responsibilitykaixin.com

P.
L.
P.
Rau(Ed.
):CCD/HCII2014,LNCS8528,pp.
351–360,2014.
SpringerInternationalPublishingSwitzerland2014HowtoPromotePatientSafetyinSocialMedia:AComparisonbetweenMessagesinSocialMediaandNewspapersNaSunandPei-LuenPatrickRauInstituteofHumanFactorsandErgonomics,TsinghuaUniversity,Beijing10084,China{sun-n10,rpl}@mail.
tsinghua.
edu.
cnAbstract.
Theobjectiveofthispaperistoevaluatethefeaturesofpatient-safetyrelatedmessagesinsocialmediaandprovidesuggestionsforhowtopromotepatientsafetyinsocialmedia.
Ahierarchicalframestructureespeciallyforpa-tientsafetymessageswasproposedforcontentanalysis.
45messagesfromFa-cebookand45reportsfromnewspapersaboutpatientsafetywereevaluated.
Theresultsindicatedthatmessagesinsocialmediaprovidedmorecomprehen-siveinformationaboutpatientsafetythannewspapers.
Andsocialmediapaidmoreattentiononindividuals,includingdoctors,nursesandpatients.
Further-more,socialmediamessagesprovidedmorepositiveinformationthannewspa-pers,suchastreatmentrecommendationsandprogramintroductions.
Keywords:patientsafety,socialmedia,newspaper,contentanalysis.
1IntroductionPatientsafetyisanewhealthcaredisciplinethatemphasizestheabsenceofpreventa-bleharmtopatientduringtheprocessofhealthcare(WHO).
Theimpactofmedicalerrorshasbeenwidelyreportedanddiscussed.
InstituteofMedicineoftheNationalAcademiesreportedthatatleast44,000Americansdieeachyearasaresultofmedi-calerrors(2000)[1].
Morepeopledieinagivenyearasaresultofmedicalerrorsthanfrommotorvehicleaccidents(43,458),breastcancer(42,297),orAIDS(16,516).
Theimprovementofpatientsafetyrequirestheinvolvementofeveryone,becauseanyonecouldbecomeapatientsometimeandthepatients'awarenessofpatientsafetyissuesisamajorsteptoavoidmedicalerrors.
Traditionalwaytospreadpatientsafetyinformationisvianewspapers.
Thereportsinnewspapersusuallytalkedaboutpatientsafetyissueswithunexpectedaccidentsorannouncementsmadebyorganization.
Forinstance,intheMarch2010,theUKMedicinesandHealthcareproductsRegulatoryAgency(MHRA)releasedaMedicalDeviceAlertinrelationtoallsiliconegelfilledbreastimplantsmanufacturedbyaFrenchcompanycalledPIP.
Aftertheannounce-ment,alargenumberofreportsaboutPIPimplantfailureappearedinmanynewspa-pers.
Somereportsinterviewedanunsuccessfulimplantsurgery,andsomereportspresentedtheattitudeofgovernmentandhealthinstitutions.
352N.
SunandP.
-L.
PatrickRauAnotherwaytospreadpatientsafetyinformationisviasocialmedia.
Socialmediaspreadsinformationfasterandmoreflexiblethantraditionalmedia,especiallyamongyoungpeople.
Comparedwithnewspapers,socialmediahaslesslimitationoftime-sensitive,sotheinformationcanbemorecomprehensiveandcovermoretopics,suchasmedicaltreatmenttipsandpatient-safetyrelatedprogramintroduction.
Therefore,manypatient-safetyrelatedorganizationshavesetuptheirownpublichomepageonFacebook(themostpopularsocialmedia),suchasNationalPatientSafetyFoundationandCanadianPatientSafetyInstitute.
Theyusesocialnetworktointroducetheiror-ganizationdevelopment,tosendannouncements,andmostimportantly,toenhancepeople'sawarenessofpatientsafetyissues.
Socialnetworkhasbecomeanimportantwayforpatientsafetyeducationandpromotion.
2ObjectiveTheinitialobjectiveofthispaperwastofindoutthedifferenceofpatient-safetypro-motioninsocialmediabetweenChinaandAmerica.
However,thistopicdidnotworkout,becausetherewerenopublicpagesonpatientsafetyinthemostpopularsocialnetworkwebsitesinChina,includingRenren.
com,Kaixin.
com,Weibo.
comandT.
qq.
com.
Whenusingpatientsafetyasthekeyword,onlyblogssentbyindividualreturned,andmostofthemessageshadnothingtodowithpatientsafetyissues.
EventhoughtherewereseveralorganizationsconcentratingonhealthcarepromotioninChina,noneofthembuiltpublicpagesinsocialmedia.
Patient-safetypromotioninsocialmediaisfarfromsatisfactoryinChina.
Similarsituationcanbeobservedinotherdevelopingcountries.
Therefore,suggestionsonhowtousesocialmediatopromotepatientsafetyareneededespeciallyforpatientsafetyorganizations.
Theobjectiveofthisstudyistoevaluatethefeaturesofpatientsafetymessagesinsocialmedia,i.
e.
Facebook,andthereforeprovidesuggestionsonpatient-safetypromotioninsocialmedia.
3Methodology3.
1HierarchicalContentAnalysisOverthepastdecades,thestudyofcontentanalysismethodologyhasbeenrapidlyexpanding.
Aquantityofapproacheshasbeenproposedtomeasuretheframeofcontent.
Alltheapproachescanbedividedintofivecategories[15].
Theyareherme-neuticapproach,linguisticapproach,manualholisticapproach,computer-assistedapproachanddeductiveapproach.
Inthisstudy,amixtureofhermeneuticapproachanddeductiveapproachwaspro-posed.
Firstly,contentframesweredividedintofourdimensionsaccordingtopre-viousliterature[12]:problemdefinition,causalinterpretation,moralevaluation,andtreatmentrecommendation.
Problemdefinitionistheaccidentorpotentialriskmen-tionedinthemessages.
Causalinterpretationidentifiesthereasonthatleadstotheproblem.
MoralevaluationdescribeswhoshouldtaketheresponsibilitytosolvetheHowtoPromotePatientSafetyinSocialMedia353problem.
Andtreatmentrecommendationreferstothesuggestionsprovidedforpa-tients,physicians,hospitals,orgovernment.
Secondly,afurtherexplorationofkeywordsineachdimensionwasconductedbyhermeneuticapproach.
Acoderreadallthesamplemessagesandidentifieskeywordsforeachdimension.
Thequalitativeassessmentofframingdependsoncarefulreadingofallmessages.
Forexample,afterreadingareportaboutspainfectionrisks,thecod-eridentifiedthekeywordsas"infection"forproblemdefinition;"lackofregulationforspaservice"forcausalinterpretation;"patient,service-provider&government"formoralevaluation;and"patientinvolvement"fortreatmentrecommendation.
Sometimesamessagementionedlessthanfourdimensionsofcontentframes.
Afteridentifyallthekeywords,thecoderstartedtointegratethekeywordsintosystematicvariables.
Theintegrationfollowedtwoprinciples:first,choosevariablesthatcoverasmuchasproblemsmentionedinthemessages;second,variablesareindependentwitheachother,i.
e.
avoidingoverlapamongvariables.
Thesystematicvariablesformedthehierarchicalframestructure.
Thehierarchicalframestructureservedasacodebookforcontentanalysis.
Atlast,thecoderreorganizedthesampleframewithhierarchicalframestructurebyadjustingkeywordsintovariablesineachdimension.
Thefrequencyofvariablesrevealedthefeatureofthesamplemessages,andthereforeindicatedthefeatureofthemedia.
Byusingsamplemessagesfrombothnewspapersandsocialmedia,thispapercomparedthedifferencesofmessagesbetweennewspapersandsocialmediaandrevealedthefeaturesofpatientsafetymessagesinsocialmedia.
3.
2SamplingThereareanumberofpublicpagesaboutpatientsafetyonFacebook.
Mostofthemaresetupbynon-profitorganizations,communitiesandinstitutes,whereassomeareinterestgroups,publishersandeducationgroups.
ThemostpopularpageisNationalPatientSafetyFoundation,with1,137"likes"and31people"talkingaboutthis"inthepast7days.
ThesecondpopularpageshouldbeCanadianPatientSafetyInstitute,with511"likes"and9people"talkingaboutthis",followedbythepageofLouiseH.
BatzPatientSafetyFoundationwith258"likes"and4"talkingaboutthis".
Theotherpagesattractrelativelylessattention,nomorethan100"likes".
Thusthesethreemostpopularpublicpageswerechosenasthesourceofsamplemessages.
Atotalof45messagesaboutpatientsafetywereselectedfromtheirFacebookpagesasthesamplemessages.
Mostofthesemessagescontainedalinktoanewwebpagerepresentingapatient-safetyrelatedreport.
Onlyonemessageprovidednofurtherreportbutonlysometipsforpatients.
AnothersetofsamplemessageswereselectedfromanonlinedatabasecalledLi-brary.
PressDisplay.
com,whichprovidesmorethan800onlinenewspapersfromaround70countriesallovertheworld.
Inordertoavoidtoomanyreportsaboutthesameaccident,twotimeperiodswerechosenassampletime,whichareJan.
5toJan.
9in2012andFeb.
18toFeb.
22in2012.
Atlast,45reportsfromdifferentnewspa-perswereselectedasthesecondsetofsamplemessages.
354N.
SunandP.
-L.
PatrickRau4Results4.
1HierarchicalFrameStructureByintegratingallthekeywordsabstractedfromsamplemessages,thehierarchicalframestructurewasgenerated,asshownintable1.
Table1.
HierarchicalFrameStructureProblemDefinition:MedicationerrorClinicalnursingerrorComplicationEmergencytreatmentMisdiagnoseAmbulatorycareerrorSurgeryerrorNosocomialinfectionPatientidentificationDelayFatigueFallCausalInterpretation:LackoftrainingLackofresourcesProfessionalattitudeFinanceLackofknowledgeExternaldistractionHospitalmanagementLackofpolicyBadfacilityHospitalcultureCooperationfailureMoralEvaluation:HospitalService-providerGovernmentPatientTreatmentRecommendation:PatientinvolvementBettermedicalsystemHospitalcultureMedicaldatareleasingTechnologysupportMoreresearchesTrainingPolicyMoreprivacyImprovinginstructionBettercooperationAccordingtothehierarchicalframestructure,samplemessagesweredividedintodifferentproblems,causes,responsibilities,andrecommendations.
ThefollowingpartsexplainthecomparisonresultsofFacebookandnewspapers,whichrevealthedifferencesofpatientsafetymessagesinsocialmediaandintraditionalnewspapers.
4.
2TypesofSampleMessagesBeforesystematicallyanalyzingtheframeofsamplemessages,ageneraldescriptionwasputforwardtoevaluatewhatkindofinformationthemessagesprovide.
Fourtypesoftypicalinformationwerechecked,whicharestatisticfact,practicalaccident,patient-safetyrelatedprogramorprojectandrecommendation.
TheevaluationHowtoPromotePatientSafetyinSocialMedia355procedurecamewithquestionssuchas"Isthereanystatisticfactinthemessage""Doesthemessagementionanyprogramorprojecttoimprovepatientsafety"Some-timesapieceofmessagecontainedmorethanonetypeofinformation.
TheresultsareshowninTable2.
Thefrequencyreferstohowmanymessagescontaincertaininfor-mationandthepercentagerepresentsthepercentageofmessagesthatcontaincertaininformation.
Table2.
InformationTypesAnalysisofPatientSafetyMessagesInfoTypeFacebookPressDisplayFreq.
%Freq.
%statistic220.
49220.
49recommendation220.
49180.
40accident170.
38280.
62program170.
38130.
29Theresultsindicatedthatfourtypesofinformationappearaveragelyinsocialme-dia,comparedwithnewspapers.
Socialmediamessagescontainedmoreprogramintroduction(38%)thannewspapersreports(29%),sowastherecommendation(49%to10%).
Butnewspapersreportedmoreaccidents(62%)thansocialmedia(38%).
Thiscanbeeasilyexplainedbythedifferenceofnewspaperindustryandsocialmedia.
Newspapercompaniesareaimedatattractingmoreconsumersbyprovidingthemostfreshandattractivenews.
ButthepublicpagesonFacebookareallsetupbynon-profitorganization,whosegoalistospreadpatientsafetyinformation.
Asare-sult,newspaperspreferredthelatestaccidentsreports,andFacebookpagesincludedasmanyprogramsandrecommendationsasaccidents.
Eventhoughthesemessagesmaynotbesoattractivetoreaders,theyaregreatprogressinpatientsafetyfieldandworthreading.
4.
3ProblemDefinition45samplemessagesfromsocialmediamentioned10problems;while45newspaperreportsonlyincluded8problems.
ThisresultindicatedthatmessagesonFacebookcoveredmoretopicsthannewspapers.
Thepercentageofeachproblemisshownintable3.
Amongsocialmediainformation,nosocomialinfection(26.
7%)drewsignifi-cantlymoreattention,followedbysurgeryerror(6.
7%)andcomplication(6.
7%).
Asfornewspapers,morethanhalfofthemessagesfocusedonmedicationerrors(37.
8%)andclinicalnursingerrors(24.
4%).
Thisresultispartlyduetotwoeventsduringthesampleperiods.
Thefirstoneisanunsuccessfulbreastimplantsurgeryaccident,whichisclassifiedasmedicationerrorandsurgeryerror.
Theotheroneisthepublica-tionofadversemedicaleventsbyHealthQualityandSafetyCommission,inwhichmedicationerrorsandclinicalnursingerrorsaretwocommonmishaps.
Asaresult,thedistributionofproblemsisunbalanced.
Therefore,newspapertopicsarerelativelyconcentrated,anddependheavilyonpracticalnewsofthesampleperiod.
356N.
SunandP.
-L.
PatrickRauTable3.
ContentFrameAnalysis:ProblemDefinitionProblemDefinitionFacebookPressDisplayFreq.
%Freq.
%nosocomialinfection120.
26770.
156surgeryerror30.
06790.
200complication30.
06730.
067medicationerror20.
044170.
378clinicalnursingerror20.
044110.
244emergencytreatmentdelay20.
04400misdiagnose20.
04440.
089fatigue20.
04400ambulatorycareerror10.
02200slipandfall10.
02240.
089patientidentification0020.
0444.
4CausalInterpretationThereweretotally12interpretations,asshowninTable4.
Themostcommoncausewashospitalmanagementfailure,followedbylackofmedicalresourcesinbothme-dia.
Butnoneoftheaccidentsinnewspaperswerecausedbyhospitalculture,whichwasthethirdfrequentreasononFacebook(8.
9%).
Facility(2.
2%)andfinanceissues(2.
2%)weretwointerpretationsinnewspapers,buttheywerenotmentionedonFace-book.
Facebookaimsatimprovingtheawarenessandknowledgeofpatientsafetyissues,soitpaysmoreattentiononhospitalandindividuals.
Thatiswhy4of45mes-sagesonFacebookascribedproblemstohospitalculture.
Table4.
ContentFrameAnalysis:CausalInterpretationCausalInterpretationFacebookPressDisplayFreq.
%Freq.
%hospitalmanagementfailure110.
24470.
156lackofresources50.
11150.
111hospitalculture40.
08900lackofpolicy30.
06720.
044professionalattitude20.
04410.
022patientknowledgeshortage10.
02220.
044cooperationfailure10.
02210.
022lackoftraining10.
02240.
089externaldistraction10.
02200badfacility0010.
022financeissues0020.
044HowtoPromotePatientSafetyinSocialMedia3574.
5MoralEvaluationMoralevaluationidentifiescausalagents,whicharetakingresponsibilityfortheacci-dentsorrisks,asintable5.
Hospitalwasthemajorcausalagentsinmessagesfrombothmedia.
ButpatientstookhigherproportionofmoralevaluationinFacebookmes-sages(11.
1%)thannewspaperreports(4.
4%),whichmeansmessagesonFacebookfocusedmoreonindividualresponsibilityregardingpatientsafetyproblems.
Newspa-perspaidmoreattentionongovernmentresponsibilityforpatientsafetyproblems(31.
1%)thanFacebook(15.
6%).
ThisresultindicatedthatFacebookfocusesonindi-vidualswhilenewspapersfocusonmacro-environment.
Table5.
ContentFrameAnalysis:MoralEvaluationMoralEvaluationFacebookPressDisplayFreq.
%Freq.
%hospital270.
600250.
556service-provider90.
20090.
200government70.
156140.
311patient50.
11120.
0444.
6TreatmentRecommendationTreatmentrecommendationreferstothesuggestionsforimprovingpatientsafetywhenacceptingmedicaltreatment.
Table6shows11recommendationsabstractedfromthesamplemessages.
Thedatarevealedthatmessagesfromnewspaperspro-videdrelativelylesstreatmentrecommendationsthanFacebookmessages.
ThemajorrecommendationinFacebookmessageswaspatientinvolvement(33.
3%),followingbytraining(15.
6%)andbettermedicalsystem(15.
6%).
Threeofthesetreatmentrec-ommendationswereconsideredastreatmentsforindividuals:patientinvolvement,trainingandbettercooperation.
20of45Facebookmessagesprovidedtreatmentrec-ommendationforindividuals,butonly10of45newspaperreportsprovidedtreat-mentsforindividuals.
Table6.
ContentFrameAnalysis:TreatmentRecommendationTreatmentRecommendationFacebookPressDisplayFreq.
%Freq.
%patientinvolvement150.
33360.
133training70.
15620.
044bettermedicalsystem70.
15660.
133policyimprovement40.
08920.
044bettertechnology40.
08900bettercooperation40.
08910.
022hospitalcultureimprovement20.
04410.
022358N.
SunandP.
-L.
PatrickRauTable6.
(continued)moreprivacy20.
04410.
022informationpublicity10.
02230.
067Bettermedicalinstruction10.
02210.
022moreresearches10.
02220.
0445ConclusionThisstudyproposedahierarchicalcontentanalysismethodtoevaluatethefeatureofpatientsafetymessagesinsocialmedia,i.
e.
Facebook.
45messagesfromFacebookand45reportsfromnewspaperswerechosenassample.
Theanalysisresultsindicatedthefollowingfindings.
1.
Messagesinsocialmediaprovidedmorerecommendationsandprogramintroduc-tionsthannewspaperreports,whilenewspaperreportscontainedmorepracticalaccidents.
Socialmediahasfewerrequirementsfortimeliness,sosocialmediacoversmorekindsofinformationandprovidemorecomprehensiveknowledgeofpatientsafety.
2.
Facebookmessagescoveredalargerscopeofpatient-safetyrelatedproblemsandrisks.
However,theproblemsinnewspapermessagesweremoreconcentratedandusuallyrelatedtorecentaccidentsorpolicies.
Asaresult,socialmediaismoresuitabletopromoteoverallpatientsafetyknowledgeeventhoughsomeofthemseldomhappens.
3.
Themostfrequentcauseswerehospitalmanagementfailureandlackofresourcesinbothmedia.
Hospitalculturewasamajorcauseinsocialmedia.
Butnewspapersdidnotincludecultureissues,andinsteadnewspaperreportstalkedmoreaboutfinanceandfacilityissues.
4.
Facebookmessagesemphasizedtheroleoftheindividualinimprovingpatientsafetymorefrequentlythannewspaperreports.
NewspaperspaidmoreattentionongovernmentresponsibilitythanFacebook.
Besides,bothmediaindicatedthathos-pitalshouldtakethemajorresponsibilityofpatientsafetyproblems.
5.
Facebookmessagesprovidedmoretreatmentrecommendationsthannewspaperre-ports.
Andalargeproportionoftreatmentswereforindividuals,includingdoctors,nurses,paramedicandpatients.
6DiscussionThispaperaimstofindoutthedifferencesofpatient-safetyrelatedmessagesinsocialmediaandnewspapers.
Theresultsrevealedfourfeaturesofpatientsafetymessagesinsocialmediaandprovidedsuggestionsonhowtodesigneffectivemessagesthataresuitableforsocialmedia.
Firstly,socialmediaisaperfectplatformfornon-profitorganizationstopromotecommonwealconceptandknowledge,becauseanyonecansendmessagesinsocialHowtoPromotePatientSafetyinSocialMedia359mediaforfree.
What'smore,messagesinsocialmediacanbespreadrapidlythroughsocialnetwork,andpeoplewiththesameinterestcanbeassociatedtogetherbythesamemessage.
Organizationsshouldmakefulluseofsocialmediaadvantagestomakethemessagesmoreattractiveandmeaningful.
Forexample,theycanprovidemorecolorfulpicturesandvideosontheirpublicpages,whichseldomappearinnewspapers.
Inaddition,theycansetupsomeactivitiesamongusersaboutpatientsafetytopic,suchaspatientsafetyknowledgecontest,emergencytreatmentpracticeandmedicalenvironmentsurvey.
Secondly,therearemoreprogramintroductionsandrecommendationsinsocialmediathaninnewspapers.
Thuspositivemessagesaremoreappropriatetospreadinsocialmediathantraditionalmedia.
Negativemessagesareusuallyastonishingandattractmorereaders,sotraditionalmediaismorelikelytoreportnegativeinformationtoattractconsumers.
Incontrast,whenspreadingpatientsafetyinformationinsocialmedia,organizationsareaimedatprovidingcomprehensiveknowledgetoreaders,includingbothnegativeandpositivemessages.
Sotherearemorepositivemessagesinsocialmediathaninnewspapers.
Besides,userswhovisitpatientsafetypublicpagesinsocialmediaareconcernedwithpatientsafetyissues,sothepublicpagesshouldprovidemoreinspiringandpracticalinformationforthem.
Eventhoughsomenewsisabouttragicaccident,furtherimprovingactionorrecommendationshouldbeattached.
Thisprincipleisnotonlyforpatient-safetypromotion,butalsoforallthecommonwealconceptpromotion,suchasenergyconservation,familyplanningandpreventionofchildmaltreatment.
Anothercharacteristicofsocialmediamessagesiscomprehensiveness.
45samplemessagesinsocialmediacovered10patientsafetyproblems,while45samplenews-paperreportsonlymentioned8problems.
Socialmediapaylessattentionontimeli-nessandreaders'interest.
Thussocialmediaisabetterplatformtopromoteoverallknowledgeofpatientsafety.
Thisisalsotrueforothertopics,suchasnon-smoking,healthydietandanimalprotection.
Whenspreadingknowledgeviasocialmedia,spe-cialattentionshouldbepaidonunusualknowledgetoopenusers'minds.
Forexam-ple,NationalPatientSafetyFoundationreportedaprogramonambulatorypatientsafety,whichisararetopicintraditionalmedia.
Thelastcharacteristicofsocialmediaisindividuallyoriented,whichmeansmes-sagesinsocialmediaconcentrateonindividualconceptandknowledge.
Forexample,socialmediaprovidesmoretreatmentrecommendationsforindividualsthannewspa-pers.
Traditionalmediausuallyhasthefunctionofpublicopinionandisabletoinflu-encethegovernment,soitfocusesmoreonmacro-environment.
Incontrast,socialmediamainlyaimstoincreasetheawarenessandknowledgeofindividualusers,in-steadofimpactingthemacro-environment.
Asaresult,morerecommendationsforpatients,doctorsandnursesshouldbeprovidedinsocialmedia.
Therearesomelimitationsinthisstudy.
Firstly,thesamplesizeissmall,whichpartiallyduetothecomplexityofhermeneuticcontentanalysis.
Secondly,theproce-dureofcontentanalysisissubjective,soindividualdifferencesmayimpactthere-sults.
Infurtherstudies,moreobjectivemethodologyneedstobeexplored.
Andtheanalysisofsocialmediamessagesshouldbeextendedtoothertopics.
360N.
SunandP.
-L.
PatrickRauReferences1.
Kohn,L.
T.
,Corrigan,J.
M.
,Donaldson,M.
S.
:ToErrisHuman:BuildingaSaferHealthSystem.
NationalAcademyPress,Washington,DC(2000)2.
Boyd,D.
M.
:Socialnetworksites:definition,history,andscholarship.
JournalofComput-er-MediatedCommunication13,210–230(2008)3.
Lenhart,A.
:TeensandSocialMedia:AnOverview(2009)4.
Kaplan,A.
M.
,Haenlein,M.
:Usersoftheworld,unite!
ThechallengesandopportunitiesofSocialMedia.
BusinessHorizons53,59–68(2010)5.
Ehara,A.
:Arelongphysicianworkinghoursharmfultopatientsafety.
PediatricsInterna-tional50,175–178(2008)6.
Ramanujam,R.
,Abrahamson,K.
,Anderson,J.
G.
:Influenceofworkplacedemandsonnurses'perceptionofpatientsafety.
NursingandHealthSciences10,144–150(2008)7.
Chuang,S.
,Pan,C.
,Huang,C.
:Asystem-orientedanalysismodeltoenhancepatientsafe-tyinhealthcareorganizations.
SystemsEngineering12(3),218–232(2009)8.
Wachter,R.
M.
:Theendofthebeginning:Patientsafetyfiveyearsafter"ToErrIsHuman".
In:HealthAffairs,pp.
w4.
534–w4.
545(2004)9.
Feng,X.
,Bobay,K.
,Weiss,M.
:Patientsafetycultureinnursing:adimensionalconceptanalysis.
JournalofAdvancedNursing63(3),310–319(2008)10.
Colla,J.
B.
,Bracken,A.
C.
,Kinney,L.
M.
,Weeks,W.
B.
:Measuringpatientsafetyclimate:areviewofsurveys.
Quality&SafetyinHealthCare14,364–366(2005)11.
Semetko,H.
A.
,Valkenburg,P.
M.
:FramingEuropeanpolitics:Acontentanalysisofpressandtelevisionnews.
JournalofCommunication50(2),93–109(2000)12.
Entman,R.
M.
:Framing:Towardclarificationofafracturedparadigm.
JournalofCommu-nication43(4),51–58(1993)13.
Akhavan-Majid,R.
,Ramaprasas,J.
:Framingandideology:AcomparativeanalysisofU.
S.
andChinesenewspapercoverageoftheFourthUnitedNationsConferenceonWom-enandtheNGOForum.
MassCommunicationandSociety1(3-4),131–152(1998)14.
Graneheim,U.
H.
,Lundman,B.
:Qualitativecontentanalysisinnursingresearch:concepts,proceduresandmeasurestoachievetrustworthiness.
NurseEducationToday24,105–112(2004)15.
Matthes,J.
,Kohring,M.
:Thecontentanalysisofmediaframes:Towardimprovingrelia-bilityandvalidity.
JournalofCommunication58,258–279(2008)16.
Segvic,I.
:Theframingofpolitics:AcontentanalysisofthreeCroatiannewspapers.
TheInternationalJournalforCommunicationStudies67(5),469–488(2005)17.
Hsieh,H.
F.
,Shanno,S.
E.
:Threeapproachestoqualitativecontentanalysis.
QualitativeHealthResearch15(9),1277–1288(2005)