Guidelines for the emergency department management of traumatic brain injury : an impact assessment and development of a prognostic model to inform hospital admission decisions

Marincowitz, Carl

August 2019

Thesis or dissertation

© 2019 Carl Marincowitz. All rights reserved. No part of this publication may be reproduced without the written permission of the copyright holder.

1.4 million patients attend English and Welsh Emergency Departments (ED) annually following head injury. 95% attend with a high level of consciousness, of whom 1% have life-threatening traumatic brain injuries (TBI), whilst 7% have TBI on CT imaging.
National guidelines were introduced in England and Scotland to improve TBI outcomes and reduce hospital admissions. The impact of these guidelines has not been rigorously assessed. They recommend patients with injuries on CT imaging be admitted to hospital in case they deteriorate. Accurate prediction of deterioration could identify patients safe for discharge from the ED.

Assess the impact of national guidelines on deaths and admissions.
Develop a prediction model for deterioration in patients with injuries identified by CT imaging.

Interrupted time series analyses using national data for England and Scotland were conducted to evaluate guideline impact.
A systematic review was completed to identify candidate prognostic factors for deterioration. Multivariable logistic regression was used to develop prognostic models using these factors in an English multi-centre retrospective cohort of patients.

Guideline impact varied by age group. Associated reductions in hospital admissions and mortality were found in those aged 16-64. In older patients, an increase in TBI mortality was observed, which was unaffected by guideline introduction.
A prognostic model and decision rule was developed, using data from a cohort of 1699 patients. It achieved a sensitivity of 99.5% (95% CI: 98.1% to 99.9%) and specificity of 7.4% (95% CI: 6% to 9.1%) to a measure of deterioration encompassing need for admission.

This first national evaluation of head injury guidelines to use quasi-experimental methods suggests guideline impact varied by age. This first empirically derived prediction model to inform admission decisions suggests a small proportion of patients could be safely discharged from the ED. External validation is required before clinical use.

Hull York Medical School, The University of Hull and The University of York
Allgar, Victoria; Lecky, Fiona; Sheldon, Trevor
Qualification level
Qualification name
5 MB
QR Code