893 The maximum abbreviated injury scale as a predictor of severe/fatal injuries in Belgian Road victims. (1st September 2016)

• Record Type: Journal Article
• Authors: Nuyttens, Nina; Belleghem, Griet Van; Dupont, Emmanuelle; Leblud, Julien; Devos, Stefanie; Putman, Koen; Berghe, Wouter Van den
• Is Part Of: Injury prevention. Volume 22(2016)Supplement 2
• Page Start: A318
• Page End: A318
• Publication Date: 2016-09-01
• Subjects: MAIS -- ISS -- NISS -- ICISS -- injury severity -- road traffic victims; Children's accidents -- Prevention -- Periodicals; Accidents -- Prevention -- Periodicals; 617.1
• DOI: 10.1136/injuryprev-2016-042156.893
• Languages: English
• ISSNs: 1353-8047

Abstract

Abstract : Background: In 2013, the EC (European Commission) adopted a new definition of seriously injured road victims. All road victims with a MAIS score of 3 or more (MAIS3+) are considered as severely injured. This new definition will coexist along with the conventional definition of severely injured, namely persons who stay at least 24 hours in hospital. Most EU countries will calculate the number of MAIS3+ victims on the basis of (national) hospital data. In Belgium, the Belgian Road Safety Institute and the Vrije Universiteit Brussel have been granted jointly access to national hospital data with detailed injury information for three consecutive years (2009–2011). Consequently, Belgium is able to express the severity of injuries in terms of MAIS, but also in terms of other severity scales such as ISS ( Injury Severity Scale), NISS (New Injury Severity Scale) and ICISS (ICD-9- Based Injury Severity Score). The purpose of this project is to make a mutual comparison of these scales. Methods: The different injury severity scales will be tested and compared as predictors of severe and fatal injuries, based on a dataset of approximately 70.000 road victims. The comparison between the different injury severity scales will be achieved both by exploratory analyses (i.e. descriptive tables and scatterplots) and by fitting generalised linear models with in-hospital mortality as dependent variable and each severity scale separately as a predictor variable. Other independent variables such as age, gender, road user type will also be investigated. The estimates, discrimination ability and calibration of the model containing the MAIS scale will be compared to models containing the other severity scales. Results: Preliminary results show little differences in the predictive performance of the different severity scales. The accuracy of the model improves significantly when age is added as a predictor. Conclusions: Conclusions will be drawn on the final results.