Exploring the Validity of the Short Version of the Problem Behaviours Assessment (PBA-s) for Huntington's disease: A Rasch Analysis

Background: The short version of the Problem Behaviours Assessment (PBA-s) is the recommended outcome measure for behavioural symptoms in Huntington's disease. Rasch analysis was used to further investigate the measurement limitations of the PBA-s. Objectives: 1) To assess the psychometric properties of the 11 severity and frequency items within the PBA-s and 2) to determine the construct validity of using a total PBA-s score as a clinical outcome measure. Methods: PBA-s data for 517 participants from Enroll-HD were included in the Rasch analysis. Separate analyses were conducted for the severity and frequency items of the PBA-s, using RUMM2030 software. Achieving fit to the model provides supporting evidence that all items contribute to a single underlying latent trait. This property is defined as internal construct validity. Results: The total PBA-s severity score demonstrated several important limitations, including disordered response categories for all 11 severity items, local dependency and poor targeting. However, modifying the original five-point scoring system to a four-point system resulted in ordered response categories for seven of the severity items and achieved a good overall fit to the Rasch model. For the total PBA-s frequency score, fit to the model was not achieved even after amendments to the scoring system. Conclusions: This study suggests that with reduction to a four-point scoring system, the total PBA-s severity score may be considered a valid clinical outcome measure. This study also suggests limitations in the use of a total PBA-s frequency score.