TilEATTRACTIVENESSOF RETURN-FaR-SERVICEBURSARY
PROGRAMSTO MEDICA LSTUDENTSIN NEWFOUNDLA NDANDLABRADOR
by
©Shelley-MayGreenaway
Athesis submitted tothe School of GraduateStudies in partialfulfillmen toft he requirementsfixthedegree of
Master ofScience
DivisionofCommunity Health and Humanities(Medicine) Memorial UniversityofNewfoundland
May 20ll
St.John's Newfoundland
Abstrac t
Objcctiv cs and I\lctbods :Tbis study comparedRcturn -f or-scrvicc(RFS)programs availablcfro m prov incial/te rritorialgovcm ments,de te nn ined tennsof inte rcstand predict orsofacceptanceinNe wfound landand Labrador(NL),anddeserib ed experienees ofRFS-holders.Researchmeth odsinclud eddocum entanalysis,anonlinesurvey,and telephoneinter views.
Results : RFSprogramswere apopul armeansof improvin gphysiciandistribution.
Stude ntsrated mone ta ry valuc(37.3%)a ndloeatio no fscrviecreturn (34.9%) as thcmostimportantfeaturesin theirdecisionto acceptan RI'S.Trainees withfi na ncial ccncc rnsund thosc whoplunncd toremain inNLwere 4.8and27.7timesmorelikelyto acceptabursary .
Expe rieneeso f RI'S -ho lde rswcrc positive;eo1l11l1unieation d iftie u ltiesandalaek ofa ct iveree ruitmc ntwc rci dcntilie d asproblcms
Concl usio ns:TheRI'S shows so1l1epromiscforincrcasingphysicianrccruit rncnt, howeveri t doesnot appea rtohethemosteffectivemeans;morebursaricsfund trainccs who plantorem aininNLalrea dy(XOtY,,)than atlractsnoveltrai nccs( 20 %).
Acknowledge me nts
Itis withsince reappreciatio n thatI thankmysuperviso r,Dr. Maria Math ews, forhcr cxccll cnt gu ida nce,paticncc an dsupportthroug ho utmyMastcrsprogram.Ma ny thank s arcalsoncccssary tomy comm ittccmcmbcrs,Dr. Jam csValcour and Dr.Victor Madd alcnaforpro vidin gthci r cvcr-insight fulp crspcctivc and ad vicc.
Iwouldlikctothankthc N LDcpartm cnt o fll calth andCommunityScrvices,
and Memori al Un iversity 'sStudentAffairsand PostGrad ua te Medi cal Educatio nOffice whosc coopcra tio n andassistance mad c thiswork possible,In pa rticularl would likct o acknow lcdgc thceffortsandcollaborat io nofDr.LarryAlt ccn,Dan Fitzgcra ld,Jo hn- Pa ul Gordo n,Sca rlc tt Harm,and Dr.AsokaSama rascna
Iwould likct othankmyfriends,collcagucsandcoworkcrsatMcmorial Unive rsity tillthcir comp anionship and laughtcr duringthcpasttwoycars.Parti c ularl yI would likcto cxprcssappreciation to Crysta l Wang, Jessica Zhao,KennyIlamm ond and NoniC howdryfilrhcl pingto kccp lifc in pc rspcc tivcthro ug ho ut thisproccss.TolllY wondcrfulific ndsti'olll ho mcw ho providc dsilllilarcnco uragclllc nt,butarctoo n umcro us tona mc,l sinccrcly than kyou.
Iwouidl ikctothankm ypar cn ts, wh ot aught lllc cducati on wasal ifcl ong journ cy.ToIllytamily.Jcxtcndmya ffccti on andgratitud cfortheirencouragement,lovc andsupport. Finally, toJames,forhis cons tantpaticnccand undc rstanding,and with o ut who m nothing wouldbcpossiblc.
Table ofContents
Abstrac t Acknowledgements Table of Contents Listof Tables Listof Appendices ListofAbbreviations
I.Introduction 1.1ResearehProblem 1.2ResearehQuestions 1.3Purpose andObjec tives
1.3.1Hypoth esis 1.4StudyRationale
2.Buckgruund and LiteratureReview 7
2.1DefinitionsandTcrmsofRFS 7
2.2Dcscription of CanadianRFSPrograms 9
2.3RFSE ITcctivcncss II
2.3.1PhysicianRccruitmcnt andScrvicc Commitmcnt II Fulfillment
2.3.2PhysicianRetenti on 13
2.3.3PhysicianPost-RFSProgramPer cept ions 15
2.4Summary 16
III 18 18 18 19 19 19 20 21 21 23 23 24 25 26 27
3.3.2Samp lc 27
3.3.3 DataCollcctionlnstrumcnts 28
3.3.4 Data Preparation and Analysis 28
3.3.5EthicalConsidcrations 29
3.4 KnowlcdgcTransfcr 30
4.Results 31
4.1Docum entAnalysisResult s 31
4.2Medi calTrainee SurveyResults 39
4.2.1SamplcCharactcristics andRcprcscntativcncss 39
4.2.2SamplcDcmographics 41
4.2.3 CharacteristicsofTraineesWho Do and Do Not Wantan 46 RFSBursary
4.2.4 SlIrvcy Comm cn ts 51
4.3lntcrvicwRcsults 52
4.3.1 IntcrvicwSamplcCharactcristics 52
4.3.2ImprcssionofRFSBllrsary Elcmcnts 52
4.3.3RFS Expcricnccs 53
4.3.3.1Hcarin gabout the Rf'S 53
4.3.3.2Th cBursaryApplicationl'roccss 54
4.3.3.3 Service Return Experienc es 55
4.3.4RFSasaPart of PhysicianRecruitm ent 56
4.3.5 SlImmary 59
5.Discu ssi on 60
5.1RFS Tcrms 60
5.2RFSAttractivcncsstoNL Trainccs 61
5.3ln tlucntialRFST cnns 62
5.4 Promotion and Awaren essof RFS Program 63
5.5RFSProgram Evaluatio n 64
5.6Strcngthsand Limitatio ns 66
8.Appendic es
Listof Tables
Year of Originof Cana d ianRFSProgramsbyPro vince and Territory
Co mpa risonofProvin cialandTerritorialRFSFinancial Ince ntivePayment Details
Ca na d ia nRFSProgra m Availability byPro vince and Ter rito ry
Com pa risonofCa nad ia n RFS Valuesand Retu rn Requir ement sAccordingtoProvinceand Recip ien t Type Com parisono fCa nad ia n RFSProgram Lea ve,Dcfcrr al, and Pcnalt yt()r Non-fu ltillmcnt byProvincc and T erritory CanadianProvincialandTerritorialRFS Availabilityand Fulti llmcntRate
Dcm ograph ic Charactc risticsoft heR FS SlIrvcySample (n=228)
Fina nc ia ICharacter isticsofthc RFSSu rvcy Sam plc(n=22 7) 42 Future PracticePlansofthcRFS SurvcySam plc (n=223) RFSSurveySamplcKnowlcdge,Prefcrcncc,and Perception ofthcNL RI'SProgram(AllRcspondcntsn=222) Imp0l1anccof VariablcsonChoiccto AcccptNL RI'S Bursaryto SurveySampIc(AllRespondcnt sn=2( 9) DijTc rcn ccsinCharactc risticso f Trainccsw hollold/Planto Accept anRFS(n=69)andThose Who DoNotHold/Planto 47 Ap plyFora n RFS(n=79 )
Logi sticRegressionAnalysisPredictingWhc thcr RespondcntsWouldAcceptan RFS8ursary(n=148)
CodingScheme forCanad ian RFS ProgramsAnalysis Coding Schemefor RFSProgramsAnalysis Chi-Squa re TestComparing Sexin Sample and Actual Studen tPopulation
Chi-SquareTestCompati ng Ycarof Study in Sampleand ActualStudentPopulation
Chi-SquareTest ComparingHo rn eProvince of Sampleand EntireStudentPopulation
Chi-S quareTestComparingSex in Sampleand Actual Resid entPop ulation
Chi-Sq uare TestComparingYearofStudy ofSa m picand Entire ResidentPopulation
Chi-Squa re TcstComparingIlo mcProvinceOfSam ple And Entire ResidentPopulation
TableJI: Impo rtance of VariablesonChoice toAcce ptNL RFS BursarytoSu rvey Stud cnt Sam ple (n=135) TableJ2:Impo rtance of VariablesonChoice toAcceptNL RFSBursaryto SurveyR esidcntSample(n=74) DitferenccsinCharacteristicsofStude ntswhoHoldor Planto Applyfor an RFS(n=47) and ThoseWhoDoNot Hol d or Planto ApplyFo r anRFS(n=40)
DifferencesinCharacteristicsof Resident swhoHo ldorPlan toAcceptan RFS(n=32)and ThoscWho Do NotIioi d or PlantoApplyForanRFS(n=29)
LogisticReg ressionAnalysisPredi ctingWhetherStudents WouldAcccptanR FSBu rsary(n=83) Table L2 :Log istic Regressio nA na lysisPredic tingW hc thcr Resid ent sWouldAcceptanRFSl3ursary(n=59)
Listof Appendices
Appendi xA AppendixB
Appendix C Appendix D Appendix E AppendixF AppcndixG AppendixH Append ixI
AppendixJ
AppendixK
AppendixL
AppendixM
DocumentAnal ysis Collection Tool
CodingScheme forCana dian RFSProgram sDocument Analysis
SurveyCom m unica tio n Plan Surv ey Questionnaire Cod ing Seheme fo r R FS BursaryS urvey Qualitati veInterv ie w Invit ati on QualitativeIntervie wGuide Qualit ativeInter vie w Coding Templat e Chi-S q uar e Re sult s Sho w ingSam ple Repr esentati veness
Ill1poJ1ance o fY ar iahlesonthe Ch oieeto Ae cept an NL R FS Bursa ryto SurveySam p le by Stude ntand Re siden t Status
DifferencesinCh a ract c ristic s of StudcntsWhoHoldor Plan toApplytor anRFShy Stud cntandRcsid cnt Status
LogisticRegr c ssion Analys is Prcdict ing R FS Acc cp tance hy Stude ntand Resident Status
Trainee SurveyCom me nts
Listof Abbreviation s
CaRMS-Canadia n ResidentMatchin g Servic e CIHI-CanadianInstitutefor HealthInform ation FM-FamilyMe dic in e
HIC-HumanInvestigationCommittee
MBa-ManitobaAboriginal-Specific MUN-MemorialUn iversityof Newfoundland
NLCAHR-Newfoundland andLabradorCentrefor Applie dIlea lth Research
NIISC-National llealthService Corps
NRFMS-Northe rn RemoteFamilyMedicineResidency Stream
QUAI'-OntarioUndcrsc rviccd AreaProgram
PG-Postgraduate
PGM E-Postgraduate Med ical Education PRcs-Psych iatry resid ent QC-Qucbce
RNI-Ruralorthcm lni tia tivc RPAP-RuralPhysician Action Plan
SMA-Saskatchewan MedicalAssociation SPSS-Statistical Packageforthc SocialSciences TF-T ravcllingFcllowship Uof A-Unive rsity of Alberta Uof'C-UniversityofCalgary UG-Undcrgraduate US-UnitcdStatcs(ofAmc riea) WHO-Wo rldHealth Organization
Chapter I:Introduction
WithCan ad a' s"ph ysicianshortag e"beingthcfo cusofmuch policyand m ed ia attention,it isclearphysiciansupp ly isatop ico fimp oJ1an ceto Canadians.W h ile physiei annumbersmaybe increasing(CanadianInstituteforHealthInfonnation[CIHI], Z008),the d istrib utio nof physiciansalsoplaysan import ant rolein theelTective ness of healthcar edc1iver y . Canadai saspaciouscountr y,with vari abl epopulati ondistribution;
itsmanyrural,rem ot e,and north ernpopul ationsmake it difficultto cnsureeq ui ta ble
AccordingtotheZ007 Cana d ia nCommunity Ilealth Survey,abo ut4.1 million Canad ia nsaged IZor older(15%), rep 0J1edthatth eydidnothave a regularmedi eaI doctor,either becausetheywere unabl e tofind oneorbecausetheyhad no t looked (Statisties Canada,ZO(8).Sixpercentof thepopul ati on aged IZorolderrepor ted they could notfindaregulardoctor inZ007(S tatisticsCanada,Z008).Ruralor remoteareas may sufTerdisprop ort ionatclycomp ar edtomoreurban areas;asofZ004,"o n ly 9.4%o f all physicianswere locatedinruralareas,comparedwithZI.I%ofCanadians" (Pong and Pitblado, Z008p.16 )
NL isac o mpa rativclys pa rsc1y populatedprov ince;thepopulation densityo fN L wasestima tedinZ009tobe1.5pcrsons/knr',while thenationalpopulation density was 3.68persons/km2(StatisticsCa nada,Z009 ).This posesapartic ular challengewith regard tophysieia n d istrib ution .l\largeportio noft he provinee· spopulationis consider ed tobe rural(61.9'%), yeto nly3Z%ofN L physiciansprac ticeinruralarcas(Reamy,1994 ).
Barer,Woodand Schnc idcrrcportthatthcrc arc"v. .many arc asofIhc provincewithout
adeq uate.o rany,local spccialislscrviccs"( 1999 p.108).
Prcvious studi csfoun d NL tohave oncof thc highestprop orti ons ofindividu al s
witho uta regular physician (Talbotctal.,200I).however, more recen tresearchfind sNL tohavc a ratio of ind ividu als without a tilln ily physician to bc sim ilart o lhcn ation al
avcragc(StatisticsCanada.2009b).Whilc phys ician numbcrshavc bccnincrcasingi nthc pro vince(CIHI. 2008).thcrc remains apro blemwithphysiciandistributi on.Mathew s and Ed wa rds(2004 )foundrcsidcnt s ofru ral communiticswcrclcss Iikciyt oh avc a rcgul ar doctor thanresidentsofurban orsemi-urbancom munitics; 74 .4%0fstudy rcspondcnts withoutarcgular phys icianwcrc individualsrcsiding inaruralcommunity.
Thcprovincchascstablishcda sct o fini tiativcsto atlcmpttosolvcitsphys ici an distributionproblcms.Tn 1992.thcNLDcpartmc ntofllcalthandCommunityServices' establishedabursaryprogram(alsokno wn asarcturn-for-serviccprogram. orRFS).
dcsigncdlopaYllnivcrsitymcdicalt rainccsbllrsaricsinrctum for lhcir col11 milmcnlto pracli ccin anllndcrscrvcd arca (Rcamy.1 994).Stlldcntsandm cdi calrcsidcnts who accept these bursaries sign a contractwiththc MinistcrofllcalthandCommunity Servicesagreeing to work onereturn-o f...service year in an area dcsignatedastiu necd for cachlimdcdycar( l'raclicc NL.2010).Dcspitc thc ta ct that RFSprogramshavc bccn llscd by manyprovincesand countricstxem po ws ki. 2004).fcwhavcbccncvaluatcdlor thcir cflcctivcncss.orhavcbccnwclldocumcntcd(GroblcrctaI..2009 ;Scmpoll'ski .2(04) .
IPrio rto1998. the Newfo u nd landandLah radorDep artmentof Healt h becam ethe Departm ent01 Heal th andCommu nity Serv ices (Newfoundlandand LabradorHealthBoa rdAssoci ation [NI.IIBAI.n.d.).
1.2Resear chQuestion s
Th is study willana lyzetheRFSbursaryprog ram formedicaltraineesinNL.The studyaddressedthe followingquestions:
I.What are thereturn-lo r-se rvice progra msacrossCanada?
2.WhatRFSprogr amte rmsareimportant to NL trai nees?
3.Whatarc thecxperiencesofNLRFSbursa ryholders?
1.3PurpnscandOhj cctiv cs
Thepurposeof th is studywastodescribeRFSprogramsand tcnnsinCanada,and toassessthcimpactnfR FSprogramsonrccruitm entbyunderstanding the program elem entsthat arcilllp ort antto pot entialand pastRFShursaryholders.Thestudywillalso idc ntify thecharacteristicsof trai nceswhooptfor an RFSbursarytoundc rstand who th c progra mattracts,andwhether theprogramattrac tsmedi caltraineeswhowouldnot othe rw ise work inthe province.
Thcrcsca rchobjcctivesarcdctin cd asfollows :
I.Documentthe terms and conditionsofR FSbursaryprogramsoffere d to Illcdicaltrainccsbyprovincialgovcrnlll cn tsof Canada.
2.Assessthcpro por tio nof mcdical tra inccsinNLwho plantoacccpt RI'S
3.Dcscribc tcnns of RI'S agrcementsofintcrcst to mcdica ltrainccsinNL.
r.;Ullypolh csis
The study hypoth esis isthatthe NLRFSprogram is rcwardingindiv idual swho already plannc d towo rk in thc provinec,rathcr than attrac ting prcviously un intc rcs tcd physician st o workin NL.
1.4S Iudy Rationaic
TheCana da HealthActstat es theprimary ohj ec tiveof Canadian hcalth carcpolicy isto"protect,prom ote and restor ethcphysicaland ment al well-beingof res ide ntso f Canadaa nd to tilcilitatc reaso nab icacccssto hca lthscrvieeswitho ut finaneia lo ro ther barriers"(Go vernm ent ofCanada, 19 85p.5).Th cuncqualdistribut ion ofhealth profession alsinN Lp osesa suh stanti alobstaclct othc goal of aehi eving opt im alh ealth
Both theshortage and distributi on ofhcalthprofcssion alsaffcc tsthchcalth of Canadians.l la ving a rcgularfamilyphysicianisstro ng lycorrelated tobett crhcalth outcomes (Ch ao,1988;Dietrich ,19 82;Sudhakar- Krishnanand Rudo\l: 2( 0 7). l fthcRFS bursaryprogramisimprovingphysician recruitme ntwitho ut improving physici an retention,physiciantul11ovcr willcontinu c.l'hysicianturtl ovcrisexpen siveand disruptiveto the popul ationsthey se rv e,andcontinuityofcare is stronglyassociatcdwith patientsatisfa ctio nwith thcirhcalt hca rc(Fan.ct al.,2( 0 5).W ithcontinuityo f carc, physiciansar cnotonlyahlet oha vcani m provcd relat ionshipwithpaticntsbutth cyhave also bcc nshow n towork morcc l'ICctivclyand havc im provcdcli nica lou tcomcs (Sudhakar-Krishnanand Rudo\l:2( 07 ).
Bursnr icstakc manyformsandrequirediffe renttermsor conditions.hutin gcncral,th roug h RFS agrccmcnts " m cdicals tudentsand rcsidc ntsreccivcgra nts,!oansor
bursarie sin exchan geforagreeing tolocateinadesignated geog raphic areafora spec ifiedperiod upo n co mpletion of thcir trai ning"(l3arera ndS toddart,1999p.19 ).RFS progr am smakeupalarg eportionofaprovin ce 'srecruitmentand retcntioneffort,and miliionsofdoliarsfundtraineesaerossCanada.lnthe2009- 2010 ye ar , 88bursarieswere distribu tedinNL alone; exciuding the travelling fcllowships.jhisamounts toaspend ing of$2 ,125,000 on as ingle prog ram inasing leyear (personalcom m unication,J. P.
Gordan,May 20JO).Thisis a considera bleamountof moneyto beinvested into a program that,asofye t,rem ainsformall yune valuated (perso na l communieation,J.P.
Gordan,May 2010 ).
Thisstudyaddressesacriticalgap in theliteratureregard ingRFSbursariesand theireffcetonphysi ci and istr ibulion .Itwill provideprogramplannerswith eviden ceto improvetheRFSprogram ,andcomplem en tsastudy currentlyunderwa yinNLthat examin esthe impact of the RFSprogr amon physician retentio n (Newfoundland and Labrador Centre fo rAppliedHealthResear ch [NLCA HR], 20 10 ).Usingadministra tive dat a from themedi calreg istrar,theotherstudy willcompare theIcngthoftime physic ians with and withoutRFSbursaries work inNL(pe rso nal com m un ication,M.Math ews.
October20 10).Togeth er,thesetwo studieswillprovide amo re com plete understandin g of theroletheRFSprog ramplaysin rccruiting andretainingphysicia nsinN L.
Ifthe NLR FSprog ramissuccessfu lat rccr uitmenthutnotrctention,physician turnover willcontin ue,and theprovincewillcontinue to relyon thisphysician recruitmentprogram.If we can discernwhic h individualsarc takingadvantageof RFS programsand theirreasonsfordoingso,wc willhe able to explore the attractivenessof thebursaryprogramto potc ntialsubscribe rs .The resultsof thestudywillprovide
progr amdesignerswithvalua bleinformation toim pro vetheprogramand ultimatcly, im prove thedistribution ofphysiciansinNL.
Chapter 2: Back ground and Lite ratureReview
Rcturn -f o r-scrvicc(R FS) a gr ccmcnt sm ay al tcrn at ivcl ybc rcfc rrcd to asrcturn - of-scrvicc ( Ba rn ig ha usena nd Bloom,2009;Sempowski2004)orrcturn-in-servi c e agrccmc nts ( Bar cr,WoodandSc hnci dc r,1999), support - to r-scr vicc(!'athm an cl al., 2004),or Joanfor gi vcn csstl'athm an ctaI.,2000a).Theseagrcc mcntsa llscck to im provc physiciandistributi onbyprovidi ngphy sici an s(orphysici an s in trai ning)wi tha ti na ncia l inccnti vctopracti ccin adesignat cda rc a.Accordin gto Scm po wski (2004 p.83),thc goals ofRFS prog ra msarcto"provid eshort-term ruralphysici anmanpower (recruitme nt),"
and to"retain physici anspast thcirma nd atcdtcnn (rc tc otio n)."
Rcci pic ntsofthcsc agr ccmcnt sm ayhct argct cd asundcrgradu atc orpostgr adu atc students,orwork ingphysicians(Bar c r andStoddart,1999;Jac kso n ctaI., 2003;!'athm an
ctaI.,2000a:!'athm an ctaI.,2(04 ). RFSagrccmcntsprovid c diffc rcnt typcsofmon ct ary inccntivc s,usuall ydcpcnding on th c car ccrstagc ofthc rccipi cn t.F o rth c ir commitmcnt to workin an und crscrvcd arca,m cdi calsludcntsmayhc aw ard cdh ursarics, schol arsh ips, orgrantsfor thedurati on ofthc ired uca tio n/residency(Bassand Cop cman,1975: Fitzct al.,1977 ;Mason,1971).Altcrna tivcly ,rccc ntgrad ua tcsmay agrccto recei vetuit ion reimbu rsement , or havetheirexisti ng student loansfor givenafter complet ingate rm of scrvicc (!'athman ctal., 1994; Pathmanctal.,2000h;Ros c nh latt ctal.,1996: Wilso n ct
al.,1998).Sig ninghonuscspro vidc physicians with unrcstrictcdfundsuponscu ing up practicein an und crser vcd arcainexcha ng eforthcir cornmi tm cn ttorcm ai ninpra ct ice thcrc f(lra t lcastoncycar(13arc r,WoodandSchncidcr.1999;!'at hmanctaI..2( 04 ).
Eachyca ro ffundi nggcncrallyrcq uircsoncycarofrct umc dscrvicc inan underscrved arca.Th c dcfi nition ofu ndcrscrvcdvaricsaccordin gtoprogram,and cl igibl c undcrscrved arcasmayrange from a wide definitio n,e.g.an entirestatc(Navinand Nichols,1977) or a "smalltown" or"ruralarca"(Rabinowitzcta l.,2005),toavery narrow dcfinition,c.g.undc rscrvcdarcasdc tincd throughcommi ttccS,consideringf 'lctors beyondpopulationsize ,suchasnumbe randcompositionof physiciansupply, socioeconomics, demographics,populatio n nccdsand dcmand(l3assand Copeman,1975;
l'athma nctal.,1992;Roscnblatte ta l.,1996).
ManyR FSprogr amsincludcrbuy-outoptions,"throug hwhic hphysiciansmay rcpaythcirbursary/loan/grant inli cu of lilltill ing thcirscrvicc commitment.Thisoption maysimply requirerepaymen t (Navin andNichols,1977) or mayimposc additional intcrcst chargcsaswcll (Cop cman,1979;Jacksonct al.,2003;Mat sumoto,lnouc,and Kaji i, 2008b).
RFS agrccmcntsarc notuniquct o physicians,nor arc thcyuniqucto onc co untry . RFSprogramsarcavailablcfor manyot hcrhca lthprofcssionsand havcbccndcscribcd in thcl itcraturc for dcntists,pharmacists,nurses,nursc praetition crs, physieian assistantsand midwi vcs (Bradbury,1963; l'olitzcr ct al.,2000).WhilcUSR FSprogram s arcmost frcquc ntlydiseussedin thc literature,researchonRFSpro gram sfi 'omCanada(A nderson andRosenberg,1990;l3assandCopeman,1975;Copeman,1979;Wilson ctaI., 1998), Japan (Matsumoto,ct al.,2008a;Matsumoto,et al.,2008b;Matsumotoet al.,2010), SouthAfrica (RossandCouper,2( 04),andAustralia (Dunbabin,MeEwinandCamero n, 2( 06) havcalsobcen publishcd.
2,2 DescriptionofCa nad ia n RFSProgr am s
RFSagreements are extensivelyused acrossCanada; Alberta(AB),Quebec(QC), Manitoba(MB),Ontario(ON),New foundlandand Labrador (NL),theNo rthwes t Territories(NT), and Saskatchewan(SK) have all usedundergraduate or postgraduate studentloansfbursarieswithretumofserviee as a strategy to improv ephysici an distribution(Barer,Wood, Schneider, I(99).Whilesome RFS program shaveheenin place sineelhe60's,Barer,Woo dandSchneider' sI999Iileralurereview ,'Toward Impro ved Aeeess10MedicalServicesfor Relatively UnderservedPo pulations:Canadian Approac hes,Foreign Lesso ns",is thcfirstandonlypublicationthat hasattemp ted10 identify anddescribeCanadian RFSp rograms.
QC,ON,NL,andSK RFSprogramsutilizebursariesastheirfinancial incentive.
Establishcd in1992, NLh asmadc availahlcup lo$25.000 hursaricsl()r mcdicalstudcnls inthcir"latcryearsofresidcncy"(Barcr,WoodandSchne idcr,19(9).AccordingtoBarer et al.(1999). sincethe1990·sRFS hursaricsworth$IS.OOOperycararc availahlclo SK mcdicalsludcnlsinlhesccond .lhirdor fourth ycaroflhcirundcrgradual c study,aswell asrcs idcnts.Studcnlslrcs idcnts who bccomcg cnc ralpractitioncrs musl practicc ina rural areaof lcss thanIO.OOOpcople,andspecialists maypractice anywhere in the province except Saskat oonand Regin a(Barer,Wood andSchnei de r,1( 99).
QC sRFSprogram .initiatcd inl 97S.pro vidc slhird andfourth ycarmcdical sludcnls w ilh hursaricsof$ IO.OOOpcryca r.a nd t:lln ily mcdicinc rcsidcn lsarc eligihlc t(lI ana dd itiona l twoyea rsof funding.ln addition.rcgiona l hcalth boards(asopposcd lo lhe province)provideresidcnts slud yin gdcsignalcdspecialtieswith bursar iesbctwccn
SIO,OOOandS25,OOO peryearforthelasttwo yearsoftheir residency (Barer,Wood and Schneider,199 9).
From 1969to1996,the OntarioUnderscrvieedAreaProgram(OUAP) administeredabursaryofS7,500 tostude nls in their thirdandfourthyea rof me d iea l schoo lin exchan gefortheircomm itme ntto workoneyear inadesignatednorth ern communityperyear offunding.Barcrcta1.(1999)repo rt thatthisprogrumisnolon gcrin
QC, ON, NL,and SK RFSprogram s allrequireoneyearofserviceinexcha nge foreachyearoffund ingreceived .Betwe en1978 and1985returnedservicewas mandatoryinQC; if aphysiciandid not fulli ll his or her commitment,the physicia n would not rece ive abi llingnumbe r. l lo we ver, sinee1985,arepayment optio nhasbeen available fo r allprogr ams.Inallpro v inces,physicianswhotailto retum thei r com m itted servicearc req uire dtopay theirbursa ries back withadd itio na Iinterest(Barer,Wood and Schneider, 1999).
MBandAB"sRFSprogram sbothutilizedloanforgiveness/remission. In MB, Sl5,OOOloansareavailable tostude ntsintheir thirda ndfourth undergradu ateyea r. These loanswouldbe lorgi veno na year-for-y earbasispostgradua tion( Barer,Wood and Schneider,1999 ).Alberta'sRuralPhysician Action Plan(RPAI'),establishedint985, includco a loan rcmissionp rogram,wh crcinn cwph ysicianscouldagree to practice one year inadesignatedarea (physician:populationratioat least0fl:I,OOO)inretu rn l(lra remissionofSIO,OOOofexi sling studenlloans.Ifphysieianspraeticcdinanunderser ved area fortwoyea rs,they wouldreceiveanothcr SIO,OOO atierco mplotionof thesecond year (Barer ,Wood andSchne ider,199 9 ). TheA B program appea rstohave been
ineffe cti ve,however, andwasdiscontinu ed in 1998duet olow particip ation(Wilson et aI.,1( 98),later replaced witha signing bo nuscou pledwith rcturn -of-scrvicetlsarer, Wood,andSchneider 1( 99) .
TheNT docsnot appea r tohave a formal RFSprogram;rather,StantonRegional Hospitalwasemployingfunding fromtheTerritorialDepartment ofHealth andSocial Services topaysalariesforresidentswhoprovidedreturn-in-sorviee (l3arer,Wood and Schneide r, 1( 99).
Severalautho rsha velamented the lack of literature add ressingthe effectiveness of progra msaimed toincr easephysician recruitmentand retention(l3areran d Wood, 1999;l3arer,Wood andSchneider,1999;Curranetal.,2007 ;Gro blcrct aI.,2009;
Sempowski,2004;Simoc ns,2004;Wilso n et aI.,2009;WorldIlealth Organizations [WHO],2009). Despitethe widespread usc of RFSstrategies,l3arer andWoodreportthat
"there ha[s]not been anyformalevaluationof thestudent loan programswhiehre lyupon return-o f-serviceguarantees tosecureserviceprovisionsforunderservieed areas"(1997 p.6).
2.3. 1 I'hys icia n Rccrui tmcul andScr viec Commi tm cn tFulfillment
There islimitedpubl ished work,evaluating CanadianRFS programs.Two ON articles,the mostrecentbein g from 1979,describethe province' sundergraduatebursar y
prugram fCopcman, 1979;BassandCope man, 1(75).Thisdatedanalysisofthe RFS program found that 50%ofstudents hadhon ou red theircommitment(as opposedto repaidtheirbursary),andtwo thirdsofthesephysicians stayedint hc ircom munityoncc
they hadfin ished theirserviee (Copeman,1979).Lessthan10%of female RFSbursary recip ientscompleted theirservice (Cop eman,1979 ).
Barer,Wood andSchne ider(1999)describe the defaultratesin bothMBandQC.
At the limeof publication,MB had 102 students accept bursaries,of which 51 (50%) completed theirRFS serviceor werein the processof doingso,29(28.4%)students were in training ,and22(21.6%)had repa id their bursa ry.QCreportedsimilardc faultratcs.
Since thein trod uetionofthe huy-outoptioninl985 ahout50students acceptedbursaries eachyea r,withprogramplannersexpeeting only50%of these studentstoreturnserviee.
Ofthe QC studentswhobegantheirreturnofserv ice,50% generallyrepaidthe remainder of the ir bursari esafleronc yearofservice (Barer, WoodandSchneide r,1999). This eonccrnthatal argeproporti on ofphysician saret ak ingad vantage of available buy-out
options rathcrthancompl eting scrviee hasbeenexpre ssedinlit eratur edcscribing UxRf'S programs as well(Mason,1971;Scmpowski,2004; Simocns, 2004;Strosbcrg,Mullan andWins he rg, 1982).
Theutilization ofRf'Sprogramsin theUS issubstantial;according to asurvey carriedoutbyPathman ct al. (2000ap.265),24.6%o f asamplc of 468practicing family physieians fllndedall orpart ofth cirt rainingwithsllpp ortfrom"fcd eral, state,or commllnity-sponsorcdseholarship.loanrepaymentandsimilar program swithservice obligatio ns."Thisuptakesugge ststhcrcisa stro ngde mandfor RFSprograms.
Evaluationsof USRFSprogramshavebeenmorethoroughl yreport edinthe litcraturc.undingencral reporthigherservice comp let ionratesthanexistingCanadian data.Maso n's1971evaluationofRFSprogramsrepo rtcd60%ofphysieianseompl cted thcirRFSse rvicec ommitment (38%rcpay.2%dcfillllt).Al atcrstud y byPatlun anctal.
(20 00b)revie wed USsupport-for-serviceprograms again,andfoundasimilarrat e o f completion fo runderg rad uatestudentsc holarship progr ams(66.5%). lnterestingly,thc autho rsfoundprogramsthatcomm itphysiciansin resid en cy(or later)hadasignificantly high erc omplct ion rate (92%).
Whiletheuptak eand compl ctionratcsofR FSprogramsmaybchigh,itisalso important to conside rwhether thescprogramsattractphysic iansto underserved a re asor if physicianswho intendtoworkinundcrser ved arcasare the oneswho accept bursaries.A recentsurvey -basedstudyof three Colora do healthcarepro vide rIcan-repayme n t programsfoundthat ofthc93surveyrespond ents/prog ra mparticip ants, "74%w er e alreadyworkinginor intendingto workin aneligiblecommunitywhen theywere mad e awarco f theloan rep aym cnt program"(Renn cr ct al., 20101'.1) .Oftheindividu alsnot alreadyworking inarural comm un itywhen they applie d fortheprogram,"69%report ed thatth c opp ortunitY!llrl oanr epaym entwasan import antintluen ce on theircho iceof practice,'however , 66%alsostated they intendedto work inaruralcomm un ityalrea dy (Renner et aI.,20101'.5).
2.3.2Physici anRet ent ion
physicians,researchhas shownits effe ctson retentiontobel ess succcssful.ln19 92 Pathm an, Konra dand Rickettspublishe danine-year foll ow-upstudy of the USNatio nal IlcalthServiceCorps(NHSC) RFS scho larshipprogram.Thca uthorsrcportcd that rct cntion o fschol arship-obligatcd physicianswas " statisticallyand mcaningfull y shorter"
thannon-obligatedphysicians(p.1556) .Atlcreightyears,NHSCfunded physicia nswcrc signifi cantly lcsslikelytoremainin thci rcommunitytha n non-obligatcdphysi cians( 12%
vs. 39% ),orin anyruralco m munity (29%vs.52%) (Pathman,Konrad and Ricketts, 1992). While thelossofn on-NIISCphysiciansoccurredatasteady rate,NHSC physicianswereretaincdf'lftheirobligated years, and werel ostmor esharpl y on ee obligationswere completed .MorerecentNHSCresearchbyRabino witzctal. (2001) agreeswith the1992 findings,and reportedthatparticipationin theNII SCprogramt obe unrclatedt olong-tennph ysieianretention.
Rosenblatt etal.(1996)analyzedthe long-termcaree r pathsof famiIyphysieia ns post-NHS Cobligation.Asof 1994, 20.9%ofNHSCphysicianswhoserved betwe en 1980 and1983 werestill pract icingin their obliga ted-community.AsinthePathmanet al. study ( 1992), many physiciansleftonc e theirobligation wascomplete;of thosethat didleave,"most assigneeswholefltheir assignmentcountiesdidso withinmonthsofthe co ncl usion of theirobligati ons"(Rosenblattetal.,1996 p.26),ho w ever,longer ob ligatio nswerecorr elatedwith higher retentio nrates(Rosenblattet al.,1996).
Canadian RFSprogramsd onotsecmtofarc muchbctterthanthcirUx countcrparts.Longer f(,llow-upo fO nta rio physician distributio n bcf,)rca ndaflcr thc institution of thcO UAPf'lUnd alackoflong-tcnnret ention.Theallth orsconcludedthe OUAPdidnllth avethcd csiredimpacton physiciandistribution;"the reappearsto be littleimpro vementin physiciandistributionin thenorth compa red to southern Ontari o"
(Ande rsonand Rosenb erg ,19901' .43).
RFSobligatcdphysi cianshavcbccnfoundtohavc asignificnntlygreaterconcern abo lit their fi nancesin thefirst ycarsfollowingreside ncythannon-ob ligatedphysicians;
93%lof survcyrcspondcnt s statcd thcir nccd f()rtinanc ialassistancehad amoderate or majorinllucncconthcirdccisiontoapplyf(lfanRFSprogram(Jackson ct al.,200 3) .Th e
highcostof medicaltrainingmoti vates studentsto commit to RFS agree me nts,and thereforepro m otesphysicianemploymentin underserviced areas(l'athm anet al.,2000a), however,thismotivationmayaccountfor thelackof ret ention0fphysici ansintheir obligatedareasaswell.Oblig atedphysicianswere morelikelytoreport choosinga practicelo cation thatwouldhelpthem topayofflo ansquickly,while non-obligated physicianswe re morelikelyto report choosingapracti ce sitewithlong-te rmsettlement in mind (Jack son etaI.,2003 ).
2.3.3 Physici anPost-RFSPro!(raml'crccptions
In1996,Roscnblatt etal.condu ctedanana lysisoflong-tenn eareer pathsaswell asthe re trospeetive impressio nsofa fami ly physic ia ncohort that had re turnedservice for NIISCfundingduring1980 and1983.Ofthe258physicianswhorespondedtothe survey ,41%hadmixedopinionsof their experie nce,wh ile33%describ ed theirNHSC experience positively,20% negatively,and 6%neutr ally.Therewasnosignificant differenc einresponsesbet weentho sewhofulfilledtheirserviceand thosewho did not.
Theauthorsreport, "themost com mo nsentime ntoffered wasthatthe NHSCplacement had been asatisfyingandvalua bleexperiencethatresultedin anapprec iation fo r rural life andculture" (p.27),whiletheseco nd mostcommo ncom me nt"revolvedaro und displeasure withsome aspec tof the organizationor ad mi nistrationof theNSI ICand the processof match ing withand beingplacedatacommunity"(1'.27).Whilecom me nts evide ncct hatt hccxpcricnccwas "aftmnativeandwor thwhileexpcric ncc"(p.27) ftlr manyphysi cians.itwasclea r the reweremanyprob le ms withtheprogram as wel l. The authorshigh lighttheimportan ceofappropriatcmatch ingto ensurcbetterphysician experiencesandpotentiallyretentio n(Rosenblatt.1996).
Itisnotyet knownhow RFSprogram s affect physiciandistribution. l3are r and Stoddart (1999) state thatrather than att rac ting newphysician storural areas for recruitment,"it can be argucdthatthemain effcct ofincrca sed Ievels ofrcmun er ati on of various typeslorrura land rcmote practi ceistorewardthose who mightlocat ethere anyway,o rw ho have already doncsolargelyfornon-fin anci alrcasons"(p.15).Jaek son et al. (2003)assess ed WestVirginia'sIinan cialincentive progr am sto rruralphysici ans,and found90% of programrecipientsresp ond edthatthe program allowe d themto work in theirpreferr edsetting.According toJackson,this "indic atjcs]that theprogram sdid not attracts ignitica nt numbcrs of rccipicntsw howcre notalreadyinterestcd in underserved rural areas but,possibl y,madethese areasmore appealin g"(Jacksoneta l.,2003 p.337 ).
Otherresearch agreesthatfinancialincentivesmay ratherwork toreinforce orfacilitate the choice to wo rk in a rural area instead ofattracting uninterestcdphysicians(Math ews, Seguinan dCa rd,2 0( 9).
2.4Summa ry
Canadian RFS programs have no t been well docum cn tcd.Them ostrecent (and Iirst)attemp t atapan- Canadiansurvcyof programswasundertakenbyl3 arer,Wood, and Schneider IIyearsago ( 1999),andthe o nlyavailableCanadia n RFSart ielescome li-om ON andAB-describingprogramsthat arc nolongerineffect (Barer,Wood and Schneider,1999 ).Barcr andWo odstatethat outcomcsrdonotseem tohavebeentraeed"
in halfof the availableCanadian RFS programs(SK,ON,orNL)(1999).Healthpolicies and programshavemostlikelychangcdoverthistime,andthisthesiswillrecordwhat is currcntlya vailab le,aswellastheirtermsandconditio ns,to providcacohcsivc and currentaccountof Canadian RFSprograms
Qualitative analysis ofRFSprogramshasbccnrccommcnd cdthroughoulthc litcraturc(Jackson,Shannon andPathman,2003;Palhmanet al., 20OOa) yct the Roscnbl att(1 996)isthconlypublicationto inclu dcqualitativcdataonthc subjcc t lo datc.
Canadian mcdical,cducational,and RFS prog rams arcvcrydifTc rcnt from thosco fthc US,and isnotknown if HSCexperiences arcgeneralizabletotheCanadianorL contex ts.Thisthcsiswilladd ressthis gap, andwillqualitativc lycxplorcthc expcrienccs of bursaryrccipients.
Astudy iscurrentl ybeingcarriedout inNLthataims todctcnnin cthcpropo rtio n ofphysicianswhohavefulfilledtheirserviceob ligations,andasscssthc rclcntionof obligated physiciansagainstnon-obligatedphysicians(NLCAIIR,2(10).Bydcti ning Illotivalionsforbursaryacccpl anccandfcaturcsthalarcattra cti vctostudcnls,this stud y willcomp lement thisongoingwork.providingthccontext to undcrstand thcs ucccssor failurcofr ctcntionof Rl-S obligated physician s.
Chapter 3: Methods
Thisprojectconsistsof threesub-studies:ado cumentanalysis,anelectro nic survey ofcurre ntmedicaltraineesat Memorial University,and qualitati ve inter viewsof pastNL RFSbursary recipients.
3.1 Docu ment Analysis
Toad d ressresearch questionone,wecondu ct ed adocument analysis to createa cross-Cana dacomparisonof the RFS bursaryprogramsavailable tostudentsand residents.Theanalysis describestheten nsand cond itio nsof RFSprogramsoffere d bythe provincesandterritoriesinCanada andpreviousevalua tionsof theseprogramson the recruitm ent and retentionofphysiciansin thep rovin ee orterrit ory.
To docum entexistingprovineiallterritorialRFSbursaryprograms, wcbs itcsof government,stude ntaid,and provincial/territorialhcalthministrics wcre scarc hcdfor RFSprog raminfilIToation.lfthercquiredinfonn ati on wasnot availahleonlin e, appropr ia tcprogramcontacts were identitied thro ugh in tcme tsources andwcr econtacted foraninterviewbyteleph one orema ilinEnglish betweenJanuaryandMayo f2 010.
3. 1.2Sa m p le
Each provin ce andterrito ryisinchargeof de velopingandmaintaining thcir own recruitme nt and retentionstrateg ies. Onlypro vinc ia l/terr ito rialgovernmen tfun de d bursaryprog ramswith retu rn- for-service components wereincludedin this study,we did notcomparetheince ntivesofTcred hyhospitals,regionalorprivatefirms.
A data collection 1001wasinitiallycrea ted based oninfo rma tion fou ndinthe literature.After afewinitialinterv iews,thctoolwasmodifi edtobetterrepr esentthe
Usingthedat a collcc liontoolinAppcndix A,ti llcc na ttrihu tcswcrccollcctcd fromevery program:thcprogramtit lc,the awarding bod y,dat c of or igin,fund inglcvcis, awardworth,te rm s andeligibilityreq uireme nts,detailsahout paym ent andcomm itme nt (tim c andioc at ion),thcnumbcr of aw ardsa vailablc and acccpt cd cachycar,commitmcnt fultillmcntratcs,a ndfina lly infonnationa ho ut progra mcvaluatio n.
3.1.4.Data Prcparation an dAna lys is
Collcctcd in f(mna tionwasdocumc ntcd and thcnc ntc rcd intoa nS PSS da ta hasc.
We reviewedthcdatato grouprecurringrespo nsesandcodcddata according toAp pc nd ix B.Tostudy the comm onalityandvariatio n between prog rams,descripti vestatistics were
3.2Cr oss-Scc t io na IS n rvcy
To addressrcscarcho hjcctivctwo and three, a cross- sectionalsurvcyo f M cm or ial Univcrsity undc rgraduatc mcdical stlldcntsand rcsidcntswascondllctcd.Thc sllrvcyhas twomain ohj cct ivcs:li rst,to dctcn n inc th c proportion o f mcdi calstudcntsthat ar c considcringor havcaircadyacccptcdanRFShursary ;andsccond,to idcntifyRFS hllrsarytcnnso f intc rcst.Thcsur vcywill answcrth cqucstions"wh o arc thc RFS bursariesattracti ngandwhy?"
Survcyrcscarch providcsaqllantitativcdcscriptio noftrc nds.attitlldcs or opinions ofa popu!atio n,a ndaimslocollcctthcsam c sctofdata ll.))'cvcr y ·ca sc'int hcstlld y
(G re en andThorogoo d,20 09).They arcthe"designof choi ce"forresearehq uestio nstha t will describeprevalen ce , orassociationsbet w eenmeasurabl e variabl es(G reen an d Thoro good , 2009).
An invit ationto ourano nymousema il-d istri b uted web-b ased survey was senton ourbeh alftothe entir epopul ati on o fundergradu a temed iealstudentsat Mem o rial Un ive rsity bytheFac u ltyof Medi cine'sStude ntAffa irsoffice,and to reside ntsin trainin gbythePostGrad ua te Medical Ed uca tio n(PG ME)offic e.(Ap pe nd ixC).
Thesurveywasentailed byStuden tAffairsofficeon Janu ary14, 201l,andbythe PGM E offi ee onJanuary19,takingadva ntage oft hemed.mun.ea ema ilaccounts given to all Memorial Unive rsity med ical trainees.Theema ilcontainedanexp la natio nofthe survey and itspurpose,aswellas alinkto thesurveyonSurvey M onkey.eom (SurveyMo nkey,20l l) .An inereased nlllTIbe r o feontaetshasbeensho wntostrong ly influen c eresponse rate(Cook,Heath,and Thompson, 2000; Kaplowitz,Hadlockand Levin e,2004), there for e oneand two weeksaft er theinitialemail,foll ow-up emails were sent tostudents remindingthe mof thesurvey andthe deadlinebywh ichthesurveymust be comp leted (App endi xC).
Resear ehhasf(llIndthe response ratebyresidentsto cma ilsurveystobeslightly lowerthanpostalsurv ey(A kletaI., 2005 ),however,responsestoemailand postma il surveysby student po p ulationshave beenl(llIndto beeomparable ( Kap lowitz,Hadlock andLevine, 2004).Liter atu resuggestsan incentiveshould be providedt omotivate partie ipantstorespond; "poten tial respon d entswillwe igh thevalueofthe incentive againsttheirperceived costintime andeffort" (SueandRitter, 2007p.95).Entryintoa
draw for a$SOgiftcertifica tewasofferedasincentiveto partieipatein andcom p lete the
3.2.2 Sample
AllofMemorial University'scurre nt train ees withvalidemai laddresse sweresent the survey;262 studentand239resident emailswere contac ted.
Accord ing totheSurvey Mo nkeySmartSurveyDesignguide,it isexpectedthat surveys ad m in iste red through emai l tohave anaverageresp onserate o f 4 0% ,whil e surveyresp onseratesof SO%to 60%arcconsideredgoodandverygoodrespecti vci y (2010 ).Other web-based surveystud ies carriedoutonsimilar populationshavereeeived responseratesof42%( Coup er etal. ,2001) and 47%(Sh eehan,2001).
Thesurveyquestiooswere developedbased on themesidentified fromthe literatur e,theresearchteam,andinconsultationwithNLRFS andrecrui tme nt program planners.Su rvcyqucstionswcr cin fo rmcd from thcfi rstrcsea reh objcetivc, and thc surveyquestionsreflectedsomc o f the otherRFS bursarytcrrns/optionsthatarcavailable to other provinces.Thesurvey wascreatedwith the online survey program SurveyMon keyandconsistedof26questions(A ppe ndixD).
Thesurvey openedwith multipl e-choicequestionsthatcollected financ ia land socio-dcmograph ieinfonn ation theliteraturesugges tsmayafiect thc trainccsT mc dica l students andrcs idcnts) likelihoodo f aeeepting a bursa ry.lt then askedwhethe r ornot they acecp tcd,appli edfor,orintendto applyfor a bursar y .The questionsabo ut marital statusandperso na l incom e weremodi fie dfrom the Ca nad ianCommunityIlc althSurvey (CCHS) Cy cle1.2(Statistics Canad a,2002). Questi on s10andII, describin gthelevelof
financi alconcern,were modified fromapreviou sstudyon medicaltrainingdebtand servicecommitments(Pathmanetal.,2000a). An open-ended question asked trainees abo ut their primarymotivationfor acccptin g or choosingnotto aeceptabursary.
Finally,thesurveyaddressedthe attrae tivenessoftheterms and conditions ofthe Lbursary program .Participantswere asked to rateNL RFS term s,aswellasdiffe rent tenn sotTercdby oth erCanadianpro vinees.
Before ereatingour onlinesurvey,we pretesteda papervers ionofoursurvey.
I'retestinginvol ves"theinitial testingofoneor more aspec tsof thestudydesign,sue has thequestionnaire"(Babbie,1990,p.220).Topretest the questionnaire,asmallsam pleof seienceundergraduateandCommunityIlcalthgraduatestudents eomplctcd thc survcy.
and provideduswith feedb ackthatallowe d usto adj ustour q uestionresponse opti onsand instructionsforclarityandcomprehensive ness,aswellastogive anestimateof thetime req uired tocompletethesurvey.
Theprctestedsurveyw aslhcncreatcdonlinconSu rvc yM onk cy.comusingthc website'sbuilt-insoft ware,wherewecarriedoutapilotstudyof ourresearchins trument.
Accord ingtoDillman,SmythandChristia n.apilotstudy refersto"amini-stud yinwhich theproposcdquestionnaire and allimplemcnt ationproccd urcsaret cstcd onthesurvcy populati onin anattemp ttoidentifyprobl emswithIhe questionn aire andrelated implementationproccdurcs"(2009,p.228).Thesurveylinkwasdeliveredbycm ailtothc samcgroupofgrad uate and undergrad uatestudents whothcncomplctcdthconline qucstionnaire.Thepilot tcst allowcdustoco rrectanyproblcms w iththc onlinc tormatting, aswcllastotesttheemaildelivery oft hesurvey link.
3,2.4 Data Prep ar ati on
Resultswere submi ttcdand collectedelectronicallyby Surv ey Mo nkey"s sofiwa rc,
and expo rtedasanexccl filc.Student namesandidentifierswcrcrcmov cdand replaced
withastudynumber.Data wereimportedandcodedintoSPSS (versio n16.0for
Wind ows) . Tocleanthc data.freq uenciesandcrosstabswereusedtoidentifyimp lausible
orinco rrectanswers.Errorswerecorrected byconsulting otherqucstio ns in thesurvcy if
po ssi blc or changedto 'missingd ata'.
Vari ablesfromthc su rv ey were codedaccordingto Appen dixE.Someofthe
dcmograph icvariab lcsrcq uirc dsomemodifieatinnbefore analysis.Tncalculate
particip ant's age, wesubtracted thercpo rtcd birthyear fromthesurvey yc ar,20 Il.Tn
catcgo rizcthcvariab lc"ho m ctown :'wcuscd thcSta tistics Canada2006ccnsu s
Comm unityProfiles(Statistics Canada.2006 )andNew foundla ndand Lab rador 's
Comm uni tyAccountswebsite(Governm entof Newfoundland andLabrado r.20II)to look up thcpopul ationsofcntc rcdcomm unities.Consistcnt witbothcr Lrcsear ch
(Mathe ws andEdwa rds.20(4 ). lhcsizeofone 'shom eto wn was codcdas"rural(ha ving apopulationof<10.OOOpco plc). "scmi-urban"(10.00 0-99.999).and"urba n"
(>I00.000residents).Wecode dindivid uals'horneprovince asNL(I).andNon-NL (0).
Questio n5responses ofv m arricd"and"livingcommon-law"werecodcd as0,for
..partncrcd.Twhilcrcspo nscs ..wido wed... ..se parated,"..divor ccd .t'undvsinglc'twcrc
codcdastforvnon-purt ncrcd."
Foran indiv id ual'sbursary status10be coded as O."d ocsnot hold.anddocsnot
planto holda bursary:thcymusthaveanswcrcd "no"toq uestion I7( ind ica ting that thcy
havcnotrec ei ved a bursa ry)andalso havc answcr cd "no" toqucstion 18(indic atin gthat thcyd onotplanto apply foroncinthc future). Fo r anindividu al' sbursarysta tus to be codcdas I, " cu rrcntl yhold sorint cnd s to applyforabursary ,"thcy musthavcanswcrcd
"yes"toquestion17(indicatin gthat thcyh avcr ccci vcd abursary),orq ucs tio n l8 (indicat in gthat thcyint cndto ap ply fo r abursary).
Vari a blc slo r stud cnt cdu cation al and ovcralld cbt w crc ask cd as op cn -cnd cd question s. On ccthesurveys w cr c compl cted,catcgori cs w cr c crcat cdbascd onlhc quartilcs ofsub mittcdrespo nses,andcoded forana lysis.Thc prima rym ot iva tio n for a studc ntscho ice toaccept ordeclin e an RFSbursary wasinvcstigated thro ug hanopen- ended questi on on thcsurvcy;thc sc rcsp onseswcr c catego rizedand codcd fo ran alysis accordi ng to Ap pe ndixE).
3.2.(,Ana lys is
Toasscs sthc rcprcscnta tiv c ncssofthcsam p]c,ch isq ua rc tcstswcrc uscd to comparc thcscx,ycar,and ho mcprovi ncco fs ludc ntsa ndrcsidcn ts.Thcc harac tcr istics of thcs tudcn tsa m plcframcwcrcbascdon the (pub liclyavai lable) mcd ica lstudc nt class photos,wh ic harc arrange d byclassand includehom ct o w n/provinceinform ati on, Charac te r isticsoftheresident samp leframewereprovide d tousbythePGM E office.
Wcused dcscript ivcstatisticstfrcqu cncics,mc a ns andstanda rddeviations)to descri bethc ch ar actc risticso ft hesam plc,thc prop orti on o fr cspond cntswh o w er c awarc ofthebursaryandwhowere inte restedinRFSb ursari cs,andto identifythemost import ant RFSterms.
Chisqu a retestswere usedto comparc the cha rac tcristicsofstudc ntsa ndresidcn ts todeter m inew hc thcra ll respo nd en tscouldhcana lyzcd se pa rateIyortogcth c r. Because
ofsignificant diffcrcnccsbctwccnthctwo gro ups, all analyseswe rcdoncforstud cnt s and rcsidcnt sbothscparat clyandtogcth cr.
Totcst ourhypoth csiswcuscdmultiplclogisticrcgrcssiontoidcntifyprcdict o rs ofholding an RFS bursar y.Thc vari ablc'"planncdpracticcprovinccfivcycarsaftcr rcsidcncy"wasthc indc pendcnt va riablc a nd b ursarysta tuswasthc dcpcndcntvaria bl c.
Ch is quarc tcs ts wc rc uscd to idcntify d iffc rcnccs in thccha rac tcris ticsof trainccswho held/plannedto holdan RFS bursaryand thosewho did not. Potcntialcovariatcsincludcd va riablcsthat wc rcstatisticallysignili ca nt inthcsc bivariatc (ch is q uarc)a nalysc s.In addition,wcincludcdpotcntial intcractiontcrmsidcntificdusing chisquarct cststo comparcthccharact c ristic softrainccswho planncd and didnot plantostayinNL aft cr fiveyears.
Twovariableswereexcludedfrom thclogistic regre ssionanalysistoavoid potcntialmulti-colincarity.Thcrcspondcnt' s "planncdpracticcprovinccimmcdi atcly after residency"variablcwasnotinclud ed in thcrcgrcssionmodelasitwashighl y corrclatcdwiththcvariablc"plann cdpractic cprovinccfivcycar saft cr rcsidcn cy."Thc variablc " plan tofllndcd llcationwith R FS"wasnot incl ud cd in thca nalys isbcca usc thosc who plan tolundtheir education withan RI'Scith crholdorplan toacccpt a n R FS.
EthicalapprovalwasrcccivcdfromtheMemorialUnivc rsity ll umanlnvcstigati on Co m mittcc (IIICrefe renc e#10.215),McmorialUnivcrsitysPostgradu atcMedi cal Educa tiono fticc a ndS tudc nt Aftilirsoniccbc f()fcca rryingout thcslIrvcy.Thc initial cma il rcqucsting sllrvcypa rticipationc xplainc d thcstudyand providcdstudentswith
informationabo ut howtheir inform ationwastobeused .Participationin and completi on ofthe surveyimplied con sent.
To protectconfident iali ty,email add ressessuppliedasentryfortheincenti vedraw were remo vedimmedi atelyupon receiv ing thesurvey, andentered into a separate docum ent. Anyother identifyin g informa tion wasrem o vedfromthesu rveydata,and each surveyrespo nsewasgive nastudynumber.
No person alidentif yin ginformationwasrecordedfor ana lys is,and individu als wer enot indcntificdinanyre portor present ation.Since individualsw er e abl eto abortth e surveya ta ny time,wef()resawverylittle pote ntia lf()r hann cause d bythisresearch . Stude nts parti cipatin ginthesurveywere notlikelytobenefitdircctlyfromthisrcsear ch.
Data were store d inalocked room (CommunityHealth, 2847A),withelectro nic tilespasswordprotected.On lymysupervisorand I have accesstothcm .All datatiles will rc mainthcrefor fiv eyca rs,atlerw hich thcywillbedestroycd.
3.3Qualitati veInter vie w s
To addressresea rch objec tivethree,we cond ucted qualitativ eint er viewsto gain insig ht intothe experienceandsatisfactionof actualprogramusers.Qualitative inter vie ws.ingenera l,uscopen-end edquestion sthat seckto gain info rm ati on andare
"wellsuited fordescribingboth program processe sandoutcom esIi'omthe perspecti ve of thct ar gct audi cnce or kcy stak chold cr" (Guion,2006, p. I).
Sem i-structurcd intc rvicwsofferalcssrig idty pco f interviewt ha ns tructurcd interviews.Whilcth isintc rvic wstylcinvolvcsasking pre-dctennin cdqu cstions"t yp ically askedofcachi nterv iewccinasystcmatica ndco nsi stcntordcr",thci ntcrvicwcr sare
"allo wed freedomtodig ress :that is,theinterviewer sarepermitted (in fact expcctcdjto
probe farbeyondtheanswersto their preparedandstandard ized question s"(Ber g,1995 p.34).
To recruitparticipants,our program eonlac t DanielFitzgerald(Policy,Planning&
ResearchAnalyst,Department ofHealth&Comm unitySer vices),sent anem ailto c1igiblepastprogram-usersexplainin gthe study and requcstin gtheirparticipation(see Append ixF).On March 15and16, 201 1, theletterwasemailedtol77lra ineeswho received abursar ybelween 2004and2010.Whilethisnum ber represent sthemajorityof bursaryrecip icntsforthetimepcri od of2004-2010,it docsnot inc1udcevcryon c.
Ninetee nema ilswerereturn ed asundeliverable,leavin g159cmails successfully
Physicianscallcdour tcle phone numbcrorrespondedbyema ilto cxpresstheir interest,alierwh ichconsentformswere emailed orfaxedtothem.Theirsig nedcopies werelilxed orseanned andem ail edbacktous.Ph ysici answereth encont actcdby telcph onetosetupand carryouttheinterviewsatmutuallycon venientinterviewt imes.
Sem i-str uctured inter viewswe recarried out in Mar ch andApril20 1 1.Inte rviews lastedbet weentenand20 minutes,andwerecarried outover the telephone asphysicians were work ingin and outsidetheprovince.
3.3.2Sam p le
Tohcinc!udedinthcintervie wportionof thc study,participantsm us t havc previously accept ed anNLRFSbursar y agrecmenl,and completedtheirservicc repaymc nt,rc paid their bursary,or dclillllted.Toavoidanypereeptionofco erc io n,
physician swho were in the processof training,repa yin gorcom1'1ctingtheirserv icewere notelig ible.
Initiallyweexpected toconduct betweeneight and teninterviewstoreac h saturation (i.e.the po int uponw hich the fu ll ra ngeof idcashasbe e n reached and no new informationisbeingcollected(Krueger andCasey , 2000)). Mor eo ver,t oensurea representativesam ple,we employedpurposiveandstratiliedsam p ling(Berg,1995).We plann edtostratifyoursam plc bythetypeof bursary(familymcdi cine,psychiatry, spccialist and travelling ),and intcndedto interviewtwotothreephysiciansforeachtype of bursary(ineluding,where possible,atleastone individualwho fultill edtheirbursar y com mitment through service,andoneindividualwhohadrep aidtheir bursary).
Theintervie wsweresemi-structured, followingtheinte rvi ewguidefoundin Appendix G.Inte rviewquestion swere devel op edbased on thelite ratu rereview,and in consultationwith my supervisorycommi tte eandprogramstakeholders.Questions explorcdthe experien ccsofpast-programuscrs,andspccili callyaddressed the attractivenessof theprogram,themotivationsforand meth od of b ursaryrepayment,and theeffectsthat the RFSc xper icncehad o n thcphysicians 'car eers(AppendixG).
3.3.4l>ala Prcp a r at ion andAna lysis
Intcrvic ws were ca rricdoutovc rthctcl c phonc and rccordcdona computcras .1l14a tilcs (dig italaud ioti les). The intc rv ic wswcre thcn transc ribcdverbatim.Names and ident iiyin ginfo nnation wcrc rcmovcd,andastudynumberwas givento each intervi ew .
Thematicanalysi swascmploycd to expl oretheintervie wdataand dcrivcmcaning from thetranscripts .Thematicanalysis isamcthodformak ing scnscoft hcdatu
(Boyat zi s,1998), and"ide ntifying,analyzingand rep ortin gpatterns(them es)withindata"
(Braun andClarke,20061'.79 ).Aronsonexpla ins thatthemes emergefrom many fragment s oftheinterview ces experiences,andare "piece d togethe r tofonna eom pre he nsive pietureoft he ireo lleetiveex per ie nee " ( 1994 p. I) . The productofa themati c ana lysis invol vesnot onlyadescript ion ofthe data,butal soincludesa description of themeanin g ofthethemes/patt ern s(Braun andClarke,20 06).
Dat a wer ee odedth rou ghaniterativep roeess;my supervisorandlind epend en tly readthe firstthreeinter vie wtran scriptsandcreated acodingsehemebyclustering reeur ring wordsorideasinto catego ries,that we thenorganized intothem es.Togeth er,we negoti at edthe coding templ at einAppendix H,whichwasthenusedto codeandana lyze all rem aininginterviews.
Ourcoding processensuredourfindi ngsinclude dnaturallyarisin gcategories, rathe r tha n thoseintluenee d bype rsonalb ias.Kruegerand Caseyexpla in thatusingtwo ormoreindepend ent coderssho uld produce a"precise, rcliablc andrep rodu eibl c eoding systelll" (2000, p.42).Toilllprov ethe ered ibility o f ourstudy,detailedrecordswere kept , inc ludinginte rviewtapes.tra nsc ripts,fi eld notesando bservationsmad ethroughoutthe inter vie ws,aswellasthroughoutthedata cod ing process(Maysand Pop e,1995).We re fe rredto our fieldnot esandobservationsafterwriti ngourresults section,to ensurewe interpr et ed quotescorrec tly.
Ethica lapprovalfortheinterviewswasreceive dfromMemoria lUniversity's Iluma nInvestigationCo mm ittee (HICrefe ren ce#10.216) beforestarti ngthestudy.
Intervie w participants 'persona l identifierswere remo vedfromthcd a ta and werereplaced
withstudyIDnum bcrs.Topro tcc t thcidcnt iticsof th cphysicianswhopartici pa tcdin thc interviews,dat athatmightidentifythc individualphysi cian swere edited inthe quotation s.Individu alsparticip atin gin thcqualitati vcintc rvicw wcrcnolpcrso nall y idcntilicdinan ypuhlicationorprcscntation.
Asnopcrsonal idcntifyinginfi.mnationwasrccordcd oranalyzcd , andthc rcscarch wasl argcl y op ini onbascd,thcrcwaslinlcpotcntialforharm causcdbythisrcscar ch.
Participantswcrcnot obligatcdto an swcrqucstions,andwcrcad viscd of thisbefore the
Data weresto red inalockedroom(Communit yHealth, 2847A), andelectronic tilcswcrcpasswordprotcctcd ,with o nly mysupcrviso r andmch avingacccss.Thctilcs willrclllai nhcrcfortivcycars, allcrwhichthcywillbcdcstro ycd.
3.4 Knowl ed ge Tr a nsfer
Oursurvc y andqu alital ivc intc rvi cwsw crcd c vel op cd with input lr o mk cy programpl annc rs .Th cscprogramplanncrswcrc ourlinktothcstudcnts andphysicians studicd,andcontact cdthcscindiv idualsonour bchalf.
Theresults ofthis studywill bcofintcrcstt othcDcp art mcnto fH caIthand CommunitySe rvic es . thcNL HealthBoardAssociat ion (LHBA),the Facu ltyof Medicine,physician supplyand distributionrcscarch crsandothcrpro vin cialD cp artmcnts ofl lcalth .Torcachthcsc audicn ccsiwi lldissclllinatcth cstud ytindingsbydistributing summaryrc por ts.wri tingarticlcslilrpccr-rcvicwcdjou rnals.and fina lly,byprcscnt ingat seminarsand thc Zu ltCanadia nAssociation forll calthServicesand Poli cyResear ch (CAIISI'R)confcrcncc.
Chapter 4: Result s 4.ll>oeument AnalysisR esults
Thisanalysiswascarriedout todocumentthe return- for-service(RFS)bursa ry programsavailablefrom provincialand territori algovernments tomedicaltraineesacross Canada.The goalwastocomp areandcontrastavailable programs and their differing characteristics,terms,andsuccessrates.
Throughuscof both online and tclephonesoureesweascerta ined that each Canadian province and territoryhas some formof medicaltrainee fundingprogram,with all hut theYuko n (Y U) req uiring an RFScom mitment. Thirteengo vernment fundedRFS programswere identifi edforinclusio ninthisstudy,one from everyprovinceandterritory exceptYU, andtwo fro m MB.
Whilemedi calRFSbursaryprogramshaveexisted inCanadasincetheI960s, theyhavebeen suhjeettomany changes.Allcurrent program versions wcrccstablishcd after theyear2000,with themajoritystartingbetween2000and20OI.Themllstrec cnt programcomesfr omAlberta,whichcom me nced in2005(personalcommunication,D Kay,Marchf,20fO)(Tahle4.1).
Table4.1Yearof OriiinofCurrentCanadian RFSPro'ramsbProvinceandTerrit or
Nt. PEl NS Nil QC ON Mil Mila SK All IlC YUNI NU
2000 ./ ./ ./ ./ ./ ./
2001 ./ ./ ./ ./
2002 ./
-2003 ./
2004 2005 ./
Themajorityo fR FSprogramsarcgivenin theform ofhursaries(Table4.2).
Otherf()rnlsof inccnt ivcinciudegrants,loa nf ()rgiveness, scho larshipandt uition
reimbursement.Allprovincesdeliverfundsdirectly10thestudent exceptBC , in whieh
the physic ian'sprovinci al stud ent loansare forgi veninstead(StudentAid BC,n.d.).
Theseincentivesare consideredta xableincom eino verhalf ofprograms,but depending onthe province and typ eof bursary,maybeinel ud ed aseither tax-exe m ptor tax- deduetib leineome.TaxstatusofRFSin eenti vesi sdependentonthepar am eters set bythe Cana da RevenueAgenc y , andisnotcont rolledbythe programitself.
Whileall program sincluded in thisstudy are provinciallyfunded,theyare
ad m iniste redthro ugh differentorganizations.Mostfinanci alincentivesareawarded thro ughtheprovineeor territ o ry' sdepartment of health,howevertheyeould alsobe deliveredthroughstudentaid programs(BC,MB),oran independentbody(SK,AB).The
Saska tc he wa n MedicalAssoc iati on(SM A) adm in istersthe Saskatche wa n bursary prog ram, and in Alberta ,bursariesare distributedthroughtheAlbcrta RuralPhysician ActionPlunI'Tablc-t.Z)
Table4.2Com arisonofl'ro vin eiaLrr enit ori alRF SFinaneialln eentivel'amentDetail s NL PEl NSNil QCON Mil Mila SK AllIlC N'! NU
~~~(': BursarY. ./ ./ ./ ./ ./ ./ ./
Grant ./ ./ ./
_~-2_~~~~~g~~:~~s ./ ./
./
cw su-din eHedv:
he, lIh ./ ./ ./ ./ ./ ./ ./ ./
Pro vin cia!~:~~~~:1:~~ ./ ./ ./
Indecndent bodx ./ ./
i\leth od ofl)a\'lllcllt:
Directltostudent./ ./ ./ ./ ./ ./ ./ ./ ./ ./ ./ ./
Forgiveness of Loan ./
I
IIICClltivc T a' StaIIlS: _~_...Ta"-", e_,em",,,-l,( , ( , (
Ta,deductible ,( ,( ,( ,( ,(
Taxablcillcomc , ( , ( ,( ,( ,(
Note:MBaRefersto anAboriginal-spec if icRFSprogram
As scc ni nTable4.3,all b utonc RFS bursaryprog ramo ffcr funding to undergradu at estud ents,withsix programsofferingfund ingtostudcntsbcf or cthcir thi rd ycarof mcd icaltra ining.Elcvc no f l3RFS programsp rov idc postgraduatc fund ing.
Table4.3Canad ia n RFS Pro 'ram Availab ilit b Provinc e and Tcrrit orv
NLPElNSNilQCIONIMBMBaSK AB BC Nl NU
~:~se:~:::hI:lIC
YearI ,( ,( ,( ,( ;r
Year 2 ,( ,( ,( ,( ,/ ,(
Year3 ;r ;r ;r ,( ;r .r ;r ,( ;r
Year4 ,( ,( ,( ,( ,( ,( ,( ,/ ,( ,( ,(
~~~t;.;::~~:atc
FamilMcdicin e:IUIII ;r "7 "7 "7 "7 ,( ;r ,( "7
Famil Medicine :Rill ,/
Specialist General ;r ,( ,( ;r ,/ "7 "7 ,(
Seci alist:Psc hiatry ,/ ,/
TravelingFellowshi** ,/
Thevaluco fRFSfunding variesbyprovince,pro g ra m,traine cycar, and specialty. Thcrequiredservice limemay vary accordingtoprogram,year, andtheareaof servicereturn.Awardvaluesand returnrequirements arcsum marizedinTab le4.4.The mean value ofa singlc ycarofund ergradu at c RFSfun dingi s $15,423 acrossCanada,with amedianofSI5,OOO.Undergra d uatc RFS valucsr angeinworthfrom56,000(Ne w I3runswiekHealth,n.d.)to S25,OOOper ye ar(person aleommu nieation,J.I'c1ly,May4, 20 10;Manitoba Health, 20 11-12;I'raetieeNL , 2010)
The mean valueof oneyear'sfund ing for a familymed icin e residentis $22, 04 5 acrossCanada,withamedi an of$20,000(Tab le 4.4).Familymedicin e resident bursa ry valuesrangeinworthfrom$10,000 inOntar io (Ministry of Health and Long-Term Care, 2009),toup to $50,000,whichisavailableto studen tsin their secondyearoffam ily medi cineresiden cy whoarc takingpart inManit ob a' s Northern Remo teFamilyMedi cine Resid en cy Strea m (NR FM S)program(per son al communica tio n,R.Parkin son, Marc hII, 20 10).Boththemean and medianvalues for oneyearof specia listRFS fundin gi s
$20,000,withaminimumvalueof$IO,OOO(Gouveme me ntduQuebec,2005), anda ma ximum of$25 ,000(perso nalcomm unica tion,1.Pclly,May 4,2010; Practice NL, 20 10;New Brunsw ic k Health,n.d.; Saskatchewan Medical Associat ion, 20 10).
RFS fundingspcc itic to psychiatry resid entsisavailah leonlyinNLandNova Scotia (NS), andarc worth$25,000(Prac ticeNL,20 10)a nd$15,00orespectively(Nova ScotiaHealth,2(10). Trave llingfellowshi psarc offered to rcsidcntsorphysician swho scekspccialty rcs id cncy trainingthat is in nccdinNLbut isnot o tlc red in thc provincc.
Thcscph ysieiansrecci vetheirtuition an drcsidc nt wageswh ilc co mpl ctingrcsidcn c y outsideNL inexcha ngefor commi tting to practiceoneycarofscrviccinNl.Jorcach ycaroffunding ( Prae ticeNL,2010).
Rceipic ntsmay receivemore tha nonch ursaryover thecourscofthcir cdu cat ion.
Thc highest potcntia lcum ulativefundingis $13 1,000, availab lc tofamilymcdicine traineesin Manitoba,ifan individualreceivesthe maximum fourundcrgraduatcgrants and pa rticipatesinboththeRuralNorthe rnInitiativc (RNI) and thcNorthernRemote FamilyMedicincReside ncyStream(NRFMS)(Manitoba Health. 2011-2(12).Thelowcst
pot enti al cum ulativefundi ng isavailableinOntar io ,withstudentsrecei vin g a maximum of$40 ,000overtouryears (personalcommun ica tio n,R.Parkinson,Ma rch11, 2010).
As seen in Table 4.4, theservice comm itme ntaccompanyinganRFS agreeme nt variesbetween andwithinprovinces.Themajori tyofprogramsrequir ephysiciansto work 12mon thsofservice inexchangefor oneyear of funding. Howe ve r,particularly rural or re mote areasaswell asrcl iefp rogram s oHenrequireshorterdu rat ion o f retum ed servicc.MostRFS programs req uire physieiansto \Vorkinaruralcom m unity, oronc that isclassifie das "in need"or"undcrscrved;" howeve r,se veralprograms simply require physicianstoremain inthe irprovince.
Tablc 4.4 Compa risonofCana dia nRFS Valuesand ReturnReq uirem c ntsAccording to Pro vincc andRcci ientTve
ProvinceIRecipientIRFS,,: : :::($) Servicelocation/specialtyre quircmc l1l
(111~~1
year:Lg
NL 25,000
NL F~ 25,000
NL 25,000
NL 25,000
NL Salarv/ruition
PEl F~ 15,000
1'1'1 20.000
NS 15,000
Nil 6.000
Nil 12.500
Nil FMRes 25,000 icto Is
Nil SP Res 20.000 IS
QC ;3-4 15,000 fnc 12
OC SPIes 20,000 12
OC 'MRes 20,000 12
C SPRes 10,000 t2
ON UG 4 10.000 12
ON FMRes 10.000 12
ON SP Res 10.000 12
Mila
\~~i~;~a!l
7,000 Ruralcommunity 6Mil 1-4 12,000 oca idirectedby province 6
Mil JG3 ** 25,000 ura nmunit 12
Mil UG4** 15,000 lani s comrnunir 12
Mil UG4(RNI) 25,000 irec znrovmce 12
Mil 'M Rcs 20,000 lith , orovmcc 12
Mil
(~~/~~~)
50,000 nee 24Mil 'Res 20,000 itv
""S)
SK UG2-4 15,000
UIllS)
SK FM Rcs 25,000
UIllS)
UG 1-4 (UofA)
UG 1-3 (U of e)
Phy sici ans 1-4
'Res
11,540 reg rmunitv
14,384 reg unmunitv
33,3% ofall I'ubliclyfundcd lilcilil yinundcrserved studentloan s area
1
,000 Within thctcrritorv
000 Withintheterritorv 25,000 Within theterritory
iycarstotal
Leaveisgra nted from allbursaryprogramsformat ernity,medicaland compassiona tereasons,butrequ irephysicianstomovethe end dateof theirserv iceto honourtheir com mit me nt totheprovince.Deferr al ofserv icedueto illnessor compassionatcr casollsorfurthcrtrainingmaybcconsidcrcd inmost programs.
The penal ty forno thon ouringRFS comm itme ntsvariesbyprovin ce,andranges from nopenaltyatall toheft y feesand int e rest(Table 4.5). The maj orit y ofprograms req uire physieians/traineeswhodonot fuIIiII thei r eommitm ent to repaytheir bursa ry with interestthat acc umu latesfrom whenthe bursary wasreceive d.Someprograms insteadrequi rephysicians and traineesto repay thei r bursa ry withinterest that accu mu latesfrom whe nthey decided not to fu llilltheirservice commitment.The provinceof Ontarioalsorequ irespaymentof anadditional tincto cover thecosts associate dwithdefault(M inistry ofHealthand Long-TermCare,2009).
a c. ompansol1o ana Ian rogram cave cerraan cnayor 011- fulfi llme nt bProvineea ndTerri tor
INLPEl NSINilIQC
I
ONIMili MllalSKI
AllI
IlCI
NTINU 1..(';1\'('./ ~ ./ ./ ./ ./ ./ ~ ~ ~ ~ ./ "7
Matcmit
_ _1"1cdica llcompa~~i.Q.Il.£I_IC ./ ./ ./ ./ ./ ./ ./ ./ ./ ./ ./ ./ ./
Dcrcrr ntConsid er cd
./ ./ ./ ./ ./ ./ ./ ./ ./ ./ ./
Forfurthertruininu ./ ./ ./ ./ ./ ./ ./ ./ ./ ./
Penal
;>' None
tcnavtundine ./ ./ ./
fundine from default ./ ./
fundinufromreceipt ./ ./ ./ ./ ./ ./ ./
Fees ./
Nose:1\1B,1Refersto anAboriginal-specificRISprogram
Someprovincesandprogram shaveadefi nite numberofawa rds that arc awarded eachyear,whilein otherprovinces,thenumberaward ed variesbasedon budget.Despite thea llolled numhe rofhursarie sorhudgetavailab le,as seen in Table 4.6 thenumher of hursariesaetuallydistrihu tedtostudentshytheprovineesand territor iesvaried dramati cally,rangingfrom zero in2010 (personal communica tion,F.Nanj i, May7, 20 10 :
personalcommunication,S.Macl. can ,March 8,2( 10),to167(personalcommunication,
R.Parkinson,March11, 2( 10 ).
Table4.6Canadian Provincial and TerritorialRFSAvail abilitandFulfillm e nt Rate Provinceand Program Numbc ro f RFSCo ntracts RFSContracts RFS Fu lfillment
Availa blePerYear Awarded''" Rate
4/FMtes Accordingto budget 48
.SP Res Accordinc to budaet 30
Psvch Accordinuto budcet 7
Aecordingto budget 3
0
0 89%
,3-4 40
,3-4 40
Accord .budaet 81 90-95%
Accordin .budee t 135
Accord .budget
,
. .
,
..
lNI' Accordin .budeet
Accord .budget 167 Accordinto budzet
10 10 10 10
25 ~
AtlcnstJ
15 15
10 8 2did not
10 ; lastvca t, complCICin2009
Be Accordingtobudget 2&t~t;~;ince
NWT Nonewintake
NU Accordina tobudzct
..
" "I'''"'' U'", Mostprogramswere unabl eto provideinform ation abo ut theproportion ofth cir contrac istha ta rc fuifillcd,asop poscd to dcl'crrcd,dc filllilcdor rcpa id.W h ilc lhc majo rity of provinccsrc porlcd that" a lo t"o r" lhcmajorily"of thc irstudcntsfu lti ll thci rscrvicc, NS andQCwere the onlyprovinces10providetheir act ualfulfillme nt rat esof89%and90-95%rcspc cti vcl y(pcrsonalcommunicati on ,A . Busq uc,May 5,20 10;pcrsona l com munica tio n,F.Nanji,May 7,20 10 ).
Despitethelargc numbcr o fbur sarie savailable and distributed acrossthecountry,
vcry fcwo f thcse c urrc nl RFSprogramsha vebeenformally evaluated (Tablc4.7).
Manitobarcportstoha ve cvaluat cdtheirp ro gramsintcrn all y(R.Parkin son ,pc rson al com munica tio n, March11,201 0 ),andcurr entresear chisinprogrc ssin NLto asscssthc effec tsof R FSbursarycommitmc ntson physicianrecruitment and retention(NLCAHR,
2010).Saska tchewa nandAlber tapJanto cvaluatc thc ir program sinthefuturc(person al com munica tio n,D.Kay,MarchI,20I0;personalcomm un ica tio n,E.Hobday ,May3, 20 10).
4.2l\Iedi cal Tra ineeSu rvey Resnlls
Thc p urposcof this survc ywaslodescribe thc pro po rtio no f mcd ica l tra inccsthat areco nsider ingor havcalrcadyacccptc dan R FSb ursaryand toi dc ntifyRFSbursa ry tcnn s that ar e of int cresttomcd icaltraio ccs. lnadd ition . th rou gh our analysiswe assessed whcther thc progra misattrac tingprevio uslyunintcrcstcdtrai ncesto commi t10work in NL post-rcs idency.
4.2.1Sam p leC ha raete r isti esa ndReprese n tative ness
Ofth c 501surveysthatw crescnttostudcnlsand rcs idcnts. 150studcnts (57.25%) and7R rcsidcnts(32 .64%) replicdwithan overallrcsponscratcof45.5% .Tw ohundred
and tw enty-eighttrai neesresponded in tota l,with208 traineescomplctingthc survey entirely.
Of ourrespondents, 223 answeredwh ichprovincethey wanted topra ctice in five years.Totestourhypoth esis,we comparedthe106 traineeswho planned tostay in NL five years after complet ing trainingtothe117who didnot. Th is sam plesizeallo w ed usto detectastatisticallysignificantdiffere nceofI8 .2%ormorcbetweentraine eswho want to andwhodonot wantto work inNL usinganalphaof 0.05andpower of 0.80 (Lenth,2006-9).
Among stude nts , oursamplesizcallowedusto detect astatisticalIy signi tic a n t differenceof23.2% or morebetweenstude nts whowanted10(n>74 )and who didno t wantto (n>71)work inNL(Lcnth,2006- 9 ). Ourresidentsam plesizeallowed usto detecta statisticallysign ificant differenceof33.7% or morebetwe entraineeswhowantto (n>32) andwhodo n'twantto (n=46)wo rk inNL(Lcnth,2006 -9) .
In tcnnso f re prescntativeness,thcsam plcofstudcntswasno t d iflerentfromthe stude nt popu!ationwith respecttosex (p=0.736)orhomeprovince(p=0.855)( Sce Appe ndixI).Howeve r the samplewas differentwith respectto year of medicalscho ol(I'
=0.005); morefi rstyearstudents respondedto the survey thancxpectcd.undfewcr seco nd to fourthyea rstude nts respondedthan expec ted. Thcrcsident sa mplc was not different trorn thcrcsidcnt so urce populati onwithrespecttosex(p=0.255),homc provinc e(p>0.172) ,oryea r of residency(p=0.5(9) .