Radiologic identification of disaster victims: A simple and reliable method using CT of the paranasal sinuses

Abstract

To assess the reliability of radiologic identification using visual comparison of ante and post mortem paranasal sinus computed tomography (CT). The study was approved by the responsible justice department and university ethics committee. Four blinded readers with varying radiological experience separately compared 100 post mortem to 25 ante mortem head CTs with the goal to identify as many matching pairs as possible (out of 23 possible matches). Sensitivity, specificity, positive and negative predictive values were calculated for all readers. The chi-square test was applied to establish if there was significant difference in sensitivity between radiologists and non-radiologists. For all readers, sensitivity was 83.7%, specificity was 100.0%, negative predictive value (NPV) was 95.4%, positive predictive value (PPV) was 100.0%, and accuracy was 96.3%. For radiologists, sensitivity was 97.8%, NPV was 99.4%, and accuracy was 99.5%. For non-radiologists, average sensitivity was 69.6%, negative predictive value (NPV) was 91.7%, and accuracy was 93.0%. Radiologists achieved a significantly higher sensitivity (p < 0.01) than non-radiologists. Visual comparison of ante mortem and post mortem CT of the head is a robust and reliable method for identifying unknown decedents, particularly in regard to positive matches. The sensitivity and NPV of the method depend on the reader's experience.