Rasch analysis of SF-Qualiveen in multiple sclerosis.

ABSTRACT

BACKGROUND: A 30-item Qualiveen questionnaire was developed to measure the impact of urinary problems on everyday living in spinal cord injury, and subsequently an 8-item SF-Qualiveen was developed for those with multiple sclerosis (MS). The validity of this short form has not been previously examined using modern psychometric techniques, such as the Rasch measurement model. OBJECTIVE: The aim of this study is to test if the short form meets the requirements of the Rasch model. METHODS: A total of 401 patients with clinically definite MS were given the questionnaire at three neuroscience centres in the UK. RESULTS: A total of 258 patients (64.3% response) completed the questionnaire. The original scale failed to meet the expectations of the Rasch model. A two-testlet solution was sought to account for local dependence, differential item functioning and disordered thresholds. After the modifications were made the scale fitted the model (χ2 = 5.93 P = 0.4305), had high internal consistency (α = 0.88) and was unidimensional. CONCLUSIONS: SF-Qualiveen is a simple and valid measure of the impact of urinary problems in multiple sclerosis, which meets the requirements of the Rasch measurement model. Summed ordinal scores can be converted to interval-level using the transformation table provided.