Psychometric properties of the Impact of Event Scale in Greek cancer patients.

Abstract

To develop the Greek version of the Impact of Event Scale-Revised (IES-R-Gr), assess its psychometric properties, and finally to examine the impact of cancer diagnosis to a palliative care patient sample, the IES-R was translated into Greek using the "forward-backward" procedure. It was administered twice, at one-week intervals, to 82 eligible patients with advanced cancer. Together with the IES-R-Gr scale, the patients also completed the Hospital Anxiety and Depression (HAD) Scale. Reliability was assessed in terms of internal consistency (Cronbach's alpha coefficients) and test/retest (Spearman's r value and Kendall's tau-b). Construct validity was demonstrated through association with the HAD Scale, and convergence and discriminative validity and interscale correlations were also assessed. The Greek version of the IES-R had Cronbach's alphas for the intrusion, avoidance, and hyperarousal scales of 0.72, 0.77, and 0.85, respectively. Overall test-retest reliability was satisfactory at P<0.0005. Satisfactory construct validity was supported by the correlation analysis between the IES-R-Gr subscales and anxiety and depression. Factor analysis yielded three factors, explaining 57.26% of the variance. Interscale and interitem correlations were found satisfactory at P<0.0005. These results demonstrate that the IES-R-Gr is an instrument with satisfactory psychometric properties and is a valid research tool for the impact of cancer diagnosis in advanced cancer patients.