Serious head and neck injury as a predictor of occupant position in fatal rollover crashes

M. D. Freeman, K. Dobbertin, S. S. Kohles, L. Uhrenholt, A. Eriksson

Research output: Contribution to journalArticlepeer-review

9 Scopus citations


Serious head and neck injuries are a common finding in fatalities associated with rollover crashes. In some fatal rollover crashes, particularly when ejection occurs, the determination of which occupant was driving at the time of the crash may be uncertain. In the present investigation, we describe the analysis of rollover crash data from the National Automotive Sampling System-Crashworthiness Data System for the years 1997 through 2007 in which we examined the relationship between a serious head and neck injury in an occupant and a specified degree of roof deformation at the occupant's seating position. We found 960 occupants who qualified for the analysis, with 142 deaths among the subjects. Using a ranked composite head and neck injury score (the HNISS) we found a strong relationship between HNISS and the degree of roof crush. As a result of the analysis, we arrived at a predictive model, in which each additional unit increase in HNISS equated to an increased odds of roof crush as follows: for ≥8. cm of roof crush compared with <8. cm by 4%, for ≥15. cm of roof crush compared to <8. cm by 6% and for ≥30. cm of roof crush compared to <8. cm by 11%. We describe two hypothetical scenarios in which the model could be applied to the real world investigation of occupant position in a rollover crash-related fatality.

Original languageEnglish (US)
Pages (from-to)228-233
Number of pages6
JournalForensic Science International
Issue number1-3
StatePublished - Oct 10 2012
Externally publishedYes


  • Biomechanics
  • Fatality
  • Forensic Epidemiology
  • Rollover crash

ASJC Scopus subject areas

  • Pathology and Forensic Medicine


Dive into the research topics of 'Serious head and neck injury as a predictor of occupant position in fatal rollover crashes'. Together they form a unique fingerprint.

Cite this