151640 Developing a Risk Index for Predicting Fatality in Aviation Crashes

Monday, November 5, 2007: 11:30 AM

Guohua Li , Department of Emergency Medicine, Johns Hopkins University, Baltimore, MD
Susan P. Baker, MPH , Center for Injury Research and Policy, Johns Hopkins School of Public Health, Baltimore, MD
Hirut Gebrekristos, MPH , Department of Emergency Medicine, Johns Hopkins University, Baltimore, MD
Objective: To evaluate a simple composite risk index for predicting survival outcome in aviation crashes. Methods: The four-point (0-3) scoring index, FIA, is numerically equal to the total number of the three risk factors present in a given aviation crash: Fire, Instrument meteorological condition, and being Away from airport. We assessed the validity of this risk index using aviation crash investigation data from the National Transportation Safety Board (NTSB) for the years 1983-2005. The analysis was first limited to pilot-in-command fatality and then replicated for any fatality. Results: During the study period, the NTSB recorded a total of 43, 535 aviation crashes, in which 8392 (19.3%) pilots-in-command were fatally injured. The pilot crash fatality rate was 3.11%, 18.03%, 61.63%, and 89.50% for FIA scores of 0 (none of the three risk factors present), 1, 2, and 3 (all three risk factors present), respectively. The FIA risk index performed consistently well in predicting pilot fatality in crashes involving different types of flight operations. The area under the receiver operating characteristic curve was 0.864 for major airlines, 0.826 for commuter air carriers and air taxis, and 0.805 for general aviation. The results were similar when the outcome was measured by whether or not the crash resulted in any fatality. Conclusions: The FIA score is a valid tool for predicting fatality risk in aviation crashes. Given its simplicity and high validity, this risk index should be readily applicable to aviation safety research and practice.

Learning Objectives:
To introduce the audience to the risk factors influencing occupant survival in aviation crashes; To illustrate the development of a simple risk index for predicting fatality in aviation crashes; To explain measurement tools such as the receiver operating characteristic curve for validation studies.

Keywords: Injury Risk, Statistics

Presenting author's disclosure statement:

Any relevant financial relationships? No
Any institutionally-contracted trials related to this submission?

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.