Catatonia is an underdiagnosed and undertreated neuropsychiatric syndrome first described by Kahlbaum in 1874 [1]. During the majority of the 20th century, catatonia was categorized as a subtype of schizophrenia due to its manifestation as a disturbance of volition [2]. However, revisions in nosological systems have unveiled a high incidence of affective disorders in association with catatonia, as well as its overlap with delirium and concurrent presence with other medical conditions [3]. The DSM-5 [4] made changes allowing the recognition of catatonic symptoms in many psychiatric and medical conditions as a specifier, and introducing the term “unspecified catatonia” [5]. On the other hand, the ICD-11 [6] criteria for catatonia included three dimensions: decreased, increased and abnormal psychomotor activity. It should be noted that the ICD-11 criteria for catatonia are broader than the DSM-5 criteria and include a wider range of symptoms. The ICD-11 criteria also include the requirement that the symptoms cause clinically significant distress or impairment in social, occupational, or other areas of functioning. Finally, the ICD-11 criteria categorize catatonia as a separate disorder, not just a specifier, representing a significant advancement compared to previous classification systems [7].

The catatonia syndrome is not infrequent. A systematic review conducted in 2017 determined that the mean prevalence of catatonia in people with psychiatric or medical conditions was 9.2% [8]. Despite the fact that identifying catatonia is not difficult and diagnostic criteria have significantly improved in the last decade, it is often overlooked [9]. It is crucial to note that an early diagnosis of catatonia can prevent complications and significantly improve the prognosis [10]. Some authors argue that there is difficulty for both psychiatrists and psychiatric trainees in identifying the different phenotypes of catatonia and significant inaccuracies in understanding of catatonic features, leading to underdiagnosis of the condition [11,12]. Therefore, it is essential to possess measurement tools that are tailored to the routine clinical practice and encompass the various dimensions of the syndrome.

In order to overcome this important issue, different catatonia scales have been proposed and improved in recent decades [13]. According to a recent systematic review [14], the Bush-Francis Catatonia Rating Scale (BFCRS) [15] was by far the most frequently utilized clinical rating scale for catatonia followed by the Northoff Catatonia Rating Scale (NCRS 40-item version) [16]. The BFCRS was validated in 1996 and it includes the screening tool, the Bush-Francis Catatonia Screening Instrument (BFCSI). A cut-off of two signs from BFCSI has been a controversial topic. Several authors argue the need for at least 3 signs from BFCSI as cut-off to diagnose catatonia [17,18] while others propose the cut-off at 2 signs from the BFCRS [9]. Numerous studies have employed factor analysis of BFCRS to discern the distinct clinical dimensions of catatonia and to differentiate its presentations in mental disorders [19]. To date, as many as seven dimensions of catatonia have been characterized [9,17,20–22], with inhibition and excitement being the most frequently reported in factor analysis [23].

The BFCRS and BFCSI have also been validated in Argentina [24] and Brazil [25]. To the best of our knowledge, the BFCRS and BFCSI, neither the original scales nor subsequent adaptations, have been fully evaluated in acute psychiatric inpatient populations. Furthermore, there are no previous validation studies with such a large sample, using ICD-11 and DSM-5 diagnostic criteria, comparing different catatonia scales or including a factor structure analysis.

This study aims to fill this gap by assessing the psychometric properties of the BFCSI and BFCRS Spain Version (BFCSI- SV and BFCRS-SV) in a large psychiatric inpatient sample using ICD-11 and DSM-5 diagnostic criteria, and other catatonia scales. The properties to be evaluated include reliability, validity, convergent and divergent validity, sensitivity and specificity, and factor structure analysis.

Catatonia is a neuropsychiatric syndrome characterized by a set of behavioural, motor, and emotional symptoms, which can occur in various psychiatric and medical conditions [34]. The BFCRS is a valid and reliable instrument for the assessment of catatonia being sensitive to changes in catatonia symptoms over time [13,14]. The purpose of this study was to translate and adapt the BFCSI and BFCRS scales into Spanish (BFCSI-SV and BFCRS-SV) and to analyze their psychometric properties. In addition, this study aimed to perform a factor analysis to identify the dimensions of the scale and establish the cut-off points required for an accurate diagnosis of catatonia. The results of our study indicate that the BFCSI-SV and BFCRS-SV are valid and reliable instruments for screening and assessing the severity of catatonia syndrome among Spanish inpatient psychiatry patients. The scales demonstrated good internal validity and inter-rater reliability.

A sample of 100 psychiatry inpatients was evaluated using the BFCSI-SV and BFCRS-SV scales. None of the previous validation studies had been fully conducted on such a large sample of hospitalized psychiatric patients [15,24,25]. The presence of two or more BFCSI-SV items for more than 24 h was detected in 51% of the patients, with a sensitivity of 100% and specificity of 67.12%. These sensitivity and specificity results are similar to those reported in the validation study of the Argentinian version of the BFCRSI, 87.5% and 66.67% respectively [24]. It should be noted that this is not a prevalence study, as patients with a higher likelihood of presenting with catatonic symptoms were selected. This is the reason why the number of patients with catatonia may be higher than in prevalence studies [8].

According to our data, the BFCSI-SV and BFCRS-SV scales have good internal consistency, as measured by Cronbach's alpha. The alpha coefficient values were 0.80 and 0.84 for the BFCSI-SV and BFCRS-SV, respectively, indicating that the scales are consistent and reliable, even when removing any of the BFCSR-SV items. This finding suggests that all items were essential for the performance of the BFCSI-SV and the BFCRS-SV. Previous validation studies of the BFCRS did not provide information on internal consistency [15,24,25]. The intraclass correlation coefficient values for the BFCRS-SV and BFCSI-SV were 0.902 and 0.903, respectively, meaning that the inter-rater reliability was found to be excellent. Even though our results have been slightly lower than in the original validation trial (Kappa= 0.93 and Kappa=0.95) [15] and in the validation of the Brazilian version (Kappa=0.93 and Kappa= 0.96) [25], the scores given by different evaluators were highly consistent and they are concordant with recent studies [35,36], indicating the clarity and applicability of the scoring procedures set by the BFRSI-SV and BFCRS-SV.

The area under the ROC curve (AUC) for the BFCSI-SV was 0.971 and for the BFCRS-SV was 0.96, both with high confidence intervals (CI95%). This suggests that both scales have a high ability to discriminate between patients with and without catatonia. The study also found that the BFCSI-SV had good sensitivity (92.3%) and good specificity (90.41%) with a cut-off point of ≥ 5 score, this implies that sensitivity is reduced by almost 8%, while specificity increases by over 23%. In our view, this cutoff point significantly enhances the diagnostic accuracy of the instrument. For the BFCRS-SV, a cut-off point of ≥ 7 score provided good sensitivity (96.29%) and moderate specificity (80.82%). Some authors have suggested a cut-off point of 3 signs for BFCSI [21,24] or 2 signs for BFCRS [9]. We hypothesized that a dimensional approach to the catatonic signs could offer greater accuracy by considering not only the occurrence, but also the severity of symptoms in the diagnostic screening of catatonia. Interestingly, working with this approach, a screening of catatonia with the BFCSI-SV would require only two positive signs if at least one of them is severe and the other one is moderate. Conversely, four positive signs in the BFCSI-SV would not be sufficient for a diagnosis of catatonia if they are all of mild severity. Overall, these results support that the BFCSI -SV and BFCRS-SV scales have high accuracy for the diagnosis of catatonia.

The BFCRS-SV and BFCSI-SV showed a strong positive correlation with the DSM-5 score, ICD-11 score, and MRS, which indicates that the scale assesses catatonic symptoms well and therefore represents an appropriate instrument to measure catatonic severity and to diagnose catatonia. In addition, divergent validity was supported by weak correlations with scales measuring dissimilar constructs such as akathisia, tardive dyskinesia and parkinsonism. The present findings, therefore, add more consistency to the validity of BFCSI-SV and BFCRS-SV.

To the best of our knowledge, this is the first validation study of the BFCSI and BFCRS including a factor analysis. Factor analysis of the scales revealed that a three-factor solution best represents the data, namely the inhibition dimension, the excitement dimension and the parakinetic dimension. The results suggest that the BFCSI and BFCRS have well-defined underlying structures, with a relatively clear separation of items into three distinct factors and they were in line with previous studies analyzing factor structure of the BFCSI and BFCRS [9,21,23,37]. It is worth noting that some of the signs of catatonia included in the various dimensions of the current factor analysis do not conform to the dimensions outlined in ICD-11. For example, ICD-11 includes ambitendency in the dimension of decreased psychomotor activity, whereas in our factor analysis it is included in the parakinetic dimension, even though it also had a high factor loading in the inhibition dimension (0.464). Similarly, ICD-11 includes rigidity, waxy flexibility, and verbigeration in the dimension of abnormal psychomotor activity. Nevertheless, in our factor analysis, the first two signs were included in the inhibition dimension, and the latter was included in the excitement dimension. A relevant finding of our study was that autonomic abnormality did not fit in any of the dimensions. Precisely, ICD-11 considers it as a specifier rather than a diagnostic criterion. As Oldham M. stated in a recent review [38], autonomic abnormality should be included as a specifier in future editions of diagnostic classifications, as it refers to malignant catatonia, the most severe form of catatonia.

To summarize, the BFCSI-SV and the BFCRS-SV maintain the metric properties of the original English version. The results of the study suggest that the BFCRS-SV and the BFCSI-SV are reliable, accurate, and valid tools for the diagnosis of catatonia and for measuring severity of the syndrome. This study found that both scales had a high accuracy in identifying catatonia, with good sensitivity and specificity levels, especially when using a cut-off score of 5 or higher for the BFCSI-SV and 7 or higher for the BFCRS-SV. The results also showed that the instrument had a strong correlation with other catatonia rating scales. Finally, this study has identified a three-factor solution that may facilitate clinical evaluation of patients and guide future research in the field of catatonia.

N/A.

There was no funding for this work.