Validity and reliability of a rating scale on subjective cognitive deficits in bipolar disorder (COBRA)

BACKGROUND: Discrepancies between bipolar patients' reports and neuropsychological testing have been described and replicated. Unfortunately, no valid, specific, user-friendly, brief instrument is available to measure cognitive deficits as reported by these patients. The main aim of this study was to validate a novel instrument named the "cognitive complaints in bipolar disorder rating assessment" (COBRA). Second, we investigated the relationship between the COBRA, objective cognitive measures and illness course variables. METHOD: The total sample (N=215) included 91 bipolar disorder patients and 124 healthy controls. The psychometric properties of the COBRA (e.g. internal consistency, concurrent validity, discriminative validity, factorial analyses, ROC curve and feasibility) were analyzed. A complete neuropsychological battery was used as objective cognitive assessment. RESULTS: The COBRA had one-factor structure with very high internal consistency (Cronbach's alpha=0.913). A high convergent validity was indicated by a strong correlation with the Frankfurt Complaint Questionnaire (ro=0.888, p<0.001). Bipolar patients experienced greater cognitive complaints compared to control group suggesting a discriminative validity of the instrument. Significant correlations were found between the COBRA and some objective cognitive measures. Furthermore, higher COBRA scores were associated with bipolar II subtype, residual depressive symptoms, hypomanic episodes and total episodes. LIMITATIONS: The cross-sectional design of the study, the influence of medication and severity of patients included. CONCLUSIONS: The COBRA showed to be a useful instrument to assess overall cognitive complaints in bipolar disorder with very satisfactory psychometric properties. Cognitive complaints were partially correlated with memory and executive function measures and with issues that may increase the subjective perception of cognitive deficits, such as subthreshold depressive symptoms and number of episodes.