Maralyn Woodford

Trends in head injury outcome from 1989 to 2003 and the effect of neurosurgical care: An observational study

BACKGROUND Case fatality rates after all types of blunt injury have not improved since 1994 in England and Wales, possibly because not all patients with severe head injury are treated in a neurosurgical centre. Our aims were to investigate the case fatality trends in major trauma patients with and without head injury, and to establish the effect of neurosurgical care on mortality after severe head injury. METHODS We analysed prospectively collected data from the Trauma Audit and Research Network database for patients presenting between 1989 and 2003. Mortality and odds of death adjusted for case mix were compared for patients with and without head injury, and for those treated in a neurosurgical versus a non-neurosurgical centre. FINDINGS Patients with head injury (n=22,216) had a ten-fold higher mortality and showed less improvement in the adjusted odds of death since 1989 than did patients without head injury (n=154,231). 2305 (33%) of patients with severe head injury (presenting between 1996 and 2003) were treated only in non-neurosurgical centres; such treatment was associated with a 26% increase in mortality and a 2.15-fold increase (95% CI 1.77-2.60) in the odds of death adjusted for case mix compared with patients treated at a neurosurgical centre. INTERPRETATION Since 1989 trauma system changes in England and Wales have delivered greater benefit to patients without head injury. Our data lend support to current guidelines, suggesting that treatment in a neurosurgical centre represents an important strategy in the management of severe head injury.

Standardized comparison of performance indicators in trauma: a new approach to case-mix variation.

An institutions trauma survival rate can be compared with that predicted by TRISS using definitive outcome-based evaluation. This examines W, the difference between actual and predicted survival rates; Z, the statistical significance of this difference; and M, a measure of the similarity of injury severity mix to the prediction data base. However, it is possible for two institutions with the same survival rate within each band of injury severity to have very different W and Z scores whilst retaining a similar M score. Clearly this is unsatisfactory. A new statistic, Ws, is therefore proposed, which is standardized with respect to injury severity mix, producing more accurate comparisons between different institutions. Confidence intervals are used to graphically illustrate the magnitude of Ws, its direction, accuracy, and statistical significance. Data from the U.K. Major Trauma Outcome Study are used to demonstrate the calculations and presentation of Ws and its advantages.

A major trauma centre is a specialty hospital not a hospital of specialties

High estimates of preventable death rates have renewed the impetus for national regionalization of trauma care. Institution of a specialist multidisciplinary trauma service and performance improvement programme was hypothesized to have resulted in improved outcomes for severely injured patients.

The Effect of an Organized Trauma System on Mortality in Major Trauma Involving Serious Head Injury: A Comparison of the United Kingdom and Victoria, Australia

Objective:To compare outcomes following major trauma involving serious head injury managed in an inclusive trauma system (Victoria, Australia) and a setting where rationalization of trauma services is absent (England/Wales). Background:The introduction of regionalized trauma systems has the potential to reduce preventable deaths, but their uptake has been slow around the world. Improved understanding of the benefits and limitations of different systems of trauma care requires comparison across systems. Methods:Mortality outcomes following major trauma involving serious head injury managed in the 2 settings were compared using multivariate logistic regression. Data pertaining to the period July 2001 to June 2006 (inclusive) were extracted from the Trauma Audit and Research Network (TARN) in the United Kingdom and the Victorian State Trauma Registry (VSTR) in Australia. Results:A total of 4064 (VSTR) and 6024 (TARN) cases were provided for analysis. The odds of death for TARN cases were significantly higher than those for VSTR cases [odds ratio = 2.15, 95% confidence interval = 1.95–2.37]. After adjusting for age, gender, cause of injury, head injury severity, Glasgow Coma Scale score, and Injury Severity Score, TARN cases remained at elevated odds of death (3.22; 95% confidence interval = 2.84–3.65) compared with VSTR cases. Conclusions:Management of the severely injured patient with an associated head injury in England and Wales, where an organized trauma system is absent, was associated with increased risk-adjusted mortality compared with management of these patients in the inclusive trauma system of Victoria, Australia. This study provides further evidence to support efforts to implement such systems.

Intracranial bleeding in patients with traumatic brain injury: A prognostic study

BackgroundIntracranial bleeding (IB) is a common and serious consequence of traumatic brain injury (TBI). IB can be classified according to the location into: epidural haemorrhage (EDH) subdural haemorrhage (SDH) intraparenchymal haemorrhage (IPH) and subarachnoid haemorrhage (SAH). Studies involving repeated CT scanning of TBI patients have found that IB can develop or expand in the 48 hours after injury. If IB enlarges after hospital admission and larger bleeds have a worse prognosis, this would provide a therapeutic rationale for treatments to prevent increase in the extent of bleeding. We analysed data from the Trauma Audit & Research Network (TARN), a large European trauma registry, to evaluate the association between the size of IB and mortality in patients with TBI.MethodsWe analysed 13,962 patients presenting to TARN participating hospitals between 2001 and 2008 with a Glasgow Coma Score (GCS) less than 15 at presentation or any head injury with Abbreviated Injury Scale (AIS) severity code 3 and above. The extent of intracranial bleeding was determined by the AIS code. Potential confounders were age, presenting Glasgow Coma Score, mechanism of injury, presence and nature of other brain injuries, and presence of extra-cranial injuries. The outcomes were in-hospital mortality and haematoma evacuation. We conducted a multivariable logistic regression analysis to evaluate the independent effect of large and small size of IB, in comparison with no bleeding, on patient outcomes. We also conducted a multivariable logistic regression analysis to assess the independent effect on mortality of large IB in comparison with small IB.ResultsAlmost 46% of patients had at some type of IB. Subdural haemorrhages were present in 30% of the patients, with epidural and intraparenchymal present in approximately 22% each. After adjusting for potential confounders, we found that large IB, wherever located, was associated with increased mortality in comparison with no bleeding. We also found that large IB was associated with an increased risk of mortality in comparison with small IB. The odds ratio for mortality for large SDH, IPH and EDH, in comparison with small bleeds, were: 3.41 (95% CI: 2.68-4.33), 3.47 (95% CI: 2.26-5.33) and 2.86 (95% CI: 1.86-4.38) respectively.ConclusionLarge EDH, SDH and IPH are associated with a substantially higher probability of hospital mortality in comparison with small IB. However, the limitations of our data, such as the large proportion of missing data and lack of data on other confounding factors, such as localization of the bleeding, make the results of this report only explanatory. Future studies should also evaluate the effect of IB size on functional outcomes.

Paediatric trauma: injury pattern and mortality in the UK

Objective: Trauma accounts for a large proportion of childhood deaths. No data exist about injury patterns within paediatric trauma in the UK. Identification of specific high-risk injury patterns may lead to improved care and outcome. Methods: Data from 24 218 paediatric trauma cases recorded by the Trauma Audit and Research Network (TARN) from 1990 to 2005 were analysed. Main injury, injury patterns and outcome were analysed. Mortality at 93 days’ post-injury was the major outcome measure. Results: Limb injuries occurred in 65.0% of patients. In infants 81.4% of head injuries were isolated, compared with 46.5% in 11–15-year-old children. Thoracic injuries were associated with other injuries in 68.4%. The overall mortality rate was 3.7% (n = 893). Mortality decreased from 4.2% to 3.1%; this was most evident in non-isolated head injuries. It was low in isolated injuries: 1.5% (n = 293). In children aged 1–15 years the highest mortalities occurred in multiple injuries including head/thoracic (47.7%) and head/abdominal injuries (49.9%). Having a Glasgow Coma Scale of <15 on presentation to hospital was associated with a mortality of 16%. Conclusions: Differences in injury patterns and mortality exist between different age groups and high-risk injury patterns can be identified. With increasing age, a decline in the proportion of children with head injury and an increase in the proportion with limb injury were observed. This information is useful for directing ongoing care of severely injured children. Future analyses of the TARN database may help to evaluate the management of high-risk children and to identify the most effective care.

Using multiple datasets to understand trends in serious road traffic casualties

Accurate information on the incidence of serious road traffic casualties is needed to plan and evaluate prevention strategies. Traditionally police reported collisions are the only data used. This study investigate the extent to which understanding of trends in serious road traffic injuries is aided by the use of multiple datasets. Health and police datasets covering all or part of Great Britain from 1996-2003 were analysed. There was a significantly decreasing trend in police reported serious casualties but not in the other datasets. Multiple data sources provide a more complete picture of road traffic casualty trends than any single dataset. Increasing availability of electronic health data with developments in anonymised data linkage should provide a better platform for monitoring trends in serious road traffic casualties.

Mortality from trauma haemorrhage and opportunities for improvement in transfusion practice

The aim of this study was to describe the prevalence, patterns of blood use and outcomes of major haemorrhage in trauma.

Trauma scoring systems and databases

This review considers current trauma scoring systems and databases and their relevance to improving patient care. Single physiological measures such as systolic arterial pressure have limited ability to diagnose severe trauma by reflecting raised intracranial pressure, or significant haemorrhage. The Glasgow coma score has the greatest prognostic value in head-injured and other trauma patients. Trauma triage tools and imaging decision rules-using combinations of physiological cut-off measures with mechanism of injury and other categorical variables-bring both increased sophistication and increased complexity. It is important for clinicians and managers to be aware of the diagnostic properties (over- and under-triage rates) of any triage tool or decision rule used in their trauma system. Trauma registries are able to collate definitive injury descriptors and use survival prediction models to guide trauma system governance, through individual patient review and case-mix-adjusted benchmarking of hospital and network performance with robust outlier identification. Interrupted time series allow observation in the changes in care processes and outcomes at national level, which can feed back into clinical quality-based commissioning of healthcare. Registry data are also a valuable resource for trauma epidemiological and comparative effectiveness research studies.

Temporal trends in head injury outcomes from 2003 to 2009 in England and Wales

Background: Case fatality rates after blunt head injury (HI) did not improve in England and Wales between 1994 and 2003. The United Kingdom National Institute of Clinical Excellence subsequently published HI management guidelines, including the recommendation that patients with severe head injuries (SHIs) should be treated in specialist neuroscience units (NSU). The aim of this study was to investigate trends in case fatality and location of care since the introduction of national HI clinical guidelines. Methods: We conducted a retrospective cohort study using prospectively recorded data from the Trauma and Audit Research Network (TARN) database for patients presenting with blunt trauma between 2003 and 2009. Temporal trends in log odds of death adjusted for case mix were examined for patients with and without HI. Location of care for patients with SHI was also studied by examining trends in the proportion of patients treated in non-NSUs. Results: Since 2003, there was an average 12% reduction in adjusted log odds of death per annum in patients with HI (n=15,173), with a similar but smaller trend in non-HI trauma mortality (n=48,681). During the study period, the proportion of patients with HI treated entirely in non-NSUs decreased from 31% to 19%, (p <0.01). Interpretation: The reduction in odds of death following HI since 2003 is consistent with improved management following the introduction of national HI guidelines and increased treatment of SHI in NSUs.