B1379
Title: Testing non-stationarity and quantifying associations in the presence of missing data in time series of mHealth studies
Authors: Charlotte Fowler - Columbia University Mailman School of Public Health (United States)
Xiaoxuan Cai - Columbia University (United States)
Linda Valeri - Columbia University (United States) [presenting]
Abstract: The use of digital devices to collect data in mobile health (mHealth) studies introduces a novel application of time series methods, with the constraint of potential data missing at random (MAR) or missing not at random (MNAR). In time series analysis, testing for stationarity is an important preliminary step to inform appropriate later analyses. Further, appropriately accounting for missing data is crucial to estimate exposure effects in multivariate time series settings validly. The augmented Dickey-Fuller (ADF) test was developed to test the null hypothesis of unit root non-stationarity, under no missing data. We propose maximum likelihood estimation and multiple imputation using a state space model approach to adapt the ADF test and evaluate associations among multivariate time-series in a context with missing data. We further develop sensitivity analysis techniques to examine the impact of MNAR data. We evaluate the performance of existing and proposed methods across different missing mechanisms in extensive simulations and their application to a multi-year smartphone study of bipolar patients.