130 Chapter 5 5 be assessed following the Dutch Guideline for Economic Evaluations in healthcare (59). Blinding (performance and outcome assessment) Due to the nature of the study, patients cannot be blinded to allocation. As unblinding may introduce performance bias (i.e., a change in patient’s behaviour caused by awareness of participation in a trial, especially around visits), the medication adherence data of 1 week before and 1 week after follow-up moments are disregarded to minimise the risk of bias. Outcome assessors cannot be blinded as it is important to carefully instruct patients on how to download and use the intervention app and to provide training to intervention practices on use of the online portal. The statistician who performs the data analyses and validates the results will be blinded to group allocation to avoid bias. Data management Data will be pseudonymised by using a code list during data collection. Collection of indirect and direct identifiable information will be minimised, and will be only collected for the purpose of this study. Identifiable information will be stored separately from pseudonymised data. All data collected on paper are stored in locked filling cabinets at the study sites. Electronic data is collected using Castor EDC, an electronic data capture and management application (60). Only investigators and research staff involved in the trial have access to participant data. For the logistic management of participants and the trial, a secured access database is used. Data handling and storage comply with the General Data Protection Regulation. Source documents, informed consent forms and investigator files are archived for 15 years at the study sites, according to the Dutch Medical Treatment Act. Video consulting software used during remote visits comply with security standards set by the study sites and applicable laws and features two-factor authentication and encrypted data. Data that are stored in the apps used in this study are encrypted, as well as data that are stored in a local database on the phone. This prevents other apps on the phone from accessing the data. Data are also encrypted when data are in transit, to protect personal information. All the information supplied through the Turbu+ Insights application will be stored on secure servers in the EU (Ireland) and managed by the Turbu+ Insights programme administrators. Data are pseudonymised when exported from the system. All data collected by Hailie Sensors and transmitted by Hailie Lite App, as well as data entered into the Hailie web portal in the course of the ACCEPTANCE study (Asthma Control through Cost Effective Primary care Treatment: AdhereNCe and E-Health feedback) by the study site personnel, are stored on secure servers in the USA: (1) on AWS servers, from study start date to 17 December 2020 and (2) on MS Azure SQL servers, from 17 December 2020 to the present, under data processing addendums including standard contractual clauses. Further data management procedures and operational details are specified in the data management plan.
RkJQdWJsaXNoZXIy MTk4NDMw