Forensic Pathologist Consensus in the Interpretation of Photographs of Patterned Injuries of the Skin

Abstract

Forensic pathologists are commonly asked to evaluate injuries on the basis of photographs. Members of the National Association of Medical Examiners were asked to participate in an online survey in which they were presented with 68 patterned injuries of the skin and asked to make a diagnosis ranging from very general (e.g., "blunt" vs. "sharp") to specific (e.g., "baton blow"). This was not the case. Consensus for general diagnosis averaged 0.77 and 0.72 for specific diagnosis. While there was a strong correlation between consensus and confidence in aggregate, individual correlations were poor. Consensus diagnosis was inversely correlated with age, and positively correlated with jurisdictional size, medical degree, and whether or not the respondent was actively performing autopsies as a job function. A subsequent survey is exploring possible reasons for lack of consensus in low-consensus questions. The high correlation between confidence and consensus at the aggregate level and low correlation at the individual level may have implications for quality assurance protocols.

Keywords: diagnosis; forensic science; image analysis; patterned injuries; quality control.