Abstract
This thesis aimed to recommend a core set of clinical measures for the assessment of balance of people with cerebellar ataxia. It comprised four studies: Narrative review, Systematic review, Delphi survey and Psychometric property testing. The narrative review revealed a wide spectrum of both laboratory-based and clinically-based assessment tools. Two strategies were then implemented in order to select the most appropriate clinical measures from those identified in the narrative review. First, a systematic review explored the psychometric properties of the measures of balance tested in people with cerebellar ataxia. The systematic review identified the Posture and gait sub-component of the International Co-operative Ataxia Rating Scale (PG of ICARS) as highly appropriate measure. Secondly, a Delphi survey was undertaken to obtain the opinion of experts as to which balance measures they would recommend for the clinical assessment of balance of people with cerebellar ataxia; the Berg Balance Scale (BBS), the Timed Up and Go (TUG) test and the Gait, Stance and Sit sub-components of the Scale for the Assessment and Rating of Ataxia (SARA Bal) were identified. These four measures required further validation prior to recommending them as a core set of clinical measures. The main study of this thesis tested the reliability, validity and interpretability of the BBS, TUG, PG of ICARS and SARA Bal. Fifty participants with multiple sclerosis and secondary cerebellar ataxia were recruited across four centres in New Zealand and the United States of America. Participants were tested by physiotherapists on four measures of balance and this testing process was video-recorded. The Barthel Index, the Expanded Disability Status Scale (EDSS), the full scales of the ICARS and the SARA were also assessed and disease duration recorded. The EDSS was completed by neurologists. To investigate the intra-rater and inter-rater reliability, a repeat assessment was performed by the same physiotherapist (intra-rater) or a second physiotherapist (inter-rater) by observing the video recording. Reliability was strong for all four measures (ICC range 0.95-0.99). Internal consistency was strong for BBS (α=0.94) and PG of ICARS (α=0.87), and moderate for SARA Bal (α=0.72). Moderate to strong correlation was found between the measures of balance (ρS >0.72) and between PG of ICARS (ρS = 0.84) and SARA Bal (ρS = 0.85) against ataxia rating scales. Measures of balance correlated poorly to moderately with disease severity (EDSS), functional independence (Barthel Index) and disease duration (ρS range -0.37 to 0.76). Discriminant validity of all measures were excellent with Area Under the Curve of >0.93. Minimal detectable change was derived for the BBS (2.8), PG of ICARS (2.25) and SARA Bal (1.70). Two items of the PG of ICARS was found to have inconsistency in scoring and in addition all items of PG of ICARS may be scored using the other three measures. Therefore the BBS, TUG and SARA Bal were recommended as the core set of standardised measures. Psychometric properties that need to be studied in future to strengthen the current recommendations such as structural validity and responsiveness of the measures of balance were proposed.