LAUSR

BACKGROUND No means exists to classify the overall credibility of physical torture allegations, which relies on qualitative forensic medical criteria. OBJECTIVES To develop procedures for processing physical data from torture allegations in order to achieve robust credibility assessments. To assess inter-observer variation in semi-quantitative evaluations. To analyse the importance of physical aspects for the overall assessment of allegations of torture. To assess the correlation between a quantitative description of physical aspects of torture and the appraisal of credibility based on the semi-quantitative model. METHODS We developed a semi-quantitative model to assess credibility of physical torture allegations and a model to describe allegations quantitatively. Information about physical aspects of ill-treatment in 202 extensive reports made by psychologists (part of a Basque Country Governmental project) was appraised individually and jointly by the two authors in the semi-quantitative model. Inter-observer variation was appraised. The assessed credibility based on physical details was compared with the psychologists' overall assessments and with the quantitative scoring of physical data. FINDINGS The inter-observer agreement was substantial with kappa values of 0.64-0.77. Peer-reviewing revealed an error; using the defined criteria in the models facilitated achievement of common opinions. Comparison of credibility assessments found by different means suggests that physical data may be given a low priority in the overall assessment of torture allegations. There was a strong correlation (r=0,86, p<0.001) between the semi-quantitative credibility assessment and the quantitative scoring of physical aspects of torture, reinforcing the validity of both means. CONCLUSION Peer-reviewing of physical aspects of torture based on a qualitative /semi-quantitative /quantitative multi-step analysis leads to robust assessments of credibility. A procedure is needed on how to merge the credibility assessments of the psychological and the physical data into an overall conclusion. We suggest that the highest level of the two evaluations should constitute the overall credibility assessment. LIMITATIONS The examinations were done years after the detentions, they were not focused on physical details and the data we used was retrieved retrospectively.