Scale for the assessment and rating of ataxia: Development of a new clinical scale

Scale for the assessment and ratingof ataxia

Development of a new clinical scaleT Schmitz-Hubsch MD S Tezenas du Montcel MD PhD L Baliko MD J Berciano MD S Boesch MD

C Depondt MD P Giunti MD C Globas MD J Infante MD J-S Kang MD B Kremer MDC Mariotti MD B Melegh MD PhD M Pandolfo MD M Rakowicz MD P Ribai MD R Rola MDL Schols MD S Szymanski MD BP van de Warrenburg MD A Durr MD and T Klockgether MD

AbstractmdashObjective To develop a reliable and valid clinical scale measuring the severity of ataxia Methods The authorsdevised the Scale for the Assessment and Rating of Ataxia (SARA) and tested it in two trials of 167 and 119 patients withspinocerebellar ataxia Results The mean time to administer SARA in patients was 142 75 minutes (range 5 to 40)Interrater reliability was high with an intraclass coefficient (ICC) of 098 Test-retest reliability was high with an ICC of090 Internal consistency was high as indicated by Cronbachrsquos of 094 Factorial analysis revealed that the ratingresults were determined by a single factor SARA ratings showed a linear relation to global assessments using a visualanalogue scale suggesting linearity of the scale (p 00001 r2 098) SARA score increased with the disease stage (p 0001) and was closely correlated with the Barthel Index (r 080 p 0001) and part IV (functional assessment) of theUnified Huntingtonrsquos Disease Rating Scale (UHDRS-IV) (r 089 p 00001) whereas it had only a weak correlationwith disease duration (r 034 p 00002) Conclusions The Scale for the Assessment and Rating of Ataxia is a reliableand valid measure of ataxia making it an appropriate primary outcome measure for clinical trials

NEUROLOGY 2006661717ndash1720

With continuous progress in the understanding ofthe molecular pathogenesis of many ataxia disor-ders novel therapies are on the horizon that willneed to be evaluated in clinical trials An essentialprerequisite for such trials is the availability of vali-dated neurologic assessment methods to measure theseverity of ataxia Although a number of ataxia rat-ing scales have been proposed1-4 there have beenonly few attempts to validate these scales A widely

used ataxia rating scale is the International Cooper-ative Ataxia Rating Scale (ICARS) Although inter-rater reliability of this scale was shown to be high5more detailed evaluations revealed problems con-cerning practicability and subscale structure thatquestioned the usefulness of ICARS for future inter-ventional trials6 Recently a new scale FriedreichrsquosAtaxia Rating Scale was devised and validated4

However this scale is not readily applicable to cere-bellar ataxias because it was designed as a disease-specific scale for the evaluation of Friedreich ataxiaand thus also considers other symptoms than ataxia

We therefore devised a new scale named Scale forthe Assessment and Rating of Ataxia (SARA) that isbased on a semiquantitative assessment of cerebellar

Additional material related to this article can be found on the NeurologyWeb site Go to wwwneurologyorg and scroll down the Table of Con-tents for the June 13 issue to find the title link for this article

See also page 1711

From the Department of Neurology (TS-H TK) University Hospital of Bonn Bonn Germany Department of Biostatistics and Medical Informatics(STM) Hopital de la Pitie-Salpetriere Assistance Publique-Hopitaux Paris France Modeling in Clinical Research (STM) EA 3974 University Pierre etMarie Curie Paris France Department of Neurology and Stroke (LB) County Hospital Veszprem Hungary Department of Neurology (JB JI)University Hospital Marques de Valldecilla Santander Spain Department of Neurology (SB) University of Innsbruck Innsbruck Austria Department ofNeurology (CD MP) Hopital Erasme Universite Libre de Bruxelles Brussels Belgium Department of Biochemistry and Genetics (RF CM) IstitutoNazionale Neurologico C Besta Milan Italy Department of Molecular Neuroscience (PG) Institute of Neurology Queen Square London United KingdomDepartment of Neurology and Hertie-Institute for Clinical Brain Research (CG LS) University of Tubingen Tubingen Germany Department ofNeurology (J-SK) University of Frankfurt FrankfurtM Germany Department of Neurology (BK BPW) Radboud University Nijmegen MedicalCenter Nijmegen Netherlands Department of Medical Genetics and Child Development (MB) University of Pecs Pecs Hungary Institute of Psychiatryand Neurology (MR RR) Warsaw Poland INSERM U679 and Department of Genetics (PR AD) Cytogenetics and Embryology Hopital de laPitie-Salpetriere Paris France Department of Neurology (SS) St Josef Hospital University Hospital of Bochum Bochum GermanySupported by grant EUROSCALSHM-CT-2004-503304 from the European Union grant GeneMove01 GM 0503 from the German Ministry of Education andResearch and grant 3 PO5B 019 24 from the Polish Ministry of Scientific Research and Information TechnologyDisclosure The authors report no conflicts of interestReceived October 10 2005 Accepted in final form March 6 2006Address correspondence and reprint requests to Dr Thomas Klockgether Department of Neurology University Hospital of Bonn Sigmund-Freud-Str 25D-53105 Bonn Germany e-mail klockgetheruni-bonnde

Copyright copy 2006 by AAN Enterprises Inc 1717

ataxia on an impairment level We report here theresults of a large two-step validation trial performedin patients with spinocerebellar ataxia (SCA)

Methods Description of the scale SARA was devised duringa meeting held in February 2004 in Bonn Germany by a groupof European neurologists Test items were selected according tospecificity for ataxia and the possibility to standardize testingand rating procedures Moreover assessment had to be close tostandard neurologic examination and not require technicalequipment In its initial form SARA was composed of nineitems including one item related to oculomotor function Basedon the results of Trial 1 SARA was modified and the oculomo-tor item was omitted In its final form SARA has eight itemsthat yield a total score of 0 (no ataxia) to 40 (most severeataxia) 1 gait (score 0 to 8) 2 stance (score 0 to 6) 3 sitting(score 0 to 4) 4 speech disturbance (score 0 to 6) 5 fingerchase (score 0 to 4) 6 nose-finger test (score 0 to 4) 7 fastalternating hand movements (score 0 to 4) 8 heel-shin slide(score 0 to 4) Limb kinetic functions (items 5 to 8) are ratedindependently for both sides and the arithmetic mean of bothsides is included in the SARA total score (see appendix E-1 onthe Neurology Web site at wwwneurologyorg)

Development of the scale The study population consisted of167 patients with SCA (SCA1 23 patients SCA2 56 patientsSCA3 23 patients SCA6 18 patients SCA7 seven patientsSCA14 two patients SCA17 one patient unknown mutation 29patients) and eight controls in Trial 1 and 119 patients with SCA(SCA1 15 patients SCA2 28 patients SCA3 26 patients SCA619 patients SCA7 eight patients SCA14 one patient SCA17three patients SCA23 one patient unknown mutation 18 pa-tients) and 110 controls in Trial 2 Clinical characteristics aregiven in table 1 Inclusion criteria for the patients with SCA were1) progressive ataxia and 2) positive genetic testing for any SCAmutation or linkage to any SCA locus or clinical evidence of auto-somal dominant inheritance The majority of patients (Trial 1 n 97 Trial 2 n 95) were rated independently by two investiga-tors In Trial 2 a number of patients were rated twice with aninterval of 1 to 34 days by as many as three investigators result-ing in a set of 15 intrarater assessments All investigators wereneurologists or senior residents in neurology training There was amaximum of three investigators at each center Investigators didnot receive formal training but were asked to closely follow theinstructions that are part of the SARA scale (see appendix E-1)

In Trial 1 SARA was compared with ataxia disease stagesICARS and the Barthel Index and in Trial 2 with ataxia diseasestages Barthel Index and the part IV (functional assessment) ofthe Unified Huntingtonrsquos Disease Rating Scale (UHDRS-IV)37-9

Disease stages were defined as follows Stage 0 no gait difficul-ties Stage 1 disease onset as defined by onset of gait difficultiesStage 2 loss of independent gait as defined by permanent use of awalking aid or reliance on a supporting arm Stage 3 confinementto wheelchair as defined by permanent use of a wheelchair7 TheICARS results have been reported elsewhere6

Thirty assessments of Trial 2 were video recorded Eleven vid-eos that were representative of the entire range of clinical severitywere selected and randomly ordered on a video disk From thesevideos SARA rating was done by two examiners who were not

involved in the clinical assessments In addition the severity ofataxia of the 11 cases presented on video was globally assessed by14 examiners on a visual analogue scale (VAS) ranging from 0 (noataxia) to 100 (most severe ataxia)

The study was approved by the ethics committees of the partic-ipating centers Informed and written consent was obtained fromall study participants

Statistical analyses Data are given as mean SD and rangeInterrater and test-retest reliability was expressed using intra-class coefficients (ICCs) internal consistency using Cronbachrsquos coefficient Calculation of ICCs was based on a model with onlyrandom effects Coefficients exceeding 080 are considered accept-able for scales used to make group comparisons In order to ana-lyze internal validity a principal component analysis wasperformed using varimax rotation

In order to test the linearity of SARA a regression analysis ofthe video ratings with the VAS assessments as a reference wasperformed To make SARA and the VAS values directly compara-ble SARA values were multiplied with a factor of 25 to yield arange from 0 to 100 As SARA will be used to measure differencesbetween two assessments ie before and after treatment we alsoverified the linearity of differences between two measures To thisend all differences between patient ratings of a given examinerwere computed For both analyses a linear model was tested Forthe SARA scores a model with no intercept was used and for thedifferences a model with an intercept was used

To test for correlation between SARA scores and Barthel In-dex UHDRS-IV and disease duration Pearsonrsquos correlation testwas used The relation between SARA scores and disease stageswas analyzed with analysis of variance All tests were two sidedp 005 was considered significant Statistical analyses wereperformed using the SAS 82 statistical package (SAS InstituteCary NC)

Results Analysis of Trial 1 showed high interrater reli-ability of SARA total score (ICC 097) and of the itemsrelated to gait stance sitting fast alternating hand move-ments and heel-shin slide In contrast the items relatedto finger chase nose-finger test speech and oculomotorfunction were less successful with ICC 080 (range 070to 079) Internal consistency was high with Cronbachrsquos of 090 it increased when the oculomotor item was omit-ted A factorial analysis showed that the results of allitems with the exception of the oculomotor item weredetermined by a single factor Based on these results in-structions for the items with interrater reliability 080were modified and the oculomotor item was omitted

Analysis of Trial 2 determined the metric properties ofSARA in its final form (see appendix E-1) In the patientgroup the mean SARA score was 159 85 (range 15 to40) Ceiling effects were negligible with one patient(084) receiving the maximal possible score Skewness ofthe distribution was 074 The mean SARA score of thecontrol group was 04 11 (range 0 to 75) and 79 of

Table 1 Clinical characteristics of the study population

Patients in trial 1 Patients in trial 2 Controls in trial 2

No 167 119 110

FM 7289 6158 6743

Age y 492 142 (13ndash80) 503 132 (22ndash81) 471 150 (19ndash78)

Disease duration y 108 44 (0ndash46) 133 83 (1ndash43)

Barthel Index 889 200 (5ndash100) 826 258 (0ndash100) 1000 00 (100ndash100)

UHDRS-IV nd 167 80 (0ndash25) 250 00 (25ndash25)

Age disease duration Barthel Index and UHDRS-IV are given as mean SD The range is given in parentheses

UHDRS-IV Unified Huntingtonrsquos Disease Rating Scale Part IV (functional assessment) nd not determined

1718 NEUROLOGY 66 June (1 of 2) 2006

controls had a score of 0 Positive ratings in controls weremainly obtained in items 5 and 8 left Mean time to admin-ister SARA was 142 75 minutes (range 5 to 40) inpatients and 72 26 minutes (range 3 to 13) in controls

Based on the results obtained in 95 patients assessed bytwo examiners interrater reliability was very high withan ICC of 098 All single items had good interrater reli-ability with ICC 080 with exception of item 6 left(ICC 076) and item 8 left (ICC 074) Based on alimited set of 15 intrarater assessments test-retest reli-ability was high with an ICC of 090 Internal consistencywas high as indicated by Cronbachrsquos of 094 it did notincrease when any item was deleted (table 2)

Factorial analysis revealed that the rating results in allitems were determined by a single factor with an eigen-value of 64 which explained 80 of the variance All otherfactors had eigenvalues far less than 1

Linearity of SARA was tested by a regression analysisof video ratings of 11 patients with global assessments as areference The SARA ratings and the differences betweenmeasures fitted a linear model (SARA p 00001 r2 098 differences p 00001 r2 072) (figure 1)

The SARA score increased with the disease stage (p 0001) (figure 2A) The SARA score was closely correlatedwith the Barthel Index (r 080 p 00001) andUHDRS-IV (r 089 p 00001) while it had only aweak correlation with disease duration (r 034 p 00002) (figure 2 B through D) Further analysis showedthat the correlation between SARA score and disease dura-tion was better when only SCA-1 (r 074 p 0003) orSCA3 patients (r 059 p 0002) were considered Thiswas not found for SCA2 (r 008 p 068) and SCA6(r 026 p 028)

Discussion SARA the new ataxia scale reportedhere includes eight items reflecting neurologic man-ifestations of cerebellar ataxia SARA only ratesataxia-related symptoms and does not consider non-ataxia symptoms that often occur in patients with

SCA Therefore it is possible that disease severity incertain diseases with extracerebellar features mightnot be faithfully reflected in the SARA score How-ever consideration of nonataxia symptoms wouldhave precluded the development of a scale thatmeets basic biometric requirements To account fornonataxia symptoms we have devised an Inventoryof Non-Ataxia Symptoms (INAS) that provides struc-tured and semiquantitative information on non-ataxia symptoms The items selected for SARAfollow the steps of a standard neurologic examina-tion Oculomotor functions which were part of theinitial SARA version are not considered in the finalSARA version because the first validation trial indi-cated two underlying constructs one globally reflect-ing cerebellar ataxia and the second reflecting solelyoculomotor dysfunction

Table 2 Reliability and internal consistency of SARA

ItemInterrater

reliability (ICC)Test-retest

reliability (ICC) Cronbachrsquos

SARA 1 096644 nd 0930805

SARA2 095106 nd 0929620

SARA3 092363 nd 0933156

SARA4 085085 nd 0934190

SARA5 right 085446 nd 0936336

SARA5 left 085367 nd 0935705

SARA6 right 080507 nd 0932309

SARA6 left 075753 nd 0933658

SARA7 right 087335 nd 0931609

SARA7 left 081716 nd 0933208

SARA8 right 080890 nd 0933362

SARA8 left 074417 nd 0933398

SARA total 097776 090032 0938361

ICC intraclass coefficient SARA Scale for the Assessmentand Rating of Ataxia nd not determined

Figure 1 Regression analysis of Scale for the Assessmentand Rating of Ataxia (SARA) video ratings of 11 represen-tative patients with global assessments as a referenceSARA rating was done by two examiners who were notinvolved in the clinical assessments SARA values weremultiplied with a factor of 25 to yield a range from 0 to100 Severity of ataxia was globally assessed by 14 exam-iners on a visual analogue scale (VAS) ranging from 0 (noataxia) to 100 (most severe ataxia) (A) SARA ratings (p 00001 r2 098) (B) Differences between SARA ratingsof a given examiner (p 00001 r 2 072)

Figure 2 Validation of Scale for the Assessment and Rat-ing of Ataxia (SARA) with external measures (A) SARAscores at different ataxia disease stages Data are given asbox plots (B) Correlation of the SARA score with BarthelIndex (r 080 p 00001) (C) Correlation withUHDRS-IV (r 089 p 00001) (D) Correlation withdisease duration (r 034 p 00002)

June (1 of 2) 2006 NEUROLOGY 66 1719

SARA underwent a rigorous validation procedureinvolving two clinical trials in large groups of pa-tients with SCA and controls The number of pa-tients with SCA included in these trials is far greaterthan the number included in any previous clinicalstudy of patients with SCA Clinical presentation ofmost patients with SCA is characterized by promi-nent cerebellar ataxia They therefore represent anappropriate study population for the validation of anataxia scale To demonstrate that SARA is also ap-propriate for other groups of ataxia patients we arecurrently conducting further clinical studies

There was a clear separation of the SARA ratingsin patients and controls Nevertheless about 20 ofcontrols had positive ratings in at least one itemThe items producing the majority of positive ratingswere related to kinetic functions of the nondominantside These observations show that SARA is not anideal clinical instrument to detect disease onsetElimination of these items would increase the discrim-inatory power of SARA but reduce its sensitivity

SARA was easy to administer and required lessthan 15 minutes per patient This time is consider-ably shorter than the time required to completeICARS6 SARA satisfied a number of accepted crite-ria of reliability Interrater reliability was high bothfor the total score and for most of the items At thesingle-item level SARA performed better thanICARS6 Although based on a small number of obser-vations our data suggest acceptable test-retest reli-ability of the total SARA score However meaningfulanalysis of the test-retest reliability of each item wasnot possible Because the calculation of ICCs wasbased on a model with only random effects the re-sults are independent of the investigators and can begeneralized to investigators with the same level oftraining as the investigators of this study The inter-nal consistency of SARA was high as indicated byCronbachrsquos of 094

Most importantly SARA results were determinedby a single factor which indicates that SARA mea-sures a common underlying construct cerebellarataxia In this respect SARA favorably contrastswith ICARS the results of which were shown to beinfluenced by four different factors that did not coin-cide with the ICARS subscales6

Linearity of a scale and particularly of the differ-ences between ratings is an important prerequisitefor the use of a scale in clinical trials that includepatients with a wide spectrum of clinical severity Todemonstrate linearity we performed a regressionanalysis of the SARA ratings of two examiners withglobal assessments of 14 examiners as a referenceOur analysis showed that SARA ratings and the dif-ferences between ratings fitted a linear model A lim-itation of this analysis comes from the fact that theseratings were done on videos

To further demonstrate the validity of SARA we

used external measures SARA scores increased withataxia disease stages as defined in a previousstudy7 In addition SARA scores decreased with theBarthel Index which measures activities of daily liv-ing and is routinely used to evaluate stroke pa-tients8 SARA also decreased with UHDRS-IV whichprovides a functional assessment of patients withHuntington disease (HD)9 Because patients withataxia and patients with HD face qualitatively simi-lar functional limitations in daily life it appearedappropriate to use this scale as an external valida-tion criterion These observations are consistent withthe hypothesis that SARA is a valid method to mea-sure the severity of ataxia Based on the clinicalknowledge that SCAs are progressive but that theprogression rate is variable depending on the geno-type repeat length and unknown genetic and nonge-netic factors710 only a moderate correlation betweenthe SARA score and disease duration was expectedThis was indeed found in the present study Sepa-rate analysis of the four most common genotypesshowed that the correlation between SARA score anddisease duration was better when only patients withSCA1 or SCA3 were considered This was not foundfor SCA2 and SCA6 However the small number ofpatients in each subgroup did not allow us to drawany final conclusions

AcknowledgmentThe authors thank the probands for participating in this study Thework of MEEC Glazowski (Warsaw) in videotaping of assessmentsis gratefully acknowledged They also thank all other clinical investi-gators who contributed to data acquisition S Bonato (Brussels) MAbele M Minnerop B Viebahn and I Vogt (Bonn) S Romano(Milan) WF Abdo and HJ Schelhaas (Nijmegen) C Charles (Par-is) and J Zaremba and E Zdzienicka (Warsaw)

References1 Fillyaw MJ Badger GJ Bradley WG et al Quantitative measures of

neurological function in chronic neuromuscular diseases and ataxia JNeurol Sci 19899217ndash36

2 Klockgether T Schroth G Diener HC Dichgans J Idiopathic cerebellarataxia of late onset natural history and MRI morphology J NeurolNeurosurg Psychiatry 199053297ndash305

3 Trouillas P Takayanagi T Hallett M et al International CooperativeAtaxia Rating Scale for pharmacological assessment of the cerebellarsyndrome J Neurol Sci 1997145205ndash211

4 Subramony SH May W Lynch D et al Measuring Friedreich ataxiainterrater reliability of a neurologic rating scale Neurology 2005641261ndash1262

5 Storey E Tuck K Hester R Hughes A Churchyard A Inter-raterreliability of the International Cooperative Ataxia Rating Scale (IC-ARS) Mov Disord 200419190ndash192

6 Schmitz-Hubsch T Tezenas du Montcel S Baliko L et al Reliabilityand validity of the International Cooperative Ataxia Rating Scale (IC-ARS) A study in 156 spinocerebellar ataxia (SCA) patients Mov Disord200621699ndash704

7 Klockgether T Ludtke R Kramer B et al The natural history of degen-erative ataxia a retrospective study in 466 patients Brain 1998121589ndash600

8 Mahoney FI Barthel DW Functional evaluation the Barthel index MdState Med J 19651461ndash65

9 Huntington Study Group Unified Huntingtonrsquos Disease Rating Scalereliability and consistency Mov Disord 199611136ndash142

10 Schols L Kruger R Amoiridis G Przuntek H Epplen JT Riess OSpinocerebellar ataxia type 6 genotype and phenotype in German kin-dreds J Neurol Neurosurg Psychiatry 19986467ndash73


DOI 10121201wnl000021904260538922006661717-1720 Neurology

T Schmitz-Huumlbsch S Tezenas du Montcel L Baliko et al Scale for the assessment and rating of ataxia Development of a new clinical scale

This information is current as of June 12 2006

ServicesUpdated Information amp

httpwwwneurologyorgcontent66111717fullhtmlincluding high resolution figures can be found at

Supplementary Material

tmlhttpwwwneurologyorgcontentsuppl2006060766111717DC1hSupplementary material can be found at

References httpwwwneurologyorgcontent66111717fullhtmlref-list-1

This article cites 10 articles 4 of which you can access for free at

Citations httpwwwneurologyorgcontent66111717fullhtmlotherarticles

This article has been cited by 42 HighWire-hosted articles

Subspecialty Collections

httpwwwneurologyorgcgicollectiongait_disorders_ataxiaGait disordersataxia

study_design_httpwwwneurologyorgcgicollectionclinical_trials_methodology_Clinical trials Methodologystudy designfollowing collection(s) This article along with others on similar topics appears in the

Permissions amp Licensing

httpwwwneurologyorgmiscaboutxhtmlpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpwwwneurologyorgmiscaddirxhtmlreprintsusInformation about ordering reprints can be found online

Online ISSN 1526-632X1951 it is now a weekly with 48 issues per year Copyright All rights reserved Print ISSN 0028-3878

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology

ataxia on an impairment level We report here theresults of a large two-step validation trial performedin patients with spinocerebellar ataxia (SCA)

Methods Description of the scale SARA was devised duringa meeting held in February 2004 in Bonn Germany by a groupof European neurologists Test items were selected according tospecificity for ataxia and the possibility to standardize testingand rating procedures Moreover assessment had to be close tostandard neurologic examination and not require technicalequipment In its initial form SARA was composed of nineitems including one item related to oculomotor function Basedon the results of Trial 1 SARA was modified and the oculomo-tor item was omitted In its final form SARA has eight itemsthat yield a total score of 0 (no ataxia) to 40 (most severeataxia) 1 gait (score 0 to 8) 2 stance (score 0 to 6) 3 sitting(score 0 to 4) 4 speech disturbance (score 0 to 6) 5 fingerchase (score 0 to 4) 6 nose-finger test (score 0 to 4) 7 fastalternating hand movements (score 0 to 4) 8 heel-shin slide(score 0 to 4) Limb kinetic functions (items 5 to 8) are ratedindependently for both sides and the arithmetic mean of bothsides is included in the SARA total score (see appendix E-1 onthe Neurology Web site at wwwneurologyorg)

Development of the scale The study population consisted of167 patients with SCA (SCA1 23 patients SCA2 56 patientsSCA3 23 patients SCA6 18 patients SCA7 seven patientsSCA14 two patients SCA17 one patient unknown mutation 29patients) and eight controls in Trial 1 and 119 patients with SCA(SCA1 15 patients SCA2 28 patients SCA3 26 patients SCA619 patients SCA7 eight patients SCA14 one patient SCA17three patients SCA23 one patient unknown mutation 18 pa-tients) and 110 controls in Trial 2 Clinical characteristics aregiven in table 1 Inclusion criteria for the patients with SCA were1) progressive ataxia and 2) positive genetic testing for any SCAmutation or linkage to any SCA locus or clinical evidence of auto-somal dominant inheritance The majority of patients (Trial 1 n 97 Trial 2 n 95) were rated independently by two investiga-tors In Trial 2 a number of patients were rated twice with aninterval of 1 to 34 days by as many as three investigators result-ing in a set of 15 intrarater assessments All investigators wereneurologists or senior residents in neurology training There was amaximum of three investigators at each center Investigators didnot receive formal training but were asked to closely follow theinstructions that are part of the SARA scale (see appendix E-1)

In Trial 1 SARA was compared with ataxia disease stagesICARS and the Barthel Index and in Trial 2 with ataxia diseasestages Barthel Index and the part IV (functional assessment) ofthe Unified Huntingtonrsquos Disease Rating Scale (UHDRS-IV)37-9

Disease stages were defined as follows Stage 0 no gait difficul-ties Stage 1 disease onset as defined by onset of gait difficultiesStage 2 loss of independent gait as defined by permanent use of awalking aid or reliance on a supporting arm Stage 3 confinementto wheelchair as defined by permanent use of a wheelchair7 TheICARS results have been reported elsewhere6

Thirty assessments of Trial 2 were video recorded Eleven vid-eos that were representative of the entire range of clinical severitywere selected and randomly ordered on a video disk From thesevideos SARA rating was done by two examiners who were not

involved in the clinical assessments In addition the severity ofataxia of the 11 cases presented on video was globally assessed by14 examiners on a visual analogue scale (VAS) ranging from 0 (noataxia) to 100 (most severe ataxia)

The study was approved by the ethics committees of the partic-ipating centers Informed and written consent was obtained fromall study participants

Statistical analyses Data are given as mean SD and rangeInterrater and test-retest reliability was expressed using intra-class coefficients (ICCs) internal consistency using Cronbachrsquos coefficient Calculation of ICCs was based on a model with onlyrandom effects Coefficients exceeding 080 are considered accept-able for scales used to make group comparisons In order to ana-lyze internal validity a principal component analysis wasperformed using varimax rotation

In order to test the linearity of SARA a regression analysis ofthe video ratings with the VAS assessments as a reference wasperformed To make SARA and the VAS values directly compara-ble SARA values were multiplied with a factor of 25 to yield arange from 0 to 100 As SARA will be used to measure differencesbetween two assessments ie before and after treatment we alsoverified the linearity of differences between two measures To thisend all differences between patient ratings of a given examinerwere computed For both analyses a linear model was tested Forthe SARA scores a model with no intercept was used and for thedifferences a model with an intercept was used

To test for correlation between SARA scores and Barthel In-dex UHDRS-IV and disease duration Pearsonrsquos correlation testwas used The relation between SARA scores and disease stageswas analyzed with analysis of variance All tests were two sidedp 005 was considered significant Statistical analyses wereperformed using the SAS 82 statistical package (SAS InstituteCary NC)

Results Analysis of Trial 1 showed high interrater reli-ability of SARA total score (ICC 097) and of the itemsrelated to gait stance sitting fast alternating hand move-ments and heel-shin slide In contrast the items relatedto finger chase nose-finger test speech and oculomotorfunction were less successful with ICC 080 (range 070to 079) Internal consistency was high with Cronbachrsquos of 090 it increased when the oculomotor item was omit-ted A factorial analysis showed that the results of allitems with the exception of the oculomotor item weredetermined by a single factor Based on these results in-structions for the items with interrater reliability 080were modified and the oculomotor item was omitted

Analysis of Trial 2 determined the metric properties ofSARA in its final form (see appendix E-1) In the patientgroup the mean SARA score was 159 85 (range 15 to40) Ceiling effects were negligible with one patient(084) receiving the maximal possible score Skewness ofthe distribution was 074 The mean SARA score of thecontrol group was 04 11 (range 0 to 75) and 79 of

Table 1 Clinical characteristics of the study population

Patients in trial 1 Patients in trial 2 Controls in trial 2

No 167 119 110

FM 7289 6158 6743

Age y 492 142 (13ndash80) 503 132 (22ndash81) 471 150 (19ndash78)

Disease duration y 108 44 (0ndash46) 133 83 (1ndash43)

Barthel Index 889 200 (5ndash100) 826 258 (0ndash100) 1000 00 (100ndash100)

UHDRS-IV nd 167 80 (0ndash25) 250 00 (25ndash25)

Age disease duration Barthel Index and UHDRS-IV are given as mean SD The range is given in parentheses

UHDRS-IV Unified Huntingtonrsquos Disease Rating Scale Part IV (functional assessment) nd not determined










ItemInterrater



SARA 1 096644 nd 0930805

SARA2 095106 nd 0929620

SARA3 092363 nd 0933156

SARA4 085085 nd 0934190

SARA5 right 085446 nd 0936336

SARA5 left 085367 nd 0935705

SARA6 right 080507 nd 0932309

SARA6 left 075753 nd 0933658

SARA7 right 087335 nd 0931609

SARA7 left 081716 nd 0933208

SARA8 right 080890 nd 0933362

SARA8 left 074417 nd 0933398

SARA total 097776 090032 0938361









































Reprints












ItemInterrater



SARA 1 096644 nd 0930805

SARA2 095106 nd 0929620

SARA3 092363 nd 0933156

SARA4 085085 nd 0934190

SARA5 right 085446 nd 0936336

SARA5 left 085367 nd 0935705

SARA6 right 080507 nd 0932309

SARA6 left 075753 nd 0933658

SARA7 right 087335 nd 0931609

SARA7 left 081716 nd 0933208

SARA8 right 080890 nd 0933362

SARA8 left 074417 nd 0933398

SARA total 097776 090032 0938361









































Reprints








































Reprints




















Reprints




Documents

Scale for the assessment and rating of ataxia: Development of a new clinical scale