Use this URL to cite or link to this record in EThOS:
Title: Prognosis in traumatic brain injury
Author: Perel, Pablo Andraes
ISNI:       0000 0004 2702 6219
Awarding Body: London School of Hygiene & Tropical Medicine
Current Institution: London School of Hygiene and Tropical Medicine (University of London)
Date of Award: 2009
Availability of Full Text:
Access from EThOS:
Access from Institution:
Introduction: The general purpose of this thesis was to study prognosis in traumatic brain injury (TBI) patients, with the aim of providing useful and practical information in clinical practice and clinical research. The specific objectives were: to develop and validate practical prognostic models for TBI patients and to assess the validity of the Modified Oxford Handicap Scale (mOHS) for predicting disability at six months. Methods: A survey was first conducted to understand the importance of prognostic information among physicians. A systematic review of prognostic models for TBI patients was then carried out. Prognostic models were developed using data from a cohort of 10,008 TBI patients (CRASH trial) and validated in a cohort of 8,509 TBI patients (IMPACT study). Two focus groups and a survey were conducted to develop a paper-based prognostic score card. The correlation between the mOHS and the Glasgow Outcome Scale (GOS) was assessed, the validity of different mOHS dichotomies was assessed, and the discriminative ability of the mOHS to predict GOS was evaluated. Results: Doctors considered prognostic information to be very important in the clinical management of TBI patients, and believed that an accurate prognostic model would change their current clinical practice. Many prognostic models for TBI have been published, but they have many methodological flaws which limit their validity. Valid prognostic models for patients from high income countries and low & middle income .countries were developed and made available as a web calculator, and as a paper based score card. The mOHS was strongly correlated with and was predictive of GOS at six months. Conclusion: The prognostic models developed are valid and practical to use in the clinical setting. The association between mOHS and GOS suggest that the mOHS could be used for interim analysis in randomised clinical trials in TBI patients, for dealing with loss to follow-up, or could be used as simple tool to inform patients and relatives about their prognosis at hospital discharge.
Supervisor: Roberts, I. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral