It all looks so precise and irrefutable on TV shows such as “Bones” and “CSI,” but a study in the August issue of the American Journal of Roentgenology suggests at least one arm of forensic medicine may be more interpretive than exact science.
For the study, “Objective determination of standard of care: Use of blind readings by external radiologists,” researchers at the University of North Carolina at Chapel Hill's radiology department set out to determine if the diagnostic findings of radiologists who act as expert witness typically match up to those of radiologists with no involvement in or knowledge of the case. The researchers selected and sent six CT studies to 31 radiologists participating in the study. None were expert witnesses. The six studies consisted of three randomly selected images taken of emergency department patients, one control image and two emergency room images performed on trauma patients.
One of the trauma-case images was used during a legal action to support a plaintiff's injury claim. During the case, four expert-witness radiologists identified three findings—T3 and T10 vertebral fractures and a 1 millimeter widening of the T10 facet joints—to support the plaintiff's charges. But when the 31 radiologists from the North Carolina study read the court case image along with the five others they where given, none of them diagnosed the injuries cited by the expert-witness radiologists.
The finding, researchers say, prompts questions about the objectivity of paid medical-expert witnesses, and suggests that the use of blinded readings might be a more objective method of evaluating legal-injury cases.