Use this URL to cite or link to this record in EThOS:
Title: Derivation and validation of a clinical prediction rule to predict the likelihood of massive transfusion in military major trauma
Author: Mclennan, Jacqueline
ISNI:       0000 0004 2740 657X
Awarding Body: University of Manchester
Current Institution: University of Manchester
Date of Award: 2013
Availability of Full Text:
Access from EThOS:
Access from Institution:
Introduction: Coagulopathy in trauma increases the likelihood of death. Various approaches to correcting this coagulopathy are being investigated but the decision to use them is often haphazard. The guidelines on their use are limited. This project aims to produce a decision rule to rule in the need for massive transfusion in military trauma, and therefore to rule in the use of methods for treating the coagulopathy of trauma.Methods: A Delphi Study was performed to ascertain which variables a panel of experts felt were most predictive of the need for massive transfusion. This encompassed experts for the fields of emergency medicine, critical care, trauma surgery, pre-hospital care and haematology. Data from the British Defence Joint Theatre Trauma Registry, from January 2007 –July 2010 were obtained and divided into two groups to provide a derivation and validation dataset. Blood timings for people receiving 5 or more units were obtained from clinical records. Regression analysis of potential predictive factors either indicated by the Delphi panel or by literature review were analysed to confirm their value in prediction of Massive Blood Transfusion. A range of clinical prediction rules were produced using parameters deemed appropriate by the Delphi panel and shown to be predictive through regression analysis. Three of these models were then tested in the validation dataset.Results: Ethical approval has been found not to be required following decisions by the relevant military and civilian ethical boards. The Delphi panel was conducted over 3 rounds with a panel of 33 members. Response rates of 94%, 80% and 77% were achieved in rounds 1, 2, and 3 respectively. 195 statements were produced; agreement was achieved at the 80% level in 97 (49.7%) of statements.Regression analysis produced multiple factors that were highly predictive of Massive Blood Transfusion. These were formulated into 22 potential rules combining evidence of injury, clinical observations and pre-hospital care received to produce rules with high sensitivity and specificity. Overall 3 rules were deemed to provide the best balance of sensitivity and specificity, while remaining clinically valid. These were then validated, in a second dataset. The simplest of these rules has a sensitivity of 83.3% and a specificity of 85.5% with an AUROC of 0.907 in the derivation dataset. In the validation dataset sensitivity improved to 87.65% with a specificity of 80.45% with an AUROC curve of 0.91.Discussion: A clinical decision tool which ruled in the use of a massive transfusion protocol allowing early and aggressive resuscitation, and early provision of blood products, would result in better care for severely injured military trauma patients. Although several prediction models are available, they all require either weighted parameters or blood tests so limiting their utility. Further, sensitivity and specificity is poor. This project, through the use of expert opinion, and the production of a validated decision rule, has provided such a tool with improved sensitivity over the currently available prediction models. By using consensus from a Delphi panel, it is hoped that any tool will be acceptable to clinicians therefore improving quality of care. This rule now needs to be assessed prospectively for validity, ease of use and clinical acceptance.
Supervisor: Mackway-Jones, Kevin Sponsor: Not available
Qualification Name: Thesis (M.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: haemorrhage ; prediction rules ; trauma ; military