Modelling the consent acquisition time in organ donor management through clustering and mixture probability models

Time is a critical variable in organ transplantation. Brain-dead donors may be or become unstable. Organs are subject to degradation and must be transplanted as soon as possible. An unavoidable source of delay in the process is the time needed to get consent to the donation by the potential donor's relatives. Information about the likely consent acquisition time would allow us to predict and minimize the time needed for the overall harvesting and transplantation process. In this paper, by using a real dataset of consent acquisition time from a local transplantation centre, we identify three clusters associated with early, regular, and late consent-givers and provide a probability model for the consent acquisition time based on a mixture of three gamma components.