Indications for the use of video gait analysis in children with cerebral palsy (CP) include pre-treatment assessment and documenting change in gait pattern over time following interventions such as multilevel surgery, orthotics, botulinum toxin injections, serial casting and intensive therapy (Thomason et al. 2009). In an effort to make video gait assessment more objective and reliable a number of observational gait scales have been developed over the last few decades.

Development of the Assessment: The original Physician Rating Scale (Koman et al. 1993) was used to evaluate the effects of botulinum toxin-A in calf muscle injections on the function of the lower legs of children (4-11 years old) with CP. The Physician Rating Scale examined the hip, knee, and the foot position in the sagittal plane at defined phases of gait cycle and the Speed of gait. The scale rated the gait in six sections: Gait pattern, Hind foot position during foot strike, Knee position during stance phase (degree of recurvatum), Maximum foot/floor contact during stance phase, Degree of crouch (hip, knee and ankle) and Speed of gait, scoring each on scales ranging from 0-3 and 0-4. It was used to examine gait of children with CP, either with naked eye examination alone or together with video recordings.

Corry et al. (1998) detected lack of sensitivity and reliability in detecting specific changes after treatment with BTX-A. The investigators reduced the PRS to four sections by removing Gait pattern, Hind foot position during foot strike, and Speed of gait. The authors also and added a section on Change which aimed to improve discrimination between the two treatment groups in their study.

To improve the sensitivity of this scale, especially the assessment of the relation between foot and knee position in midstance, modifications were made by team of M. Delgado from Texas Scottish Rite Hospital for Children, Dallas USA. Boyd and Graham (1999) made additional changes to Delgado’s version of the scale and renamed the assessment the Observational Gait Scale. Alterations included: (1) the removal of three of the original sections (a) Gait pattern, (b) Hind foot position during foot strike, and (c) Speed of gait, (2) detailing foot contact in midstance, (3) adding new sections (initial foot contact/ hind foot position midstance, timing of heel rise, base of support, gait assistive devices) and (4) incorporating position “Change” proposed by Corry et al. (1998).

• Reliability and validity: Mackey et al. (2003) showed that first four sections of OGS have acceptable level of reliability and validity. The OGS was found to have acceptable inter-rater and intra-rater reliability for knee and foot position in mid-stance, initial foot contact, and heel rise with weighted kappas ranging from 0.53 to 0.91 (intra-rater) and 0.43 to 0.86 (inter-rater). Comparison with 3-dimensional gait analysis (DGA) suggests that these sections might also have high validity (range 0.38–0.94). Base of support and hind foot position had lower inter-rater and intra-rater reliabilities (0.29 to 0.71 and 0.30 to 0.78, respectively) and were not easily validated by 3-DGA. Mackey et al. (2003) cited that, “Correlation of 3-DGA data with OGS scores for the first four sections showed high validity for our most reliable observer.” This comparison with 3-DGA suggests that the sections of knee and foot position in mid-stance, initial foot contact, and heel rise have high validity (Mackey et al. 2003.). In Rathinam et al. (2014), the OGS was reported to have very good inter-rater reliability, however only the sagittal plane (ankle/foot and knee joints) items scored maximum agreement. Overall, data from this study suggests that the first four sections of the OGS scale have an acceptable level of reliability and validity when assessing gait in children with spastic diplegia and problems principally in the sagittal plane. Zanudin et al. (2017) evaluated the methodological quality and the strength of the evidence of studies that reported an evaluation of the psychometric properties of observational assessments of gait quality and walking performance in children with CP. The authors were using the modified COSMIN checklist and found conflicting findings for inter-rater reliability and no quality data for intra-rater reliability for OGS.

Pros

• OGS is a short, easy to use observational scale which assists in the recognition of gait problems that can be treated in rehabilitation process.
• The observational (video) gait assessment does not require costly equipment.
• The OGS is less costly and time consuming than 3DGA and can be used in children with ambulatory problems requiring less complex decision-making, where 3DGA is not required or available, for monitoring interventions and for younger and cognitively or behaviorally impaired children (Mackey et al. 2003; Harvey & Gorter 2011).
• The test can help distinguish between different foot alignment patterns (like true equinus, apparent equinus) or different knee abnormalities (recurvatum, crouch).
• By defining certain gait abnormalities, this test can assist in predicting changes in motor function related to gait with age progression in children with cerebral palsy (eg. recognization of mild crouch).
• The OGS has high reliability and validity at least in most sections. It includes specific measuring properties that are geared toward the gait impairments in children with cerebral palsy. It looks at change over time which enhances its use as a good outcome measure for rehabilitation. Video recordings can be assisted with computer software to make accurate assessments (Wren et al. 2005, Grunt et al. 2010)

Cons

• The assessment does require adequate training and experience.
• A lack of clear definitions for items can increase reliability issues.
• The literature for this assessment provides limited evidence on validity in children with cerebral palsy.
• Assessment of young children is often more difficult due to lack of compliance with gait instructions.
• The reproducibility of measurements of the different joint/segment angle and between the phases of the gait cycle can vary (Grunt et al. 2010).
• Precise discrimination between different percentages of heel rise (section four) may be difficult without additional tools.
• Patients with significant transverse plane problems or major foot deformities make the assessment of sagittal plane gait deviations more difficult and could potentially compromise the validity of the scale (Mackey et al. 2003).
• When two different observers perform the observational gait measurements, the SDD (smallest detectable difference) is more than 10 degrees in majority of joints. For the assessment of treatment outcomes, the same observer should perform assessment before and after the intervention (Grunt et al. 2010). But detecting 5 degree differences for knee joint positions (section one) remains difficult (Wren et al. 2005).