Page 69 - Demo

Methylphenidate for ADHD in Smith-Magenis syndrome673held by the researcher will be stored securely with access restricted and limited to nominated research staff recorded on the delegation log. A data sharing agreement between Amsterdam UMC and ‘s Heeren Loo will manage additional access for investigators.The CRFs will be set up in Castor Electronic Data Capture (EDC) in which weekly assessments will be entered. Questionnaires can be filled out digitally using the m-Path app on smartphones,44 on computers (Castor EDC) or by using paper forms. Data from the app will be collected at the end of each trial and will be loaded into Castor EDC. In advance, participants will be recommended to download the m-Path app to easily and confidentially answer the daily questionnaires, although the use of different ways is allowed to enlarge feasibility for raters. For the sake of participant retention, automatic reminders will be sent to raters when questionnaires have not yet been filled in. Participant burden will be limited as much as possible by having contact moments by video-conference or phone instead of a visit. The investigator can also decide to withdraw a subject for urgent medical reasons. A participant who withdraws consent for an assessment of one outcome may be willing to continue with assessments for other outcomes.A subject identification code list will be used with unique participant identifiers not deducible to patients. Only two investigators will have access to the key. In addition, two methodologists and biostatisticians will have access to the source data for methodological and statistical purposes. Data will be stored for 15 years according to the Amsterdam UMC regulations.Statistical methodsAn individual treatment effect for each participant will be determined based on summary statistics. A mixed model analysis will be applied for analysing the effectiveness of the intervention at the population level combining data from the individual N-of-1 trials.The mean treatment effect on the primary outcome will be estimated and tested for significance using a linear mixed model with a fixed effect for treatment (MPH or placebo) and random effects for patient, cycle within patient, and treatment (within patient). The mixed model will account for Annelieke Muller sHL.indd 67 14-11-2023 09:07