Development and proof of concept of a blended physiotherapeutic intervention for patients with non-specific low back pain

(1)

Amsterdam University of Applied Sciences

Development and proof of concept of a blended physiotherapeutic intervention

for patients with non-specific low back pain

Kloek, C.J.J.; van Tilburg, M.L.; Staal, Bart ; Veenhof, C.; Bossen, D.

DOI

10.1016/j.physio.2018.12.006

Publication date

2019

Document Version

Final published version

Published in

Physiotherapy

License

CC BY-NC-ND

Link to publication

Citation for published version (APA):

Kloek, C. J. J., van Tilburg, M. L., Staal, B., Veenhof, C., & Bossen, D. (2019). Development

and proof of concept of a blended physiotherapeutic intervention for patients with non-specific

low back pain. Physiotherapy, 105(4), 483-491. https://doi.org/10.1016/j.physio.2018.12.006

General rights

It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations

If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please contact the library:

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied Sciences), Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

(2)

Development

and

proof

of

concept

of

a

blended

physiotherapeutic

intervention

for

patients

with

non-specific

low

back

pain

C.J.J.

Kloek

a,b,c,∗

,

M.L.

van

Tilburg

a,c

,

J.B.

Staal

d,e

,

C. Veenhof

a,c

,

D. Bossen

f

a_Department_of_{Rehabilitation,}_{Physiotherapy}_Science_and_Sports,_Brain_Centre_Rudolf_Magnus,_University_Medical_Centre Utrecht,Utrecht,TheNetherlands

b_Netherlands_Institute_for_Health_Services_Research,_Utrecht,_The_Netherlands

c_Research_Group_Innovation_of_Human_Movement_Care,_HU_University_of_Applied_Sciences,_Utrecht,_The_Netherlands d_Research_Group_{Musculoskeletal}_{Rehabilitation,}_HAN_University_of_Applied_Sciences,_Nijmegen,_The_Netherlands e_Radboud_Institute_of_Health_Sciences,_Radboud_University_Medical_Centre,_IQ_Healthcare,_Nijmegen,_The_Netherlands f_ACHIEVE_Centre_of_Expertise,_Faculty_of_Health,_Amsterdam_University_of_Applied_Sciences,_Amsterdam,_The_Netherlands

Abstract

Objective Todevelopablendedphysiotherapeuticinterventionforpatientswithnon-specificlowbackpain(e-ExerciseLBP)andevaluate

itsproofofconcept.

Design Focusgroupswithpatients,physiotherapists,andeHealthandLBPexpertswereconductedtoinvestigatevaluesaccordingtothe

developmentofe-ExerciseLBP.Proofofconceptwasevaluatedinamulticentrestudy.

Setting Dutchprimarycarephysiotherapypractices(n=21therapists).

Participants Adultswithnon-specificLBP(n=41).

Intervention e-ExerciseLBPwasdevelopedbasedonclinicalLBPguidelinesandthefocusgroups,usingtheCenterforeHealthResearch

Roadmap.Face-to-facephysiotherapysessionswereintegratedwithawebapplicationconsistingof12informationlessons,video-supported exercisesandaphysicalactivitymodulewiththeoptiontograduallyincreaseindividuals’levelofphysicalactivity.Theinterventioncould betailoredtopatients’riskofpersistentdisablingLBP,accordingtotheSTarTBackScreeningTool.

Mainoutcomemeasures Functionaldisability,pain,physicalactivity,sedentarybehaviourandfear-avoidancebeliefs,measuredatbaseline

and12weeks.

Results After12 weeks,improvementswerefoundinfunctionaldisability[QuebecBackPainDisabilityScale:meandifference(MD)

−12.2/100;95%confidenceinterval(CI)8.3to16.1],pain(NumericPainRatingScale:MD−2.8/10;95%CI2.1to3.6),subjectivephysical activity(ShortQuestionnairetoAssessHealthEnhancingPhysicalActivity:MD11.5minutes/day;95%CI−47.8to24.8)andobjective sedentarybehaviour(ActiGraph:MD−23.0minutes/day;95%CI−8.9to55.0).Smallimprovementswerefoundinobjectivephysicalactivity andfear-avoidancebeliefs.Theoptiontograduallyincreasephysicalactivitywasactivatedforsixpatients(15%).Onaverage,patientsreceived sevenface-to-facesessionsalongsidethewebapplication.

Conclusions Theresultsofthisstudyprovidethefirstindicationoftheeffectivenessofe-ExerciseLBP,particularlyfordisabilityandpain

amongpatientswithLBP.Futurestudieswillfocusonend-userexperiencesand(cost-)effectiveness.

Keywords:Lowbackpain;Physiotherapy;e-Health;Telemedicine

∗_{Corresponding}_author_at:_HU_University_of_Applied_Sciences,_Utrecht,

TheNetherlands.

E-mailaddress:corelien.kloek@hu.nl(C.J.J.Kloek).

Introduction

Inmostcountries,lowbackpain(LBP)istheleadingcause ofdisability[1].LBPcontributestohighdirectcostswithin

https://doi.org/10.1016/j.physio.2018.12.006

(3)

484 C.J.J.Kloeketal./Physiotherapy105(2019)483–491

healthcare,aswellashighindirectcostsintermsoflossof productivity[2].LBPcanbecausedbyaspecificpathology ortrauma;however,anunderlyingdiseaseisabsentin90%of cases[3].Theclinicalcourseofthisso-called‘non-specific LBP’varies; somepeoplerecoverwithin acoupleofdays or weeks,andotherpeopleexperiencepersistent disabling symptomsleadingtochronicLBP[4–6].IntheNetherlands, backpainisoneofthemostcommonreasonsforvisitinga physiotherapist[7].

ClinicalLBPguidelinesrecommendthatphysiotherapists, aswellasothercaregivers,shouldidentifypatients’riskof persistent disabling symptoms at an early stageand strat-ifytreatmenttoindividualneeds[8–11].Anexampleofan instrumenttoassessanindividual’sriskofchronicityisthe STarTBackScreeningTool[10].Generally,patientsatlow riskbenefitfromeducationandgeneralexercise recommen-dations.Patientsatmediumorhighriskneedapersonalised and supervised exercise programme, in combination with cognitivebehaviouralcomponentsifnecessary[3,8,12,13].A crucialpre-conditionfortheeffectivenessofexercisetherapy ingeneralispatientcompliancewithexerciseprescriptions [14,15]. Stimulation of exercise compliance is one of the greatest challenges for physiotherapists, as 45%–70% of patientsdonot(completely)followtheirexercise recommen-dations[16,17].

Online applications, such as websites and apps, pro-videnewsolutionstostimulatepatients’abilitytomanage theirconditionindependently,andincreaseexercise compli-ancebetweenface-to-facesessions[18].Theintegrationof supervised physiotherapy withaweb application is called ‘blendedphysiotherapeuticcare’[19].However,asthisisa newfield,blendedphysiotherapeuticinterventionsforLBP remain understudied [20]. Recently, the authors’ research groupdevelopedandevaluatedablendedphysiotherapeutic interventionforpatientswithosteoarthritisofthehipand/or knee (e-Exercise Osteoarthritis) [21,22]. Patients treated withe-ExerciseOsteoarthritisexperiencedimprovementsin physicalfunctioning,pain,tiredness,qualityoflifeand self-efficacy[23]. Moreover,patients were highly positiveand satisfied with the availability of information and assign-ments regardless of time or place [24]. The treatment of non-specificLBPissimilarinnaturetoosteoarthritisofthe hipand/orknee,ascoretreatmentelementsincludephysical activity,exerciseandinformation.Assuch,itwasanticipated thattheintegrationofface-to-facecarewithonlinesupport wouldalsobeapplicableinpatientswithnon-specificLBP. Therefore,thisstudyaimedto:(1)developablended physio-therapeuticinterventionthatmatchesthevaluesofend-users; and (2) investigate the proof of concept of e-Exercise in patientswithnon-specificLBP.

Methods

The Center for eHealth Research (CeHRes) Roadmap – a five-step development, evaluationand implementation

approach – was used for development of e-Exercise LBP (AppendixA,seeonlinesupplementarymaterial)[25].The model is based on the principles of participatory design, which means that e-Exercise LBP was developed col-laboratively with physiotherapists, patients, developers, a commercialeHealthentrepreneurandresearchers[26].The first threestepsof theCeHResRoadmapwerefollowedin thedevelopmentofe-ExerciseLBP:contextualinquiry,value specificationanddesign.Proofofconceptofthee-Exercise LBPprototypewastestedinafeasibilitystudy.Theresults of thisstudyandtheexperiences ofend-userswillbeused toimproveandfurtherevaluatee-ExerciseLBPinafuture phase.

Steps1and2.Contextualinquiryandvaluespeciﬁcation

Theaimofthecontextualinquirywastoidentifythehealth problem.First,nationalandinternationalclinicalLBP guide-lineswerestudied.Second,theresultsandexperiencesfrom the previous study on e-ExerciseOsteoarthritis were used [21–23].

Next, three focus groups were conducted: one with patientswithLBP(recruitedinaphysiotherapeutictraining group ata primary carephysiotherapy practice),onewith physiotherapistsandonewithexpertsinthefieldofeHealth (i.e. physiotherapists with experience in blended therapy, researchersandeHealthentrepreneurs).Theprimaryaimof thefocusgroupswastoidentifythevaluesandrequirements ofend-usersinconsiderationwithe-ExerciseLBP.Allfocus groups lastedfor 60minutes. Atopic guide (AppendixB, seeonlinesupplementarymaterial)wasusedtodiscussthe following subjects: the content of usualphysiotherapy for patientswithnon-specificLBP,problemsincurrent physio-therapyfor LBP,interventionrequirementsofstakeholders andend-users, andintegrationof online carewithin phys-iotherapeutic care. All focus groups were audio-recorded andtranscribed.Twoauthors(CKandMvT)independently identifiedthemes,codedthemintomeaningfulsections,and categorisedthemintogeneralthemes.Codesandthemeswere discussedbetweenCKandMvTuntilconsensuswasreached.

Step3.Design

During the design phase, the researchers cooperated withHCHealth,acommercialeHealthentrepreneur,which supplied an online physiotherapy platform tooffer video-supportedexercisesandinformationtopatients.Theaimof thiscollaborationwastofacilitatealastingimplementation of e-Exerciseinphysiotherapeuticcare.Theresearchteam provided inputforthe contentof the webapplication,and HCHealthintegratedthiscontentinto their platform. Con-tent ofthe webapplicationwasbasedonSteps 1and2 as describedaboveandaDutchself-helpbookforpatientswith backpain[27].

(4)

Feasibilitystudy

Proofofconceptoftheprototypewastestedina multi-centrefeasibilitystudy.

Procedureandparticipants

Fifty physiotherapists working in primary care were invited toparticipate in the study. Most of them had par-ticipatedinthepreviouse-ExerciseOsteoarthritisstudy[22], andotherswererecruitedfromtheauthors’professional net-work.Physiotherapistswereeligibleiftheytreatedatleast 10 patients with non-specific LBP per year. In total, 21 physiotherapistsparticipated.Allphysiotherapistsattendeda half-dayface-to-facegrouptrainingsessionaboutthestudy proceduresandthe e-ExerciseLBP intervention,andwere providedfreeaccess totheonline partof the intervention. Astheaimofthisstudywastoincludeatleast40patients, physiotherapistswereaskedtorecruitatleasttwo patients withLBP.Allpatientsreceivedtheintervention.The inclu-sioncriteria forpatientswere:(1)age18–65yearsand(2) non-specificLBPaccordingtothephysiotherapists. Exclu-sioncriteriawere:(1)contraindicationsforphysicalactivity withoutsupervisionaccordingtothePhysicalActivity Readi-nessQuestionnaire[28];(2)receivedphysiotherapyforLBP inthelast6months;(3)noaccesstotheinternet;or(4)no mastery of the Dutchlanguage.Physiotherapists informed eligible patients about the study, and provided interested patientswithaninformationletter.Allpatientswereaskedto signaninformedconsentform.Patients’contactinformation wasstoredseparatelyfromtheresearchoutcomes.

Outcomemeasures

FunctionaldisabilitywasassessedusingtheQuebecBack PainDisabilityScale(QBPDS),whichconsistsof20items aboutthedifficultyofperformingdailyactivities.Thetotal scorerangesfrom0to100(0=noproblemsindailyactivities; 100=maximallydisabledindailyactivities)[29].

Painwasassessedusing theNumericPainRatingScale (NPRS)(0=nopain;10=worstpossiblepain)[30].

Globaleffectwasassessedusingafive-pointLikertscale aboutthedegreeofchangeinLBPsymptomssincebaseline measurements(1=muchworse;5=muchbetter).

Physicalactivitywasmeasuredobjectivelywiththe Acti-Graph GR3X tri-axial accelerometer [31]. Patients were asked to wear the monitor for five unspecified consecu-tivedays, andtocompleteashortactivity diaryabout the times when the accelerometer was taken on and off and reasonsfordoing so.The accelerometersanddiarieswere returnedbypost.Freedson’sthresholdswere usedfordata analysis[32]:0–99countsperminuteforsedentary activi-ties,100–1951countsperminuteforlightphysicalactivity, 1952–5724countsperminuteformoderatephysicalactivity, 5725–9498counts per minutefor vigorousphysical activ-ity,and≥9499countsperminuteforveryvigorousphysical activity.Datawererecordedat1-minuteintervals.The aver-agenumberofminutesofmoderateandvigorousactivityper

daywasusedforanalysis.Next,physicalactivitywas mea-sured subjectively withthe Short Questionnaire toAssess HealthEnhancingPhysicalActivity[33].Thisquestionnaire measureslight,moderateandvigorousphysicalactivityover anormalweek.Foranalysis,theaveragenumberofminutes ofmoderateandvigorousactivityperdaywasused.

Pain-relatedfearwasassessedusingtheFear-Avoidance Beliefs Questionnaire, which consists of items related to physical activity (range 0–24) and work (range 0–42). A higher score indicates greater fear and avoidance beliefs abouthowphysicalactivityandworknegativelyaffectLBP [34].

Otheroutcomemeasures

Physiotherapists were asked to complete a registration formfor eachpatientabout thesubgroupof LBP(i.e.low, medium or highrisk for persistent disablingsymptoms as assessedusing the STarT BackScreeningTool), the num-berofface-to-facesessions,andtheinterventionsusedinthe physiotherapysessions.

Onlineadherence(i.e.numberoflog-insandnumberof activated graded activity modules)was assessed based on objectiveweb-usagedataprovidedbyHCHealth.

Patientcharacteristics(i.e.age,sex,height,weight, edu-cationallevelandcomorbidities)wereassessedaspartofthe baselinequestionnaire.

Analysis

Descriptive statistics wereused todescribe the general characteristics of the study population, patients’ website usage,globaleffect,physiotherapists’recruitmentrates,and the content andnumberof physiotherapy sessions.As the studyhadnocontrolgroupandwasnotpoweredsufficiently to test for significance, descriptive statistics were used to describeproofofconceptintermsofmeanvaluesandmean differences (MD)for disability, pain,physicalactivity and pain-relatedfear.MeanvaluesandMDbetweenbaselineand follow-upwere providedfor completecases.Results were described for the entire group and for each risk group as determinedwiththeSTarTBackScreeningTool.All anal-yses wereperformed using SPSS Version24 (IBM Corp., Armonk,NY,USA).

Results

Steps1and2.Contextualinquiryandvaluespeciﬁcation

Values and requirements that emerged from the focus groupswithpatients,physiotherapists andexperts are pro-videdinTable1.Patientsinthefocusgroupindicatedthat physicalactivityandexerciseswereanimportantpartofthe treatment of LBP. Onepatientstated:‘Walking isthe best medicationformybackpain’.Furthermore,patientsexpected thataplatformwithvideo-supportedexercise

(5)

recommenda-486 C.J.J.Kloeketal./Physiotherapy105(2019)483–491

Table1

Resultsoffocusgroupswithpatients,physiotherapistsandexpertstodeterminevaluesandrequirementsofe-ExerciseLBP. Theme Patients(n=4) Physiotherapists(n=5) Experts(n=4)

Averageage:53years Averageage:39years Averageage:40years Male/female:2/2 Male/female:2/3 Male/female:4/0 Physicalactivity Physicalactivitypromotion

shouldbeanimportantpartofthe treatmentofLBP

Gradedactivityisnotapplicableforevery patientwithLBP.Gradedactivityprinciples shouldbeanoptionalmodule

Physicalactivitypromotionshould beanimportantpartofthetreatment ofLBP.Gradedactivityprinciples shouldbeanoptionalmodule Strengthandstability

exercises

Video-supportedexercise recommendationsaresupposed tosupportexercisecompliance

Thespecifictypeofexercisesdependsonthe individualpatient.Apre-selectionofexercises pertreatmentobjective(e.g.mobilisation, stabilisation)wouldsavetime

Exercisesareimportant.Evidence aboutmoreorlesseffectiveexercises isabsent.Ideally,thephysiotherapist selectsspecificexercisesperpatient Informationmodules Thereisaneedfortrustworthy

information

Allinformationthemesarerelevantforallrisk groupsofpatients.Ideally,the

physiotherapistsselectwhichinformation modulesarepresentedatwhichtime

Physicalactivityshouldbethe cornerstoneofallinformation modules.Videoswithpatient experiencesshouldalsobeincluded Stratificationofcare Notapplicable RecommendationforSTaRTBackScreening

tool

RecommendationforSTaRTBack Screeningtool

Stratificationof blendedcare

Notapplicable Anextratoolforstratificationofblendedcare takestoomuchtime

Anextratoolforstratificationof blendedcaretakestoomuchtime LBP,lowbackpain.

tionswouldsupportthemtoaccomplishexercisesathome, as they often forgot how to execute specific exercises, as well as the number of repetitions.Overall, patients had a positiveattitudetowardsblendedcare.Physiotherapists sug-gestedtheintegrationofagradedactivitymodule,similarto e-ExerciseOsteoarthritis,butonlyas anoptionalfunction. Agradedactivitymodulecanbeusedtograduallyincrease physicalactivity,inaccordancewiththeprinciplesofgraded activity.Asonephysiotherapistexplained:‘gradedactivity isindicatedonlyinaminorityofLBPpatients,namelythose whoavoidphysical activity duetoLBP’. Physiotherapists andexpertshadnopreferencesforspecificstrength, mobil-ityorstabilityexercises.Accordingtolimitedavailabletime, apre-selectedsetofexercisesthatcouldbetailoredto individ-ualneedswouldbepreferable.Physiotherapistsandexperts recommendedthat thesameinformationcontentshouldbe availabletoallthreesubgroupsofpatientswithLBP.Asone expertillustrated:‘AsingleeventofLBPisariskfactorfor recurrenceofLBP.Allpatientsneedtoknowwhichfactors arerelatedtoLBPinordertopreventrecurrence’.

Step3.Design

Thee-ExerciseLBPinterventionisanintegrationof face-to-facephysiotherapysessionswithawebapplication.Fig.1 providesanoverviewoftheintervention.

e-ExerciseLBPprotocol

Duringthefirstface-to-facesession,thephysiotherapist identifiedeachpatient’srisk(i.e.low,mediumor high)for persistentdisablingsymptomsusingtheSTarTBack Screen-ingTool[11].Basedonpatients’physicalcapacityandrisk profile,physiotherapistshadtheability totailore-Exercise LBPtomeettheirspecificneeds[10].Asrecommendedby

theexperts,patientsatlowriskforpersistentdisabling symp-tomshadfourface-to-facephysiotherapysessions,patientsat mediumriskhad12sessionsandpatientsathighriskhad20 sessions.However,accordingtothephysiotherapists’clinical knowledge,physiotherapistswere freetodeviatefromthis protocol. The application consistedof three modules. The firstmoduleconsistedof12weeklylessons(textandvideo) aboutthe aetiologyofLBP, physicalactivity, patient expe-riences, painmanagement andpsychosocial factorsrelated to LBP. The second module consistedof video-instructed exercises. Physiotherapists could choose the pre-selected exercises or create an individualised exerciseprogramme. Thethirdmoduleconsistedofphysicalactivity recommen-dations.Basedona3-daybaselinetestwithinthismodule, thephysiotherapistscouldcalculatethepatients’currentlevel ofphysicalactivity.Ifpatientswereinsufficientlyphysically active, the physiotherapistcould activatethe graded activ-ity functionality andset agoal toreach within 12 weeks. This graded activity functionality with tailored feedback has been studied previously in two osteoarthritis studies [23,35].Patientsreceivedweeklyreminderstovisittheweb application.Theonlinemoduleslastedfor12weeks. Physio-therapistscouldwatchpatients’usageofthewebapplication, monitor evaluatedassignments, select othertypesof exer-cises,andcommunicatewithpatientsusingamessenger ser-vice.Aprintscreenoftheonlinee-ExerciseLBPapplication isgiveninAppendixC(seeonlinesupplementarymaterial).

Feasibilitystudy

Participants

Intotal,46eligiblepatientswererecruitedbetweenMay andOctober2016.Ofthesepatients,41signedinformed

(6)

con-Fig.1.Overviewoftheblendedinterventione-ExerciseLBP.

sentformsandcompletedthebaselinequestionnaire.After12 weeks,37patients(90%)completedthefollow-up question-naire.Nodifferencesinbaselinecharacteristicswerefound betweenrespondersandnon-responders.Accelerometerdata wereavailablefrom30patients(73%)atbaseline,from22 patients (54%) at follow-up, andfrom 20 patients at both baselineandfollow-up(45%).Fig.2providesanoverview of thestudy process.Patient demographicsat baselineare presentedinTable2.

Physiotherapyandonlineadherence

Overall,theaveragenumberofface-to-facesessionswas 6.8[standarddeviation(SD)4.0].Face-to-facesessionsmost frequentlyconsistedofprovidinginformation,exercisesand active/passivemobilisation.Theaveragetotalnumberof log-inswas28(SD27)over12weeks.Sixpatients(15%)received thegradedactivitymodule;mostofthemwereclassifiedas

lowrisk. Table3 providesanoverviewof thenumberand contentoftheface-to-facesessions.Thenumberof face-to-facesessionsdifferedbetweenriskgroups[lowrisk:5.7(SD 2.9); mediumrisk: 6.5(SD 4.9);highrisk:9.6 (SD 3.2)], but the content of these sessions and website usage were comparable(Table3).

Proofofconcept

Levels of functional disability, physical activity, fear avoidance,painatbaselineandat12-week follow-up,and global effect are presented in Table 4. An improvement was found in functional disability [MD −12.2/100; 95% confidence interval (CI) 8.3 to 16.1], pain (MD −2.8/10; 95% CI 2.1 to 3.6), subjective physical activity (MD 11.5minute/day; 95% CI −47.8 to24.8), objective physi-calactivity(3.1minute/day;95%CI−8.8to2.7),objective sedentary behaviour(MD−23.0min/day; 95%CI−8.9to

(7)

Fig.2.Flowchart.

Table2

Patientdemographics,entiregroupandsubgroups.

Totalgroup Lowrisk Mediumrisk Highrisk Numberofrespondents n=41 n=18 n=15 n=8 Gender,n(%)

Male 18(44) 10(56) 7(47) 1(13)

Female 23(56) 8(44) 8(53) 7(88)

Ageinyears,mean(SD) 44.3(10.4) 44.9(10.5) 42.0(12.0) 47.5(5.8) BMI,mean(SD) 25.9(4.7) 25.7(3.8) 27.0(6.4) 24.4(2.8) Educationlevel,n(%) Low 4(10) 1(6) 1(7) 2(25) Middle 17(42) 9(50) 5(33) 3(38) High 20(49) 8(44) 9(60) 3(38) Diseaseduration,n(%) 0to6weeks 5(12) 2(11) 2(13) 1(13) 6to12weeks 6(15) 3(17) 2(13) 1(13) 12weeksto12months 9(22) 4(22) 4(27) 1(13) >12months 21(51) 9(50) 7(47) 5(63) Comorbidities,n(%) None 18(44) 8(44) 7(47) 3(38) 1 15(37) 7(39) 6(40) 2(25) >1 8(20) 3(17) 2(13) 3(38)

(8)

Table3

Descriptionofface-to-facesessionsandusageofthewebapplication.

Overall(n=41) Lowrisk(n=18) Mediumrisk(n=15) Highrisk(n=8) Numberofface-to-facesessions,mean(SD) 6.8(4.0) 5.7(2.9) 6.5(4.9) 9.6(3.2) Mostfrequentcontentofface-to-face

sessions,n(%)

Information:97% Information:94% Information:100% Information:100% Exercises:75% Exercises:63% Exercises:85% Exercises:86% Mobilisation:75% Mobilisation:69% Mobilisation:70% Mobilisation:86% Totalnumberoflog-ins,mean(SD) 28(27) 26(25) 27(28) 37(32)

Numberofweekswith≥1log-in,mean(SD) 7.3(4) 7.4(4) 7.0(4) 7.5(4) NumberofpatientswithGAmodule,n(%) 6(15) 4(22) 1(7) 1(13) SD,standarddeviation;GA,gradedactivity.

Table4

Proofofconceptofe-ExerciseLBP(baseline,12weeksandmeandifference). Descriptivestatistics

Outcomemeasure Baseline,n=41,mean(SD) 12weeks,mean(SD) Meandifference(95%CI) Functionaldisability(0to100),n=37 33.4(15.7) 21.2(17.2) −12.2(8.3to16.1) VAS–pain(0to10),n=37 6.1(1.9) 3.3(2.4) −2.8(2.1to3.6) Physicalactivity,subjective(minutes/day),n=37 125.2(93.5) 136.6(132.5) 11.5(−47.8to24.8) Physicalactivity,objective,moderateandvigorous

(minutes/day),n=20

30.7(21.0) 33.8(17.9) 3.1(−8.8to2.7) Sedentarybehaviour(minutes/day),n=20 490.7(78.0) 467.8(62.9) −23.0(−8.9to55.0) Fear-avoidance,activity(0to30),n=33 14.5(5.3) 13.0(6.1) −1.5(−1.2to4.2) Fear-avoidance,work(0to66),n=33 16.1(14.5) 16.0(16.9) −0.1(−3.1to3.2) Globaleffect(1to5),n=36 3.7(1.0)

VAS,visualanaloguescale;CI,confidenceinterval;SD,standarddeviation. 55.0),fearavoidanceactivitysubscale(MD−1.5;95%CI −1.2to4.2)andfearavoidanceworksubscale(MD−0.1; 95%CI−3.1to3.2).Duetopracticalreasons,notall partic-ipantsworeanaccelerometer(n=20).Theaveragelevelof globaleffectwas3.7(SD1.0)onafive-pointscale. Differ-encespersubgroupweresomewhatcomparabletotheentire group(datanotshown).

Discussion

This study described the participatory development of ablended physiotherapeutic intervention for patients with non-specificLBP(e-ExerciseLBP).Face-to-face physiother-apysessionswereintegratedwithawebapplicationwhich consistedof 12weekly information, exerciseandphysical activitymodules.TheSTarTBackScreeningToolwasused tostratify theofflineandonlinepart ofe-ExerciseLBP to patients’individualneeds[11].Proofofconceptofe-Exercise LBP was evaluatedina feasibilitystudy whichsuggested that e-Exercise LBP is effective, especially for disability (QBDPS: MD −12.2/100; 95% CI 8.3 to 16.1) and pain (NPRS:MD−2.8/10;95%CI2.1to3.6).

The CeHRes Roadmap provided guidance during the developmentofe-ExerciseLBP. Thisapproachisnot tech-nologydriven,buttakesthe complexityofhealthcareinto account [25].The aim wastodevelopan interventionthat meetsthe values of patients, physiotherapistsand experts. Patients’ recommendations for the design of the interven-tionwereless concretethanphysiotherapists’andexperts’

recommendations.Thismayhavebeenduetothedifficulty ofprovidingfeedbackataveryearlystageindevelopment, when there isonlyan unfamiliar non-material concept.In futuredevelopmentstudies,itisrecommendedthatpatients shouldreceivedraftsorpreviewsoftheinterventioninorder tomakeitmoreconcrete,andtomakeiteasierforthemto providefeedbackaboutthedesign.

Theimprovementsseenindisabilityandpaincanbe con-sideredasclinicallymeaningfulasbothoutcomesdecreased by >30%[30].ACochrane reviewonthe effectivenessof exercise therapy compared with no treatment described a substantially smaller pain reduction of 0.59 points/100 in acute LBP, −1.89points/100 in subacute LBP and−10.2 points/100 in chronic LBP [36]. In the current study, the sample sizes were too small to perform inferential statis-tics as wellas subgroupanalyses. The averagenumberof face-to-facesessionsine-ExerciseLBPwasslightlylower thanthe general averagenumberof treatment sessionsfor patients withback pain inThe Netherlands (7 vs 9) [37]. Therefore,theresultsof thisstudy tentativelyindicatethat e-ExerciseLBPmayleadtoareductioninphysiotherapeutic costs.Theeffectivenessofe-ExerciseLBPshouldbefurther investigatedinarandomizedcontrolledtrial.Areductionin physiotherapeuticcostswasalsoseeninthecomparisonof e-Exerciseandusualphysiotherapyinpatientswithhipand kneeosteoarthritis,wherethenumberofface-to-facesessions couldbereducedfrom12tofive[38].

Stratifyingblendedcaretopatientsatlow,mediumorhigh risk ofpersistent LBPappearedtobechallenging. Physio-therapistsandexpertsrecommendedintegrationoftheSTarT

(9)

BackScreeningToolincombinationwithrecommendations fortheaveragenumberoftreatmentsessionsperriskgroup. Withinthecurrentstudy,patientsinthelow-riskgroup vis-itedthephysiotherapistsmoreoftenthanthee-ExerciseLBP protocolrecommended,whereasthemedium-andhigh-risk groups received fewer sessions than the protocol recom-mended.Thesefindingsareinlinewitharecentinvestigation ofstratifiedLBPcareintheNetherlands,whichindicatedthat low-riskpatients weregenerallyover-treated andhigh-risk patientsweregenerallyunder-treatedcomparedwith recom-mendationsinclinicalguidelines[39].Stratifyingtheonline partofe-ExerciseLBPwaspossiblebyapplyingthegraded activityfunctionality.Clinicalguidelinesspecifically recom-mended thisbehavioural approachin patients at highrisk forpersistentLBP[8,9].Remarkably,thisfunctionalitywas onlyappliedinoneoftheeightpatientsathighrisk.A forth-comingmixed-methodsstudywillprovidemoreinsightinto patients’andphysiotherapists’experienceswiththe stratifi-cationof e-ExerciseLBP,as wellas howthisstratification canbeimproved.

Limitationsandrecommendations

Asthisstudywasprimarilyfocusedonthedevelopment ofe-ExerciseLBPanditspotential,acontrolgroupwasnot incorporated.Therefore,causeandeffectcannotbeclaimed anditisuncertainwhethertheimprovementsinhealth-related outcomeswereduetonatural recoveryortheeffectiveness of e-ExerciseLBP. Moreover,the study sample wassmall andlong-termfollow-upmeasurements werenotincluded. Anotherlimitationofthisstudywastheabsenceofan out-comemeasurementrelatedtoexercisecompliance,although itishypothesisedthat blendedcarecanstimulateanactive roleofthe patientwithinthe dailysituation. Theinclusion of measurements for exerciseadherence, self-management skillsorcopingstyleisrecommendedforfuturestudies. Fur-thermore,afuturelarge-scalerandomizedcontrolledtrialis neededtostudythe(cost-)effectivenessofe-ExerciseLBP comparedwithusualphysiotherapy.

Overall,e-ExerciseLBPdemonstratedinitialindications foreffectivenessinreducingdisabilityandpain.Thenextstep will bethe investigation of end-user experiences withthis prototype.Theseexperiences willbe usedtofurtheradapt the intervention to the needs of patients and physiothera-pists.Hereafter,arandomizedcontrolledtrialonthe(cost-) effectivenessofe-ExerciseLBPcomparedwithusual phys-iotherapywillbeconducted (Steps4and5of theCeHRes Roadmap).

Ethical approval: The medical Ethical Committee of the Utrecht University Medical Centre declared that the e-ExerciseLBPproject isnotcoveredbytheDutchMedical ResearchInvolvingHumanSubjectAct(WMO;number 16-231).

Funding:Thisstudy wasfunded by theScientific College

Physical Therapy (WCF) of the Royal DutchSociety for PhysicalTherapy(KNGF).

Conﬂictofinterest:Nonedeclared.

AppendixA. Supplementarydata

Supplementary data associated with thisarticle can be found, in the online version, at https://doi.org/10.1016/ j.physio.2018.12.006.

References

[1]GBD2015DiseaseandInjuryIncidenceandPrevalence Collabora-tors.Global,regional,andnationalincidence,prevalence,andyears livedwithdisabilityfor310diseasesandinjuries,1990–2015:a sys-tematicanalysisfortheGlobalBurdenofDiseaseStudy2015.Lancet 2016;388:1545–602.

[2]VosT,FlaxmanAD,NaghaviM,LozanoR,MichaudC,EzzatiM,etal.

Yearslivedwithdisability(YLDs)for1160sequelaeof289diseases andinjuries1990–2010:asystematicanalysisfortheGlobalBurden ofDiseaseStudy2010.Lancet2012;380:2163–96.

[3]KoesBW,vanTulderMW,ThomasS.Diagnosisandtreatmentoflow backpain.BMJ2006;332:1430–4.

[4]ChouR,ShekelleP.Willthispatientdeveloppersistentdisablinglow backpain?JAMA2010;303:1295–302.

[5]Hayden JA, Chou R, Hogg-Johnson S, Bombardier C. System-aticreviewsoflowbackpainprognosis hadvariablemethodsand results: guidance for future prognosis reviews. J Clin Epidemiol 2009;62:781–96e1.

[6]HaydenJA,DunnKM,vanderWindtDA,ShawWS.Whatisthe prog-nosisofbackpain?BestPractResClinRheumatol2010;24:167–79.

[7]LLJ.BDK.Zorgdoordefysiotherapeut–top-10 gezondheidsproble-men(DCSPH).Uit:NIVELZorgregistratieseerstelijn;2017.Available at:www.nivel.nl/node/4677.[Accessed17January2017].

[8]WongJJ,CoteP,SuttonDA,RandhawaK,YuH,VaratharajanS,etal.

Clinicalpracticeguidelinesforthenoninvasivemanagementoflow backpain:asystematicreviewbytheOntarioProtocolforTrafficInjury Management(OPTIMa)Collaboration.EurJPain2017;21:201–16.

[9]KoesBW,vanTulderM,LinCW,MacedoLG,McAuleyJ,MaherC. Anupdatedoverviewofclinicalguidelinesforthemanagementof non-specificlowbackpaininprimarycare.EurSpineJ2010;19:2075–94.

[10]FosterNE,MullisR,HillJC,LewisM,WhitehurstDG,DoyleC,etal.

Effectofstratifiedcareforlowbackpaininfamilypractice(IMPaCT back):aprospectivepopulation-basedsequentialcomparison.AnnFam Med2014;12:102–11.

[11]HillJC,WhitehurstDG,LewisM,BryanS,DunnKM,FosterNE,etal.

Comparisonofstratifiedprimarycaremanagementforlowbackpain withcurrentbestpractice(STarTBack):arandomisedcontrolledtrial. Lancet2011;378:1560–71.

[12]StaalJB,HendriksEJM,HeijmansM.KNGF-richtlijnLagerugpijn. [Dutchphysiotherapeuticclinicalguidelineforlowbackpain]. [13]HaydenJA,vanTulderMW,TomlinsonG.Systematicreview:

strate-giesforusingexercisetherapytoimproveoutcomesinchroniclow backpain.AnnInternMed2005;142:776–85.

[14]vanGoolCH,PenninxBW, KempenGI,RejeskiWJ, MillerGD, vanEijkJT,etal.Effectsofexerciseadherenceonphysical func-tionamongoverweightolderadultswithkneeosteoarthritis.Arthritis Rheum2005;53:24–32.

[15]MannionAF,HelblingD,PulkovskiN,SprottH.Spinalsegmental stabilisationexercisesforchroniclowbackpain:programmeadherence anditsinfluenceonclinicaloutcome.EurSpineJ2009;18:1881–91.

(10)

[16]SluijsEM,KokGJ,vanderZeeJ.Correlatesofexercisecompliance inphysicaltherapy.PhysTher1993;73:771–82,discussion83–6.

[17]FriedrichM,GittlerG,HalberstadtY,CermakT,HeillerI.Combined exerciseandmotivationprogram:effectonthecomplianceandlevel ofdisabilityofpatientswith chroniclowbackpain:arandomized controlledtrial.ArchPhysMedRehabil1998;79:475–87.

[18]NichollBI,SandalLF,StochkendahlMJ,McCallumM,SureshN, Vas-seljenO,etal.Digitalsupportinterventionsfortheself-management oflowbackpain:asystematicreview.JMedInternetRes2017;19: e179.

[19]WentzelJ,vanderVaartR,BohlmeijerET,vanGemert-PijnenJE. Mixingonlineandface-to-facetherapy:howtobenefitfromblended careinmentalhealthcare.JMIRMentHealth2016;3:e9.

[20]DarioAB,MoretiCabralA,AlmeidaL,FerreiraML,RefshaugeK, SimicM,PappasE,FerreiraPH.Effectivenessoftelehealth-based inter-ventionsinthemanagementofnon-specificlowbackpain:asystematic reviewwithmeta-analysis.SpineJ2017.

[21]BossenD,KloekC,SnippeHW,DekkerJ,deBakkerD,VeenhofC. Ablendedinterventionforpatientswithkneeandhiposteoarthritisin thephysicaltherapypractice:developmentandapilotstudy.JMIRRes Protoc2016;5:e32.

[22]KloekCJ,BossenD,VeenhofC,vanDongenJM,DekkerJ,deBakker DH.Effectivenessandcost-effectivenessofablendedexercise interven-tionforpatientswithhipand/orkneeosteoarthritis:studyprotocolofa randomizedcontrolledtrial.BMCMusculoskeletDisord2014;15:269.

[23]KloekCJJ,BossenD,SpreeuwenbergPM,DekkerJ,deBakkerDH, VeenhofC.Effectivenessof ablended physicaltherapist interven-tionin peoplewith hip osteoarthritis, kneeosteoarthritis, or both: a cluster-randomized controlled trial. Phys Ther 2018;98:560–70,

http://dx.doi.org/10.1093/ptj/pzy045.

[24]deVriesHJ,KloekCJJ,deBakkerDH,DekkerJ,BossenD,Veenhof C.Determinantsofadherencetotheonlinecomponentofablended interventionforpatientswithhipand/orkneeosteoarthritis:amixed methodsstudyembeddedinthee-Exercisetrial.TelemedJEHealth 2017.

[25]vanGemert-PijnenJE,NijlandN,vanLimburg M,OssebaardHC, KeldersSM,EysenbachG,etal.Aholisticframeworktoimprove theuptakeandimpactofeHealthtechnologies.JMedInternetRes 2011;13:e111.

[26]KensingFBJ.Participatorydesign:issuesandconcerns.Comput Sup-portCoopWork1998;7:167–85.

[27]CoeneEK,ErlandssonS.Zorgboek Rugpijn.Stichting September; 2011.

[28]ThomasS,ReadingJ,ShephardRJ.RevisionofthePhysicalActivity ReadinessQuestionnaire(PAR-Q).CanJSportSci1992;17:338–45.

[29]Kopec JA, Esdaile JM, Abrahamowicz M, Abenhaim L, Wood-DauphineeS,LampingDL,etal.Thequebecbackpaindisabilityscale. Measurementproperties.Spine(PhilaPa1976)1995;20:341–52.

[30]OsteloRW,DeyoRA,StratfordP,WaddellG,CroftP,VonKorffM,

etal.Interpretingchangescoresforpainandfunctionalstatusinlow backpain:towardsinternationalconsensusregardingminimal impor-tantchange.Spine(PhilaPa1976)2008;33:90–4.

[31]KellyLA,McMillanDG,AndersonA,FippingerM,FillerupG,Rider J.Validityofactigraphsuniaxialandtriaxialaccelerometersfor assess-mentofphysicalactivityinadultsinlaboratoryconditions.BMCMed Phys2013;13:5.

[32]Freedson PS, Melanson E, Sirard J. Calibrationof the Computer ScienceandApplications,Incaccelerometer.MedSciSportsExerc 1998;30:777–81.

[33]Wendel-VosGC,SchuitAJ,SarisWH,KromhoutD.Reproducibility and relative validity of the short questionnaire to assess health-enhancingphysicalactivity.JClinEpidemiol2003;56:1163–9.

[34]Waddell G,Newton M, HendersonI, Somerville D,Main CJ.A Fear-AvoidanceBeliefsQuestionnaire(FABQ)andtheroleof fear-avoidance beliefs in chronic low back pain and disability. Pain 1993;52:157–68.

[35]BossenD,VeenhofC,VanBeekKE,SpreeuwenbergPM,DekkerJ,De BakkerDH.Effectivenessofaweb-basedphysicalactivityintervention inpatientswithkneeand/orhiposteoarthritis:randomizedcontrolled trial.JMedInternetRes2013;15:e257.

[36]HaydenJA,vanTulderMW,MalmivaaraA,KoesBW.Exercisetherapy fortreatmentofnon-specificlowbackpain.CochraneDatabaseSyst Rev2005;3:CD000335.

[37]LeemrijseCJBJ,KooijmanMK,VerberneLDM,VeenhofC,Swinkels ICS.Omvangenkenmerkenfysiotherapeutischezorgbijpatiëntenmet rugpijn.Uit:NIVELZorgregistratieseerstelijn;2017.Availableat:

www.nivel.nl/node/4394.[Accessed10July2017].

[38]Kloek CJJ,vanDongenJM, deBakker DH,BossenD,DekkerJ, VeenhofC.Cost-effectivenessofablendedphysiotherapy interven-tioncomparedtousualphysiotherapyinpatientswithhipand/orknee osteoarthritis:aclusterrandomizedcontrolledtrial.BMCPublicHealth 2018;18:1082,http://dx.doi.org/10.1186/s12889-018-5975-7. [39]BierJD,Sandee-GeurtsJJW,OsteloR,KoesBW,VerhagenAP.Can

primarycare forbackand/orneckpainintheNetherlandsbenefit fromstratificationforriskgroupsaccordingtotheSTarTBack Tool-classification?ArchPhysMedRehabil2017.

Availableonlineatwww.sciencedirect.com

Development and proof of concept of a blended physiotherapeutic intervention for patients with non-specific low back pain

Amsterdam University of Applied Sciences