The Critical care ICU had the highest incidence of HAIs (37.0/1000-admissions), followed by the burns unit (24.8/1000-admissions), neurology (8.8/1000-admissions), and the general surgery ICU (8.0/1000-admissions)(Table2).
Theaveragelengthofhospitalizationofpatientshavingoneor moreHAIswas31days.Theaverageintervalbetweenadmission andonsetofthefirstHAIforpatientshavingoneormoreHAIswas 13days.Theaveragelengthofhospitalizationofallhospitalized patients during thestudy period was8 days. According tothe Rhame and Sudderth formula, the calculated prevalence and calculated incidence were 5.45% and 2.41/1000-admissions, respectively, for all departments of the hospital. In the study, the observed prevalence and calculated prevalence, and the observed incidence and calculated incidence were foundto be almostthesame by WPSand PIS, accordingtotheRhameand Sudderthformula.
4. Discussion
Inthisstudy,only trends andtypes of HAIwereexamined usingWPS, andthe results were compared withPIS. Further-more,thedataofbothmethodswereconvertedfromonetothe otherusingtheRhameandSudderthformula.Previousstudies havegenerallybeenmademonthly,havebeenmulticenter,and havebeenspecifictoasingletime.1–4,7,8,11,15–17,21,22However, thesestudiesdidnotcompareWPSwithPIS.Onlythestudyof Petittietal.23wasperformedweekly,butitwasnotcompared withPIS.
Inthisstudy,wefoundalmostthesameresultsforbiweekly andmonthlyobservedmeanprevalenceratesofHAIsrecorded byWPS.Theseresultsmayindicatethatbiweeklyandmonthly point-prevalence surveys have equal validity and that point-prevalence survey studies can be performed at monthly intervals. In the present study, the observed prevalence rate of HAIs detected by WPS was similar to the prevalence rate calculatedbytheRhameandSudderthformulausingthedataof PIS.Inaddition, thefrequency andtypeof HAIs showedclose similarity between WPS and PIS. The HAI rates in the chest surgery,neurosurgery, burnsunit,Criticalcare ICU,neurology, hematology, andnephrology clinics,where HAIs are themost frequent, showed similar frequencies by the two methods. Furthermore,wefoundaclosesimilaritybetweentheobserved incidenceofHAIsdetectedbyPISandtheincidencecalculated by the Rhame and Sudderth formula using the data of WPS. Results similar to those of our study have been reported by Gastmeier et al.22 who also used the Rhame and Sudderth formula. This similarity may demonstrate that the repeated point-prevalencesurveyisaproper,reliable,andvalidmethod for following HAIs. Furthermore, the Rhame and Sudderth formulaissuitableforconvertingdatabetweenprevalenceand incidence. In contrast, Haore et al.15 and Rossello-Urgell and Rodriguez-Pla24reportedthattheRhameandSudderthformula is not acceptable forconverting data between prevalenceand incidence, andthus,they didnot recommend convertingdata betweenprevalenceandincidence.
In previous point-prevalence survey studies,4,6,8,21–23 the observed prevalence of HAIs hasbeen reported to bebetween 3.5%and11.6%. Wefoundtheobservedprevalencetobe5.42%, which is compatible with previous studies. During the study period,thehighestprevalenceratesofHAIs,showninFigure1,in thesurgicalclinicswereduetotheaccumulationofpatientswith HAIsintheCriticalcareICUandtheburnsunit.Theseunitshadthe highest prevalence rates over the study period. Similarly, the reasonsforthehighprevalenceratesintheinternalclinicswere thehighprevalenceratesofHAIsintheneurology,hematology, nephrology,andoncologyclinics.Generally,HAIsarefrequently seeninthesesurgicalandinternaldepartmentsbecausepatientsin poorgeneralconditionfromtertiaryreferralhospitalsareaccepted here.Ontheotherhand,thelowestprevalenceratesinthesurgical clinicswereduetothelowprevalenceratesofHAIsintheCritical
Table1
Infection type, prevalence rate (%), and incidence (per 1000-admissions) of hospital-acquiredinfections(HAIs)accordingtoweeklypoint-prevalencesurvey (WPS)andprospective-activeincidencesurvey(PIS)
TypeofHAI WPS PIS
HAI count Prevalence rateb ,range (min–max)
HAIcount Incidence
Pneumonia 397 0.94(0.2–1.9) 160 0.64 Urinarytractinfection 169 0.37(0.0–1.2) 93 0.37 Surgicalsiteinfection 149 0.33(0.0–1.1) 54 0.22 Bacteremia 137 0.35(0.0–1.1) 51 0.20 Burninfection 116 0.27(0.0–1.1) 49 0.20 Woundinfection 106 0.26(0.0–1.3) 60 0.24 Sepsis 92 0.27(0.0–0.9) 49 0.20 Catheter 31 0.07(0.0–0.8) 53 0.21 Othera 90 0.19(0.0–1.1) 33 0.13 Total 1287 602 a
Other:intraperitonealinfections,abscess,empyema,meningitis,andprosthesis infections.
b
TheprevalencerateofHAItypewascalculatedasthemeanvalueofweekly prevalenceratesofHAItypes.
Table2
Themeanprevalencerates(%)ofhospital-acquiredinfections(HAI)accordingto weeklypoint-prevalencesurvey(WPS),andtheincidences(per1000-admissions) ofHAIaccordingtoprospective-activeincidencesurvey(PIS)foreachhospital department. Clinic WPS PIS Prevalence ratec ,range (min–max)
HAIcountAdmissiondaysIncidence
Surgical Pediatricsurgery 1.9(0.0–12.5) 13 6858 1.9 Chestsurgery 5.2(0.0–17.8) 26 7662 3.4 Cardiovascularsurgery 1.8(0.0–14.3) 9 4014 2.2 Orthopedic 2.8(0.0–10.8) 33 11295 2.9 Neurosurgery 7.1(0.0–22.6) 29 5659 5.1 GeneralsurgeryICU 8.4(0.0–23.3) 26 3258 8.0 Burnsunit 27.6(0.0–55.0) 54 2181 24.8 Plasticsurgery 8.3(0.0–25.0) 22 4080 5.4 CriticalcareICU 49.4(14.3–75.0)61 1651 37.0 Othera 0.7(0.0–8.3) 56 65255 0.9 Internal Breastinfection 1.1(0.0–6.7) 8 10174 0.8 Pediatric 1.2(0.0–3.2) 55 36904 1.5 Neurology 10.5(0.0–20.0) 82 9276 8.8 Physicaltherapy 3.1(0.0–16.8) 13 5342 2.4 Infectiousdiseases 2.3(0.0–16.7) 5 5412 0.9 Hematology 4.4(0.0–20.8) 35 9536 3.7 Nephrology 6.6(0.0–16.7) 37 9043 4.1 Oncology 2.4(0.0–13.1) 7 5052 1.4 Otherb 0.6(0.0–6.7) 31 44971 0.7 ICU,intensivecareunit.
aOther:gynecology,ophthalmology,otorhinolaryngology,urology,andgeneral
surgery.
b
Other:dermatology,cardiology,psychiatric,endocrinology,gastroenterology, andhepatology.
c
Theprevalencerate ofHAI was calculatedasthemean valueofweekly prevalenceratesofHAIsforeachclinic.
C.Ustunetal./InternationalJournalofInfectiousDiseases15(2011)e684–e687 e686
careICUandburnsunit.Similarly,thelowestprevalenceratesin theinternalclinicswereduetothelowprevalenceratesofHAIsin theneurology,nephrology,andoncologyclinics.Allthesurgical and internaldepartments mentioned aboveshouldbecarefully followedintermsofHAIs.
Theresultsofthepresentstudyindicatethatshorterintervals, suchasweeklyorbiweekly,mayprovideabettermeanstoobserve fluctuationsinoroutbreaksofHAIs.Thus,WPSmaybeanavailable methodtodetermineepidemicsofHAIs.Inthepresentstudy,no epidemicofHAIswasfoundduringthestudyperiod.Accordingto ourobservationsofthedataforeachdepartmentoverthestudy period,theprevalenceratesdidnotshowremarkableelevations thatcouldbeconsideredasoutbreaksofHAIs.
ThisstudydemonstratesthatWPSisanalternativemethodfor evaluatingthetrendand extentofHAIs.Itmayalsobeusedto evaluatethetrendofHAItypes.Urinarytractinfections,surgical siteinfections,hospital-acquiredpneumonia,andotherHAIsmay beinvestigated by this method. Gastmeier etal.22investigated urinary tract and surgical site infections using this method. However,onlytrendsandtypeofHAIswereinvestigatedinour study. In addition, WPS may provide data on whether or not seasonalalterationsaffectthetrendofHAIs.Overthestudyperiod, no significantfeatures resultingfrom seasonalalterations were found.
PatientswithHAImaybereportedasmarkedbyrepetition whena repeatedpoint-prevalence surveyis performedat less than 4-week intervals. In this study, these cases were not resolved. If new recorded cases were detected by WPS, new infectiontypes,rates,andcasescouldeasilybedetected,which couldprovideabetteranalysisofHAIssuchastheaccumulation ofHAItypes.
The results of this study indicate that the repeated point-prevalencesurveymay bean alternativemethodfor following HAIsbecauseitiseasilyappliedin hospitals,especiallywhere PIS cannot be performed. Moreover, the repeated point-prevalencesurveyismorecost-effectivethan PIS.Forinstance, thisstudywasperformedbytwophysiciansandtwonursesin an 1150-bed tertiary referral hospital. Petitti et al.23 also reported that WPS is a reliable surveillance method for followingHAIs.
Thelimitationofthis studyisthat theRhameand Sudderth formulawasnotusedtoconvertthedataofweekly,biweekly,and monthly prevalence and incidence. The Rhame and Sudderth formula was used to convert only the data of prevalence and incidencefora48-weekperiod.
Inconclusion,thisstudyhasshowntheaccuracyandvalidity of WPS. The Rhame and Sudderth formula is suitable for convertingdata between prevalence andincidence.WPSis an effective and practical method for evaluating the trend and extent of HAIs. However, although some authors do not recommend that WPS be routinely performed, this method can be used in developing countries, especially those with limitedresources.
Conflictof interest:Allauthorsdeclareno conflictofinterest, ethicsrulesinfringements,oranyfinancialsupportrelevanttothis study.
References
1.GravelD,MatlowA,Ofner-AgostiniM,LoebM,JohnstonL,BryceE,etal.A point-prevalencesurveyofhealthcare-associatedinfectionsinpediatricpopulations inmajorCanadianacutecarehospital.AmJInfectControl2007;35:157–62. 2.HadjuA,SamodovaOV,CarlssonTR,VoinovaLV,NazarenkoSJ,TjurikovAV,
etal.Apoint-prevalencesurveyofhospital-acquiredinfectionsand antimicro-bial use in a pediatric hospital in north-western Russia. J Hosp Infect 2007;66:378–84.
3.3BalkhyHH,CunninghamG,ChewFK,FrancisC,Al-NakhliDJ,AlmuneefMA, etal.Hospital-andcommunity-acquiredinfections:apoint-prevalenceandrisk factors survey in a tertiary care center in Saudi Arabia. J Hosp Infect 2006;10:326–33.
4.HarbartS,RuefC,FrancioliP,WidmerA,PitterD.NosocomialinfectionsinSwiss universityhospitals:amulti-centresurveyandreviewofthepublished expe-rience.SchweizMedWochenschr1999;129:1521–8.
5.ErbayH,YalcinAN,SerinS,TurgutH,TomatirE,CetinB,etal.Nosocomial infectionsinintensivecareunitinTurkishuniversityhospital:a2-yearsurvey. IntensiveCareMed2003;29:1782–8.
6.SartorC,DelchambreA,PascalL,DrancourtM,DeMiccoP,SambueR. Assess-mentofthevalueofrepeatedpoint-prevalencesurveyforanalyzingthetrendin nosocomialinfections.InfectControlHospEpidemiol2005;26:369–73. 7.GastmeierP,SohrD, RathA,ForsterDH,WischnewskiN,LacourM,etal.
Repeated prevalence investigatoronnosocomial infectionsfor continuous surveillance.JHospInfect2000;45:47–53.
8.GravelD,TaylorG,OfnerM,JohnstonL,LoebM,RothVR,etal.Point-prevalence surveyforhealthcare-associatedinfectionswithinCanadianadultacute-care hospitals.JHospInfect2007;66:243–8.
9.HarbarthS, SaxH,GastmeierP.Thepreventableproportionofnosocomial infections:anoverviewofpublishedreports.JHospInfect2003;54:258–66. 10.GikasA,PediaditisI,RoumbelakiM,TroulakisG,RomansonJ,TselentisY,etal.
Repeatedmulti-centreprevalencesurveysofhospital-acquiredinfectionin Greekhospital.JHospInfect1999;41:11–8.
11.LizioliA,PriviteraG,AlliataE,Antonietta-BanfiEM,BoselliL,PanceriML,etal. PrevalenceofnosocomialinfectionsinItaly:resultfromtheLombardysurvey in2000.JHospInfect2003;54:141–8.
12.HoranTC,LeeTB.Surveillance:intothenextmillennium.AmJInfectControl 1997;25:73–6.
13.GastmeierP,SohrD,JustHM,NassauerA,DaschnerF,Ru¨denH.Howtosurvey nosocomialinfections.InfectControlHospEpidemiol2000;21:366–70. 14.Rossello-UrgellJ,Vaque-RafartJ,Villate-NavarroJI,Sanchez-PayaJ,
Martinez-GomezX,Arribas-LlorenteJL,etal.Exposuretoextrinsicriskfactorsin preva-lencesurveysofhospital-acquiredinfections:amethodologicalapproach.J HospInfect2006;62:366–71.
15.HaoreHG,MullerA,TalonD,BertrandX.Estimationofthecumulative inci-denceofhospital-acquiredbacteremiafromprevalencedata:aformula.Infect ControlHospEpidemiol2005;26:415–7.
16.HumphreysH,SmythET.Prevalencesurveyofhealthcare-associated infec-tions:whatdotheytellus,ifanything?ClinMicrobiolInfect2006;12:2–4. 17.GastmeierP,KampfG,WischnewskiN,HauerT,SchulgenG,SchumacherM,
etal.PrevalenceofnosocomialinfectionsinrepresentativeGermanhospitals.J HospInfect1998;38:37–49.
18.GarnerJS,JarvisWR,EmoriTG,HoranTC,Hughes JM.CDCdefinitionsfor nosocomialinfections.AmJInfectControl1988;16:128–40.
19.National Nosocomial Infections Surveillance System.National Nosocomial InfectionsSurveillance(NNIS)System Report,datasummaryfromJanuary 1992throughJune2004.AmJInfectControl2004;32:470–85.
20.RhameFS,SudderthWD.Incidenceandprevalenceasusedintheanalysisofthe occurrenceofnosocomialinfections.AmJEpidemiol1981;113:1–11. 21.GravesN,NichollsTM,WongCG,MorrisAJ.Theprevalenceandestimatesof
cumulativeincidenceofhospital-acquiredinfectionsamongpatientsadmitted toAucklandDistrictHealthBoardHospitalsinNewZealand.InfectControlHosp Epidemiol2003;24:56–61.
22.GastmeierP,BrauerH,SohrD,GeffersC,ForsterDH,DaschnerF,etal. Con-vertingincidenceandprevalencedataofnosocomialinfections:resultsfrom eighthospitals.InfectControlHospEpidemiol2001;22:31–4.
23.PetittiT,Sadun B,DicuonzoG. Usefulnessandaccuracy ofweekly point-prevalencesurveysinactivesurveillanceforhealthcare-associatedinfections. InfectControlHospEpidemiol2005;26:335–6.
24.Rossello-UrgellJ,Rodriguez-PlaA.Behaviorofcross-sectionalsurveysonthe hospital setting: a simulation model. Infect Control Hosp Epidemiol 2005;26:362–8.
