Assessing catatonia using four different instruments: Inter-rater reliability and prevalence in inpatient clinical population.

Abstract

The present study aimed to assess inter-rater reliability and prevalence of catatonia according to four diagnostic methods: Bush Francis Catatonia Rating Scale (BFCRS) both screening and complete scale, Braunig's Catatonia Rating Scale (CRS), ICD 10 and DSM5. For inter-rater reliability, different raters evaluated patients using the definitions provides by the four scales: BFCRS Screen and Total, CRS, ICD10 and DSM5. Kippendorff'α was used to compute the inter-rater reliability. Concordance between different systems was assessed using spearman correlation. Prevalence of catatonia was studied using the four definitions in a clinical sample of consecutive adult admissions in a psychiatry ward of a tertiary care hospital. The inter-rater reliability was found to be good for BFCRS Total (α=0.779), moderate for DSM5 and BFCRS screen (α=0.575 and α=0.514 respectively) and low for CRS and ICD10 (α=0.111 and α=0.018 respectively). BFCRS Total and DSM5 definitions of catatonia had highest concordance (rs=0.892 p<0.001). In the prevalence sample of consecutive hospital admissions, the prevalence was found to be highest with the definitions of BFCRS Screen and ICD 10 (10.3%, confidence intervals [CI] 3.9% to 16.7%), followed by BFCRS Total and DSM5 definitions 6.9%, CI 1.6% to 12.2%) and while CRS yielded the lowest prevalence rate (3.4%, CI 0% to 7.2%). Different methods used to determine catatonia in the clinical sample yield different prevalence of this condition.