APEX Proficiency Tests Results

The document here provides a summary of results from a proficiency testing agency. The names of the participating laboratories are not revealed but the results were shocking to me. The test is a relatively simple sexual assult: a known sample from the victim, a known sample from the suspect and a mock vaginal swab made up of 0.5 million white blood cells from the victim and 1.4 million sperm from the suspect. As an example of how proficiency tests differ from real case work consider how the evidence sample was handled after preparation (air dried, vacuum sealed with desiccant and stored at -20o C). There was also a hair sample from the victim provided for mitochondrial DNA analysis. In this test it is clear that the major error that can be committed when comparing the sperm fraction of the evidence to the suspect is a false negative NOT a false positive. Nevertheless, 3 out of 135 labs reported that the sperm fraction was only consistent with the victim. Obviously these labs reveal how a switch of the knowns (as in the Kocak case) could then lead to a false positive. Since the hair sample could theoretically match either the victim or the suspect this test provided an opportunity for a false positive (e.g. if the lab declared a match to the suspect). Only 14 laboratories did mitochondrial DNA tests but among this group one lab made a false match. For those of you without a calculator this is an error rate of 7.1%! APEX documents