Evaluation of Relationship Between Negative Symptoms Scales in an IRT Framework

Anzalee Khan, PhD, Senior Director of Data Science at VeraSci, presented “Evaluation of Relationship Between Negative Symptoms Scales in an IRT Framework.” at ISCTM’s 2018 autumn conference. There are many assessments of negative symptoms in schizophrenia with no clear consensus on the primary scale to use for this population as the constructs and general utility of each scale varies. This presentation used Item Response Theory to assess the quality and utility of the CAINS (Clinical Assessment Interview for Negative Symptoms), BNSS (Brief Negative Symptom Scale), PANSS Negative Factor (Positive and Negative Symptoms Scale), and the NSA-16 (Negative Symptom Assessment-16). Results showed that the BNSS, CAINS and NSA-16 were more informative for subjects with moderate severity, compared to those with low or severe severity levels. In contrast, the PANSS NSFS was equally informative for all levels of severity, although the BNSS and CAINS have distinct conceptual advantages over the PANSS NSF and NSA-16 and captures a broader spectrum of negative symptoms.