Patient mobile computing system (1) and method for predicting exacerbations in a patient illness related status. The method comprises collecting patient sensor data using sensors (8), and executing a local prediction model. The local prediction model comprising a Bayesian network (6) to obtain a probability of an exacerbation of a patient status. The Bayesian network (6) receives patient input data, patient sensor data, and providing probability data, and the local prediction model provides guidance data based on the probability data. The local prediction model (6) is implemented both in a portable device (1), such as a smart phone, and in a centralized patient data processing system (20).
