Back

RAPEX HARM from AIS 2015 Coded Injuries

Krampe, J.; Junge, M.

2026-03-10 health informatics
10.64898/2026.03.04.26346267 medRxiv
Show abstract

The European Unions Safety Gate Rapid Alert System (RAPEX) requires Hazard and Risk Assessment Methodology (HARM) evaluations addressing both injury lethality and long-term consequences (LTC). This paper developed a post-processing method to use AIS 2015-coded trauma data directly for RAPEX HARM assessments. AIS 2015 utilizes two metrics: the AIS Code (AIS-CD) for threat to life and the predicted Functional Capacity Index (pFCI) for LTC. While the AIS-CD has been validated on numerous datasets, the pFCI values are based on a theoretical framework that is pending validation. To counter coding variability and poor alignment with clinical diagnoses, initial AIS identifiers (AIS-IDs) were aggregated to a robust level of detail for both metrics. Individual injury severities (AIS-CD/FCI-CD) were aggregated to the person level using a conversion derived from the three most severe injuries (triples), mirroring the concept of the New Injury Severity Score (NISS). The final HARM Level is the most severe outcome derived independently from either the AIS-CD or FCI-CD triple conversion. Analysis showed over 70% of injuries in the aggregated codebook had no LTC. While AIS-CD dominated lower HARM scores, LTC became more defining with increasing HARM severity for the GIDAS sample. At HARM 4 (highest severity), AIS-CD accounted for 53% of cases, FCI-CD accounted for 16%, and both were equally severe in 31% of cases. This method successfully assigns HARM values to AIS 2015 injuries, providing a more holistic severity measure than the current AIS-CD-only approach. HighlightsO_LINovel method assigns RAPEX HARM values to AIS 2015-coded injuries. C_LIO_LICombines lethality (AIS-code) and long-term consequences (FCI-code) for injury severity assessment. C_LIO_LIAggregates injury severity using the three most severe injuries per person. C_LIO_LILong-term consequences account for 13% and 16% of the highest two HARM ratings, respectively. C_LI

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

1
BMJ Open
554 papers in training set
Top 1%
13.1%
2
Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences
15 papers in training set
Top 0.1%
10.6%
3
Scientific Reports
3102 papers in training set
Top 8%
8.8%
4
PLOS ONE
4510 papers in training set
Top 23%
7.5%
5
International Journal of Environmental Research and Public Health
124 papers in training set
Top 0.6%
7.5%
6
JMIR Public Health and Surveillance
45 papers in training set
Top 1%
2.2%
7
BMC Medical Research Methodology
43 papers in training set
Top 0.5%
2.0%
50% of probability mass above
8
Journal of Neurotrauma
27 papers in training set
Top 0.3%
2.0%
9
Frontiers in Artificial Intelligence
18 papers in training set
Top 0.2%
1.8%
10
Computer Methods and Programs in Biomedicine
27 papers in training set
Top 0.3%
1.7%
11
Frontiers in Neurology
91 papers in training set
Top 3%
1.7%
12
Journal of Biomechanics
57 papers in training set
Top 0.4%
1.7%
13
Computers in Biology and Medicine
120 papers in training set
Top 2%
1.6%
14
Scientific Data
174 papers in training set
Top 1%
1.6%
15
Journal of the American Heart Association
119 papers in training set
Top 3%
1.4%
16
BMJ Health & Care Informatics
13 papers in training set
Top 0.5%
1.4%
17
Emergency Medicine Journal
20 papers in training set
Top 0.3%
1.4%
18
JMIR Medical Informatics
17 papers in training set
Top 1.0%
1.3%
19
Frontiers in Digital Health
20 papers in training set
Top 1.0%
1.0%
20
npj Microgravity
11 papers in training set
Top 0.2%
1.0%
21
Journal of Clinical Epidemiology
28 papers in training set
Top 0.4%
1.0%
22
npj Digital Medicine
97 papers in training set
Top 3%
0.9%
23
Royal Society Open Science
193 papers in training set
Top 4%
0.9%
24
JAMIA Open
37 papers in training set
Top 1%
0.8%
25
Philosophical Transactions of the Royal Society B
51 papers in training set
Top 5%
0.8%
26
International Journal of Drug Policy
11 papers in training set
Top 0.3%
0.8%
27
International Journal of Medical Informatics
25 papers in training set
Top 2%
0.8%
28
Data in Brief
13 papers in training set
Top 0.5%
0.8%
29
BMC Medical Informatics and Decision Making
39 papers in training set
Top 3%
0.7%
30
The Lancet Digital Health
25 papers in training set
Top 1%
0.5%