Our PIs Nadine Binder and Holger Dette have produced a flexible new tool to address the identification of similar patient pathways in clinical healthcare with funding from the CRC 1597 Small Data. Specifically, they considered parametric competing risk models, where transition intensities may be specified by a variety of parametric distributions. They assessed the similarity between two such models by examining the transitions between different health states. This enabled them to introduce a method to measure the maximum differences in transition intensities over time, leading to the development of a test procedure for assessing similarity. They proposed a parametric bootstrap approach for this purpose and evaluated its performance through a simulation study, considering a range of sample sizes, differing amounts of censoring, and various thresholds for similarity. They also demonstrated the practical application of their approach with a case study from a urological clinical routine practice.
We caught up with Nadine to ask her about the potential and primary motivation of this research:
“Multistate models form a powerful class of statistical models for conveniently describing pathways with several possible state types for each individual. In combination with statistical testing, this can enable to identify valid best-practice treatment strategies from routine clinical data.”
Check out the article here.
Institute of Medical Biometry and Statistics,
Faculty of Medicine and Medical Center –
University of Freiburg
