warwick.ac.uk/lib-publications

Original citation: Grove, Amy L., Johnson, Rebecca E., Clarke, Aileen and Currie, Graeme. (2016) Evidence and the drivers of variation in orthopaedic surgical work : A mixed method systematic review. Health Systems and Policy Research, 3 (1). Permanent WRAP URL: http://wrap.warwick.ac.uk/78073 Copyright and reuse: The Warwick Research Archive Portal (WRAP) makes this work of researchers of the University of Warwick available open access under the following conditions. This article is made available under the Creative Commons Attribution 4.0 International license (CC BY 4.0) and may be reused according to the conditions of the license. For more details see: http://creativecommons.org/licenses/by/4.0/ A note on versions: The version presented in WRAP is the published version, or, version of record, and may be cited as it appears here. For more information, please contact the WRAP Team at: wrap@warwick.ac.uk

iMedPub Journalshttp://www.imedpub.com/ HEALTH SYSTEMS AND POLICY RESEARCH

2016Vol. 3 No. 1: 25

1© Under License of Creative Commons Attribution 3.0 License | This article is available from: http://www.hsprj.com/archive.php

Review Article

Grove A1, Johnson R2, Clarke A2 and Currie G3

1 Warwick Medical School, University of Warwick, Gibbet Hill Road Coventry, CV4 7AL, England

2 CollaborationforLeadershipinAppliedHealth Research and Care West Midlands, Warwick Medical School, University of Warwick, Coventry, CV4 7AL, England

3 Entrepreneurship&Innovation,Organizing Healthcare Research Network,WarwickBusinessSchool,University of Warwick, Coventry, CV4 7AL, England

Corresponding author: Amy Grove

A.L.Grove@warwick.ac.uk

MScBScMBPsS,NIHRDoctoralResearch Fellow, Warwick Medical School, University of Warwick, Coventry, CV4 7AL, England.

Tel: +44 (0)2476151584

Citation: Grove A, Johnson R, Clarke A, et al.EvidenceandtheDriversofVariationinOrthopaedicSurgicalWork:AMixedMethodsSystematicReview. Health Syst Policy Res.2016,3:1.

Introduction Inthisstudywesetouttodiscoversomeofthepossibleunderlyingdriversforvariation inorthopaedicsurgicalworkdespitethegrowing evidence from researchers and policymakers onbest practice in orthopaedics. Internationally, the pressureon orthopaedic services is growing. Musculoskeletal (MSK)conditions are a major burden on individuals, health, andsocialcaresystems.IntheUKforexample,theannualbudgetfor treatment ofMSK disorders is approximately £10 billion

per year [1]. The increasing demands on orthopaedic servicesbrought about by the aging population and rising incidence of arthritismeanthatthereisamountingrequirementforincreasingorthopaedictreatmentswithinarestrictedbudget.Thispressuretodelivermoreforlesshasledtopatternsofvariationincurrentpracticeandprovisionoforthopaedicsurgerywhicharedrivenbyfactorsunrelatedtopatientneed[2].

Levels of variation are substantial and deemed unacceptable,hencepolicymakingorganisationssuchastheNationalInstitute

AbstractThe increasing demands on orthopaedic services brought about by the agingpopulationandrisingincidenceofarthritis,togetherwiththefinancialconstraintsinthehealthcaresystemmeanthatorthopaedicdepartmentshavetodelivermoreforless.Thisiscoupledwithpatternsofvariationincurrentpracticeandprovisionof orthopaedic surgery which are not justified by variation in patient needs.Researchers and policymakers set out to develop evidence-based guidelines,interventions, and recommendations to address the balance and ensure thatservicesarebothclinicallyandcost-effectivewhilstbeinghighquality, safeandequitable. However, it is not clear to what extent these external sources ofevidenceandguidelinesareacceptedandimplementedinorthopaedicpractice.

We conducted a systematic review using an interpretative mixed methodssynthesis and thematic analysis. We aimed to understand: i) Approaches andtechniquestoknowledgeandevidenceuse,andii)Factorsinfluencingdecision-makingandpracticevariationwithinorthopaedicsurgery.Twosystematicreviewsand24primarystudieswereincluded.Evidencesourcesweregroupedintoeightthemesandpresentedinaconceptualframework(formalcodifiedandmanagerialknowledge, medical socialisation, cultural, normative and political influence,trainingandformaleducation,experientialfactors,individualpatientandsurgeonfactors).Formalcodifiedknowledgeintheformofclinicalguidelinesappearedtoplayasmallpartinclinicalpracticedecision-makinginorthopaedicsurgery.Moresignificantdriversofvariationincludemedicalsocialisation,culturalandnormativefactorsandtheexistenceofcomplexandcompetingsourcesofevidence.Thesedriversneedtobeincorporatedintoanewmethodtoimplementevidence-basedknowledgeinpracticewhichtakesaccountofthecontextinwhichthedecisionismade.

Keywords: Evidence-basedmedicine;Orthopaedics;Decision-making;Systematicreview;Mixed-methods

Evidence and the Drivers of Variation in Orthopaedic Surgical Work: A Mixed

Methods Systematic Review

Received: December15,2015; Accepted: March06,2016; Published: March 09, 2016

2 This article is available from: http://www.hsprj.com/archive.php

2016Vol. 3 No. 1: 25HEALTH SYSTEMS AND POLICY RESEARCH

forHealthandCareExcellence(NICE)intheUKandtheAgencyfor Healthcare Research and Quality (AHRQ), the PreventiveServices Task Force (USPSTF) in theU.S. setout to standardisecare across healthcare systems through the implementationof evidence [3,4]. Evidence-based medicine informs clinicalguidelineswhichareproducedtoestablishclinicaleffectivenessand criteria for treatments at acceptable levels of expenditure[5,6].Theimpactoftheseevidence-basedpolicies,interventionsandrecommendationswithinorthopaedicsisunknown,asarethebarriersandfacilitatorstotheuseofevidence-basedmedicineindecision-makingfororthopaedicsurgery.

Previousresearchdatingbacktothe1930'shighlightedvariationin surgicalpracticeacrossandwithingeographical areas [7-12]andwithindifferentsurgicalspecialties[13,14]andsub-specialties[15,16].Ratesofsurgicalinterventiondonotalwaysaligntoratesofdisease.Many reasonsaregivenprovided forperformingorlimitingsurgerywhichmaynotdirectlylinktotheclinicalneedsofpatientorpopulations[17].Thispaperisfocusedspecificallyon orthopaedic practice where the rate of joint replacementsurgeryhasbeen reported to varyby geographical population,hospital and by surgeon [18,19]. Inconsistencies in proceduresconducted and the prostheses selected for joint replacementhavealsobeendescribed,aswellassignificantvariation inthelevelofexpenditureandpatientreportedoutcomes[2,20-22].

Thesedifferencescannotbeexplainedbydifferencesintheclinicalneedsof patients andhence it is important to understand thedrivingforcesbehindthisvariation.Thisresearchwasdesignedtoreviewandsummarisethepublishedliteraturesystematicallyi) tounderstandapproachesand techniques toknowledgeandevidenceuse,andii)toidentitythefactorsinfluencingdecision-making and practice variation within orthopaedic surgery.Throughthissystematicreviewwehighlightsourcesofevidenceand knowledge that have been important in decisions madewithinorthopaedicsurgery,andexplainhowthesecomplexandcompetingsourcesofevidencemayleadtounjustifiedvariationin practice. We propose a conceptual model to highlight therelationshipbetweenevidencesourcesandpracticeinhealthcare.

Methods Protocol registration WeusedCochraneCollaborationguidelines topreparea studyprotocol for thesystematicreview[22]andregisteredwiththePROSPEROdatabase:CRD42015016792.

Inclusion and exclusion criteria We included all study designs and orthopaedic surgicaltreatments.Allhealthcareprofessionalsormemberofthesurgicalteaminvolvedindecision-makingfororthopaedicpatientswereincluded. The main outcomes were evidence sources used indecisions and the approaches and techniques to evidence useanduptake.NocountryrestrictionswereappliedbutonlyEnglishlanguage papers were included. We excluded basic researchstudies, abstract or conference proceedings, editorials, lettersandcommentaries;individualcasestudiesandstudiesincludingchildren.

Information sources and selection of studiesIndividual studieswere identified using a structured electronicsearch strategydeveloped forMEDLINE and adapted for otherdatabases. The electronic databases and web resources arelisted in Table 1.Wescannedreferencelistsofincludedstudiesandcontactedkeyexperts inthefield.Asupplementarysearchtechniquewasusedtosearcharoundclustersofpapersbasedoncitations,relevanttheoryandconcepts,suchas'evidence-basedorthopaedics' [23,24]. Searches were undertaken in February2014 and includedpapers published between1946 to January2014(SupplementaryfileS1).

Data extraction and quality assessment Two reviewers (AG/RJ) simultaneously extracted the following core data from the included studies: aims and objectives,research methods, participant characteristics, intervention,evidence sources and use in decision-making and approachesandtechniquestoevidenceuseanduptake.Otherrelevantdatawereextracted,suchasanystatedtheoreticalunderpinningsormechanismsofaction.Duringthisprocessprovisionalcategoriesemerged from the data and were iteratively ‘sense checked’with the review team. This enabled us to begin to develop aconceptual frameworkthatwouldbringthestudiestogether inameaningfulnarrative[25].StudieswerequalityappraisedusingtheCriticalAppraisalSkillsProgramme(CASP)toolforqualitative,quantitativestudiesandsystematicreview[26].Theriskofbiaswasconsideredacrossallstudies,andresultswereexaminedformissingdatawithinindividualstudies.

Methods of analysisAnalysis was undertaken using interpretative mixed methods

Databases Web resources and grey literature Medline[OVID] King’sFundLibraryDatabase

MedlineIn-ProcessandOtherCitations[OVID] AgencyforHealthcareResearchandQualityNationalGuidelineClearingHouse

Embase[OVID] TheJoannaBriggsInstituteLibraryASSIA[ProQuest]CochraneLibrary[Wiley]includingCDSR,DARE,CENTRAL,NHSEED,HTADatabaseCINAHL[EBSCO]PsycINFO[ProQuest]ScienceCitationIndexandSocialScienceCitationIndex[WebofScience]

Table 1Databasesandotherwebresourcesusedinthesystematicreview.

3© Under License of Creative Commons Attribution 3.0 License

2016Vol. 3 No. 1: 25

HEALTH SYSTEMS AND POLICY RESEARCH

synthesis and using thematic analysis and narrative synthesis[27].Theinitialanalysisinvolvedfamiliarizationandimmersioninthedata,andcomparisonbetweentheincludedstudies.WeusedthePillarIntegrationProcess(PIP)toincorporatebothqualitativeandquantitativedatainameaningful,yettransparentway[28].PIP enabled the articles to be drawn together systematicallywhilstalsoprovidinganindependentviewofthecurrentstate-of-the-artandcumulativeknowledgeonthesubject[29-31].

Studies were integrated using a convergence coding matrixand synthesised narratively. The first three stages of PIP (thelisting,matching and checking)were followed to construct thefoundations of the PIP codingmatrix as displayed in Figure 1. Subsequently, we completed the more creative fourth ‘PillarBuilding’ stage. Using thematic analysis we coded the datafromallpaperstodevelopnewconstructsbyidentifyingrelatedconcepts in the studies. We then summarised the findings ofdifferent studies under thematic category headings, reworkingand reformulating themtogether into thecentral ‘pillar’.Usingtheemergentcategoriesandconceptualframework,anarrativesynthesiswasdevelopedtoexaminerelevantthemes,identifyingpatternsandanomaliesacrossthestudies.

ResultsThe flow chart outlining the process of identifying relevantliteraturecanbefoundinFigure 2[31].Atotalof14,316articleswerefoundofwhich13,804didnotmeetthe inclusioncriteriaandwereremoved.Titleandabstractscreeningremovedafurther392studies,leaving120articlestobeinvestigated.Fulltextsofthesestudieswereexaminedforinclusionbytworeviewers(AG/RJ). Disagreements were resolved by consensus or third partyreview.Ninety-fourstudieswereexcludedwithreasonsresultingin 26 final articles (24 primary studies, 2 systematic reviews).Included studies were heterogeneous in methods used andphenomena studied.Hence,meta-analysis of quantitativedatawasnotconsidereduseful.

Characteristics of the studies selected Theprimaryaimofmanyoftheincludedstudieswastoidentifytheinfluenceofafactororvarietyoffactorsonaparticularaspectofsurgicalpracticeortreatmentdecision.Studiesvaried inthefactors they assessed over a range of orthopaedic procedures

includinghipandkneearthroplasty,hipfracture,upperextremityandspinesurgery,anteriorcruciateligamentsurgeryandtheuseofassociatedproceduressuchasbloodtransfusionanddrainage,steroidinjectsandphysicaltherapy.

Summary of primary study evidenceTwenty-four primary studies were included, themajority werequantitative (n=20) the remainder were qualitative (n=4). Webroughttogetherthestudiessystematicallyusingsummarytablespresented in SupplementaryfileS2.A summaryof SPIDERandPICOcharacteristicsisprovidedinSupplementaryfileS3toordertheevidenceandproduceadescriptionofstudycharacteristics.A large number of the studies reported cross-sectional surveydata (n=16). Other study designs included: prospective (n=3),retrospective (n=3) and qualitative studies (n=4) including:interviews(n=2)focusgroups(n=1)observation(n=1)andacasestudy(n=1).

Summary of systematic review evidence Twosystematicreviewstudieswereincludedinthereview.Thesewere undertaken by Barr et al. and Bederman et al. [32,33].The first addressed the drivers of transfusion decision-makingin orthopaedic surgery and the second aimed to discover thedecision-making drivers for degenerative hip, knee and spinesurgery.

Pillar Integration Process The initial stages of PIP resulted in the identification of 44prominentcodesintheincluded26studies.ThePIPcentralpillarintegrationprocessresultedineightthemes(Table 2).

AnoverviewdiagramofthedatastructureispresentedinFigure 3.DisplayingthedatastructureinthiswayisrecommendedbyPrattetal.2006,2009[34,35]Prattarguesthatthechallengeofqualitativeandquantitativeresearchisthattherearenoaccepted“boilerplate” forwriting upmethods and determining quality[35] andobjectivity. Theboilerplate refers to a standardisedlanguageandformatforpresentingresearchfindings,andthisisnot somethingqualitativeandmixedmethods researchersstrive to achieve. This equifinality can make it extremelyproblematic to portray qualitative and qualitative research[35].However,presentingthedatausinganoverviewdiagramas in Figure 3,enablesustohonortheworldviewofthearticlesthat were included in the review, provide sufficient evidenceforclaimsmade,andallowsus tocontributes toextant theorythroughtheconclusionsmade[35].

Discussion The review identified sources of evidence, or the knowledge types reported as important for orthopaedic decision-making.These sources are influential in determining patient treatmentandhelptoexplainhowandwhythereisunwarrantedvariationin orthopaedic surgical practice. Factors were identified and categorisedintoeightthemeswhichreflectthemicro-levelpatientandclinicaldrivers;andmesofactorssuchascharacteristicsoftheorganisationorsurgeonthroughtotheimpactofformaltraining.Atthemacro-levelweidentifiedtheinfluenceofevidence,policy

A diagrammaticrepresentationofthepillarintegrationprocess. Figure 1

4 This article is available from: http://www.hsprj.com/archive.php

2016Vol. 3 No. 1: 25HEALTH SYSTEMS AND POLICY RESEARCH

Flowchartoutliningtheprocessofidentifyingrelevantliterature.Figure 2

andguidelines.Eachthemeisdescribedbelowandaconceptualmodelispresentedwhichdemonstratestherelationshipbetweenthesesourcesofevidenceandknowledgetypes.

Formal codified knowledgeFormal codified knowledge is explicit, written down and thusavailabletoeveryonetousealongsidepersonaljudgment[34].Inourreview,formalcodifiedknowledgerepresentsthemacro-levelclinicalguidelinesandscientific literaturetowhichorthopaedicsurgeonscan(andareexpectedto)referwhenmakingevidence-baseddecisions. This knowledge ishard, factual, speltout andeasytotransferbetweenindividuals.Hencetheassumptionisto‘policymakersisthat’standardisedknowledgeinguidelinescanbedisseminatedacrossclinicalpopulations.

Formalcodifiedknowledgewas reportedto influencedecision-makingintenofthe26studies.Itincludedreferencetoguidelines[36-38], evidence-based medicine [39-44] and independentpeerreviewed literature[43,45].The includedstudiesreporteda low influence of this type of knowledge. This low influenceis in contrast towhatwould be expected by the advocates of

evidence-basedmedicine and the significance attached to the hierarchyofevidenceintheclinicalfield.

Managerial knowledgeManagerial knowledge represents an important component ofclinical decision-making within orthopedics as it can underpintheroutinesandcapabilitiesofpractice–i.e.,thescopeofworkconducted.Inthisarea,theliteratureoftenreferredtoresourceissues such as time, cost and safety or quality of services butwithout definite or consistent criteria of what is consideredacceptable. For example, a “treatment is too expensive” wasconsideredafactorthatinfluencesclinicaldecisionsbutwithoutproviding a context within which to benchmark the conceptof ‘expensive’ [36]. Managerial knowledge is subjective andexperiential and is often notwritten down for healthcare stafftoaccess.Thismakesitdifficulttotransferbetween and across organisational, departmental and also across professionalboundarieswithinthesameorganisation[35].

In the six studies which mention managerial knowledge, cost [36,46-51] and availability of resources [47,48] were mostinfluential in clinicaldecision-making.The knowledge and skills

5© Under License of Creative Commons Attribution 3.0 License

2016Vol. 3 No. 1: 25

HEALTH SYSTEMS AND POLICY RESEARCH

Table 2Pillarintegrationprocessconvergencecodingmatrix.

1 2 3 4 5

QUANT STUDIESCITATION

FACTOR(S)/SOURCES OF EVIDENCE PILLARTHEMES

FACTOR(S)/SOURCES OF EVIDENCE QUAL STUDIESCITATION

deBoeretal.[34],Wattsetal.[35],Zielinskietal.[36]

Marxetal.[37],Khanetal.[38],Krahnetal.[39]

Marxetal.[37],Kumaretal.[40]

Schulzetal.[41]

Tejawanietal.[42]

Cantyetal.[43],Khanetal.[38],Tejiwanietal.[42]

Practiceguidelinesarepresent

More agreement when there is more evidence,useevidencewhenitexists

Poordisseminationofevidence,difficulttoaccessEBM

Independentpeerreviewedpapers

Literature(formal)

Believe it is evidence based medicine,BeliefinRCTs

Formal codified knowledge

deBoeretal.[34]

Hagemanetal.[44]

Kegaletal.[45]

Lingardetal.[46]

Okikeetal.[47]

Treatmentistooexpensive

Treatmentischeapest

Priceofthemedicationandavailability

Belief costs and availability of extendedcarefacilities

Costknowledgeassociatedwithuseofimplants

Managerial knowledge

deBoeretal.[34]

deBoeretal.[34]

deBoeretal.[34]

Krahnetal.[39],deBoeretal.[34]

Lingardetal.[46]

Lingardetal.[46],Hagemanetal.[49]

Wrightetal.[48]

Lackofequipmentorfacilities

Rotalimitedactions

Supportstaffunabletodowhatisrequired

Time(pressure)

Waitingtimes/lists

Fundingstatusofhospital,reimbursementorsurgeons

Organisationalstatusassociationtoamedical school

Organisational knowledge

Constraints of the healthcare system/institutionalconstraintssuchasavailableoperatingtheatres

Extensivewaitinglists

Suitabledoesn’tguaranteesurgeryunfair/discriminate/bestratherthanthemostappropriate

Prioritisationofpatientseveryonewho needs one is not going to get one

Linkwithteachinghospitalisimportantingettingevidenceintopractice

Availabilityofresources(imagining)

Hudaketal.[49]

Hudaketal.[49]

Hudaketal.[49]

Hudaketal.[49]

Ferlieetal.[8]

Bedermanetal.[33]

6 This article is available from: http://www.hsprj.com/archive.php

2016Vol. 3 No. 1: 25HEALTH SYSTEMS AND POLICY RESEARCH

deBoeretal.[34]

Hagemanetal.[44]

Hagemanetal.[44]

Hagemanetal.[44],Kegaletal.[45]

Kumaretal.[40]

Schulzetal.[41]

Vashitzetal.[50]

Supervisorpreventeduse

Whatmymentortaughtme

Burnsfewerbridgeswithcolleagues

What others are doing, what colleaguesuse

Morelikelytodiscusswithcolleagues

Meetingswithcolleaguesandconferences

Influencedbyopinionofothersandpreviousopinions

Socialisation and association with

colleagues

Clinicianplaysarole/cliniciansastheexperts/assignedandadoptedRoles

Relationshipbuildingbetweensurgeonandpatient/negotiatingrelationshipswithotherprofessionals/maintainingprofessionalnetworks/maintainpositionintheorganisation/maintainpeernetwork/socialandcultural

Changeinopinionisfast,decisionmakingandopinionwasvolatile

NottrusttheRCTsfromoutsidetheirgroup/alwaysfindapapertosupportyouridea,orthopaedicjournalsmostpowerfulposition

Indirectchannelsofinformationtransfer

Decisionsmadelocally/negotiated/acoregroupofprofessionalshavelegitimacyandarebelieved

Influentialactorsworkismorelikelyto be believed

Theorthopaediccommunityisimportant/professionalcommunitywithdistinctnorms(resilient/embedded/retaincontrol)/internalcommunity/groupknowledge

Gooberman-Hilletal.[51]

Hudaketal.[49]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Wattsetal.[35]

Wattsetal.[35]

Schulzetal.[41]

Medico-legalconcerns

Uniformity in the region in which surgeonpractices

Pharmaceuticalcompanies,implantmanufactures

Cultural, normative and political

influence norms of the sector

Highlevelofgroupandindividualautonomyoverwork,Highlyprofessionalisedformofclinicalwork

Orthopaedicsisdifferent/separate/autonomous,

Pharmaceuticalcompanies

Orthopaedicsareactorsnotreflectors,personalitycharacteristicsaredistinct/certainpeopleMedico-legalconcerns

Authorityfigure/judgment/controlofspecialists/experts

Supplyanddemand/gamingthesystem/havingastrategy

Lackofhomecare/postoperativesupport/judgmentbasedonaftercaresupport

Playingthemiddleman/medicalbroker

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Gooberman-Hilletal.[51]

Hudaketal.[49]

Hudaketal.[49]

7© Under License of Creative Commons Attribution 3.0 License

2016Vol. 3 No. 1: 25

HEALTH SYSTEMS AND POLICY RESEARCH

Cantyetal.[43]deBoeretal.[34]Irwinetal.[52]Kumaretal.[40]Kumaretal.[40]Schulzetal.[41]Tejawanietal.[42]Kegaletal.[45]

LearnedinpracticeDidnotlearnitDifferenttrainingbackgroundAcademicqualificationsinEBMFormal training in EBMTrainingcourseGainedduringtraditionaltrainingEducated,learnedinfellowship

Training and formal education

Cantyetal.[43]

Cantyetal.[43],Dusiketal.[53]

Wrightetal.[48]

Wrightetal.[45]

Tejiwanietal.[42]

Kumaretal.[40]

Personal reasons

Don’tbelieveinevidence(based)

Surgeonsopinionsabouttreatment

Surgeonsenthusiasmforprocedure

Personalexperiencegainedinpractice

BelieveitisdifficulttoadheretoEBM

Informal experiential Implicit knowledge

Surgeonsappraisalofthepatientscomplexity,variousexplicitfactorsplayapartinthejudgmentincludinglifestyleandclinicalinfluencesSkillscomefromexperienceandinstinctSurgeonsstyleofapproachtopatient,surgeonsconfidenceintheirownability/moreconfidentmorelikelytoperformsurgery/enthusiasmPersonalitycharacteristics/paternalisticdiplomaticconservativeassertive/stylesinfluencebehaviourBeliefwhoisableandbetter,notwho is in needImplicitdefinitionofpatientcandidacy/definitionsarerarelyexplicit/notknowingtherulesofthe gamePatientprioritisationinfluencesbehaviourHighlyprofessionalisedformofclinicalwork,disagreewithscience/differentviews/skepticalaboutformal scienceControversyintreatmentoptions/nottrusttheRCTsfromoutsidetheirgroup/alwaysfindapapertosupportyouridea,orthopaedicjournalsmostpowerfulpositionTheindividualconsultantontheinsideismoreimportantthanexternalsourceofevidenceOrthopaedicsisalearntcraft/tacitandexperientialmoreimportantPatientsinresearchdonotmatchtherealpatientsinpracticePracticeislearntslowly,itdoesnottransferacrossgroupseasilySurgeonbelievessurgery‘Worksinmyhands’IndividuallearningadreflectionisfavouredoverEBM/experience

Gooberman-Hilletal.[51]

Gooberman-Hilletal.[51]

Gooberman-Hilletal.[51],Bedermanetal.[33]

Gooberman-Hilletal.[51]

Hudaketal.[49],Bedermanetal.[33]Hudaketal.[49]

Hudaketal.[49]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]

Ferlieetal.[8]Ferlieetal.[8]

Ferlieetal.[8]

8 This article is available from: http://www.hsprj.com/archive.php

2016Vol. 3 No. 1: 25HEALTH SYSTEMS AND POLICY RESEARCH

Bhandarietal.[54],Irwinetal.[42],Wrightetal.[48],Zielinskietal.[36]

Bhandarietal.[54],Curtisetal.[55],Wattsetal.[35],Cantyetal.[43]

Bhandarietal.[54],Curtisetal.[55]

Borkhoffetal.[56]

Cantyetal.[43],Dusiketal.[53],Kegaletal.[45],Zielinskietal.[36]

Cantyetal.[43],Hagemanetal.[44]

Curtisetal.[55],Cantyetal.[43],Dusiketal.[53],Irwinetal.[52]

deBoeretal.[34]

Dusiketal.[53],Hagemanetal.[44],Kegaletal.[45]

Irwinetal.[52],Marxetal.[37]

Marxetal.[37]

Borkhoffetal.[56]

Patientsage

Patientsmedicalcondition

Patientlifestylefactors,socialcircumstance

Patientssex

Clinicaltreatmentandmedication

Practicalorpragmaticreason,shorterprocedure

Symptoms,pain

Needtohavepatientwiththeproblem

Yearsinpractice,levelofexperience

Surgeonsage

Surgicalvolume

Surgeonssex

Individual Patients and surgeon factors

Clinicalfactorsmostimportant/symptomseverity/pain/stiffness

Patientlifestylefactors/social/workcontext,familysupport/Post-operativesupport

Pragmaticreasons

Personalitycharacteristicsaredistinct/certainpeople

Patientsage

Patientsmedicalcondition

Patientssex

Surgicalvolume

Patientsweight

Surgeonssex

Patientsinsurancecoverage

Gooberman-Hilletal.[51],Barretal.[32],Bedermanetal.[33]

Gooberman-Hilletal.[51]

Ferlieetal.[8]

Ferlieetal.[8]

Barretal.[32],Bedermanetal.[33]

Barretal.[32]

Barretal.[32],Borkhoffetal.[56]

Barretal.[32]

Barretal.[32]

Bedermanetal.[33]

Bedermanetal.[33]

of individuals who manage healthcare organisations wereconsideredvaluablebutintangibleintheorganisations.Thisledtouncertainly,forexampletreatmentcostsinfluenceddecisionswhenbothexpensiveoptions[36]andcheapertreatmentoptionswere available [44]. Surgeons’ knowledge of treatment andorthopaedicimplantcostswasalsoassociatedwiththeiruseinpracticebutthedefinitionofacceptablecostwasnotexplicit[49].Theweightassignedtomanagerialknowledgeinorthopaedicsisincreasingly importantduetotherisingdemandfortreatment,andreinforcedbypressurestoreduceresourceuse.

Organisational knowledgeOrganisationalknowledgehasawiderstructuralemphasis. It isanecdotallyreferredtoas“thewaywedothingsaroundhere”asitshapestheperspectivesofcliniciansworkinginanorganisation.Itisembeddedintheprocessesofhealthcareorganisationsandinfluencesthebehaviorofitsmembers.Organisationalprocessesbecomenormativeandreflect thecommoneducation, trainingandcareerstructuresofparticularorganisations[50].Thistypeof knowledge is ingrained in the routines of the orthopaedic

departmentsandhospitalsbutnotnecessarilyacknowledgedbytheindividualsthemselves[47].

Organisational constraints such as theatre availability, surgicalwaiting lists and patient prioritisation acted as forms oforganisational knowledge in the included studies [8,47,51].Variation in thepracticeofpatient categorisationor treatmentdelay resulted from organisational knowledge that does notdiffuse but becomes stickywithin the organisation.One paperstated“prioritisationofpatientsmeansthateveryonewhoneedsone is not going to get one” [51]. Time pressures and staffinginfluenced clinical decisions when planning surgery [36,41].These organisational factors develop over time and becomeentrenched,sothatknowledgeexistsintheprocessesandclinicalpathwaysthemselvesnotintheindividualactors.Thelikelihoodof finding a common ground for collaboration and knowledgesharingwithinbutnotacrossorthopaedicdepartmentstoreducevariationinthehealthcareorganisationsislimited.

Socialisation and association with colleaguesThesocialisationofindividualsintodifferentclinicalprofessions

EBM=Evidence-BasedMedicine

9© Under License of Creative Commons Attribution 3.0 License

2016Vol. 3 No. 1: 25

HEALTH SYSTEMS AND POLICY RESEARCH

plays an important role in their decision-making processes.Orthopaedicsurgeryrepresentsahighlyprofessionalisedareaofclinicalworkasaspecialtywhereanelitecommunityofpracticeis strongly embedded [8,9]. This community has socialisedknowledge that impacts on theway decisions aremade by itsmembers.Theknowledgeistreatedasavitalsourceofevidencewhich isheld in thegroupbutnot sharedwithoutsiders.Ninepapers reported socialised knowledge influenced decisions.Examples include “what my mentor taught me”, restrictionsplacedonpracticeby supervisors [31,43] “whatmy colleaguesaredoing”[46,47,51]andthenotionthat“decisionsmadelocallyandnegotiatedbyacoregroupofprofessionalshavelegitimacyandarebelieved”[8,51-53].

There was a distinction between knowledge that came frominsideoroutsidethedefinedgroup,inthiscasetheorthopaediccommunity [8,52]. Surgeons reportedlydidnot “trust theRCTsfromoutsidetheirgroup”[38]andweremoreinfluencedbyworkthat came from “influential actors” in the field [8] Particularlyimportant were knowledge and evidence gained at localprofessional meetings and conferences [42,43] Decisions aresaidtobedrivenbysocialisedknowledgebecauseorthopaediccommunities share common values, language, procedures andknow-how,andhenceactasasourceoflearningandpracticeforeach other.

Culture, norms and political influence of the sectorThisthemedemonstrateshowthewiderorthopaedicprofessioncan influence clinical practice and drive decision-making forpatients. For example, the presence of professional societies,such as the British Orthopaedic Association in the UK andthe American Academy of Orthopaedic Surgeons in the USA,enabledthememberstoretainsubstantialautonomy,authorityand control over their work practices and to resist externalintervention [8,51,53]. This external intervention might be intheformofclinicalguidelinesandregulationwhicharecodifiedevidence produced outside the orthopaedic sector. The use ofthistypeofevidenceindecision-makingwaslikelytobecomplexand fraughtwithpoliticalchallenges. Itwas toacertainextentlinked to how surgeons maintained their elite position in thewider clinical field by privileging their normative professionalknowledge over clinical guidelines that can be accessed byanyone[51,54].Therefore,itcaninfluencedecisionsatalllevels:the micro individual surgeon deciding to operate or not, themesocommunitiesofpracticewhoplanandallocateinterventionthresholds, and the macro professional groups who buildconsensusstatementsandestablishorthopaediccriteria.

Medico-legal challenges to practice drove patterns of a- todecision-makingtominimiselegalactionintwopapers[8,35].Thepowerofpharmaceuticalandimplantmanufacturingcompanieswithinthissectorwasalsohighlighted[8].Aswasthe“struggletobalancesupplyanddemand”ofservices[49]andbeliefsabouthowalliedhealthprofessionscanorcannotcontributetoservicedelivery [51,55]. Inonestudy,adecision for surgerydependedonthesurgeonssubjectiveviewof“appropriatepost-operativesupport” [50] Attitudes towards the ability and or use of non-

surgeonprofessionalsvaried[53].Thiscouldbeassociatedwithincentives to maintain professional control and power overclinical decisions. The process of prioritising, negotiating andjugglingthesefactorsduringdecision-makingwasreferredtoas“medicalbrokering”anddefinedas“aproceduralstrategyusedbyphysicianswhendemandsinasystemareperceivedtoexceedresources”inonepaper[51].

Training and formal educationIt would be expected that a surgeons’ training and formaleducationmighthaveanimpactuponhowtheymakedecisionsforpatients.Sevenpapers reported thatsomeformof traininginfluenced surgeons’ clinical decisions [36,42-45,47,54]. Thetraining and formal education theme covered the standard“academicqualifications”[40]suchasmedicaltrainingbutalsothe apprenticeship style training gained through fellowshipprogrammes and practice-based learning [45,51,54]. Trainingcourses undertaken later in surgical practice were consideredmore important to surgeons [43].This ismaybebecause thesereflect the subspecialist training that the surgeons were mostattracted to. Formal training in evidence-based medicine wasreported its to increase its use in clinical decisions, and theperceivedimportanceofevidencetopractice[42].Trainingandformal educationbecomes an important foundationwhich canbebuiltonovertimeusingelementsfromallotherknowledgetypes.

Informal experiential knowledgeThisrepresentsthetacitknowledgethatsurgeons‘know’whichhas built up over time but which can be difficult to describe.Thistypeofknowledgeistreatedasoneofthemostimportantsourcesofevidence indecision-making inorthopaedicpracticebecause it represents a surgeon’s lifetime’s work, and in turntheiridentityasasurgeon.Whenconsideringthetacit–explicitknowledge spectrum, informal experiential knowledge sits inopposition to formal codified knowledge. It cannot easily beexplained, transferred and understood by another person,particularly someone outside the orthopaedic community.Individual knowledge gained from their experience does notexistintheactivityalone,butintheknowledgethatindividualsuse to perform the activity [56]. Hence, a surgeon possessestacitknowledgeofhowtoperformanoperationwhentheyareoutsideoftheatre.

The included studies reported several examples of informalexperientialknowledgewhichdrivedecisions[8,51,53].Referencetosurgeon’s judgment,skill,craft,experienceandinstinctwereall reported as important [8,51,53]. As were confidence, styleandapproachtopatients’aftertreatmentalongsidebeliefsaboutpatients ingeneral [55]. Forexample surgeonshada “belief inwho is able and better (for surgery) not who is in need”[49]whichcontrastswiththeconceptofclinicalneedorapatients’requirementforsurgerywhichmightbeexpectedwhenapplyingtheprinciplesofevidence-basedguidelines.

Individual patients and surgeons factorsThe final theme clusters all factors that were directly relatedto thecharacteristicsof thepatientor surgeon that influenced

10 This article is available from: http://www.hsprj.com/archive.php

2016Vol. 3 No. 1: 25HEALTH SYSTEMS AND POLICY RESEARCH

Overviewofdatastructure.Figure 3

First-Order Codes Theoretical Categories Aggregate Theoretical Dimensions

Statements about guidance from NICE , the BOA and professional societies in orthopaedics and sub-specialties acting as a source of evidence National guidelines

Formal codi�ied knowledge

Managerial knowledge

Organisational knowledge

Socialisation and associate with

colleagues

Cultural normative, and political

in�luence norms

Training and formal education

Informal experiential implicit

knowledge

Individual patient and surgeon

factors

Academic journal articles

Findings of RCT studies

• Statements about what RCTs mean for orthopaedics , why they are a useful source of evidence .

• Comments about beliefs and trust in data that comes from RCTs

• Statements about “independent peer reviewed” papers in trusted journals • Academic literature that is published in reputable forms

Statements about organisational �inances in general that in�luence clinical practice and service delivery decisions

Healthcare �inance

Speci�ic reference to orthopaedic implants and treatment or pharmaceutical costs that linked to decision made about patient treatment “treatment is too expensive”

Cost of orthopaedic provision

• Statements about hospital equipment and facilities availability such as operating theatres “constraints caused by theatre”

• Staff availability and scheduling of staff that cause problems in planning surgery for patients

• Statements about high demand for orthopaedic procedures • Comments about growing orthopaedic waiting lists in�luencing decisions

Statements about the need to prioritise orthopaedic patients treatment for reasons not linked to clinical need “being suitable does not guarantee surgery”

Availability of equipment and resources

Demand for surgery

Patient prioritisation

Statements about surgeons trusting other surgeons in the profession as a valid source of evidence. Other Surgeons “decisions have legitimacy”.

Statements about learning from mentors and other colleagues that in�luence decisions. Surgeons are more likely to exchange knowledge as evidence with other surgeons and “discuss with colleagues” when making decisions.

Trust in other surgeons

Knowledge exchange between surgeons

• Statements about high pro�ile and expert orthopaedic surgeons acting as a source of evidence.

• Surgeons were “in�luenced by the opinion of others” and signi�icant individuals were more “likely to be believed”.

In�luential experts

• Statements about the impact of medical malpractice and liability claims that in�luence or restrict clinical decisions.

• The in�luence of pharmaceutical and implant manufacturer companies upon surgeons decisions.

External in�luence

Surgeon characteristic norms

• Statements about orthopaedic surgeons as a particular type of medical professional displaying certain behaviours. “Surgeon uniformity”

• Surgeons were “in�luenced by the opinion of others” and signi�icant individuals were more “likely to be believed”.

Statements about supply and demand issues across healthcare in general that impact on orthopaedic service delivery, such as “lack of post operative support”.

Healthcare supply and demand

Statements about academic education and formal quali�ications from medical school that in�luence later orthopaedic practice.

Surgeon training

Surgeon education

Statements about training in later career, such as course and “formal training in EBM” that drive clinical decisions for surgeons.

Statements about surgeons’ personal beliefs and attitudes about their practice that inform decision making.

• Statements about surgeons’ clinical experience built up over years of clinical practice. “Personal experience gained in practice”

• Comments that re�lect how surgeons’ instinct and tacit knowledge was an important source of evidence.

• The importance of selecting and trusting personal experience over other evidence sources.

Statements about orthopaedic surgeons not trusting information that comes from “outside” of their own profession

Surgical experience

Attitudes and beliefs of surgeons

Surgical evidence is credible evidence

Statements about factors that relate to individual patient factors, including patient demographics. For example “patient sex” and “patient age”

• Statements about routine surgical procedures, treatments and medication that limit or facilitate clinical decisions. The “surgical volume” of a surgeon

• Statements about factors that relate to individual clinical actors, including surgeon demographics. For example “surgeons sex”.

Clinical demographic and procedural information

Patient demographic information

11© Under License of Creative Commons Attribution 3.0 License

2016Vol. 3 No. 1: 25

HEALTH SYSTEMS AND POLICY RESEARCH

clinicalpracticedecisionsintheincludedstudies.Surgeonswitha greater surgical volume in joint replacementwould bemoreinclinedtoconductajointreplacementcomparedtonon-surgicalmanagement.Patientfactorsincludedage,[38,41,51,56]medicalcondition,[36,45,35,56]sex,[33,56]lifestyle,[51,54,55]treatmentoptions,medicationandsymptoms[39,43,49,51,53,55].Surgeonfactors included age, [39,54] sex, [33] personality type [8]and surgical volume [39]. Some pragmatic factors were alsoimportant,suchastimetakentoperformsurgerywherecertainprocedureswereselectedbecausetheywereshorterthanotheroptions[8,45,46].

What does this mean for orthopaedic practice? We have identified many sources of evidence which competefor space and prominence in the process of decision-making.Thecompetitionmaybesubconsciousasmedicalprofessionalsmay broker various evidence sources and knowledge typeswithincurrentorganisationalcontingencies.Thiselementoftacitpracticecamethroughstronglyinmanyoftheincludedstudies[8,51-58] and has been recognised elsewhere in the medicalliterature[14,63].

It is important that surgeons and orthopaedic departmentsdevelop an awareness of this subjective and subconsciousbrokering process. It is essential in clinical practice to enablesurgeons to be cognizant ofwho andwhat is influencing theirclinical decisions and the contingencies and constraints theyareworkingwithin.Once thedriversofdecisionsareapparenttothedecision-makers, itwillbepossibletoadjust, improveoreliminate factors which can be considered inappropriate andwhich may lead to unfair or inappropriate and unwarrantedvariationinpractice.Acrosstheliteratureeachofthesedrivers,canbeconsideredaviablesourceofevidenceorknowledgeforsurgeons,onaparwith formal codifiedclinical guidelines, andour review demonstrates that they influence clinical practicedecisions[59-62].

Our findings highlight the power of local clinical communities,networks and hierarchies within the orthopaedic profession.Theyresult indecision-makingthat isgrounded in“that’swhatwedoaroundhere”evidence, rather than theevidence-basedmedicinewhich isofproveneffectivenessforpatients.Utilisingthese strong clinical networks for positive knowledge sharingmaybeanavenueforfutureinvestigationparticularlyrelevantforthisclinicalgroup.Wehavefoundfromthisreviewthatalthoughclinicalevidence,suchaspatients’functionalscore,isimportantandnecessaryfordecision-makingtheyarenotsufficienttodriveaclinicaldecision.Othersourcesofevidenceforexample,currentwaiting lists and operating room availability (organisationalknowledge) or the necessity to maintain good relationshipswith colleagues and superiors (socialised knowledge) can takeprecedence.Thedifferentwaysinwhichsourcesofevidenceandknowledgearebrokeredandprivileged inpracticeresult inthevariation in delivery of orthopaedic services and in potentiallyunsatisfactoryoutcomesforpatients.

Conceptual framework of themesWehavenotsetouttopresenttheeightthemeshierarchically,as theways they interact depend on the contingencies of the

currentcontext.Thismaychangeovertime for individualsandorganisations as and when the different evidences come intofocus. Instead we propose a conceptual framework (Figure 4) structured around the wider definition of an evidence-baseddecisiontodescribehowthedifferentevidenceandknowledgesourcescanactasmicro,mesoormacrodriversofvariation inorthopaedicsurgicalwork.

The next step is to test this conceptual framework in practiceand to understand its applicability to modern healthcareorganisations, we need to understand whether there aresituationsinwhichthealignmentandhenceimportanceofthedifferentthemeschange.Eachsourceofevidencehasstrengthsand weaknesses and contributes to variation in orthopaedicpracticedependingonwho ismakingthedecisionand inwhatcircumstance.Decisionsmade from a financial stancemay notleadtoeffectiveoutcomesforapatientontheward,andhenceare reported to carry less weight for the individual surgeon[49,63-65].However,whenexaminingtheleveloforganisationalpractice (clinical commissioning)ornationalpolicy (clinical andcost-effectivenessanalysis)financialdecisionsplayacentralrolein decision-making.What is important here is howevidence iscombinedinthereal lifepragmaticprocessofdecision-making,wheredecisionsarecomplexandchallenging.

ConclusionWesystematicallyreviewedandsummarisedpublishedliteratureon the approaches and techniques towards evidence use, andthe factors that influence to evidence-based decision-makingand practicewithin orthopaedic surgery. Datawere integratedusing the Pillar Integration Process and analysed thematicallytoproduceeightthemeswhichcanbetreatedasthesourcesofevidence that influence thedecisionsoforthopaedic surgeons.These factors may be explicit, or may act outside consciousdecision-making processes. They are brokered directly orindirectlyincompetitionwitheachother.

Inthisreviewwehavefoundthatformalcodifiedknowledge,e.g.,clinicalguidelinesproducedbyNICEandothernationalbodies,appearstoplayasmallpartinorthopaedicdecision-making.Moresignificantdriversofvariationincludethesocialisationofmedicalprofessionals,theculturalandnormativefactorsatplayandtheexistenceofcomplexandcompetingsourcesofknowledge.

Patient candidacy does not in itself account for the reportedvariation in the delivery of orthopaedic services. Individualsurgeons, their clinical communities and wider professionalsocieties coupled with the constraints and contingencies ofhealthcare systems influence and complicate the decision-makingprocess.As a result, the factors feeding intodecisions,contribute to the widespread variability in surgical practice.Services are not as high quality; cost-effective and equitableas they could bewhichmeans that “not everyone who needs oneisgoingtogetone”.Untiltheprocessofdecision-makingisfullyexplored,andstrategiesaredevelopedtomitigateagainstthis incongruous evidence use, interventions, policies andrecommendationsaimedatachievingequitable,clinicalandcost-effectiveorthopaedicdecisionswillbelimitedintheirreachandeffectiveness.

12 This article is available from: http://www.hsprj.com/archive.php

2016Vol. 3 No. 1: 25HEALTH SYSTEMS AND POLICY RESEARCH

Funding Amy Grove is supported by the National Institute for HealthResearch(NIHR)DoctoralResearchFellowshipProgrammegrantnumber DRF-2013-06-064. The views expressed are those ofthe author(s) and not necessarily those of the NHS, the NIHRor the Department of Health. Rebecca Johnson, Aileen ClarkeandGraemeCurrie are supportedby theNational Institute forHealthResearch (NIHR)CollaborationforLeadership inAppliedHealthResearchandCareWestMidlandsatUniversityHospitalsBirminghamNHSFoundationTrust.

Acknowledgements Theauthorswould liketoacknowledgeSamanthaJohnsonandRachelCourt,InformationSpecialistsattheUniversityofWarwickwhoassistedinthedevelopmentofthesearchstrategiesusedinthis review.

Competing Interests All authors declare that we have no competing interests. AllauthorshavecompletedtheICMJEuniformdisclosureformanddeclare: no support from any organisation for the submittedwork;nofinancialrelationshipswithanyorganisationsthatmighthave an interest in the submitted work in the previous threeyears; no other relationships or activities that could appear tohaveinfluencedthesubmittedwork.

Data Sharing Noadditionaldataavailable.

Ethical Approval Noethicalapprovalwassoughtforthissystematicreview.

Conceptualframeworkofthesourcesofevidenceandknowledgethemesidentifiedinthesystematicreview.Figure 4

13© Under License of Creative Commons Attribution 3.0 License

2016Vol. 3 No. 1: 25

HEALTH SYSTEMS AND POLICY RESEARCH

References1 BriggsT(2015)GettingItRightFirstTime(GRIFT).BritishOrthopaedic

Association.

2 JudgeA,WeltonNJ,SandhuJ,Ben-ShlomoY(2010)Equityinaccessto total joint replacement of the hip and knee in England: crosssectionalstudy.BMJ341:c4092.

3 Amick HR, Gartlehner G, Gaynes BN, Forneris C, Asher GN, et al. (2015) Comparative benefits and harms of second generationantidepressants and cognitive behavioral therapies in initialtreatment of major depressive disorder: systematic review andmeta-analysis.BMJ351:h6019.

4 USPreventiveServicesTaskForce(USPSTF)(2015)AbouttheUSPSTF.

5 Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS(1996) Evidencebasedmedicine:what it is andwhat it isn't. BMJ312:71-72.

6 KellyM,MorganA,EllisS,YoungerT,HuntleyJ,etal.(2010)Evidencebased public health: A review of the experience of the NationalInstituteofHealthandClinicalExcellence(NICE)ofdevelopingpublichealthguidanceinEngland.SocSciMed71:1056-1062.

7 GloverJA(1938)TheIncidenceofTonsillectomyinSchoolChildren:(SectionofEpidemiologyandStateMedicine).ProcRSocMed31:1219-1236.

8 FerlieE,WoodM,FitzgeraldL(1999)Somelimitstoevidence-basedmedicine:acasestudyfromelectiveorthopaedics.QualHealthCare8:99-107.

9 FerlieE,FitzgeraldL,WoodM,Hawkins,Cetal.(2005)Thenonspreadof innovations: themediating role of professionals. AcadManag J48:117-134.

10 Wennberg J,Gittelsohn (1973)Smallareavariations inhealthcaredelivery.Science182:1102-1108.

11 WennbergJE(2002)Unwarrantedvariationsinhealthcaredelivery:implicationsforacademicmedicalcentres.BMJ325:961-964.

12 Mulley A, Wennberg J (2011) Reducing unwarranted variation inclinical practice by supporting clinicians and patients in decisionmaking. Chp 3. In Gigerenzer G, Gray MJA eds. Better Doctors,BetterPatients,DetterDecisions:EnvisioiningHelathCare2020.MA:MassachusettsInstituteofTechnology.

13 MoritzS,BatesT,HendersonSM,HumphreysS,MichellMJ(1997)VariationinmanagementofsmallinvasivebreastcancersdetectedonscreeningintheformersoutheastThamesregion:observationalstudy.BMJ315:1266-1272.

14 PopeC.Contingencyineverydaysurgicalwork.SociologyofHealth&Illness2002:24:369-384.

15 BirkmeyerJD,SharpSM,FinlaysonSR,FisherES,WennbergJE(1998)Variationprofilesofcommonsurgicalprocedures.Surgery124:917-923.

16 DunnWR,SchackmanBR,WalshC,LymanS,JonesEC,etal.(2005)Variationinorthopaedicsurgeons'perceptionsabouttheindicationsforrotatorcuffsurgery.JBoneJointSurgAm87:1978-1984.

17 McPherson K (2008) Reprints and reflections: The incidence oftonsillectomyinschoolchildren.IntEpidemiolAss37:9-19.

18 SouthEastPublicHealthObservatory,AtlasofVariation(2010)NHSAtlasofvariation.

19 Appleby J (2010)WhichEnglishhospital isbestathips?TheKingsFund.

20 ApplebyJ,DevlinN(2010)GettingthemostoutofPROMS.Puttinghealth outcomes at the heart of NHS decision-making. The KingsFund.

21 NationalJointRegistry(NJR)11thAnnualReport.

22 HigginsJPT,GreenS(eds).(2001)CochraneHandbookforSystematicReviewsof InterventionsVersion5.1.0 [updatedMarch2011].TheCochraneCollaboration.

23 ChalmersI,GlasziouP(2009)Avoidablewasteintheproductionandreportingofresearchevidence.Lancet374:86-89.

24 BoothA,PapaioannoD,SuttonA(2012)SystematicApproachestoaSuccessfulLiteratureReview.UK:SagePublicationsLimited.

25 GoughD,ThomasJ,OliverS (2012)Clarifyingdifferencesbetweenreviewdesignsandmethods.SystRev1:28.

26 CASP (2015) Critical Appraisal Skills Programme. Critical appraisaltools.

27 Dixon-Woods M, Agarwal S, Young B, Jones D, Sutton A (2004)Integrative approaches to qualitative and quantitative evidence.London:NHSHealthDevelopmentAgency.

28 Johnson J, Grove A, Clarke A (2015) Pillar Integration Process: Atechniquetomixmethodsinthehealthsciences.JMMR

29 BrittenN,CampbellR,PopeC,DonovanJ,MorganM,etal.(2002)Usingmetaethnographytosynthesisequalitativeresearch:aworkedexample.JHealthServResPolicy7:209-215.

30 Campbell R, Pound P, Morgan M, Daker-White G, Britten N, etal. (2011) Evaluating meta-ethnography: systematic analysis andsynthesisofqualitativeresearch.HealthTechnolAssess15:1-164.

31 Moher D, Liberati A, Tetzlaff J, Altman DG (2009) The PRISMAGroupPreferredReportingItemsforSystematicReviewsandMeta-Analyses:ThePRISMAStatement.PLoSMed151:264-269.

32 Barr PJ, DonnellyM, Cardwell C, Alam SS,Morris K, et al. (2011)Driversoftransfusiondecisionmakingandqualityoftheevidenceinorthopedicsurgery:asystematicreviewoftheliterature.TransfusMedRev25:304-316.

33 Bederman SS, Rosen CD, Bhatia NN, Kiester PD, Gupta R (2012)Drivers of surgery for the degenerative hip, knee, and spine: asystematicreview.ClinOrthopRelatRes470:1090-1105.

34 Pratt MG, Rockmann KW, Kaurmann JB (2006) Constructingprofessional identity: the roleofworkand identity learning cyclesin the customization of identity among residents. Academy ofManagementJournal49:235-262.

35 PrattMG(2009)Forthelackofaboilerplate:tipsonwritingup(andreviewing) qualitative research. Academy ofManagement Journal52:865-862.

36 deBoerPG,BuckleyR,SchmidtP,FoxB(2012)Barrierstoorthopaedicpractice--why surgeons do not put into practice what they havelearnedineducationalevents.Injury43:290-294.

37 WattsAC,BrenkelIJ(2005)Apostalsurveyofcurrentthromboprophylacticpractices of consultant orthopaedic surgeons in the treatment offractureofthehip.JBoneJointSurgBr87:296-300.

38 ZielinskiSM,MeeuwisMA,HeetveldMJ,VerhofstadMH,RoukemaGR, et al. (2013)Adherence to a femoral neck fracture treatmentguideline.IntOrthop37:1327-1334.

39 MarxRG,JonesEC,AngelM,WickiewiczTL,WarrenRF(2003)BeliefsandattitudesofmembersoftheAmericanAcademyofOrthopaedic

14 This article is available from: http://www.hsprj.com/archive.php

2016Vol. 3 No. 1: 25HEALTH SYSTEMS AND POLICY RESEARCH

Surgeons regarding the treatment of anterior cruciate ligamentinjury.Arthroscopy19:762-770.

40 KhanH,HussainN,BhandariM(2013)Theinfluenceoflargeclinicaltrials in orthopedic trauma: do they change practice? J OrthopTrauma27:e268-274.

41 Krahn J, Sauerland S, Rixen D, Gregor S, Bouillon B, et al. (2006)Applyingevidence-basedsurgeryindailyclinicalroutine:afeasibilitystudy.ArchOrthopTraumaSurg126:88-92.

42 Kumar M, Gopalakrishna C, Swaminath PV, Mysore SS (2011)Evidence-basedsurgery--evidencefromsurveyandcitationanalysisinorthopaedicsurgery.AnnRCollSurgEngl93:133-138.

43 SchulzAP,JönssonA,KaschR,PritheeJ,BhandariM(2013)Sourcesof informationinfluencingdecision-makinginorthopaedicsurgery-aninternationalonlinesurveyof1147orthopaedicsurgeons.BMCMusculoskeletDisord14:96.

44 TejwaniNC,ImmermanI(2008)Mythsandlegendsinorthopaedicpractice:areweallguilty?ClinOrthopRelatRes466:2861-2872.

45 Canty SJ, Shepard GJ, RyanWG, Banks AJ (2003) Do we practiceevidence based medicine with regard to drain usage in kneearthroplasty?ResultsofaquestionnaireofBASKmembers.Knee10:385-387.

46 HagemanMG,GuittonTG,RingD,ScienceofVariationGroup(2013)Howsurgeonsmakedecisionswhentheevidenceisinconclusive.JHandSurgAm38:1202-1208.

47 KegelG,MarshallA,CatalanoL,BarronOA,GlickelSZ,etal.(2013)Steroid injections in the upper extremity: experienced clinicalopinionversusevidence-basedpractices.Orthopedics36:e1141.

48 Lingard EA, Berven S, Katz JN, Kinemax Outcomes Group (2000)Managementandcareofpatientsundergoingtotalkneearthroplasty:variationsacrossdifferenthealthcaresettings.ArthritisCareRes13:129-136.

49 Okike K, O'Toole RV, Pollak AN, Bishop JA, McAndrew CM, et al.(2014)Surveyfindsfeworthopedicsurgeonsknowthecostsofthedevicestheyimplant.HealthAff(Millwood)33:103-109.

50 Wright JG,HawkerGA,BombardierC,CroxfordR,DittusRS,et al.(1999)Physicianenthusiasmasanexplanationforareavariationintheutilizationofkneereplacementsurgery.MedCare37:946-956.

51 HudakPL,GrassauP,GlazierRH,HawkerG,KrederH,etal.(2008)"Noteveryonewhoneedsoneisgoingtogetone'':theinfluenceofmedicalbrokeringonpatientcandidacyfortotal jointarthroplasty.MedDecisMaking28:773-780.

52 VashitzG,PliskinJS,ParmetY,KosashviliY,IferganeG,etal.(2012)Dofirstopinionsaffectsecondopinions?JGenInternMed27:1265-1271.

53 Gooberman-Hill R, Sansom A, Sanders CM, Dieppe PA, HorwoodJ, et al. (2010)Unstated factors inorthopaedicdecision-making: aqualitativestudy.BMCMusculoskeletDisord11:213.

54 Irwin ZN, Hilibrand A, Gustavel M, McLain R, Shaffer W, et al.(2005)Variationinsurgicaldecisionmakingfordegenerativespinaldisorders.PartI:lumbarspine.Spine30:2208-2213.

55 DusikCJ,BuckleyRE,Robertson-MoreC(2013)Orthopedicsurgeonperspectivesonappropriatereferraloftraumapatientstophysicaltherapy(PT).ArchOrthopTraumaSurg133:603-608.

56 Bhandari M, Devereaux PJ, Swiontkowski MF, Schemitsch EH,Shankardass K, et al. (2003) A randomized trial of opinion leaderendorsementinasurveyoforthopaedicsurgeons:effectonprimaryresponserates.IntJEpidemiol32:634-636.

57 CurtisAJ,WolfeR,RussellCO,ElliottBG,HartJA,etal.(2011)Determiningpriority for joint replacement: comparing the views of orthopaedicsurgeonsandotherprofessionals.MedJAust195:699-702.

58 BorkhoffCM,HawkerGA,KrederHJ,GlazierRH,MahomedNN,etal.(2008)Theeffectofpatients' sexonphysicians' recommendationsfortotalkneearthroplasty.CMAJ178:681-687.

59 TsoukasH,VladimirouE(2005)‘Whatisorganizationalknowledge?.ManagingKnowledge:AnEssentialReaderpp:85.

60 CurrieG,SuhomlinovaO(2006)TheimpactofinstitutionalforcesuponknowledgesharingintheUKNHS:thetriumphofprofessionalpowerandtheinconsistencyofpolicy.PublicAdministration84:1-30.

61 RobertsA(2015)VeryShort,FairlyInterestingandReasonablyCheapBookAboutKnowledgeManagement.UK:SagePublications.

62 GroveA,ClarkeA,CurrieG(2015)Thebarriersandfacilitatorstotheimplementationofclinicalguidanceinelectiveorthopaedicsurgery:aqualitativestudyprotocol.ImplementSci10:81.

63 Lave J, Wenger E (1991) Situated learning: Legitimate peripheralparticipation.CambrigeCambridgeUniversityPress.

64 Polanyi M (2012) Personal knowledge: Towards a post-criticalphilosophy.Chicago:UniversityofChicagoPress.

65 GabbayJ,leMayA(2004)Evidencebasedguidelinesorcollectivelyconstructed "mindlines?" Ethnographic study of knowledgemanagementinprimarycare.BMJ329:1013.