Elsevier

Gait & Posture

Volume 27, Issue 4, May 2008, Pages 648-652
Gait & Posture

Reliability and validity of the Visual Gait Assessment Scale for children with hemiplegic cerebral palsy when used by experienced and inexperienced observers

https://doi.org/10.1016/j.gaitpost.2007.08.008Get rights and content

Abstract

This study investigated the reliability and validity of the Visual Gait Assessment Scale when used by experienced and inexperienced observers. Four experienced and six inexperienced observers viewed videotaped footage of four children with hemiplegic cerebral palsy on two separate occasions. Validity of the Scale was obtained by comparison with three-dimensional gait analysis (3DGA). The experienced observers generally had higher inter-observer and intra-observer reliability than the inexperienced observers. Both groups showed higher agreement for assessments made at the ankle and foot than at the knee and hip. The experienced observers had slightly higher agreement with 3DGA than the inexperienced observers. The inexperienced observers showed a learning effect and had higher inter-observer agreement and higher agreement with 3DGA in the second assessment of the videotapes. This scale can be used by inexperienced observers but is limited to observations in the sagittal plane and by poor reliability at the knee and hip for experienced and inexperienced observers.

Introduction

Computer aided three dimensional gait analysis (3DGA) has become the gold standard for the assessment of pathological gait. However, 3DGA is not always appropriate for very young children or those with severe disability who may not be able to complete the task [1] and is not available in many parts of the world [2]. Video recording is a cheap and accessible alternative to gait analysis and several observational scoring systems for cerebral palsy (CP) have been developed for use from video footage [1], [3], [4], [5].

The Physicians Rating Scale (PRS) was described by Koman et al. [6] and derivatives have been produced subsequently [1], [7], [8], [9], [10], [11]. As Dickens and Smith [1] have pointed out, new or modified scales require validation for inter-, intra-observer reliability and validity. Validity or reliability of the PRS [6], [11], and the Observational Gait Scale (OGS) [9] do not appear to have been published. Corry et al. [7] calculated the inter-observer reliability of a modified PRS but did not consider the intra-observer reliability. Maathuis et al. [2] found that the PRS had excellent intra-observer reliability and poor inter-observer agreement but did not consider its validity. Flett et al. [8] used a modified PRS in their study of botulinum toxin versus casting for dynamic calf tightness in CP but this was not tested for reliability or validity. Ubhi et al. [10] used a modification of the foot contact subscale of the PRS in their study of the effect of botulinum toxin on walking in CP. This was assessed for intra- and inter-observer reliability but not validity.

In 2006 Dickens and Smith [1] investigated the reliability and validity of the Visual Gait Assessment Scale (VGAS), based on the PRS, for children with hemiplegic cerebral palsy. They found poor agreement when comparing the VGAS with 3DGA to evaluate the Scale's validity. Intra- and inter-observer reliabilities had high levels of agreement and were better at the knee and foot than at the hip. Other studies have also suggested that accurate visual analysis of movement at the hip is difficult to achieve. Read et al. [3] found that proximal observations at the pelvis and trunk showed more variation between observers than assessment at more distal joints when investigating the validity and reliability of the Edinburgh Visual Gait Score. Wren et al. [5] investigated the validity and reliability of the modified PRS and found that inter-observer reliability was worst for hip and ankle position and best for crouch, foot contact and knee position. Mackey et al. [4] in their study of reliability and validity of the OGS found that assessments made at the hip tended to show more variation than distal events such as initial foot contact.

We wished to extend the work of Dickens and Smith [1] who had investigated the reliability and validity of the VGAS when using two experienced observers to considering its reliability when used by more experienced and inexperienced observers.

Section snippets

Observation of gait

Ten observers (four experienced and six inexperienced in observational gait analysis) analysed sagittal, coronal and close up video recordings of the hemiparetic leg during barefoot gait of four children using the VGAS (Fig. 1). The severity of the hemiplegia was evaluated using the Gillette (Normalcy) Index [12] and values for the hemiparetic leg were 92, 132, 156 and 485. All the children had 3DGA using a Vicon 370 system (Oxford Metrics, Oxford, UK) during the same session as the video

VGAS total scores

Table 1 shows the mean difference between the total VGAS scores (sum of scores of each of the seven parameters) for each observer compared with every other observer and the coefficient of repeatability for each data set. Difference against mean was also plotted for each group of observers and each session and showed no systematic error or bias. In session 1 the inexperienced observers had a lower coefficient of repeatability than the experienced observers but in session 2 the experienced

Discussion

Dickens and Smith [1] investigated the reliability and validity of the VGAS for children with hemiplegic cerebral palsy when using two experienced observers. The present study has extended their study by using more observers, both experienced and inexperienced. Predictably, the experienced observers showed better inter- and intra-observer reliability than the inexperienced observers although the mean differences were lower for the inexperienced observers suggesting the possibility of some

Conflict of interest statement

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

References (14)

There are more references available in the full text version of this article.

Cited by (0)

View full text