Applications and limitations of Forensic Biomechanics: a Bayesian perspective.

Forensic Biomechanics is an analytic method intended for presentation in a court of law. The method consists of the reconstruction of an injury mechanism followed by a comparison between the injury risk of the mechanism and the injury tolerance of the individual. In recent years some courts have excluded such testimony based, in part, on the inability of experts to quantify the potential error of the methods they relied upon in reaching their conclusions. The application of Bayes' Law to a forensic test of truth in a disputed matter allows for quantification of the error inherent in the method through the conditioning of the pre-test probability of the test outcome with the true and false positive rate of the test. The result of the calculation is the Error Odds (O(E)) for the test, or the ratio of correct to incorrect tests. We present an Error Odds analysis of seven previously published case studies in Forensic Biomechanics as an illustration of the utility of the O(E) as a metric for admissibility of testimony in the courts, with a minimum Error Odds ratio of 10 proposed as a threshold. The results of our analysis yielded only 1 of 7 cases of applied Forensic Biomechanics that surpassed the threshold for admissible testimony of 10, with most the cases falling below an O(E) of 3. The results of the present study suggest that the forensic application of biomechanics is potentially fraught with error. We suggest that an Error Odds analysis be incorporated in Forensic Biomechanics as part of the analysis as a form of quality control and as demonstrable evidence of the accuracy of the methodology. Copyright 2009 Elsevier Ltd and Faculty of Forensic and Legal Medicine. Published by Elsevier Ltd. All rights reserved.