An assessment of the repeatability of pubic and ischial measurements.

Forensic anthropologists frequently use measurements of the human skeleton to determine sex and ancestry. Since the establishment of the Daubert criteria of admissibility of scientific evidence to court, methodologies used by anthropologists came under severe scrutiny. It is therefore important to ensure that the osteometric standards that are used in skeletal analyses are clearly explained, repeatable and reliable. Adams and Byrd (2002) found that measurements of the pelvis that originated from a point inside the acetabulum could not be repeated accurately. The purpose of this paper was to use three different sets of pubic and ischial measurements to establish whether they can be repeated with high precision between four different observers, and also by the observers themselves. Generally, high levels of repeatability were obtained, with intra-class correlations (ICC) above 0.8. Pubic and ischial measurements using a point in the acetabulum as origin performed the worst (ICC values of 0.82 and 0.79 respectively for inter-observer repeatability), whereas other methods performed better with ICC values above 0.9. It is advised that pubic and ischial measurements should be taken using the origin of the iliac blade as landmark.