Trauma audit: clinical judgement or statistical analysis?

Comparisons have been made between two methods currently used to assess the effectiveness of management of major trauma. These are the review of fatal cases by senior clinicians and the use of statistical analysis of severity scores. The former was assessed by a re-examination of the Coroners' reports of 508 patients reviewed by senior clinicians at the request of The Royal College of Surgeons of England Working Party on the Management of Patients with Major Injuries. The latter was based on the 665 fatalities on the files of the UK Major Trauma Outcome Study. The two groups of patients had comparable age and sex profiles and broadly similar ranges of injury severity. There were major differences between and inconsistencies within the two assessments. Clinicians more frequently judged death avoidable in those with very severe injuries. In contrast, the statistical analysis suggested, paradoxically, that the proportion of avoidable deaths in those patients who had minor injuries was less than the proportion of avoidable deaths in those who had more serious injuries. These variations underline the limited values of retrospective peer review and will not encourage clinicians to adopt currently available statistical methods. Further refinements of anatomical and physiological scoring systems and their integration to provide a statistically valid and clinically acceptable measure of outcome are essential prerequisites to the wider introduction and success of trauma audit.