Informative drop-out/censoring in SAFE-TBI

Regular measurements of intracranial pressure in SAFE patients with Traumatic Brain Injury (TBI) are subject to informative censoring due to death, ICU discharge or clinician’s decision to stop monitoring.

A more complex analysis involving pattern mixture models or selection models is investigated. The objective of this modeling is to assess the evolution of this endpoint and related therapies over time and possibly explain why patients on saline resuscitation fluid survive longer than the ones on albumin.

The joint modelling of death and ICP measurement will/can also be investigated.

Related unit(s): Statistics