The ability of the ICD-AIS map to identify seriously injured patients in road traffic accidents-A study from Finland

Traffic Inj Prev. 2018;19(8):819-824. doi: 10.1080/15389588.2018.1520985. Epub 2018 Dec 13.

Abstract

Objective: In Finland, the severity of road traffic injuries is determined using the International Classification of Diseases, 10th Revision, Finnish Modification (ICD-10-FM) injury codes from Finnish Hospital Discharge data and the automatic conversion tool (ICD-AIS map) developed by the Association for the Advancement of Automotive Medicine (AAAM). The aim of this study was to evaluate the ability of the ICD-AIS map to identify seriously injured patients due to traffic accidents in Finnish injury data by comparing the severity rating generated by an expert and by the ICD-AIS map.

Methods: Our data came from the North Kymi Hospital (level 2 trauma center at the time of the study). The data included 574 patients who were injured in traffic accidents during 2 years. The severity rating (Maximum Abbreviated Injury Scale [MAIS] 3+) of each patient was recorded retrospectively by an expert based on information from patient records. In addition, the rating was generated from ICD-10 injury codes by the ICD-AIS map conversion tool. These 2 ratings were compared by road user categories and the strength of agreement was described using Cohen's kappa.

Results: The proportion of seriously injured patients was 10.1% as defined by the expert and 6.6% as generated by the ICD-AIS map; exact agreement was 65.5%. The highest concordance was for pedestrians (exact agreement 100%) and the weakest for moped drivers and motorcyclists (46.7%). Furthermore, the overall strength of agreement of the severity ratings (slightly or seriously injured) between the expert and the ICD-AIS map was good (κ = 0.70). Most (65%) of the conversion problems were misclassifications caused by the simplicity of the Finnish ICD-10 injury codes compared to the injury codes used in the ICD-AIS map. In Finland, the injuries are recorded mainly with 4-digit codes and, infrequently, with 5-digit codes, whereas the ICD-AIS map defines up to 6-digit codes.

Conclusions: For this sample of simplified ICD-10-FM codes, the ICD-AIS map underestimated the number of seriously injured patients. The mapping result could be improved if at least open and closed fractures of extremities and visceral contusions and ruptures had separate codes. In addition, there were a few injury codes that should be considered for inclusion in the map.

Keywords: AIS; ICD; Traffic injuries; injury coding; injury severity; mapping.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Abbreviated Injury Scale*
  • Accidents, Traffic / statistics & numerical data*
  • Finland
  • Humans
  • Injury Severity Score*
  • Medical Records / statistics & numerical data*
  • Retrospective Studies