Measuring resilience with the Connor-Davidson Resilience Scale (CD-RISC): which version to choose?

Abstract

Cross-sectional psychometric study. To compare psychometric properties of the Connor-Davidson Resilience Scale (CD-RISC) with 25, 10, and 2 items, and to assess the agreement between these versions in individuals with spinal cord injury (SCI). Standard psychological screening at a Dutch rehabilitation centre during the first 2 weeks of inpatient rehabilitation. Anonymous data from the psychological screening were analysed. CD-RISC outcomes were checked for floor and ceiling effects. Internal consistency was assessed by calculating Cronbach's α. Convergent validity was assessed by Spearman's correlation between resilience and anxiety, depression, passive coping, and life satisfaction. Agreement between CD-RISC versions was examined by calculating intraclass correlation coefficients (ICCs), corresponding 95% confidence intervals (CIs), and Bland-Altman plots. Total CD-RISC scores were only skewed on the CD-RISC 2 (-1.12). There were no floor and ceiling effects. Internal consistency of the 25-, 10-, and 2-item scales was good to moderate (0.90, 0.86, and 0.66, respectively). Good convergent validity was shown only for the CD-RISC 10. Agreement was highest between the CD-RISC 25 and CD-RISC 10 with an ICC of 0.90 with 95% CI from 0.85 to 0.94. Out of the three CD-RISC versions, the CD-RISC 10 showed the best combination of reliability, validity, and practicality. Therefore, this version is advised as measure of resilience in individuals with SCI in a rehabilitation setting. Measurement of resilience could be part of a psychological screening to identify individuals at risk to develop psychological problems after SCI.