P-1301 - Cross-cultural differences in the positive and negative syndrome scale (PANSS) among european and japanese raters

Introduction In clinical trials, good inter-rater reliability reduces error and influences statistical power. PANSS rater differences across cultures are rarely investigated. Objectives To conduct a cross-cultural analysis for European and Japanese PANSS raters, theorized to have unique frames of reference for schizophrenia assessments. Aims To examine reliability with Gold Standard scores (GS); to compare PANSS subscales. Methods There were 170 raters from Europe (73) and Japan (97). Analysis: scale-level for internal consistency (Cronbach α) and inter-rater reliability (ICC); item-level using Rasch and Differential Item Functioning (DIF). Results Internal consistency for PANSS Positive, Negative, and General Psychopathology subscales exceeded α = .99 for both groups. ICCs were higher for Europeans for all subscales. The average Inter-Item Correlation (IIC) with GS was highest for the Positive subscale (Europe, r = .901; Japan, r = .883). Average IIC with GS was lowest for both groups for the Negative subscale (Europe, r = .453; Japan, r = .323). For Negative items, accuracy with GS for Europeans was 46.58% to 89.04%, and 18.56% to 73.20% for Japanese. For PANSS Total, the average IIC was r = .721 (Europe) and r = .659 (Japan). European raters endorsed higher scores on P1.Delusions, P7.Hostility, N1.Blunted Affect, N5. Abstract Thinking. Japanese raters endorsed higher scores on N7.Stereotyped Thinking. Conclusions There were significant differences in several PANSS items, possibly due to disparities between original and translated scale versions, cultural differences in interpreting items, or PANSS scoring parameters. Identifying culturally problematic items can help to tailor PANSS training to diverse geographic regions.