Factor analysis of the initial 27-items produced 3 subscales: Emotional Burden, Support and Communication, after 4 items were removed due to poor fit. Rasch analysis was carried out on each of these sub-scales and a further 7 items with poor fit to the Rasch model were removed.  This resulted in a long-form 16-item (PREM-16) demonstrating good internal reliability (Cronbach’s  =0.91). Each subscale showed good internal reliability (0.91, 0.85 and 0.71).  A short-form (PREM-9) version was developed for use in clinical practice ( =0.79). Both forms of the PREM demonstrated medium to strong significant correlations with the validated measures.