Psychometric Properties of the Bush Francis Catatonia Rating Scale

Background: The organization of the signs of catatonia into dimensions is unclear, and little is known about how much information the Bush Francis Catatonia Rating Scale (BFCRS) assesses. Objective: The current study aims to use traditional and modern psychometrics to explore the dimensions of the BFCRS and the amount of information provided by the BFCRS in a catatonic sample. Methods: A retrospective chart review of patients from psychiatric and medical hospitals was conducted. Catatonia was identified in 300 patients with a Bush Francis Catatonia Screening Instrument (BFCSI) score of 2 or more signs. Principal component analysis (PCA) was used to group BFCRS signs into dimensions. Based on these groupings, Item Response Theory (IRT) was computed to estimate parameters that inform item utility and scale reliability Results: PCA identified three components interpretable as “Increased Psychomotor Activity,” “Decreased Psychomotor Activity,” and “Abnormal Psychomotor Activity.” IRT analysis shows that the items Excitement, Immobility/Stupor, Waxy Flexibility were the most representative of each factor, respectively. The BFCRS has many redundant items, and the “frequent” response option could be combined with either “occasional” or “constant” without loss of much information. Finally, the scale as a whole has low reliability at low severities of catatonia, but good reliability at moderate-high severity of catatonia. Conclusions: The BFCRS assessment of catatonia has three dimensions roughly organized by quality of the motor signs. A new or improved rating scale is required to better identify infrequent and subtle forms of catatonia.