• Rishma Maini, Martin Kirkpatrick, Aileen McCafferty, Colin Dunkley, Simon Ogston, and Fiona Williams.
• Ninewells Hospital and Medical School, University of Dundee, UK.
• Seizure. 2018 Dec 1; 63: 71-78.

PurposeTo validate a patient-reported-experience-measure, PREM, of the NHS paediatric epilepsy service.MethodsSection 1 of the PREM recorded demographic and clinical characteristics, and Section 2 collected information about the users' experience with the service. Section 2 included eighteen statements around three constructs: communication and provision of information to service users, interpersonal skills of staff, and clinic visits and accessibility to the services. Face validity, construct validity, internal reliability, and internal consistency were used to examine the robustness of these statements. The PREM was completed by parents/carers and also children/young people.ResultsPREMs were received from 145 of the 192 audit units; 2335 completed forms were returned; the attitude statements were completed by 750 children/young people and 1550 parents/carers. Face validity of the PREM was good. Construct validity was indecisive; confirmatory factor analysis of the hypothesised construct was weak. Exploratory factor analysis identified a four factor solution for the parent/carers dataset and a five factor solution for the children/young people's dataset. Internal reliability was good for the parent/carers dataset but less good for the children/young people. Internal consistency was moderately good for both datasets.ConclusionsThese findings indicate that the PREM is likely to be a valid tool with the potential to elicit a wide variety of reliable views from parents/carers of children with epilepsy. The construct validity for the PREM should be reassessed with confirmatory factor analysis in a new dataset. More work needs to be undertaken with children/young people to design statements that capture their specific needs.Copyright © 2018 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.

26 keywords...

hide…