Physical examination findings and their relationship with performance-based function in adults with

12 Pages·2017·0.48 MB·English

CORE Metadata, citation and similar papers at core.ac.uk
Provided by Springer - Publisher Connector
Iversenetal.BMCMusculoskeletalDisorders (2016) 17:273
DOI10.1186/s12891-016-1151-3
RESEARCH ARTICLE Open Access
Physical examination findings and their
relationship with performance-based
function in adults with knee osteoarthritis
Maura D. Iversen1,2,3*, Lori Lyn Price4,5, Johan von Heideken1,3, William F. Harvey6 and Chenchen Wang6
Abstract
Background: Many physical examination (PE)maneuvers exist to assess knee function, none ofwhich are specific
to knee osteoarthritis(KOA). The OsteoarthritisResearch Society International also recommends the use ofsix
functional performance measuresto assess function inadults with KOA. While earlier studies have examined the
relationship between PE findingsand self-reported function or PE findings and select performancetests in adults
withknee pain and KOA,few have examined the allthreetypes of measures. This cross-sectionalstudy specifically
examines therelationships between results of PE findings, functional performance tests and self-reported function
inadults with symptomatic KOA.
Methods: We used baseline PE data from a prospective randomized controlled trial in 87 participants aged
≥40 years with symptomatic and radiographic KOA.The PE performed by threeexperienced physical therapists
included: muscle assessment, function and special tests. Participantsalso completedfunctional performance tests
and the Western Ontario and McMaster Osteoarthritis Index (WOMAC).Multivariate linear regression identified
contributions ofPE findings towards functional performance and WOMAC scores, adjusting for age and gender.
Results: Participants’mean age was 60.4 years (SD=10.5), meandisease duration was 8.4years (SD=10.1) and 27
participants had varus knee alignment. Mean WOMAC pain and function scores were 211 (SD=113) and 709
(SD=394), respectively. Weakness was present in major hip and knee muscles. Seventy-nine participants had a
positiveEly’s,65apositiveWaldronand49apositiveGrind.Mean6-minwalkwas404m(SD=83)andmeanBerg
Balancewas53(SD=4).Regressionanalysisidentifiedpositivefindingson5specialtests(P<0.05)asindicativeof
poorer6minwalk.PositiveApley’swasassociated(P<0.05)withslower20mwalkandapositiveOberwithpoorer
balancescores(P<0.05).
Conclusions:Diminishedhipmusclestrengthandflexibility,andpatelladysfunctionwereprevalentintheseadults
withsymptomaticKOA.Resultsoffunctionalperformancetestssuggestbalanceandwalkingabilityareimpairedand
areassociatedwithPEfindingsofmusclelengthimbalance,hipmuscleweaknessandpatelladysfunction.Noneofthe
PEmeasureswereassociatedwithself-reportedfunction.Therefore,performance-basedtestresultsmaybemore
usefulininformingrehabilitationinterventions.
Keywords:Kneeosteoarthritis,Physicalexamination,Performance-basedfunction,Self-reportedfunction
*Correspondence:[email protected]
1DepartmentofPhysicalTherapy,MovementandRehabilitationSciences,
NortheasternUniversity,360HuntingtonAvenue301CRB,Boston,MA
02115,USA
2DepartmentofMedicine,SectionofClinicalSciences,Divisionof
Rheumatology,Immunology&Allergy,Brigham&Women’sHospital,Harvard
MedicalSchool,Boston,MA,USA
Fulllistofauthorinformationisavailableattheendofthearticle
©2016TheAuthor(s).OpenAccessThisarticleisdistributedunderthetermsoftheCreativeCommonsAttribution4.0
InternationalLicense(http://creativecommons.org/licenses/by/4.0/),whichpermitsunrestricteduse,distribution,and
reproductioninanymedium,providedyougiveappropriatecredittotheoriginalauthor(s)andthesource,providealinkto
theCreativeCommonslicense,andindicateifchangesweremade.TheCreativeCommonsPublicDomainDedicationwaiver
(http://creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthisarticle,unlessotherwisestated.
Iversenetal.BMCMusculoskeletalDisorders (2016) 17:273 Page2of12
Background PE findings on self-reported function (Western Ontario
Symptomatic knee osteoarthritis (KOA) is one of the ten and McMaster Osteoarthritis Index (WOMAC) pain,
mostdisablingdiseasesindevelopedcountriesaffectingan WOMAC function). It was hypothesized that special tests
estimated 19 % of women and 14 % of men in the United of pain provocation, and muscle flexibility would more
States over 45 years[1]. KOA resultsfrom progressivede- strongly correlate with functional performance than self-
structionofarticularcartilage,ligaments,andjointcapsule, reportedfunctionandpain.
synovial membrane inflammation and subchondral bone
calcification [2]. Pain from knee and muscle impairments, Methods
mainly thequadricepsand hamstrings, are associated with This cross-sectional study is a secondary analysis of data
KOA [3]. Other common symptoms of KOA include from a prospective randomized clinical trial evaluating
crepitus, reduced joint motion (both range of motion and the effectiveness and cost-effectiveness of Tai Chi versus
arthrokinematic motion quality), impaired proprioception, a physical-therapy regimen in adults with symptomatic
joint line and periarticular tenderness on palpation and and radiographically confirmed KOA [18, 19]. Institu-
mildsynovitis[2,4–6].Thesesymptomsmayproduceim- tional Human Subjects approval was obtained for this
pairments in body functions, activity limitations and par- secondary study. We included the subset of 87 partici-
ticipation restrictions [2]. In 2013, Osteoarthritis Research pants from the primary study who were randomized to
Society International (OARSI) provided a set of recom- physical therapy between January, 2010 and December,
mended performance-based measures to assess physical 2013andfor whom PEdata existed.
functioninadultswithKOA.Thesemeasuresincludetests
ofaerobicconditioning,walkingspeed,functionalmobility Recruitmentandparticipants
andlowerextremitystrength[7]. The recruitment strategy has been described previously
Radiographic findings in the absence of a physical exam- [18].Inbrief,participantswererecruitedfromthecommu-
ination,arenotveryusefulinidentifyingthesourceofpain nityusingamulti-modalrecruitmentcampaign.Studyad-
in symptomatic KOA [8, 9]. Therefore, clinicians also rely vertisements were placed in a variety of media venues
onthepatienthistory,physicalexamination(PE)procedures including: myHospitalWebsite, Craigslist, Facebook, the
andspecialteststoassistinclinicaldecision-making.Special medical center website, SAMPAN (a Chinese newspaper),
tests of muscle flexibility are used to assess muscle length localnewspapersandviaaboothata seniorexpo.Partici-
and flexibilitye.g.,Ely’s forrectusfemoris and Ober forthe pantswerealsoidentifiedfromtherheumatologyclinicpa-
iliotibial band [10]. Ligamentous tests are conducted to tient database at a large urban tertiary medical center.
examinekneejointintegrity/stability,asaproxyforchanges Inclusion criteria for the primary trial consisted of adults
in knee biomechanics, secondary to muscular tightness or aged 40 years or older with a diagnosis of symptomatic
changesinlowerlimbalignment[10].Despitethecommon KOAbasedontheAmericanCollegeofRheumatologycri-
use of these physical examination procedures in clinical teria (pain on more than half the days of the past month
practice, the psychometric properties of these procedures during at least one of the followingactivities: walking, go-
areweak[10,11].ThepoorassociationbetweenPEfindings ing up or down stairs, standing upright, or lying in bed at
and self-reported function and performance-based function night) and who had radiographic evidence of tibiofemoral
may be related, in part, to central sensitization in arthritis, orpatellofemoralOA(definedasthepresenceofatibiofe-
which can contribute to several of the “positive” findings. moral compartment and/or patellofemoral compartment
Performance based examinations are additional methods osteophyteinstandinganterior/posterior,lateralorskyline
usedtoassessphysicalfunctionandprovidecomplementary views) [20]. Weight-bearing radiographs were also scored
informationregardingKOA-associateddisability[12].Thus, using the Kellgren and Lawrence (K-L) grading system
research into optimal clinical examination sequence and (range 0–4). A K-L grade of 0 indicates no radiographic
compositionislackingandnecessarygiventimeconstraints features and higher grades indicate more severe global
inclinicalpractice. tibiofemoral radiographic structural damage [21]. Add-
WhilestudieshaveexaminedtheassociationbetweenPE itionally,asthegoaloftheprimarytrialwastorecruitpar-
findingsandself-reportedfunctioninadultswithkneepain ticipants with symptomatic KOA, all participants were
[13] and in those with KOA [14–17], few studies have required to have score of 40 or greater on the Western
examined the association between PE findings and Ontario and McMaster’s Universities Osteoarthritis Index
performance-based function in adults with symptom- pain scale (WOMAC) [22, 23]. The study rheumatologist
atic KOA [14]. This study provides a detailed clinical de- (WH)confirmedparticipanteligibility.
scription of adults with symptomatic KOA and aims to Participants were excluded if they: received physical
determine therelative contributionof PEresultsprimarily therapy in the last year, had medical conditions affecting
onperformance-basedfunction,adjustingforageandgen- their ability to safely participate, received steroid injec-
der.Secondarily,weaimedtoexaminethe contributionof tions or reconstructive surgery in the past three months,
Iversenetal.BMCMusculoskeletalDisorders (2016) 17:273 Page3of12
were non-English speakers, pregnant, had a score below Physicalexaminationprotocol
24 on the Mini-mental state exam [24], were non- A supervising physical therapist (PT) research scientist
ambulatory or had 100% dependence on an assistive de- (MDI) created a physical examination (PE) form based
vice,wereenrolledinanotherclinicaltrialinlast30days, on contemporary orthopedic physical therapy practice.
or had plans to relocate. All participants were examined She conducted training sessions with the three study
by one of three physical therapists (PTs). Human partici- PTs to review the content of the PE and to develop a
pants approval was obtained and all participants con- consensus on how to perform and document results of
sented toparticipate. the special tests in a standardized manner. Specifically,
the physical therapy team agreed upon a single method
for each special test used for muscle flexibility, ligament-
Studyknee ous and meniscal integrity and posture. The initial PE
For knee specific variables, we used data from the study was performed the week following study intake by one
knee, defined as the knee diagnosed with symptomatic of the three study PTs. All three PTs had ten or more
KOA. For participants with bilateral KOA, we defined years of clinical experience and one was also a certified
the study knee as the knee that was most severely af- Orthopedic Clinical Specialist. During each recruitment
fected accordingtoWOMACscores[22]. cycle, thesupervisingresearchscientist visitedeachther-
apist to ensure consistent documentation of examination
proceduresandfor qualityassurance.
Baselineintake ThePEincludedpastandcurrentmedicalhistoryandPE
Participantscompletedabaselineintakewithastudyteam procedures. Selected comorbidities (Heart Disease, Hyper-
member that included performance measures, a demo- tensionandDiabetes)werecollectedviaself-report.Patients
graphic survey (age, gender, race, co-morbidities, and were asked about the use of pain medication (classified as
marital/livingstatus),Bodymassindex(BMI),andpatient- Nonsteroidal Anti-inflammatory Drugs (NSAIDs) or other
reportedoutcomemeasures.Patient-reportedknee-related type of analgesics), their chief complaint, and the duration,
outcome measures included the WOMAC pain and func- locationandtypeofpain.Objectiveinformationincludedan
tion scales, completed at the time of the clinical examin- integumentary assessment focused on edema, tenderness
ation.TheWOMAC(versionVA3.1)isareliableandvalid and sensation, joint range of motion, and muscle strength
instrument specifically designed to evaluate knee and hip usingmanualmuscletesting(MMT)[10].Thetherapistob-
OA [23]. The pain subscale score ranges from 0 to 500 servedgaitandfunctionandassessedthelimbforthepres-
andfunctionscoresfrom0to1700,withhigherscoresin- ence of a leg length discrepancy. An evaluation of knee
dicating worse outcome. Performance tests included valid contour and alignment was performed; specifically genu
and reliable tests of gait speed, endurance, strength and varus(>3fingersapartatthekneewithanklestogether),val-
balance as recommended by OARSI [7]. Gait speed was gus (>3.5 in. between malleoli when knees are together) or
assessed with the timed 20-m (m) walk test. A research recurvatum (hyperextension of the knee joint greater than
teammemberdemonstratedthetestprocedurebywalking 0°),anklealignment,pesplanusorcavus.Next,specialtests
theprescribeddistanceatacomfortablewalkingpace.The were used to assess muscle length of primary hip and knee
outcome was the total time taken to walk 20 m (m). Par- musculature,suchastheEly’sTestandOberTest[30,31].
ticipants completed two trials and the average time to AsKOAisinfluencedbykneebiomechanics,testsofliga-
complete the trials was calculated [25]. The 6-min walk mentous integrity helped determine whether alterations in
test,areliablemeasureoffunctionalexercisecapacity,was ligamentsweredirectlyaffectingsymptoms[32].Ligament-
used to evaluate endurance. Participants walked as fast ous tests included were the Lachman’s, Posterior Sag and
and as far as possible within the six-minute period with theVarusandValgusStressTestsat0and30°.McMurray’s
verbal encouragement provided every minute. The dis- testwasusedtoassessmeniscalintegrity[33].Thepatello-
tance, measured in meters, was recorded [26]. The Berg femoraltests,Apley’scompressionanddistractiontests,the
BalanceScaleisareliabletestthatassessesbalanceduring Grind and the Waldron tests, were included to assess pa-
the performance of 14 functional tasks such as standing tella dysfunction and tracking issues [34]. See Table 1 for
from a seated position, standing unsupported for two mi- detailsofPEproceduresandreliabilityandvalidityinforma-
nutes,turning 360°,and standingon onefoot.Bergscores tion for these procedures. The final portion of the PE in-
range from 0 to 56 with higher scores indicating better cluded a functional activity assessment with the PT noting
balance[27,28].Thechair-standtestmeasuredlowerlimb whetheranassistivedevicewasusedduringambulation.
muscle strength and mobility. This test assesses the time
taken tocomplete 10full stands from a sittingposition as Statisticalanalysis
fast as possible, and the fastest time of two trials was re- Descriptive statistics were used to characterize the sam-
cordedtothenearest0.01s[29]. ple. As most clinical special tests (e.g., Lachman, grind,
Iversenetal.BMCMusculoskeletalDisorders (2016) 17:273 Page4of12
Table1Standardphysicalexamproceduresusedforkneeosteoarthritis
Test Descriptionofprocedureandstructuretested Interraterreliability(Kappa)
Ely’sTest Subjectprone.Examinerstandsnexttothesubject,atthesideofthelegthat 0.46[30]
willbetested.Examinerplacesonehandonlowerback,theotherholdingthe
legattheheel.Thekneeispassivelyflexedinarapidfashion.Theheelshould
touchthebuttocks.Bothsidesaretestedforsymmetry.Thetestispositive
whentheheelcannottouchthebuttocks,thehipofthetestedsiderisesup
fromthetable,orthepatientfeelspainortinglinginthebackorlegs.This
procedureassessesthetightnessofrectusfemorismuscle.
OberTest Subjectliesontheuninvolvedsidewithhipandkneeflexedat90°.Examiner 0.73[47]
placesthekneein5°offlexion,fullyabductsthelegbeingtested,andthen
allowsgravitytoadducttheextremityuntilthehipcannotadductanyfurther.
Theprocedureassessesthetightnessoftheiliotibialband.
Lachman's Examinerflexesthekneeto30°withthepatientinsupineandappliesananterior 0.36[48]
forcetothetibia,notinganyexcessmotion.Theprocedureassessestheintegrity
oftheanteriorcruciateligament.
PosteriorSag Examinerflexesthehipandkneeto90°withthepatientandassessapossible Notfound
posteriorsagofthetibia.Thetestassessestheintegrityoftheposterior
cruciateligament.
Varusat30° Theexaminerbringsthetestingkneeto30°andappliesavarusforcewhilethe Notfound
subjectsisinsupine.Theexaminernotesanyexcessmotion.Theprocedure
assessesthelateralcollateralligament,thefibularcollateralligamentandother
posteriorlateralcornerkneestructures.
Valgusat0/30° Examinerbringstestingkneeto30°andappliesavalgusforcewhilethesubject 0.16[49]
isinsupine.Theexaminernotesanyexcessmotion.Thenrepeatsthetestwith
thekneein0°.Thisprocedureassessesthemedialcollateralligamentswithor
withouttheposteriorcapsule.
McMurray's Subjectsupine.Theexaminerplacesonehandtothesideofpatellaandotherat 0.35[50]
distaltibiaandextendsthekneefrommaximumflexiontoextensionwithinternal
rotationandvarusstress.Thekneeisthenreturnedtomaximumflexionandthe
kneeextendedwithexternalrotationandvalgusstress.Thisprocedureassesses
meniscalintegrity.
Apley's(distraction Subjectispronewiththeirkneeflexedto90°.Theexaminermediallyandlaterally Notfound
andcompression) rotatesthetibia,combinedfirstwithdistractionofthelowerleg.Theexaminerthen
appliesanaxialloadthroughthekneeandrotatesthejointviathelowerleg.A
positivetestwillresultinpainorincreasedrotationrelativetotheothersidewhen
distractionisapplied.
Compressiontest–thetestispositiveiftherotationpluscompressionofpatellais
morepainfulorshowsdecreasedrotationrelativetothenormalside.Thedistraction
testassessesthemedialandlateralcollateralligament.Thecompressiontestassesses
meniscalintegrity.
WaldronSign Examinergentlycompressesthekneewhilethesubjectissquatting,notinganypain Notfound
orcrepitus.Thisprocedureassessespatellofemoraljointandcartilageintegrity.A
positivetestindicatesthepresenceofchondromalacia,patellaoranteriorkneepain
frompatellacontactpressure.
GrindTest Subjectsupinewithkneeslightlyflexed.Theexaminerprovidesdistalforceatsuperior Notfound
borderofpatellaasthepatientcontractsthequadriceps.Painproductionindicatesa
positivetest.Theprocedureassessestheintegrityoftheposteriorpatellaandthe
trochleargrooveofthefemur.
VastusMedialis Patientsupine,kneesupportedin20°flexion.Patientactivelyextendsknee.The Notfound
ObliqueTest examinerassessescontractionandmovementofpatellasuperiorlyintogrove.
Gradedasweak,nocontractionorfindingswerewithinnormallimits
etc.) are recorded as either positive or negative, these test WOMAC scores and functional performance results. For
results were treated as dichomotous variables. For tests the Ely test, a Wilcoxon Ranked Sum test was used to
that had more than 2-level responses, we collapsed them examine the association of this test with theTimed Chair
into dichotomous variables. Data from participants who Standtest.
completed 10 chair stands were included in analyses in- In preparation for regression modeling, muscle weak-
volvingchairstands.T-testswereruntodeterminetheas- ness scores, by major muscle groups (ordinal scores
sociation between the results of special tests and ranged from 0 to 5) were collapsed into as a series of
Iversenetal.BMCMusculoskeletalDisorders (2016) 17:273 Page5of12
dichotomous variables (weakness defined as a score<4) Table2Demographicfeatures,radiologicalcharacteristics,and
to indicate the presence of muscle weakness (y/n) in self-reportedoutcomesinparticipantswithkneeosteoarthritis
each major lower extremity muscle group. Scores for Variable Mean(SD) Range
muscle weakness in the study leg were use in the regres- Age,years(n=87) 60.4(10.5) 41-85
sion. Cross tabulations of special tests for similar struc- Female,n(%)(n=87) 60(69%)
tures (e.g., ligament weakness) were conducted to
Race:n(%)(n=87)
examinethe parsimony ofthe data to reduce variables in
Caucasian 50(57%)
the modeling. Data were also examined to determine
whether the occurrence of a positive test result was ex- AfricanAmerican 27(31%)
ceptionally low or exceptionally high (either<10 % or AmericanIndianorAlaskanNative 3(4%)
>90 % positive test results) to determine appropriateness Morethanonerace 4(4%)
for statistical testing. Special tests with exceptionally low
Other 3(4%)
or exceptionally high positive test results were not in-
MeanBMIa(kg/m2)(n=87) 32(6.9) 20-49
cluded in data modeling. Multivariable linear regression
DurationofKneePaininindex 8.4(10.1) <1-65
was employed to model the primary outcomes:
knee,years(SD)(n=82)
WOMAC function, WOMAC pain, 20 m walk, 6 Min
MeanComorbidities(n=87)
walk, Timed Chair Stand test and Berg Balance with PE
measures, adjusting for age and gender. Regression diag- HeartDisease 7(8%)
nostics were run on each model to determine the pres- Hypertension 42(48%)
ence of influential observations. Results are presented Diabetes 12(14%)
for the whole cohort, however, when studentized resid- NSAIDbpriortostudy,n(%)(n=86) 51(59%)
uals or Cook’s D indicated observations were influential,
Othertypeofanalgesicspriorto 28(33%)
a comment on the impact of these observations on the study,n(%)(n=86)
regression is noted. The data analysis for this paper was MeanWOMACcpainatexam 211(113) 28-490
generated using SAS software, version 9.4 for Windows [0–500](SD)(n=84)
(Copyright,SASInstituteInc., Cary, NC,USA). MeanWOMACcfunctionatexam 709(394) 47-1661
[0–1700](SD)(n=84)
Results SDstandarddeviation
In the randomized controlled, 98 adults with KOA were aBMI=Bodymassindex,≥30kg/m2isdefinedasbeingobese
bNSAID=NonsteroidalAnti-inflammatoryDrugs
recruitedandrandomizedtophysicaltherapy.Ofthese,11 cWOMAC=WesternOntarioandMcMasterUniversities.Aself-administered,
didnotappearforthePE,leaving87participantswhohad visualanaloguescalespecificallydesignedtoevaluatekneeandhip
osteoarthritis.Higherscoresindicatingmoreseverediseasesymptoms
complete PE data. Sixty participants (69 %) were female,
themeanagewas60.4years(SD=10.5)andthemeandur-
ation of knee pain was 8.4 (SD=10.1) years. Fifty-one had weakness in the knee extensors and 21 (24 %) in the
(59%)participantsusedNSAIDsand28(33%)usedother flexors. The mean Timed Chair Stand time was 31 s
types of analgesics pain medication prior to the study to (SD=11) and the average distance on the 6-min Walk
managetheirkneesymptoms.Themajorityofparticipants Test was404m(SD=83).Themean BergBalanceScore
(57 %) were Caucasian and 47 (54 %) had a Body mass was 53 (SD=4), indicating a low fall risk and the mean
index(BMI)of≥30kg/m2,whichisdefinedasbeingobese 20 m Walk was 18 s (SD=3), indicating an average gait
[25]. Eighty-one participants(91%)had radiological struc- speed of1.1m/s(Table3).
turalfindingsofK-Lgrade2ormore,indicatingmoderate
tosevereradiologicalstructuralkneedamage.Twenty-four
participants (28 %) presented with genu varus, 9 (10 %) Resultsofspecialtests
with genu valgus and 4 (5 %) with recurvatum. Examin- Special tests for muscle flexibility indicated a prevalence
ation of the foot alignment revealed that 2 (2 %) had pes of hip muscleinflexibility, especiallyin the rectis femoris
cavusand49(56%)hadpesplanus.TheaverageWOMAC (91 %). The results of the Grind and Waldron tests sug-
pain score was 211 (SD=113) and the average WOMAC gested many patients may have had patella-femoral joint
function score was 709 (SD=394). Of the 87 participants, involvement in addition to tibio-femoral KOA. Twenty-
15(17%)ambulatedwithcrutches/canes(Table2). nine participants (34 %) had a positive McMurrary test
of meniscal integrity. Nine participants (10 %) had a
Physicalexaminationandperformancedata positive Lachman test which is one test to assess anter-
Weak hip musculature was prevalent with the greatest ior cruciate ligament damage. No participants had a
impairmentsnotedinhipflexorsandextensors,followed positive Posterior Sag and few had involvement of the
by abductors/adductors. Thirty-seven participants (43 %) medial collateral ligament or posterior crucial ligaments.
Iversenetal.BMCMusculoskeletalDisorders (2016) 17:273 Page6of12
Table3Lowerextremitymuscleweaknessofstudykneeand based outcomes. With respect to self-reported function,
physicalfunctionscoresinparticipantswithkneeosteoarthritis both WOMAC pain and function scores were signifi-
Variable cantly associated with a positive Apley’s Distraction test
Hipflexor/extensorweakness,n(%)(n=87/86) 46(53)/46(53) (mean difference of 103; p=0.047 and mean difference
Hipabductor/adductorweakness,n(%)(n=87/86) 52(60)/48(56) of498;p=0.0006,respectively (Table5).
Hipinternal/externalrotatorweakness,n(%)(n=86/85) 17(20)/20(24)
Associationsbetweenphysicalexaminationfindingsand
Kneeflexor/extensorweakness,n(%)(n=87/87) 37(43)/21(24)
self-reportedfunctionandperformancetestscores
Ankleinvertor/evertorsweakness,n(%)(n=87/87) 8(9)/8(9) adjustingforAgeandgender
Ankledorsiflexor/plantarflexorweakness,n(%) 10(11)/8(9) To determine the associations between PE findings,
(n=87/87) patient-reported outcomes and performance test results,
MeanTimedChairStanda,seconds(SD)[range](n=78) 31(11)[10–69] after adjusting for age and gender, we conducted linear
Mean6-minwalkb,meters(SD)[range](n=84) 404(83)[202–645] regression analysis (Table 6). There were no significant
MeanBergBalancescorec(SD)[range](n=87) 53(4)[35–56] associations with self-reported pain (WOMAC) in the
full regression. A positive Apley’s Distraction test (collat-
Meantime20mwalkd,seconds(SD)[range] 18(3)[10–32]
(n=87) eral ligament dysfunction) was significantly associated
MuscleweaknessdefinedasaManualMuscleTestgradeof≥4onscalefrom with lower self-reported function (WOMAC) (p<0.01).
0to5 Positive findings on five special tests (Ely’s, Waldron,
Dataisprovidedforallsubjectswhocompletedthetest Ober and Apley’s compression and distraction) were as-
aThetestmeasuresthetimetakentocomplete10fullstandsfromasitting
positionasfastaspossible sociated with 6-min walk test performance (p<0.05 for
bThe6-minwalktestmeasuresthedistancecoveredduringthe6-minwalk each special test). A positive Ober test was also associ-
cTheBergBalanceScaleassessedbalanceduringperformanceof14functional ated with poorer Berg Balance performance (p<0.05).
tasks,rangefrom0to56andhigherscoresindicatebetterbalance
d20mwalktestisthetotaltimeittakestowalk20m With respect to walking speed, apositiveApley’s distrac-
tion (collateral ligament dysfunction) was associated
These ligamentous test results indicate knee structures with slower 20 m walk times (p<0.05). Individuals with
wererelativelyintactandthekneeswerestable(Table4). a positive Ely’s test (tight rectus femoris) performed the
Timed Chair test faster than those without a tight rectus
Univariateassociationsbetweenspecialtestresultsand femoris (p<0.01). A positive Apley’s Compression test
WOMACscoresandfunctionalperformance was also associated with better Timed Chair Test per-
Using t-tests, special tests of ligamentous integrity, formance(p<0.05).
muscle flexibility and patella dysfunction were examined Influential points were identified for the following out-
against the outcomes of performance tests and self- comes: WOMAC pain, Berg Balance score and the
reported function and pain. Individuals with a positive Timed Chair Stand. When the influential point was re-
Apley’sCompressiontestperformedpoorer(fewercycles moved from the model with WOMAC pain as the out-
ofrisingandsitting) ontheTimedChairTest (meandif- come, a positive Apley’s Distraction test was found to be
ference of 8; p=0.02). No other special tests were sig- significantly associated with more self-reported knee
nificantly associated with differences in performance- pain, while apositive McMurray test wasassociated with
an improvement in pain. With balance as the outcome,
Table4Positivefindingsforspecialexaminationsperformedon after removing one influential point, a positive McMur-
thestudykneeandperformancetestresults
ray test (meniscus pathology) was associated with poorer
ClinicalandPerformanceTests Number(%) balance while a positive Grind test (assess integrity of
VastusMedialisObliqueWeakness(n=87) 75(86) posterior patella and the trochlear groove of the femur)
PositiveEly’s(n=87) 79(91) was significantly associated with better balance. Two ob-
PositiveOber(n=85) 29(34) servations were found to be influential in the chair stand
model. With these two observations removed, knee liga-
PositiveLachmansTest(n=86) 9(10)
ment dysfunction as noted by positive Apley’s Compres-
PositiveVarusat30°(n=86) 2(2)
sion and Distraction test results, were associated with
PositiveValgusat0/30°(n=86) 16(19) test performance, while the association with the positive
PositiveMcMurray's(n=86) 29(34) Ely’stestbecame non-significant.
PositiveApley's(distraction)(n=85) 5(6)
PositiveApley's(compression)(n=85) 15(18) Discussion
While earlier studies have examined the relationship be-
PositiveWaldronSign(n=87) 65(75)
tween select PE findings and performance based func-
PositiveGrindTest(n=87) 49(56)
tion in adults with knee pain and others have examined
Iv
e
rs
e
n
e
t
a
l.
B
M
C
M
Table5Differenceinperformancetestfindingsandself-reportedkneepainandfunctionamongthosewithpositiveandnegativespecialexaminationresults u
sc
WOMACpaina WOMACfunctiona 20mwalkb 6minwalkc BergBalanced Timedchairstande ulo
sk
SpecialTests Meandiff(95%CI) t-stat Meandiff(95%CI) t-stat Meandiff(95%CI) t-stat Meandiff(95%CI) t-stat Meandiff(95%CI) t-stat Meandiff(95%CI) t-stat ele
Ely’s −52(−135,31) −1.3 −247(−535,42) −1.7 0(−3,2) −0.3 41(−24,106) 1.2 −1(−4,2) −0.7 +Ely’s=29 N/A tal
+-==879 –Ely’s=41.6median Disord
e
Lachman’s −26−106,53 −0.7 31(−248,311) 0.2 0(−4,1) −1.0 26(−36,88) 0.8 0(−2,3) 0.3 −5(−12,3) −1.3 rs
+=9 (2
0
-=77 1
6
McMurray's 32(−21,84) 1.2 158(−24,340) 1.7 −1(−2,1) −0.6 −6(−45,33) −0.3 −1(−2,1) −0.6 −1(−7,4) −0.5 ) 17
+=29 :27
-=57 3
Waldron −28(−84,29) −1.0 −43(−242,155) −0.4 −1(−3,1) −1.3 23(−19,65) 1.1 0(−2,2) 0.3 −1(−7,5) −0.3
+=65
-=22
Apley's(compression) 29(−36,93) 0.9 113(−111,337) 1.0 −1(−3,1) −1.2 36(−9,81) 1.6 1(−2,3) 0.7 −8(−14,−1) −2.3
+=15
-=70
Apley's(distraction) 103(1,205) 2.0 498(150,844) 2.9 3(−1,6) 1.6 −58(−131,14) −1.6 −1(−4,3) −0.3 8(−4,19) 1.3
+=5
-=80
Grind 15(−34,65) 0.6 78(−95,250) 0.9 −1(−2,1) −1.1 16(−21,52) 0.9 2(0,3) 1.7 −2(−7,3) −0.8
+=49
-=38
Ober −49(−98,1) −2.0 −163(−340,14) −1.8 1(−1,2) 0.7 −22(−61,18) −1.1 −1(−3,1) −1.2 0(−6,5) −0.1
+=29
-=56
+numberofpositivetest,−Numberofnegativetest.CIconfidenceinterval.Significantassociation(P<0.05)inbold
aWOMAC=WesternOntarioandMcMasterIndex
b20mwalktestisthetotaltimeittakestowalk20m
cThe6-minwalktestmeasuresthedistancecoveredduringthe6-minwalk
dTheBergBalanceScaleassessedbalanceduringperformanceof14functionaltasks
eThetestmeasuresthetimetakentocomplete10fullstandsfromasittingpositionasfastaspossible.MedianvaluesusedtotestEly’sversusTimedChairStand
P
a
g
e
7
o
f
1
2
Iv
e
rs
e
n
e
t
a
l.
B
M
C
M
u
sc
u
lo
sk
Table6Associationbetweenoutcomesofspecialtests,performancetestsandself-reportedkneepainandfunction ele
ta
WOMACaPainnumber WOMACafunctionnumber 20mwalkbnumber 6minwalkcnumber BergBalancednumber Timedchairstandenumber l
D
ofobservations=80 ofobservations=80 ofobservations=83 ofobservations=80 ofobservations=83 ofobservations=74 iso
F=1.74;p=0.09 F=2.27;p=0.023 F=1.44;p=0.18 F=1.93;p=0.06 F=2.06;p=0.04 F=1.93;p=0.06 rde
R2=0.09) R2=0.14 R2=0.05 R2=0.11 R2=0.11 R2=0.11 rs
(2
Explanatoryvariable B SEB B SEB B SEB B SEB B SEB B SEB 01
6
PositiveEly’s −30.2 40.3 −175.3 138.9 −1.0 1.3 65.4* 31.6 −0.7 1.5 −12.7** 4.4 ) 1
7
PositiveLachman’s −8.9 37.9 109.3 130.5 −0.8 1.2 15.1 29.3 0.5 1.4 −3.7 3.9 :27
3

PositiveMcMurray's 36.1 29.2 160.4 100.6 0.01 0.9 −24.9 21.4 −1.7 1.0 0.2 3.1
PositiveWaldronsign −56.3 28.9 −181.5 99.5 −1.4 0.9 44.7* 21.5 0.4 1.0 −2.1 3.0
PositiveApley’s 0.21 34.8 −66.3 120.0 −1.5 1.1 66.0* 25.8 0.5 1.3 −9.2* 3.7
(compression)
PositiveApley’s 86.9 52.0 489.3** 179.1 3.9* 1.7 −99.4* 38.3 −1.4 1.9 9.9 5.9
(distraction)
PositiveGrind 6.1 26.1 75.0 90.0 0.01 0.8 0.7 19.3 1.7 0.9 0.1 2.7
PositiveOber −21.6 27.8 −51.3 95.7 1.7 0.9 −44.7* 20.2 −2.2* 1.0 2.7 2.9
Age −2.0 1.1 −2.0 3.9 0.02 0.04 −0.6 0.8 −0.1 0.04 0.1 0.1
Female 37.4 27.3 170.2 94.0 1.6 0.9 −20.3 20.1 −2.6** 1.0 0.9 2.9
Multivariateregressionadjustedforageandsex.Significantassociation(P<0.05)inbold
Notationsforlevelofp-valuesasfollows:*p<0.05,**p<0.01
aWOMAC=WesternOntarioandMcMasterIndex
b20mwalktestisthetotaltimeittakestowalk20m
cThe6-minwalktestmeasuresthedistancecoveredduringthe6-minwalk
dTheBergBalanceScaleassessedbalanceduringperformanceof14functionaltasks
eThetestmeasuresthetimetakentocomplete10fullstandsfromasittingpositionasfastaspossible
P
a
g
e
8
o
f
1
2
Iversenetal.BMCMusculoskeletalDisorders (2016) 17:273 Page9of12
self-reported function and PE findings or functional per- may be less useful in clinical practice when examining
formance in adults with KOA [10–15], there is limited patientswith KOA[11].
literature providing this level of detail on the clinical With respect to self-reported outcomes, we found no
presentation of adults with symptomatic KOA and significant associations between PE test results and self-
which address the relationship between self-reported reported pain and found only one significant association
function, performance-based function and special PE with self-reported function, a positive Apley’s test.
maneuvers. Recognizing there are no specific PE maneu- Whereas, Wood et al. found moderate correlations be-
vers to detect KOA, the procedures used in this study tween WOMAC function scores and specific PE tests, in-
aretypicalPEmaneuvers routinelyused when evaluating cluding: tenderness on palpation of the infrapatellar area,
knee function. Given the time constraints in clinical maximal isometric quadriceps femoris muscle strength,
practice settings, this study focused on common PE spe- reproductionofsymptomsonpatellofemoralcompression,
cial test maneuvers to determine whether these tests and knee flexion range-of-motion. In this study, PE ma-
contributed to the variance in functional performance neuvers explained between 7 and 13 % of the variance in
andself-reportoffunction. WOMACfunctionscores,aftercontrollingforage,gender,
We demonstrated that lack of hip muscle flexibility is andbodymassindex[13].Thedifferencesinourstudyre-
prevalent, especially in the rectus femoris and iliotibial sults may be attributed in part to sample size, as Wood’s
band. These muscle groups may be less flexible due to sample included over 800 individuals and this study in-
biomechanical changes in the knee joint, knee pain, pain cluded 87 adults. Additionally, Wood’s sample included
centralization and subsequent restriction of functional adults with general knee pain while our sample consisted
activities [35]. Patients in this sample also walked at ofadultswithsymptomaticKOA.
slower speeds. According to Middleton et al., a walking Evidence partially supported our hypothesis that spe-
speed of 1.1 m/s is considered to be lower than average, cial tests of muscle strength and flexibility would have
since normalwalkingspeed ingeneralis1.2-1.4m/s,not greater association with performance test results than
basedonageorgender[36]. patient self-reported function and pain. We could deter-
Hip muscle weakness, defined as a grade of≥4 on mine the relative contribution of special test results on
MMT, of the hip flexors, extensors, abductors and ad- functional performance, specifically balance, functional
ductors was present in more than 50 % of participants. exercise capacity and gait speed. For example, rectis
These results are consistent with previous studies that femoris tightness was significantly associated with func-
demonstrate significant hip muscle weakness in individ- tional aerobic capacity (6 min Walk Test) and poorer
uals with KOA compared with asymptomatic controls balance(Bergtest). Apositive Waldronorsquat testwas
[37–39]. Zeni et al. found lower extremity weakness in not associated with patient-reported knee function al-
participants with end stage hip osteoarthritis is related though participants in this study were relatively high
to worse scores on functional performance tests functioning, as indicated by their WOMAC function
whereas, hip pain is related to worse scores on question- score. The lack of association between certain clinical
naires that capture self-perceptions of function [40]. maneuvers (special tests) and functional test perform-
Thereisapossibilitythatthisisalsotrueforparticipants ance may be explained by the fact that anatomical struc-
with KOA. tures such as the meniscus and ligaments may be
Quadriceps muscle weakness is a well-documented provoked with clinical maneuvers (forced rotation or
clinical sign of KOA and has been attributed to activity anterior-posterior displacement) but are not stressed in
limitations related to knee pain [3]. In this sample, 37 the same manner with ambulation. Also, as special tests
participants (43 %) had quadriceps weakness on manual aredesignedtoassess theintegrity ofspecificanatomical
examination and 21 (24 %) had hamstring weakness. structures, it is not surprising that a positive test for a
The combination of special tests for patella femoral in- single anatomical structure would not be associated with
volvement were positive in more than half of the partici- global self-reported knee function and pain. The lack of
pants, suggesting knee arthritis was also affecting this association between a positive Lachman’s test with func-
area of the knee joint. Degenerative meniscus disease is tional performance is not surprising as participants with
a known component of KOA and we found one third of KOA often have stiff, stable knees and special tests for
theparticipantshadapositiveMcMurray’stest.Fewpar- cruciate ligament integrity in adults with KOA have lim-
ticipantshad positive testsfor ligamentous instability,al- itedreliabilityandvalidity(Table1)[11,41].
though that is not unusualin a samplewith older adults. In regression models, after adjusting for age and gen-
Additionally, special tests for ligament instability have der, there was some evidence that special tests of knee
been reported to have poor inter-rater reliability (preva- and hip structures, are associated with functional per-
lence and bias-adjusted kappas of 0.02 to 0.3) and lim- formance but less so with self-reported knee function
ited validity in this patient population. Thus, these tests and pain. These results are also in accordance with
Iversenetal.BMCMusculoskeletalDisorders (2016) 17:273 Page10of12
Tevald and colleagues who reported assessment of lower Conclusions
extremity strength is more closely associated with per- In conclusion, this study provides a rich clinical descrip-
formancebasedtestresultsthanself-reportedkneefunc- tion of patient with symptomatic KOA. The results from
tion andpain[37]. this study are exploratory but suggest that hip muscle
Our data suggest lower extremity muscle weakness and strength and flexibility need to be formally addressed in
tightness in combination with knee pain were impacting KOA management. The study also suggests that positive
walking endurance (6 Min Walk). Several studies have findings on clinical examination maneuvers are more
concluded that pain, muscle strength, obesity and age are stronglyassociatedwithfunctionalperformanceteststhan
more important determinants of functional impairment patient self-report of knee pain and function. Although,
thanradiologicalseverityofKOA[42–44]. specialtestsofligamentsmayprovidelittleinformationof
Our findings may have clinical implications. These clinical value, as supported by studies of their reliability
preliminary data support the use of select PE proce- and validity [11, 41]. We propose the use of functional
dures, muscle flexibility and strength of the hip and performance tests during the physical examination, when
knee, when evaluating patients with KOA. The data sug- positive findings on special tests are present and as the
gest specialtestsofligamentintegrity provide clinicalin- use of these performance-based may better inform phys-
formation of little value. However, it is possible that the icaltherapyinterventions.
detection of any associations we found is limited by the
range of PE procedures selected, and their correspond- Abbreviations
ing validity and reliability in patients with KOA. We can BMI,bodymassindex;KOA,kneeosteoarthritis;MMT,manualmuscletesting;
NSAIDs,nonsteroidalanti-inflammatorydrugs;OARSI,OsteoarthritisResearch
only speculate, but we agree with Hinman et al., that
SocietyInternational;PE,physicalexamination;PTs,physicaltherapists;SD,
there might be a benefit to include hip muscle strength- standarddeviation;WOMAC,WesternOntarioandMcMasterOsteoarthritis
ening into rehabilitationprograms for patients with knee Index
OA[39].
Acknowledgements
Thisstudyhasanumberofstrengths.Theperformance- Notapplicable.
basedoutcomemeasuresused(WOMAC,Bergetc.)areall
Funding
validandreliablemeasureswithnormativevaluesavailable
ThisstudyissupportedbytheNationalCenterforComplementaryand
forpatientswitheitherKOAorofthisagegroup.ThePTs IntegrativeHealth(NCCIH)oftheNationalInstitutesofHealth.
received formal instruction in the consistent application of (R01AT00552,1K24AT007323,UL1RR025752,UL1TR000073and
UL1TR001064).Thecontentsofthismanuscriptaresolelytheresponsibility
thespecialtestsandinthedocumentationoftheirfindings
oftheauthorsanddonotnecessarilyrepresenttheofficialviewsofthe
to reduce reporting error and improve consistency of test- NCCIH.Theinvestigatorsaresolelyresponsibleforthecontentofthe
ing. Finally, patients were mixed with respect to race and manuscriptandthedecisiontosubmitforpublication.Thefundershadno
roleinstudydesign,datacollectionandanalysis,decisiontopublish,or
socio-demographics to produce a more representative pic-
preparationofthemanuscript.
tureofadultswithKOA.
Several limitations exist. The cross-sectional design Authors'contributions
Allauthorsapprovedthefinalversionforpublication.MDIdevelopedthe
does not allow for statements about cause and effect.
researchquestionandhadprimaryresponsibilityforstudydesign,
The small sample might result in a type II error. There interpretationofdataandwritingofthemanuscript.LLPconductedthedata
is also a potential for selection bias (WOMAC pain cleaning,dataanalysesandmanuscriptediting.JVHparticipatedin
interpretationofdata,andinwritingthemanuscript.WFHparticipatedin
threshold for study entry), however, there was a rela-
identifyingandrecruitingparticipants,acquisitionofdata,andmanuscript
tively large variability in clinical presentations as well as editing.CWwasprincipalinvestigatorforthemainstudyandparticipatedin
the radiological severity of KOA. These participants studydesign,identifyingandrecruitingparticipants,andmanuscriptediting.
could potentially have co-existing hip osteoarthritis,
Competinginterests
which could explain hip muscle weakness but this can- Theauthorshavenofinancialrelationshiprelevanttothisarticletodisclose.
not be confirmed. There is also a potential for variability Theauthorsdeclarethattheyhavenocompetinginterests.
in the application of the PE maneuvers despite PT train-
Consentforpublication
ing and the use of a standard protocol. Additionally, Notapplicable
manual muscle testing may not be sufficiently sensitive
Ethicsapprovalandconsenttoparticipate
to measure strength compared to other more objective
InstitutionalHumanSubjectsapprovalwasobtainedforthissecondarystudy
methods [45]. Finally, as we enrolled participants with fromTuftsMedicalCenter,TuftsUniversitySchoolofMedicine,Boston,MA,
bilateral KOA, individual’s perception of pain and func- USA.Humansubjectapprovalwasobtainedandallparticipantsconsented
toparticipate.
tional ability may impact self-reported outcomes [46].
Hips were assessed by PTs (neurologic and joint range Authordetails
of motion). However, we did not collect additional PE 1DepartmentofPhysicalTherapy,MovementandRehabilitationSciences,
NortheasternUniversity,360HuntingtonAvenue301CRB,Boston,MA
data on hip (e.g.,scour test) and lumbar in the compara-
02115,USA.2DepartmentofMedicine,SectionofClinicalSciences,Division
tive effectiveness trial. ofRheumatology,Immunology&Allergy,Brigham&Women’sHospital,