RESUMEN
OBJECTIVES: The study of preventable deaths is essential to trauma research for measuring service quality and highlighting avenues for improving care and as a performance indicator. However, variations in the terminology and methodology of studies on preventable prehospital trauma death limit the comparability and wider application of data. The objective of this study was to describe the heterogeneity in terminology and methodology. METHODS: We performed a systematic literature review and report this using the PRISMA guidelines. Searches were conducted using PubMed (including Medline), Ovid, and Embase databases. Studies, with a full text available in English published between 1990 and 2015, meeting the following inclusion criteria were included: analysis of 1) deaths from trauma, 2) occurring in the prehospital phase of care, and 3) application of criteria to ascertain whether deaths were preventable. One author screened database results for relevance by title and abstract. The full text of identified papers was reviewed for inclusion. The reference list of included papers was screened for studies not identified by the database search. Data were extracted on predefined core elements relating to preventability reporting and definitions using a standardized form. RESULTS: Twenty-seven studies meeting the inclusion criteria were identified: 12 studies used two categories to assess the preventability of death while 15 used three categories. Fifteen variations in the terminology of these categories and combination with death descriptors were found. Eleven different approaches were used in defining what constituted a preventable death. Twenty-one included survivability of injuries as a criterion. Methods used to determine survivability differed and eight variations in parameters for categorization of deaths were used. Nineteen used panel review in determining preventability with six implementing panel blinding. Panel composition varied greatly by expertise of personnel. Separation of prehospital deaths differed with 10 separating those dead at scene (DAS) and dead on arrival, three excluding those DAS, three excluding deaths prior to EMS arrival, and 11 not separating prehospital deaths. CONCLUSIONS: The heterogeneity in methodology, terminology, and definitions of "preventable" between studies render data incomparable. To facilitate common understanding, comparability, and analysis, a commonly agreed ontology by the prehospital research community is required.