Regional trauma systems

The negative results from an evaluation do not tell the whole story

In 1988 the Royal College of Surgeons published recommendations designed to improve the management of patients with major injuries.1 The Department of Health responded by supporting the development of a regional trauma service in North Staffordshire and commissioning an in depth analysis of its performance compared with the orthodox British model of care in two other centres in Lancashire and Humberside. Reviews were also undertaken of the cost effectiveness of helicopter ambulances and the role of minor injuries units. The college introduced courses in advanced trauma life support, and the Department of Health supported the extension of the major trauma outcome study, to provide comparative audit data on hospital performance.2

Although the analysis of the regional trauma service examined outcomes in 1990-3, the article by Nicholl and Turner in this issue provides the first opportunity for the general reader to review the data (p 1349).3 Some will conclude that this delay is due either to publication bias against the negative results or to uncertainty about their validity when compared with the very positive earlier report from the Stoke clinicians.4 Whatever the reason, the paper provides an opportunity to rekindle the debate on the organisation of trauma care in Britain eight years after accidents were included as a key area in the Health of the Nation strategy.

Nicholl and Turner state, “There was no reliable or consistent evidence that the developments [in North Staffordshire] improved the chance of survival from major trauma in the region,” whereas the Stoke clinicians' reported an overall reduction in mortality from 38% to 27% in five years and a saving of 17 lives a year. The shorter timescale of Nicholl and Turner's study is unlikely to invalidate their results. Most of the service reconfiguration was completed within the study period, and subsequent referral patterns have not changed significantly. There are two other possible explanations for the discrepancies. Clinicians will claim that it is difficult to ensure that every confounding variable has been addressed when the injuries sustained by the study populations are so diverse. Statisticians will point to the limitations imposed by the relatively low incidence of death after injury when mortality is the main outcome measure.

Perhaps a more robust approach to evaluating trauma systems would be to concentrate on the process of care.5 Measuring adherence to guidelines could be a surrogate measure of outcome if the guidelines had been shown to be based on accepted standards. The starting point must be the randomised controlled trial. This is slowly replacing anecdotal reports on trauma care, but most of the evidence is from overseas. For example, Bickel et al have shown that prehospital intravenous fluid therapy is associated with reduced survival rates in patients with penetrating trunk injuries in Houston.6 This is supported by good experimental and clinical evidence that the currently recommended aggressive treatment of hypovolaemic shock with crystalloid or colloid is misplaced.7 Those data will be reflected in the next version of the advanced trauma life support guidelines, to be published in 1998, which will advocate a cautious move towards hypotensive resuscitation and a renewed emphasis on early surgical assessment.

In 1995 Regel et al described how the integration of trauma services in Germany was associated with a reduction in mortality from 37% to 22% over 20 years.8 Selective use of doctors in the prehospital phase, rapid evacuation by helicopter to a designated trauma centre, the early intervention of intensivists, the ability of senior surgeons to take patients quickly to the operating theatre, and the integration of well resourced rehabilitation services into the hospital environment are considered to be the essential features of this service. One weakness of their study was the failure to analyse the comparative effectiveness of each component of the system, though the authors do emphasise the importance of integration. This concept of the “chain of survival,” so evident in the management of cardiac arrest, is not yet built in to the British response to major trauma.

Nevertheless, the British system has many good features, and it is reassuring to find recent clear evidence of the contribution of treatment to the significant reduction in deaths from trauma among under 25 year olds over the past eight years.9 Equally, we should not be unduly influenced by unfavourable comparisons with North America. The cause, frequency, and demographics of trauma vary significantly between the two countries. Moreover, the popular British understanding of the American system may be inaccurate. It is, for example, generally assumed that American centres employ resident consultants throughout 24 hours, that centres are evenly distributed across the country, and that they treat large numbers of patients. However, the American College of Surgeons Committee on Trauma accepts that on site cover can be provided by a fourth year resident, who will usually have less experience than a fourth year specialist registrar in Britain.10 Many states do not have an integrated trauma system, and some small cities have more than one competing trauma centre.11

The infrastructure in Britain may be more consistent, but it needs enhancement and integration. The rigorous analysis by Nicholl and Turner provides some useful data about how this should—and should not—be done. However, their negative results must be taken in the context of a “shire county” comparison which may not be relevant to the larger metropolitan areas. Also, they emphasise that their report is limited to a review of mortality after major trauma and refer to their unpublished work on avoidable deaths from less severe injuries, the quality of life of survivors, and the cost of the service. These are important data which must be used alongside the results of further randomised controlled studies to construct a much needed evidence based system of trauma care in Britain. They should be published without further delay.