Rasch analysis of the modified version of the Postural Assessment Scale for Stroke patients: Postural Stroke Study in Gothenburg (POSTGOT)
Institute of Neuroscience and Physiology, Rehabilitation Medicine, Sahlgrenska Academy, University of Gothenburg, Per Dubbsgatan 14, Gothenburg S-413 45, Sweden
BMC Neurology 2014, 14:134 doi:10.1186/1471-2377-14-134Published: 19 June 2014
The modified version of the Postural Assessment Scale for Stroke Patients (SwePASS) is a new ordinal outcome measurement designed to assess postural control in patients with stroke. Before implementation of SwePASS into the clinical setting, it is necessary to know its measurement properties. Thus, the aim of the study was to evaluate the measurement properties of the SwePASS.
Rasch analysis, based on data of 150 SwePASS assessments was made the first week after stroke onset. The measurement properties referred to were unidimensionality, local independence, invariance, category function, targeting of persons and items and the reliability.
The initial analysis showed disordered thresholds in four items. After adjustment of the scoring categories, this was resolved. However, analyses of local dependency revealed correlations between two of the items. These two items were collapsed into one. After adjustments, the person separation index that acts as an indicator of the whole model fit was 0.96. The adjusted SwePASS is a global scale that works the same way regardless of gender, age and location of stroke lesion. Overall, the population had better postural control than was targeted with the items in the scale.
Rasch analysis of the adjusted SwePASS showed that the scale was unidimensional. In SwePASS, equal capacity in postural control provides the same response to an individual item in patients with stroke, regardless of gender, age and location of stroke lesion. Regarding clinical implications, before introducing SwePASS in clinical routine and to confirm the results, further research including a larger sample with poorer postural control is suggested.