Behavioral assessment of unilateral neglect: study of the psychometric properties of the Catherine Bergego Scale

Abstract

Objective: To assess the psychometric properties of a scale for spatial neglect in everyday life.

Design: Validation study.

Setting: A neurologic rehabilitation unit in a university hospital.

Participants: Eighty-three consecutive right-hemisphere stroke patients.

Interventions: Not applicable.

Main outcome measures: The Catherine Bergego Scale (CBS) was used to assess neglect behavior and anosognosia. Its sensitivity was compared with that of 3 conventional tests. The inner structure of the scale was studied by principal component analysis. In addition, linearity, unidimensionality, and reliability of the scale were tested through Rasch analysis.

Results: Behavioral assessment correlated significantly to, but was more sensitive than, conventional tests. Anosognosia correlated significantly with neglect severity, although individual dissociations occurred. Factorial analysis disclosed a single underlying factor, explaining 65.8% of total variance. Rasch analysis also revealed that the 10 items defined a common, single ability continuum with widespread measurement range and quite regular item distribution, and showed a satisfactory reliability.

Conclusion: Behavioral assessment proved to be more sensitive than conventional paper and pencil tasks. Both conventional statistics and Rasch analysis suggest that the CBS is reliable and valid, and that the 10 items define a homogeneous construct.