Here, we share the first outcome of EVOTION in the form of a data-set to inspire, encourage, and motivate a data-driven analytical approach to evidence-based healthcare policy modeling using real-world longitudinal data. The data-set includes information relating to patterns of real-world hearing aid usage and sound environment exposure. Undoubtedly, many such data-sources will be
available for researchers and policy-makers in the future, and the data-set presented here can act as a first step of building and testing potential statistical models (Christensen et al., 2018, 2019). Specifically, the data-set represents a sub-sample of the data being collected in EVOTION. It contains longitudinally sampled observations from 53 individuals and includes the following measures: the sound environment, the hearing aid setting, logging time (timestamps), ID, and the degree of hearing loss on the best hearing ear of the individuals. Note that the ID (an integer between 1 and 53, randomly assigned to each individual) does not link to the real identity of the participants. Data are considered sensitive as they contain personal and health related information, and EVOTION adhere to strict data ethics by applying privacy-aware big data analytics (Anisetti et al., 2018). Here, we overcome the problem of sharing such personal data by working with a fully synthetic data-set that preserves structural and statistical properties of the original data (see section Technical Validation), without allowing the extraction of personal information (see section Data Synthesization). Thus, the synthetic data-set can readily be shared among professionals.