Outcome of a neuromuscular training program on recurrent ankle sprains. Does the initial type of healthcare matter?

ContentslistsavailableatScienceDirect

Journal

of

Science

and

Medicine

in

Sport

Original

research

Outcome

of

a

neuromuscular

training

program

on

recurrent

ankle

sprains.

Does

the

initial

type

of

healthcare

matter?

Adinda

K.E.

Mailuhu

_,

_Marienke

_van

_Middelkoop

_,

_Sita

_M.A.

_{Bierma-Zeinstra}

_,

Patrick

J.E.

Bindels

_,

_Evert

_A.L.M.

_Verhagen

a_{DepartmentofGeneralPractice,ErasmusMCUniversityMedicalCenter,Rotterdam,TheNetherlands}

b_{AmsterdamCollaborationforHealthandSafetyinSports,DepartmentofPublicandOccupationalHealth,AmsterdamMovementSciences,VUUniversity}

MedicalCenter,Amsterdam,TheNetherlands

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received1October2019

Receivedinrevisedform9March2020

Accepted16March2020 Availableonlinexxx Keywords: Ankleinjury medicalconsumption interventionoutcome re-sprains

a

b

s

t

r

a

c

t

Objectives:Tostudytheimpactofinitialhealthcareforalateralanklesprainontheoutcomeof neu-romusculartraining(NMT)onrecurrentsprainsanddescribeathletecharacteristicsreceivingdifferent typesofhealthcare.

Design:Secondaryanalysisofthreerandomizedtrials.

Method:FromthreeprevioustrialsevaluatingNMT,dataonathletecharacteristics,typeofinitial health-careandarecurrentsprain duringone-yearfollow-upwerecollected (N=705). Multilevellogistic regressionanalyseswereusedtotesttheimpactofinitialhealthcareontheoutcomeofNMTon recur-rentsprains.Potentialdifferencesinathletecharacteristicsbetweendifferenttypesofhealthcarewere examinedusingone-wayanalysisofvarianceandPearsonchi-squaretest.

Results:AfterNMT,39.7%oftheathletesvisitingparamedicalcarereportedarecurrentsprain,21.8%of secondaryand34.0%ofprimarycare.Athleteswhovisitedaphysiotherapistreportedmorerecurrent sprains,thanthosenotvisitingaphysiotherapist(adjustedOR3.15;95%CI1.88-5.23).Athletesvisiting paramedical(49.7%)andprimarycare(48.4%)usedmorebracesand/ortapeduringsportthanthosenot visitinganycare(34.2%).

Conclusions:Theinitialtypeofhealthcareofathleteswithananklesprain,thatconsecutivelyreceived NMT,seemstoimpacttheoccurrenceofrecurrentsprains.Physiotherapistsmaybevisitedbyathletes withapoorerprognosis,whichmaybeexplainedbydifferentathletecharacteristics.

©2020SportsMedicineAustralia.PublishedbyElsevierLtd.Allrightsreserved.

Practicalimplications

• Theinitialtypeof healthcarefora lateralanklesprainamong athletes,thatwerethereaftertreatedwithNMT,seemstoimpact theoccurenceofrecurrentsprains.

• Physiotherapistsshouldbeawarethattheyarelikelytoseea spe-cificpopulationofanklesprainpatientsthaymayhaveapoorer prognosis,i.e.haveahigherriskonsustainingarecurrentsprain.

1. Introduction

Anklesprainsareoneofthemostcommontraumastothe mus-culoskeletalsystem.IntheUnitedStatestheincidencerateis2.15 per1000person-yearsinthegeneralpopulationand3.29per1000

∗ Correspondingauthor.

E-mailaddress:a.mailuhu@erasmusmc.nl(A.K.E.Mailuhu).

person-yearsinpatientsattendinganemergencydepartment.1,2

Aftersustainingan anklesprain,up to33%of thepatientsstill experiencepersistingcomplaintslikepain,swellingandstiffness after12months.3_{Evenafterfive-yearsalmost20%ofthepatients}

reportpersistentcomplaints.4_{Inadditiontopersistingcomplaints,}

anincreasedriskofarecurrentsprainisespeciallypresentinthe firstyearaftertheindexsprain.3_{Studieswithlong-termfollow-up}

(5-7years)showedthatafteranindexsprainupto20%reported recurrentinjuries.5,6

Thus,ananklesprainhaslong-lastingimpactonphysical func-tionandtheuseofmedicalcare,upto12monthsaftertheinjury.7

Therefore,theseinjuriesdonotonlyposeaphysicalburdenforthe individual,butalsobringasocietalfinancialburden.Eveninasmall countryliketheNetherlands(about17millioninhabitants), the annualmedicalcostsforanklesprainsamongathleteswhoattend theemergencyroomwasin2017almostD6million.8

Provided the individual and societal burden after an ankle sprain,thereisdemandforeffectiveandevidence-basedtreatment https://doi.org/10.1016/j.jsams.2020.03.010

Table1

Characteristicsoftrialsincluded.

Hupperetsetal.(2009)* Janssenetal.(2014)† Reijenetal.(2017)‡

Totaltrialsamplesize(n) 522 384 220

Receivedneuromusculartrainingprogram(n) 258 227(withn=120incombinationwithabrace) 220

Age(years),mean(SD) 28.6(11.8) 34.1(13.4)§ _37.8(13.4)

Sex(male),n(%) 136(52.7) 110(48.5) 110(50.0)

BMI(kg/m2_{),mean(SD) 23.4(3.2)}|| _23.8(3.2)¶ _23.2(2.8)

n=number;SD=standarddeviation;BMI=bodymassindex*Hupperetsetal.Effectofunsupervisedhomebasedproprioceptivetrainingonrecurrencesofanklesprain: randomisedcontrolledtrial.BMJ.2009;339:b2684.†Janssenetal.Bracingsuperiortoneuromusculartrainingforthepreventionofself-reportedrecurrentanklesprains: athree-armrandomisedcontrolledtrial.BrJSportsMed.2014;48(16):1235-1239‡VanReijenetal.The ¨Strengthenyourankle ¨programtopreventrecurrentinjuries:A randomizedcontrolledtrialaimedatlong-termeffectiveness.JSciMedSport.2017;20(6):549-554§,||,n=1unknown

andpreventionmodalities.Oneofsuchisneuromusculartraining (NMT),whichhasbeenproveneffectiveinpreventingrecurrent sprainsafteraninitialanklesprain.9,10_{Ourgroupperformedthree}

randomizedtrialsintheNetherlands,allincludingaNMT

interven-tionarm.11–13_{Hupperetsetal.}11_{werethefirsttoshowthattheNMT}

waseffectivebutalsoshowedthatNMTwasspecificallybeneficial inathleteswhoseindexanklesprainwasself-treatedratherthanby ahealthcareprovider.Thisraisesthequestionswhetherthereisan associationbetweenthehealthcareproviderwhotreatedtheindex spraininitiallyandtheeffectofNMTonrecurrentanklesprains and,also,whetherathlete characteristicsdifferbetweenthe dif-ferenttypeofhealthcareproviders.Thesecharacteristicswillbe usefulforhealthcareproviderstotakeintoaccountwhenchoosing interventionstrategiesordetermineprognosis.

Therefore,theaimsofthisstudywereto(1)studytheimpact oftheinitialtype ofhealthcareforalateralanklesprainonthe outcomeafterNMTforthepreventionofrecurrentsprainsandto (2)describecharacteristicsof athletesvisitingdifferenttypesof healthcareandexaminepotentialdifferencesinthese characteris-tics.

2. Methods

Thecurrent studypresentsa post-hoc secondary analysisof thecombinedindividualpatient dataascollectedin three pre-viousDutchtrialsonthepreventiveeffectsofNMTonrecurrent anklesprains.Thesetrialsevaluated,inchronologicalorder(1)the effectivenessofNMTforthepreventionofrecurrentanklesprains (n=522),(2)theeffectivenessofNMTagainstbracingwithoutand incombination withNMT(n=384) and(3) theeffectiveness of deliveringNMTthroughaninteractivemobileapplicationagainst writtenexercisesmaterials(n=220)(Table1).11–13_Foreachstudy

ethicalapproval wasobtainedbythemedical ethicscommittee oftheVUUniversityMedicalCenter,Amsterdam,Netherlands.All athletesprovidedwritteninformedconsentpriortoinclusion.For athletesundertheageof18additionalparentalinformedconsent wasprovided.

Allthreetrialsincludedrecreationalathletes(12-70years),who hadsustainedalateralanklesprainwithintwomonthspriorto inclusion(i.e.indexanklesprain).Treatmentoftheindexankle sprainfollowedusualcarewiththehealthcarethattheathleteshad visitedontheirowninitiative.ThesameNMTwasprovidedineach ofthethreetrials11–13_{oncetheathleteenteredthestudy.Allstudies}

hadaprospectivedesignwitha12-monthfollow-up.Arecurrent sprainwasreportedbytheathletethroughmonthlyquestionnaires anddefinedasaself-reportedsuddeninversionoftheindexankle. Forthepurposeofthecurrentstudyweonlyincludedthose ath-leteswhowereallocatedtothestudygroupsthatreceivedtheNMT (Table1).

Baselinedatafromthethreetrialswereusedtodescribe ath-letecharacteristics,i.e.demographics(age,gender,bodymassindex (BMI)),occurrenceof anyanklesprains beforetheindexsprain (‘yes’or‘no’),initialtypeofhealthcaresettingthatathletevisited

forindexsprain(‘none’;‘paramedicalcaresetting:physiotherapist, manualtherapist,exercisetherapist’;‘secondarycaresetting:general surgeon,orthopaedicsurgeon’;‘primarycare setting:general prac-titioner(GP), sportsphysician’; ‘other’), initialtype of healthcare providerthatathletevisitedforindexsprain(GP,sportsphysician, physiotherapist),typeofsports,occurrenceofindexanklesprain duringasportactivity(‘yes’or‘no’)andtheregularuseofabrace and/ortapeduringsportactivity(‘yes,foroneankle’;‘yes,forboth ankles’or‘no’).Datafromtheoneyearfollow-upofeachtrialwere usedtoregistertheoccurrenceofrecurrentsprains(‘yes’or‘no’), timetorecurrentsprain(months)andsportsexposureuntil recur-rentsprain(hours).

In order topool the data of the three individual trials, the variablesthatwerenotlabeledidenticalbetweenthetrials,were redefinedinnewvariables(Table2).Onlyathleteswithavailable dataonbaselineandfollow-upwereusedforthepooledanalyses. Thenumberofathletesfromthereceivedoriginaldataset(n=258) ofHupperetsetal11_{didnotcorrespondwiththenumberof}

ath-letesdescribedinthepublishedarticle(n=256).Thereasonforthe discrepancycouldnotberetrievedandtherefore,thenumberof athletesin theoriginaldatasetwasusedforthepurposeof our study(n=258).

Theprimaryoutcomeofinterestwastheincidenceofrecurrent sprainsduringfollow-up,reportedasthenumberofself-reported recurrent sprains and as the recurrence incidence density per 1000hoursofsportparticipationuntiltheoccurrenceofthe recur-rent sprain. A recurrent sprain was defined as a self-reported suddeninversion of theindex ankle,not necessarily leadingto cessationfromparticipationortime-loss.

Descriptivestatisticswereusedtodescribeathlete characteris-tics,thetypesofinitialhealthcaresettingsandhealthcareproviders fortheindexanklesprain,theincidenceofrecurrentsprainsand timeinmonthstoonsetofthereportedrecurrentsprain,usingthe meanandstandarddeviation(SD)forcontinuousdataand propor-tionsforcategoricaldata.

Potentialdifferencesincharacteristicsbetweenathletesvisiting differenttypesofinitialhealthcaresettingsandbetweenathletes visitingdifferenttypesofinitialhealthcareprovidersfortheirindex anklesprainwereexaminedusingone-wayanalysisofvariance (ANOVA)forcontinuousvariablesandPearsonchi-squaretestfor categorical variables. When a significant difference was found, post-hocBonferronicorrectionwasperformedinordertoadjust formultiplecomparison.

Multilevellogisticregressionanalyseswereperformedto exam-inetheimpactoftheinitialtypesofhealthcaresettingsandinitial typesofhealthcareproviderforalateralanklesprainonthe out-come of NMT for theprevention of recurrent sprains. Data on visitstospecifiedhealthcareproviderfortheindexanklesprain wasonlyavailablein2trials(n=447).12,13_{Thedifferentstudies}

wereusedasclusterlevelinthemultilevelanalyses,11–13_theinitial

typesofhealthcaresettingandproviderswereusedas indepen-dentvariables, and theoccurrenceof a recurrentsprainduring follow-upwasusedasdependentvariable.Analyseswereadjusted

Table2

Definitionsofredefinedvariablesforpoolingdataoftrialsincluded.

Redefinedvariables DefinitionbyHupperetsetal.* DefinitionbyJanssenetal.† DefinitionbyReijenetal.‡ Definitionofvariableinpooled

data Baselinevariables

Previoussprainbeforeindex anklesprain

Previoussprainbefore inclusionanklesprain

Previoussprainbefore inclusionanklesprain

Previoussprainbefore inclusionanklesprain

‘no’

‘yes’or‘no’ ‘yes’or‘no’ ‘no’;‘yesincidental2ormore’

or‘yesfrequent,3ormore’

‘yes’(including‘yesincidental (2orless)’and‘yesfrequent(3 ormore)’)

Regularuseofbraceand/or tapeduringsportactivity

Useofbraceduringsport activity‘yesonbothankles’;‘yes onleftankle’;‘yesonrightankle’ or’no’

Useofbraceortape‘none’, ‘brace’,‘tape’or‘both’

Useofbraceduringsport activity‘yesonbothankles’;‘yes oninjuredankle’;‘yeson non-injuredankle’or’no’

Dataonbraceandtapepooled: ‘no’‘yes,onbothankles’or’yes, ononeankle’(including‘yeson leftankle’;‘yesonrightankle’; ‘yesoninjuredankle’and‘yeson non-injuredankle’)

Useoftapeduringsport activity‘yesonbothankles’;‘yes onleftankle’;‘yesonrightankle’ or’no’

Useoftapeduringsport activity‘yesonbothankles’;‘yes oninjuredankle’;‘yeson non-injuredankle’or’no’ Initialtypeofhealthcare

settingforindexanklesprain

Treatmentforinclusionankle sprain‘nomedicaltreatment’; ‘paramedicaltreatment: physiotherapist,manual therapist,exercisetherapist’; ‘intramuraltreatment:general surgeon,orthopeadicsurgeon, emergencyphysician’or ‘extramuraltreatment:general practitioner,sportsphysician’

Treatmentforinclusionankle sprain‘none’,‘sportmasseur’, ‘trainer’,‘generalpractitioner’, ‘physiotherapist’,‘sports physician’,‘orthopeadic surgeon’,‘generalsurgeon’, ‘emergencyphysician’or‘other’

Treatmentforinclusionankle sprain‘notreatment’,‘sports physician’,‘physiotherapist’, ‘sportmasseur’,general practitioner’,‘sports physiotherapist’or‘other’

‘none’‘paramedicalcaresetting: physiotherapist,manual therapist,exercisetherapist’ ‘secondarcaresetting:general surgeon,orthopeadicsurgeon, emergencyphysician’‘primary caresetting:general

practitioner,sportsphysician’or ‘other’

Initialtypeofhealthcare providerforindexankle sprain

Notavailable Treatmentforinclusionankle

sprain‘none’,‘sportmasseur’, ‘trainer’,‘generalpractitioner’, ‘physiotherapist’,‘sports physician’,‘orthopeadic surgeon’,‘generalsurgeon’, ‘emergencyphysician’or‘other’

Treatmentforinclusionankle sprain‘notreatment’,‘sports physician’,‘physiotherapist’, ‘sportmasseur’,general practitioner’,‘sports physiotherapist’or‘other’

‘generalpractitioner’,‘sports physician’or‘physiotherapist’

*Hupperetsetal.Effectofunsupervisedhomebasedproprioceptivetrainingonrecurrencesofanklesprain:randomisedcontrolledtrial.BMJ.2009;339:b2684.†Janssen etal.Bracingsuperiortoneuromusculartrainingforthepreventionofself-reportedrecurrentanklesprains:athree-armrandomisedcontrolledtrial.BrJSportsMed.2014; 48(16):1235-1239‡VanReijenetal.The ¨Strengthenyourankle ¨programtopreventrecurrentinjuries:Arandomizedcontrolledtrialaimedatlong-termeffectiveness.JSci

MedSport.2017;20(6):549-554

for potentialconfounders(i.e.athlete characteristics) thatwere determinedfromtheliteratureorbasedonaprioriassumptions, includingtheuseofbraceinadditiontoNMTinathletesfromtrial ofJanssenetal12_{(n=120)andsportsexposure(hours)until}

recur-rentsprain.14,15_{Resultsoftheregressionanalyseswerereportedin}

unadjustedandadjustedoddsratios(OR)andtheir95%confidence interval(CI).

Thesignificancelevelwasseton0.05andallanalyseswere per-formedwiththeStatisticalPackagefortheSocialSciencesV.21.0. (SPSS,IBM,Armonk,NY).

3. Results

Atotalof705athleteswhoreceivedNMT,wereincludedinthis study.AthletesfromthetrialofvanReijenetal.13_{weresignificantly}

older(37.8(SD13.4)years)thantheathletesofthetrialofJanssen etal.12_{(34.1(SD13.4)years)andofthetrialofHupperetsetal.}11

(28.6(SD11.8)years)(Table1).Themeanageofourpopulationwas 33.3(SD13.4)years(Table3).Apreviousanklesprainwasreported in70.1%oftheathletesand76%oftheindexanklesprainsoccurred duringasportactivity.

Thedistributionoftheinitialtypeofhealthcaresettingthatthe athletesvisitedfortheirindexsprainispresentedinFig.1.Thetype ofhealthcaresettingwasunknownin8.5%ofthestudypopulation. Sinceonlyoneathletereportedan‘other’healthcaresetting,dataof thissubjectwereexcludedforfurtheranalyses.Alargepartofthe athletes(n=238;37.0%)didnotvisitahealthcareproviderfortheir indexanklesprain,andofthoseathleteswhodid,most(n=257; 39.9%)visitedaparamedicalcaresetting.

Atotalof252(35.7%)recurrentsprainswerereportedduring one-yearfollow-up,withanoverallrecurrenceincidencedensity

Table3

Baselineathletecharacteristics

Totalstudy population n=705 Athletecharacteristics

Age(years),mean(SD) 33.3(13.4)*

Sex(male) 356(50.5)

BMI(kg/m2_{),mean(SD) 23.4(3.1)}†

Previousanklesprainbeforeindexanklesprain 493(70.1)‡

Top3sportactivities§

1.Athletics/running 137(19.4)

2.Soccer 117(16.6)

3.Volleyball 94(13.4)

Regularuseofbraceand/ortapeduringsportactivity|| _231(44.4)¶

Occurrenceofindexanklesprainduringsportactivity 534(76.0)#

Presentedinn(%),unlessotherwisestatedn=number;SD=standarddeviation;BMI =bodymassindex*n=1unknown;†n=2unknown;‡n=2unknown;¶n=185

unknown;#n=60unknown§Inpopulationofn=704,sinceoneathletereported

notparticipatinginsportattimeofinclusion||Includingdataontheregularuseof braceand/ortapeduringsportactivityononeankleandbothankles

of3.98per1000hoursofsportsparticipation(95%CI3.49-4.47). Almost40%oftheathleteswhovisitedparamedicalcaresetting reportedarecurrentsprain(39.7%),withaninjuryincidence den-sityof4.11per1000hoursofsportsparticipation(95%CI3.31-4.91) (Fig.2).Athletesvisitingasecondarycaresettinghadameantime toarecurrentsprainof7.1months(SD3.0),whereasathletes visit-ingothercaresettingsandotherhealthcareprovidershadamean timetoarecurrentsprainvariatingfrom2.0to4.0months.

Multilevellogisticregressionanalysesontheimpactofthe ini-tialtypeofhealthcareproviderontheoutcomeofNMTonankle sprainrecurrences,showedmorerecurrentsprainsafterNMTin

238

257

55

94

1

60

Other treatment

Unkown

Total population n=705

Fig.1.Typeofhealthcaresettingatinitialvisitforindexanklesprain(n).

Fig.2. Numberofrecurrentsprainsduringone-yearfollow-up(%).

Table4

Multilevellogisticregressionanalysesofinitialtypeofhealthcaresettingandhealthcareproviderforindexanklesprainontheoccurrenceofrecurrentsprainduringone-year

follow-up

TotalstudypopulationN=705* Multilevellogisticregressionanalysesoftypeof

healthcaresettingonoccurrenceofrecurrentsprain†

Initialtypeofhealthcare setting

Recurrent sprainduring follow-up

Incidenceofrecurrent sprains(95%CI)per 1000hoursofsport participation Meantimeto recurrentsprainin months(SD) UnadjustedOR (95%CI) p-value AdjustedOR‡_(95% CI) p-value Nohealthcaresetting(n=238) 78(32.8) 3.51(2.73-4.29) 4.0(3.4) 1.03(0.84-1.26) 0.77 0.96(0.77-1.20) 0.74§

Paramedicalcaresetting (physiotherapist,manual therapist,exercisetherapist) (n=257)

102(39.7) 4.11(3.31-4.91) 3.9(3.5) 1.09(0.95-1.26) 0.22 1.28(0.94-1.74) 0.12||

Secondarycaresetting(general surgeon,orthopaedicsurgeon emergencyphysician)(n=55)

12(21.8) 1.99(0.87-3.12) 7.1(3.0) 0.68(0.28-1.66) 0.40 0.87(0.47-1.60) 0.65¶

Primarycaresetting(general practitioner,sportsphysician) (n=94)

32(34.0) 4.79(3.13-6.45) 3.0(2.9) 0.86(0.48-1.55) 0.62 0.82(0.37-1.84) 0.63#

Unknown(n=60) 27(45.0) 7.78(4.85-10.72) 3.2(3.0) - - -

-TotalstudypopulationN=447** Multilevellogisticregressionanalysesoftypeof

healthcareprovideronoccurrenceofrecurrentsprain†

Initialtypeofhealthcare provider

Recurrent sprainduring follow-up

Incidenceofrecurrent sprains(95%CI)per 1000hoursofsport participation Meantimeto recurrentsprainin months(SD) UnadjustedOR (95%CI) p-value AdjustedOR‡(95% CI) p-value Generalpractitioner(n=70) 23(32.9) 5.01(2.96-7.05) 2.0(1.4) 0.84(0.40-1.79) 0.65 0.77(0.08-7.15) 0.81 Physiotherapist(n=118) 50(42.4) 4.43(3.20-5.66) 3.2(3.0) 1.58(1.09-2.28) 0.02 3.15(1.88-5.23) <0.01 Sportsphysician(n=11)†† 4(36.4) 4.77(0.10-9.45) 2.0(0.8) - - -

-Presentedinn(%),unlessotherwisestated.;n=number;SD=standarddeviation;OR=oddsratio;CI=confidenceinterval*Dataofoneathletewhoreported‘othertreatment’ notpresentedintableandnotincludedinmultilevellogisticregressionanalysis†Theoriginofclinicaltrial(Hupperetsetal.,2009;Janssenetal.,2014orVanReijenetal., 2017)wasusedasclusterlevelinthemultilevelanalyses‡AdjustedforallathletecharacteristicsmentionedinTable3,theuseofNMTincombinationwithabracein

athletesfromJanssenetal.(2014)(n=120)andhoursofsportexposureduringfollow-upbeforetheoccurrenceofrecurrentsprain§Referencecategoryisanyothertypeof

care(i.e.paramedical,secondaryorprimary)||Referencecategoryisanyothertypeofcare(i.e.secondaryorprimary)andnotypeofcare.Referencecategoryisanyother

typeofcare(i.e.paramedicalorprimary)andnotypeofcare#Referencecategoryisanyothertypeofcare(i.e.paramedicalorsecondary)andnotypeofcare**Dataon

treatmentbygeneralpractitioner,physiotherapistandsportsphysicianseparatelyonlyavailableindatasetofJanssenetal.(2014)(n=227)andVanReijenetal.(2017)(n=

Table5

Athletecharacteristicsbyinitialtypeofhealthcaresettingandhealthcareproviderforindexanklesprain

Totalstudypopulationn=705* Initialtypeofhealthcare

setting Numberofathletes visitingtypeof healthcaresetting Age(years), mean(SD)

Sex(male) BMI(kg/m2_),

mean(SD)

Previoussprain

beforeindex

anklesprain

Regularuseof braceand/ortape duringsport activity Occurrenceof indexankle sprainduring sportactivity Nohealthcaresetting 238(37.0) 32.3(13.1) 130(54.6) 23.6(3.1) 168(70.6) 64(34.2)† _166(69.7)

Paramedicalcaresetting (physiotherapist,manual therapist,exercisetherapist)

257(39.9) 32.1(13.3)‡ _{128(49.8) 23.0(2.9)}§ _{179(69.6) 88(49.7)}|| _205(79.8)

Secondarycaresetting(general surgeon,orthopaedicsurgeon

emergencyphysician)

55(8.5) 31.5(13.4) 23(41.8) 23.8(4.2) 37(67.3) 30(55.6)¶ 44(80.0)

Primarycaresetting(general practitioner,sportsphysician)

94(14.6) 35.5(12.5) 44(46.8) 23.6(2.8) 62(66.0) 31(48.4)# _80(85.1)

Unknown** 60(8.5) 40.4(14.2) 31(51.7) 23.9(2.8) 46(79.3)†† _18(48.6)‡‡ _38(65.5)§§

p-value|||| _{- 0.13 0.28 0.12 0.85 <0.01}¶¶ _<0.01##

Totalstudypopulationn=447*** Initialtypeofhealthcare

provider Numberofathletes visitingtypeof healthcare provider Age(years), mean(SD)

Sex(male) BMI(kg/m2_),

mean(SD)

Previoussprain

beforeindex

anklesprain

Regularuseof

braceand/ortape

duringsport activity Occurrenceof indexankle sprainduring sportactivity Generalpractitioner 70(15.7) 35.6(12.1) 29(41.4) 23.7(2.7) 48(68.8) 26(52.0)††† _62(88.6) Physiotherapist 118(26.4) 33.9(13.3)‡‡‡ _{59(50.0) 23.4(2.6)}§§§ _{87(73.7) 47(58.8)}¶¶¶ _95(80.5) Sportsphysician 11(2.5) 38.5(12.3) 7(63.6) 22.6(1.9) 6(54.5) 2(100)### _10(90.9) p-value**** - 0.42 0.29 0.34 0.36 0.35 0.28

Presentedinn(%),unlessotherwisestated.n=number;SD=standarddeviation;BMI=bodymassindex*Dataofoneathletewhoreported‘othertreatment’notpresented

intableandnotincludedinone-way-analysisofvarianceandPearsonchi-squaretestanalyses‡n=1unknown;§n=1unknown;†n=51unknown;||n=80unknown;¶

n=1unknown;#n=30unknown;††n=2unknown;‡‡n=23unknown;§§n=2unknown**Groupnotincludedinone-way-analysisofvarianceandPearsonchi-square

testanalyses||||Differencesinathletecharacteristicsbetweentheinitialtypeofhealthcaresettingforindexanklesprainusingone-wayanalysisofvarianceforcontinuous

variablesandPearsonchi-squaretestforcategoricalvariables.

¶¶_{Posthocsignificantdifferencebetween‘none’and‘paramedicalcaresetting’andbetween‘none’and‘’secondarycaresetting’(p<0.05).}

##Posthocsignificantdifferencebetween‘none’and“primarycaresetting’(p<0.05)***Dataontreatmentbygeneralpractitioner,physiotherapistandsportsphysician

separatelyonlyavailableindatasetofJanssenetal.(2014)(n=227)andVanReijenetal.(2017)(n=220)†††n=20unknown;‡‡‡n=1unknown;§§§n=1unknown;¶¶¶n

=38unknown;###n=9unknown****Differencesinathletecharacteristicsbetweentheinitialtypeofhealthcareproviderforindexanklesprainusingone-wayanalysis

ofvarianceforcontinuousvariablesandPearsonchi-squaretestforcategoricalvariables.

athletesthatvisitedaphysiotherapistfortheirinitialindexankle sprainwhencomparedtothosewhohadnotvisiteda physiothera-pist(adjustedOR3.15;95%CI1.88-5.23)(Table4).Nosignificant differences on the occurrence of recurrent sprains were found betweenthedifferenttypesofhealthcaresettingsontheoutcome ofNMT.

Characteristicsofathletesvisitingdifferenttypesofhealthcare settingsanddifferenttypesofhealthcareprovidersfortheir ini-tialindex anklesprainarepresentedin Table5.Therewereno significantdifferencesincharacteristicsbetweenthegroupof ath-letes,withmissingdataontheinitialtype ofhealthcaresetting andthegroupofathleteswithavailabledata.Significantlymore athleteswhovisitedparamedicalorsecondarycaresettinguseda braceand/ortaperegularlyduringsportactivity(49.7%and55.6% respectively),thanthosewhodidnotvisitanyhealthcareprovider (34.2%)(p<0.01).Moreofteninathleteswhovisitedprimarycare setting,theindexanklesprainhadoccurredduringsportsactivities (85.1%)whencomparedtothosewhodidnotvisitanycaresetting (69.7%)(p<0.01).Nofurtherdifferenceswerefoundinother ath-letecharacteristics,norbetweenthedifferenthealthcareproviders GP,physiotherapistandsportsphysician.

4. Discussion

Wepooleddataof705athletesfromthreetrialstoexaminethe impactofinitialtypeofhealthcareforalateralanklesprainonankle sprainrecurrencesduringone-yearfollow-upafterNMT.252 ath-letes(37.5%)reportedarecurrentsprainduringfollow-up,which iscomparablewithpreviousstudiesevaluatingrecurrentsprains duringfollow-up.3,16_{Morerecurrentsprainsoccurredinathletes}

whovisitedaphysiotherapistcomparedtothosewhodidnotvisit aphysiotherapist.Theuseofbracesand/ortapeandtheoccurrence oftheindexsprainduringsportsactivitieswerecharacteristicsthat significantlydifferedbetweeninitialtypeofhealthcaresettings.

Althoughwefounda significantdifferenceintheoccurrence ofrecurrentsprainsbetweenathletesvisitingdifferenthealthcare providersina subsampleofthestudy,wefoundnodifferences intheoccurrenceofrecurrentsprainswhenweclusteredhealth caresettings.However,wedidseeapositivetrendtowardsmore recurrentsprainsinathletesvisitingparamedicalcaresetting,than thosevisitingothertypesofcaresetting.Thismaybeexplainedby theinclusionofphysiotherapistsintheparamedicalcaresetting category.Thefactthatwefoundasignificantdifferencebetween athletesvisitinga physiotherapistandthosewhodidnotvisita physiotherapist,islikelytheresultofselectionbias.Athletesthat visitedaphysiotherapistfortheirindexanklesprainarelikelyto differfromthosevisitingotherhealthcareprovidersintheir char-acteristicsorpreferenceforthetypeofhealthcareprovider.Some oftheunmeasuredvariables,includinginjuryseverityand activ-itylevel,maydifferbetweengroups.Moreover,thedatashowed thatathletesvisitingthephysiotherapistseemtobemorelikelyto havehadaprevioussprainbeforetheirindexanklesprain. There-fore,athletesvisitingthephysiotherapistmayalreadybeatahigher riskofsustainingarecurrentsprainbeforereceivingtreatment.No informationwasavailableonthereasonforchoosingaparticular healthcareproviderinourdatasets.Itcanthereforeonlybe hypoth-esizedthatthetypeofhealthcareprovider,i.e.thephysiotherapist, islikelytoserveasaproxyvariableinourdataset:itmightnotitself directlyexplainourresult,butmightservesinplaceofavariable, i.e.athletecharacteristic,thatwasnotavailableinourstudy.

Nodifferenceswerefoundincharacteristicsbetweenthe ath-letes visiting different healthcare providers, but we did find differences in characteristicsof athletesvisiting different types ofhealthcaresettings.Athletesvisitingparamedicalcaresetting andvisitingsecondarycaresettingusedsignificantlymorebraces and/ortaperegularlyduringsportactivitiesthanathleteswhodid notseekanycaresetting.Potentially,theuseofabraceortape wasappliedbytheparamedicalorsecondaryphysician,aspartof therehabiliationoftheindexanklesprain.However,itmayalsobe hypothesizedthattheseathletes,whochosetovisitaparamedical orsecondarycaresetting,perhapssustainedamoresevereinjury orparticipatedinahigh-risksport,forwhichabraceortapewas preferred.Wehadnoinformationonwhyandwhentheathletes startedtouseabraceand/ortape.Intheliterature,littleisknown aboutcharacteristicsofpatientsattendingdifferenthealthcare set-tingsandproviders.Onlyonerecentstudydescribedcharacteristics ofpatientsattendingtheemergencydepartmentintheUK.17This studyfoundthatmoreyoungermen(14-37yearsold)thanwomen ofthesameagevisitedtheERwithanankleinjury.However,they madenocomparisonwithotherhealthcaresettings.Therefore,our findingsaredifficulttocomparewithexistingliterature.

Toourknowledge,thisisthefirststudyevaluatingtheimpactof initialtypeofhealthcareforalateralanklesprainontheoutcome ofNMTonrecurrentsprainriskanddescribingcharacteristicsof athletesattendingdifferenttypesofhealthcare.Astrengthisthat allincludedathletesreceivedthesameNMTandhadone-yearof follow-up.Wethinkthisperiodisappropriatetodetectrcurrent sprainsaftertheindexanklesprain,astheriskforarecurrentsprain isincreasedduringthefirstyearafteraninitialsprain.18,19_Also,we

believethatwehadinformationonarepresentativesetoftypesof healthcaresettingsandproviders,inbothprimaryandsecondary care,thatareinvolvedinthetreatmentofananklesprain.9,10

Nevertheless,somelimitations should beaddressed.For the purposeofourstudywefocusedontheinitialtypeofhealthcare settingand provider andwe didnot focusonthespecific type oftreatmentthatwasgivenbytheinvolvedhealthcareprovider. Differenttypes of healthcare providerscan applysimilar treat-mentmodalities(e.g.bothphysiotherapistandsportsphysiciancan applyatypeofexercisetherapy)andalso,onetypeofhealthcare providercanapplydifferenttypesoftreatmentmodalities(e.g.a GPcanapplyfunctionalorexercisetherapy).Itwould be inter-estingtoevaluateiftheoutcomeofthetypeoftreatmentdiffers betweendifferenttypesof healthcare provider.Another limita-tion,isthefactthatdataontheseparatehealthcareproviderswere onlyavailableintwotrials.12,13_{Asaconsequence,wepresented}

asmallerproportionofathletesvisiting thedifferenthealthcare providers(GP,physiotherapist,sportsphysician).Thus,wemight havefoundasmallerimpactofthehealthcareproviderfora lat-eralanklesprainontheoutcomeafterNMTonrecurrentsprains thanthetrueimpactandwemightnotdetectedexistingdifferences betweencharacteristicsofathletesvisitingtheseparatehealthcare providers.Moreover,for8.5%ofthesubjectsthetypeof health-caresettingwasunknownandthesesubjectswerethereforenot includedinthemultilevelregressionanalyses.Lastly,onlyathletes wereincludedinourstudyanditisknownthatsportsparticipants areatahigherriskforrecurrentanklesprainsthannon-athletes.9,16

Therefore,ourresultsarenotgeneralizabletoageneralpopulation whichincludesnon-athletes.However,sinceanklesprainsarea verycommoninjuryamongsportingpopulations,17,20_afirst

eval-uationoftheimpactofinitialhealthcareforalateralanklesprain inthispopulationwasausefulapproach.

Fromourstudyit seemsthatspecificeathletecharacteristics areassociatedwithavisittospecifichealthcareproviders. Conse-quently,thisseemstoinfluencetheoutcomeofNMTasevaluated inourstudy.Thisshouldaddtowardsawarenessof physiothera-piststhattheyarelikelytoseeaspecificpopulationthatmayhave

apoorerprognosis,i.e.havehigherriskonsustainingarecurrent sprain.We thinkthatourstudygivesinsightinthepathwayof healthcareconsumptionofahtleteswithananklesprainandits impactonrecurrentsprainsafterNMT.Furtherresearchwithlarger samplesizesisnecesarrytoevaluatetheimpactofinitialhealthcare ontheprognosisofanklesprains.

5. Conclusion

AmongathletesthatweretreatedwithNMT,theinitialtypeof healthcareforananklesprain,seemstoimpacttheoccurrenceof recurrentsprains.Accordingtoourstudy,physiotherapistsmaybe visitedbyathleteswithapoorerprognosisandshouldtakethisinto accountwhendeterminingrehabilitationstrategies.Thisseemsto beduetodifferentcharacteristicsofathletesvisitingdifferenttypes ofhealthcare.

Acknowledgement

This study was supported by the Netherlands Organisation forHealth Research andDevelopment (ZonMW) [grantnumber 4201.1007].

AppendixA. Supplementarydata

Supplementarymaterialrelatedtothisarticlecanbefound,in theonlineversion,atdoi:https://doi.org/10.1016/j.jsams.2020.03. 010.

References

1.WatermanBR, Owens BD, DaveyS,Zacchilli MA, BelmontJrPJ. The epi-demiologyofanklesprainsintheUnitedStates.JBoneJointSurgAm2010; 92(13):2279–2284.

2.ShahS,ThomasAC,NooneJM,BlanchetteCM,WikstromEA.IncidenceandCost ofAnkleSprainsinUnitedStatesEmergencyDepartments.SportsHealth2016; 8(6):547–552.

3.vanRijnRM,vanOsAG,BernsenRM,LuijsterburgPA,KoesBW,Bierma-Zeinstra SM.Whatistheclinicalcourseofacuteanklesprains?Asystematicliterature review.AmJMed2008;121(4),324–331e326.

4.MailuhuAKE,OeiEHG,vanPutte-KatierNetal.Clinicalandradiological predic-torsforpersistentcomplaintsfiveyearsafteralateralanklesprain:Along-term follow-upstudyinprimarycare.JSciMedSport2018;21(3):250–256.

5.KemlerE,ThijsKM,BadenbroekI,vandePortIG,HoesAW,BackxFJ. Long-termprognosisofacutelateralankleligamentoussprains:highincidenceof recurrencesandresidualsymptoms.FamPract2016;33(6):596–600.

6.KonradsenL,BechL,EhrenbjergM,NickelsenT.Sevenyearsfollow-upafter ankleinversiontrauma.ScandJMedSciSports2002;12(3):129–135.

7.vanOchtenJM,B-ZS,BindelsPJE,vanMiddelkoopM.ImpactofaLateralAnkle SpraininGeneralPractice:ComparisonBetweenPatientsWithandWithout PersistentComplaintsAfter6-12Months.JSMFootAnkle2016;1(3):1013.

8.LetselInformatieSysteem.V.Letsellastmodel2017,VeiligheidNL,2017.

9.VuurbergG,HoorntjeA,WinkLMetal.Diagnosis,treatmentandpreventionof anklesprains:updateofanevidence-basedclinicalguideline.BrJSportsMed 2018;52(15):956.

10.DohertyC,BleakleyC,DelahuntE,HoldenS.Treatmentandpreventionofacute andrecurrentanklesprain:anoverviewofsystematicreviewswith meta-analysis.BrJSportsMed2017;51(2):113–125.

11.HupperetsMD,VerhagenEA,vanMechelenW.Effectofunsupervisedhome basedproprioceptivetrainingonrecurrencesofanklesprain:randomised con-trolledtrial.BMJ2009;339:b2684.

12.JanssenKW,vanMechelenW,VerhagenEA.Bracingsuperiortoneuromuscular trainingforthepreventionofself-reportedrecurrentanklesprains:athree-arm randomisedcontrolledtrial.BrJSportsMed2014;48(16):1235–1239.

13.VanReijenM, VriendI, Zuidema V,vanMechelen W,VerhagenEA.The ¨Strengthen your ankle ¨program to prevent recurrent injuries: A random-izedcontrolledtrialaimedatlong-termeffectiveness.JSciMedSport2017; 20(6):549–554.

14.PourkazemiF,HillerCE,RaymondJ,BlackD,NightingaleEJ,RefshaugeKM. Pre-dictorsofrecurrentsprainsafteranindexlateralanklesprain:alongitudinal study.Physiotherapy2017.

15.ThompsonJY,ByrneC,WilliamsMA,KeeneDJ,SchlusselMM,LambSE. Prog-nosticfactorsforrecoveryfollowing acutelateralankleligamentsprain:a systematicreview.BMCMusculoskeletDisord2017;18(1):421.

16.AttenboroughAS,HillerCE,SmithRM,StuelckenM,GreeneA,SinclairPJ.Chronic ankleinstabilityinsportingpopulations.SportsMed2014;44(11):1545–1556.

17.AlBimaniSA,GatesLS,WarnerM,EwingsS,CrouchR,BowenC.Characteristics ofpatientswithanklesprainpresentingtoanemergencydepartmentinthe southofEngland(UK):Aseven-monthreview.IntEmergNurs2018.

18.VerhagenEA,VanderBeekAJ,BouterLM,BahrRM,VanMechelenW.Aone seasonprospectivecohortstudyofvolleyballinjuries.BrJSportsMed2004; 38(4):477–481.

19.BahrR,BahrIA.Incidenceofacutevolleyballinjuries:aprospectivecohort studyofinjurymechanismsandriskfactors.ScandJMedSciSports1997;7(3): 166–171.

20.DohertyC,DelahuntE,CaulfieldB,HertelJ,RyanJ,BleakleyC.Theincidence andprevalenceofanklespraininjury:asystematicreviewandmeta-analysisof prospectiveepidemiologicalstudies.SportsMed2014;44(1):123–140.