Abstract
BACKGROUND: Resilience to affective disorders in rehabilitating patients or in individuals with a severe disability is of special research interest. However, there is no gold standard for measuring resilience. We aimed to test the accuracy of the Dutch translation of the Brief Resilience Scale (BRSnl) and of the Resilience Scale (RSnl) in recognizing rehabilitating patients without anxiety and depression, and to determine the reliability and construct validity of both scales. METHODS: A within-subjects longitudinal study with six assessments, each one week apart. Forty residents of a nursing home rehabilitating unit were interviewed to assess resilience (BRSnl and RSnl), optimism and pessimism (LOT-R), depression and anxiety (HADS), positive and negative affect (PANAS), and pain (VAS). RESULTS: Receiver operating characteristic analyses for recognizing the absence of depression and anxiety (HADS-score</=7) revealed better accuracy (P=0.038) for the BRSnl (AUC=0.84; p<0.0001) than for the RSnl (AUC=0.68; P=0.017). The scales correlated moderately at baseline (rs=0.35; p=0.026), and at four-week follow-up (rs=0.50; p=0.004). The RSnl was positively associated with positive outcomes (optimism and positive affect), and the BRSnl positively with positive outcomes, and negatively with negative outcomes (pessimism, anxiety and negative affect). The RSnl showed a better four-week test-retest reliability (ICC, 0.94; 95% CI, 0.87 to 0.97) than the BRSnl (0.66; 95% CI, 0.29 to.83). LIMITATIONS: Short study duration, a relatively small sample. CONCLUSION: The BRSnl showed better performance in detecting people without depression and anxiety than the RSnl, and performed better on construct validity.
