The aim of this study was to evaluate interpretations and the reproducibility of interpretations when looking at colposcopic photographs in a forensic setting, as well as discussing some of the dilemmas and pitfalls of forensic colposcopic photography. A total of 316 colposcopic photographs from 51 women taken on three occasions following consensual sexual intercourse, and 78 colposcopic photographs from 39 rape victims, were evaluated by four different observers. Photographs were taken in the same setting, by the same group of investigators, before and after application of toluidine blue dye. The overall Kappa-value for the four observers' judgment of lesion vs. no lesion was 0.41 which can be interpreted as moderate agreement. Intra-observer agreement was calculated for two of the observers looking at photographs with a 10 months' time-gap, and the Kappa-values were 0.41 and 0.52. Positive and negative predictive values of the photographs were 82 and 81 % respectively. This study demonstrates relatively poor reliability of colposcopic photography. Some would argue that this makes colposcopic photography a low-quality method of evaluation and that forensic science should aim for higher standards because of its use in court. Others would argue that as long as the limitations of a scientific method are acknowledged then it is still eligible for use. The moderate agreement and accuracy stresses the need for quality control in the gynecological part of a rape examination. Colposcopic photography also provides a good option for supervision and teaching in an ethically difficult setting. It strengthens the legal rights for both victim and perpetrator.
Forensic Science, Medicine, and Pathology, 2013, Vol 9, Issue 1, p. 24-30