Determination of Consistency and Reliability of Diagnostic Accuracy Derived From Autopsy Photographs

Abstract

Forensic photography plays a significant role in investigation and recording in criminological procedures. Photographs are used as a document in many instances and sometimes becomes a part of an investigation method. Photographs taken at the scene, during the autopsy or examination, might be inspected again for education purposes or in order to revise the details and obtain consultation from other specialists. In this study, determination of the consistency and reliability of diagnosis derived from the photographs taken from firearm injuries during the autopsies is aimed. The cases were chosen randomly among the firearm injury cases, in the Morgue Specialization Department in the Institute of Forensic Medicine, Ministery of Justice, between October 2001 and April 2002. Forensic medicine specialists evaluated lesions of 18 cases at the autopsy and photographs were taken with a 35 mm camera using dia-positive films. The photographs evaluated independently by four forensic specialists twice in a month interval. In order to evaluate the reliability of diagnosis from photographs, the ‘gold standard’ findings which had been defined in the autopsy were compared with the findings of photograph examination. Self-consistency of each reviewer was estimated with kappa test. The study was designed in double-blind. The significance value was accepted as p < 0,03. The kappa median values of the four evaluators were found to be between 62,0 and 86,0. In the first evaluation, one of the reviewers had the highest mean accuracy rate 88 + 0.03, whereas the lowest rate was 78 + 0.03. In the second evaluation, the highest mean accuracy rate was 86 + 0.03 while the lowest was 79 + 0.03. When both evaluations taken in consideration for all the cases, a significant difference in diagnostic accuracies of the reviewers was not noticed (p > 0.05).

Keywords: Forensic photography, autopsy, firearm injuries.