ORIGINAL ARTICLEReliability and Validity of Visual Assessments of GaitUsing a Modified Physician Rating Scale forCrouch and Foot ContactTishya A. L. Wren, PhD,*†‡ Susan A. Rethlefsen, PT,* Bitte S. Healy, MS, PT,* K. Patrick Do, BS,*Sandra W. Dennis, PT, MS,* and Robert M. Kay, MD*†Abstract: This study evaluates the visual assessment of gait usingportions of the Physicians’ Rating Scale (PRS). Thirty children withpathologic gait were evaluated ‘‘live’’ and using full- and slow-speedvideo. Interobserver reliability (weighted kappa) was 0.57 to 0.74for foot contact, 0.69 to 0.71 for crouch, 0.30 to 0.40 for hip flexion,0.57 to 0.65 for knee flexion, and 0.42 to 0.52 for dorsiflexion instance. Intraobserver reliability (comparing the three conditions) was0.50 to 0.78 for foot contact, 0.71 to 0.80 for crouch, 0.26 to 0.44 forhip flexion, 0.60 to 0.86 for knee flexion, and 0.39 to 0.61 fordorsiflexion. Observers were correct only 12% to 32% of the timewhen reporting less than 0 degrees of dorsiflexion and 0% to 29%of the time when reporting more than 20 degrees of hip flexiondue to overestimation of hip flexion and underestimation of ankledorsiflexion. These errors could lead some clinicians to presume thepresence of contractures that do not actually exist. Visual assessmentusing the PRS does not appear to accurately measure what it is mostcommonly used to assess: ankle position in stance.Key Words: gait analysis, crouch, equinus, toe walking(J Pediatr Orthop 2005;25:646–650)Computerized gait analysis is becoming more widely usedin evaluation and treatment planning for children with gaitabnormalities.1,2Even with the less sophisticated technologyavailable decades ago, computerized motion analysis was shownto be more accurate than observational gait analysis forassessment of gait deviations.3Computerized gait analysis isnot available in all communities, since it requires specializedequipment and specially trained personnel. When computerizedgait analysis is not used, gait problems are most often assessedthrough observation in the clinic. The Physicians’ Rating Scale(PRS) is an observational tool that has been used to evaluate gaitand assess the outcome of botulinum toxin injection in childrenwith cerebral palsy.4–9The most useful sections of the PRS are itsassessments of crouch and foot contact patterns.4In the PRS,crouch is defined based on combined hip, knee, and ankleposition during gait,4,6a definition that is limited in scope andmay be a source of confusion for raters.The reliability and validity of observational tools such asthe PRS have not been established. Two studies examined theinterrater reliability of three components of the PRS—crouch,foot contact, and recurvatum—based on review of videotaperecordings.4,5The other components were deemed unlikelyto be discriminatory in one study of botulinum toxin injectionto the plantar flexors, and the knee position (recurvatum)component showed poor interrater reliability.4No studies haveassessed the reliability of ratings performed during live obser-vation, and the validity of the PRS has not been established. Itis particularly important that the reliability and validity of liveratings be evaluated because most observational assessmentsare performed ‘‘live’’ in the clinic.The current study was undertaken in effort to finda valid, reliable global assessment tool for use in the absenceof, or in addition to, computerized gait analysis. The PRSwas used with modifications made to minimize potentialsources of interrater and intrarater error. The purposes of thisstudy were to determine the agreement among modified PRSratings based on live observation, full-speed video, and slow-motion video; to assess the interrater reliability of theseratings; to determine whether interrater reliability is improvedby using video recordings and observing gait in slow motion;and to determine the accuracy of visual joint position ratingsby comparing them to corresponding three-dimensional kine-matic values.METHODSThis study involved 30 children with disabilities (ages5–20 years) referred to our laboratory for evaluation of gaitabnormalities. The study was approved by the human subjectsreview board at our institution. All subjects provided writtenassent, and their parents provided written consent to par-ticipate in the study.Each subject made three round trips on a 15-m path inthe laboratory while being observed by four raters: threeexperienced gait laboratory physical therapists and a gaitlaboratory engineer. The subjects were videotaped walkingFrom *Children’s Orthopaedic Center, Children’s Hospital Los Angeles, LosAngeles, CA; †Department of Orthopaedics, University of SouthernCalifornia, Los Angeles, CA; and ‡Departments of Radiology andBiomedical Engineering, University of Southern California, Los Angeles,CA.Study conducted at Children’s Hospital Los Angeles, Los Angeles, CA.None of the authors received financial support for this study.Reprints: Susan A. Rethlefsen, PT, Children’s Orthopaedic Center, Children’sHospital Los Angeles, 4650 Sunset Blvd., MS 69, Los Angeles, CA 90027(e-mail: [email protected]).Copyright Ó 2005 by Lippincott Williams & Wilkins646 J Pediatr OrthopVolume 25, Number 5, September/October 2005across the laboratory. For 13 subjects, the videotaped walksand the ‘‘live’’ observation walks were the same. For the other17 subjects, the videotaped walks occurred during gait testingwithin 1 hour of the live observations. For these 17 subjects,the live and videotaped walks were different since their livewalks were not videotaped.Each rater scored each subject for foot contact patternand crouch according to a modified version of the PRS4(Table1). The ‘‘gait pattern’’ and ‘‘degree of crouch’’ sections of thePRS were used. In addition to rating crouch qualitatively, theraters gave separate ratings for components of crouch (mini-mum hip flexion, minimum knee flexion, and maximum ankledorsiflexion during stance). This breakdown of the ‘‘crouch’’category of the PRS was included to allow for variations ofcrouched gait (eg, excessive hip/knee flexion with ankleplantarflexion, or excessive hip/knee flexion and ankledorsiflexion) that would not be included in the narrowerdefinition of ‘‘crouch’’ in the PRS.4,6In addition, raters wereinstructed to evaluate minimum hip and knee flexion andmaximum ankle dorsiflexion during the stance phase of gait.The PRS