Multi-Rater Agreement Using the Adapted Fracture Healing Scale (AFHS) for the Assessment of Tubular Bones on Conventional Radiographs : Preliminary Study

Better understanding of the timing of fracture healing may help in cases of interpersonal violence but also of personal identification. The intra- and inter-rater agreement for the adapted fracture healing scale (AFHS) assessing the post-traumatic time interval on radiographs were tested. This is a preliminary study, providing essential information on method reliability for upcoming studies using the AFHS. Five raters (two radiologists, a forensic pathologist, an orthopedist, and an anthropologist) were presented with a test in three parts consisting of 85 radiographs (from 30 adults) of fractures of tubular bones in different stages of healing purposefully selected from more than 1500 radiographs. The raters were firstly asked to assess 15 features describing fracture healing as present, absent, or not assessable. Thereafter, the raters were asked to choose from the AFHS a single-stage best representing the observed healing pattern. The intra- and inter-rater agreement were assessed using single-rating, absolute agreement, two-way mixed-effects intra-class correlation (ICC) coefficients. The intra-rater ICC of radiologist 1 ranged from 0.80 to 0.94. The radiologists' inter-rater ICC ranged from 0.68 to 0.74, while it ranged from -0.01 to 0.90 for the other raters. The good to excellent ICC among the radiologists and forensic anthropologist provides good foundation for the use of the AFHS in forensic cases of trauma dating. The poor to good results for the other physicians indicate that using the AFHS requires training in skeletal anatomy and radiology.