Are prehospital deaths from trauma and accidental injury preventable? A summary report 17 FIGURE 1. FLOW DIAGRAM DETAILING INCLUSION AND EXCLUSION OF CASES FOR STUDY ONE Deaths referred to the coroner due to injury: n = 564 Deaths excluded: n = 401 Hospital death: n = 243 Death occurred overseas: n = 19 Prehospital deaths: n = 163 Death due to hanging: n = 139 Further deaths excluded: n = 29 End-of-life care: n = 12 Deaths meeting inclusion Medical or natural cause of death: n = 16 criteria: n = 134 Records not located: n = 1 overall score as a result of multiple injuries. together to produce the ISS score, which Each injury was allocated to one of six ranges from 1 to 75 (TARN 2016). body regions (head, face, chest, abdomen, extremities (including pelvis), external) (Brohi > Probability-of-survival estimations were 2007b). The three most severely injured body calculated using Bull’s probits (Bull 1975).4 regions had their score squared and added 4 In 1975 Bull used the ISS to re-analyse the data on 1,333 victims of road traffic accidents treated as inpatients at Birmingham Accident Hospital in 1961. He used probit analysis to linearise the mortality data in separate age groups and showed a good correlation between the ISS and the probability of survival when this North American method was applied to a British population. ‘Bull’s probits’ have been used by several other authors to identify preventable deaths from injury.