M35. clinical validation of the six-item positive and negative syndrome scale (panss-6)

Abstract Background The development of the brief and psychometrically valid, six-item Positive and Negative Syndrome Scale (PANSS-6) holds promise to improve the treatment of schizophrenia by paving the way for implementation of measurement-based care. However, an important limitation to the existing studies of PANSS-6 is that PANSS-6 was extracted from studies in which the 30-item PANSS ratings were obtained through the Structured Clinical Interview (SCI-PANSS). Therefore, it remains unknown whether it is possible to extract sufficient and equally valid information for PANSS-6 rating via a brief and focused interview, which is a prerequisite for the utility of PANSS-6 in clinical practise. The Simplified Negative And Positive Symptoms Interview (SNAPSI) is a brief semi-structured interview, which focuses specifically on extracting information on the PANSS-6. The aim of the present study was to perform a clinical validation study of PANSS-6 ratings obtained via the SNAPSI using PANSS-30 ratings obtained via SCI-PANSS as a gold standard reference. Methods Participants were ≥18 years old, had a diagnosis of schizophrenia (ICD-10: F20.x) and were undergoing inpatient treatment at the Department for Psychosis, Aarhus University Hospital - Psychiatry, Denmark. The SNAPSI and the SCI-PANSS were conducted by trained and reliable independent interviewers, which was followed by independent PANSS-6 and PANSS-30 ratings at two time-points: as soon as possible after admission and as close to discharge as possible. The degree to which the PANSS-6 (rated independently using the SNAPSI) corresponds to PANSS-6 extracted from PANSS-30 (rated using the SCI-PANSS) was tested by means of intra-class correlation coefficient (ICC) analysis. The sensitivity to change was tested by comparing the endpoint-baseline change in the PANSS-6 total scores to the endpoint-baseline change in the PANSS-6 total scores extracted from the PANSS-30 ratings via Spearman correlation analysis. Results A total of 77 inpatients with schizophrenia (age=35.3, SD=11.8 years; males=56%, paranoid schizophrenia=79%) were included. Of these 65% (n=50) were rated at two time-points. Time to complete the SNAPSI was 18.1, SD=6.9 minutes. The mean score of PANSS-30 at baseline and at follow-up was 81.0, SD=15.9 and 71.8, SD=12.5, respectively. The mean score of PANSS-6 at baseline and follow-up was 18.8, SD=4.6 and 18.1, SD=4.0, respectively. The ICC between the PANSS-6 total scores obtained by the SNAPSI and the PANSS-6 total scores extracted from the PANSS-30 ratings was 0.77 [95% CI 0.62–0.85]. The absolute mean deviation between PANSS-6 ratings and PANSS-6 derived from PANSS-30 ratings was 0.7, SD=0.9. Three percent (n=4) of the PANSS-6 ratings deviated by more than a mean of 1 point i.e. >6 points on the PANSS-6 total score compared to the PANSS-6 derived from PANSS-30 ratings. The Spearman correlation coefficient for changes in endpoint-baseline PANSS-6 and PANSS-30 derived PANSS-6 total scores was 0.67, p<0.001. The full results of the study will be presented at the SIRS 2020 conference. Discussion We found an excellent level of correlation between the PANSS-6 total scores obtained via SNAPSI and the PANSS-6 total scores extracted from the PANSS-30 ratings obtained via SCI-PANSS. Also, the sensitivity to change reached a good level of agreement. In conclusion, the combination of SNAPSI and PANSS-6 allows for a brief and valid assessment of the severity of core symptoms of schizophrenia. These results hold promise for the implementation of measurement-based care in the treatment of schizophrenia.