Predicting duration of stay in a pediatric intensive care unit: A Markovian approach

Pediatric Risk of Mortality scores representing the intensity of illness are utilized to study the dynamics of the movement of patients in pediatric intensive care. The flow of patients is modeled as a discrete time Markov process. The course of treatment and length of stay in the intensive care is described as a sequence of ‘Low’, ‘Medium’ and ‘High’ severity of illness states rather than by a string of services and scores. The Markovian model well reproduces the shape of data and the strength of the methodology lies in its generalizability to settings where patient's length of stay and utilization of resources is highly correlated.