|
Abstract:
|
Current health policy calls for greater use of evidence based care delivery services to improve patient quality and safety outcomes. Care delivery is complex, with interacting and interdependent components that challenge traditional statistical analytic techniques. Methods for analyzing interrupted time series (ITS) do not account for potential changes in variation and correlation following the intervention, and require a pre-specified intervention time point with an instantaneous effect. This is a key limitation since it is plausible to have either anticipatory or delayed change, which can influence determination of overall effectiveness. In this paper, we describe and develop a novel 'Robust-ITS' model that overcomes these omissions and limitations. The Robust-ITS model formally performs inference on the change-point; pre- and post-intervention correlation; variance of the outcome measure, and pre- and post-intervention trajectory. We illustrate the proposed method by analyzing patient satisfaction from a hospital that implemented and evaluated a nursing care delivery model. The Robust-ITS model is implemented in a R Shiny toolbox, freely available to the community.
|
