Evaluating digital health interventions with complex designs

Digital technologies such as wearables, health apps, and mobile sensors help collect patient data, for example, on heart rate, movement, or blood pressure. They offer great potential, especially in the prevention and treatment of cardiovascular diseases: continuous data collection enables more personalized, patient-centered care. However, analyzing this data presents researchers with enormous challenges: high data complexity, outliers, missing values, or small sample sizes, such as those found in rare diseases. Eleonora Carrozzo is therefore dedicated to developing new statistical methods specifically designed for such challenging health data. "The goal of my work is to arrive at statistically sound and clinically relevant conclusions despite small sample sizes, high data complexity, or erroneous values," says Eleonora Carrozzo of Salzburg Research. The goal of Carrozzo`s research is to create innovative analysis methods based on so-called nonparametric methods. These do not require strict assumptions about the distribution of the data and are therefore particularly suitable for digital health data that is patchy, high-dimensional, irregular, or highly individualized Carrozzo aims to close existing methodological gaps and, in particular, support medical professionals: The new tools are intended to help with the sound evaluation of digital health measures and their clinically meaningful use. A particular focus is on practical implementation: The developed methods are made available in the form of easy-to-use software packages (R packages). These are intended to be used not only in research, but also in clinical practice or evaluation studies. Even complex study designs for example, with multiple groups or very small samples, many measurement points, or high-dimensional data can be better represented with the new methods. This enables well-founded decisions even when traditional statistical methods would fail due to a sparse data base. Anna Eleonora Carrozzo is a postdoctoral researcher at the Salzburg Research and the Paris Lodron University of Salzburg in the joint EXDIGIT program (funded by the State of Salzburg as part of the WISS2030 program). Previously, she worked at the Ludwig Boltzmann Institute for Digital Health and Prevention in Salzburg. Anna Eleonora Carrozzo received her doctorate in Management and Engineering from the University of Padua in 2016. She previously earned a master`s degree in statistics there. Her research focuses on biostatistics, nonparametric statistics, statistical methods in medical research, and data science in digital health.