Validation of the Italian version of the ROSIER scale for stroke patients at triage

Background: The ROSIER scale allows rapid evaluation of patients with suspected stroke; this paper presents the validation of an Italian version of the ROSIER scale. Methods: Observational study in the emergency department of a major teaching hospital. Patients aged 18 or more, admitted to the emergency department, during a period of two years were enrolled. Construct validity was investigated through exploratory factor analysis (EFA), content validity through the content validity index (CVI-I). Internal consistency, interrater reliability and test-retest stability were assessed. Sensitivity, specificity, positive and negative predictive values were calculated. Results: 539 subjects were enrolled. The Italian ROSIER scale (ROSIER-ita) showed a sensitivity of 97.6%, a specificity of 90.1%, a positive predictive value of 97.5% and a negative predictive value of 82.7%. EFA pointed out 1 factor; the ROSIER-ita explained 59.79% of total variance. Content validity was satisfying (CVI-I = 0.97). No significant differences (p = 0.90) were found in the scores of the first 50 recorded at the time of admission and after 10 min. Interrater reliability was excellent (rho = 0.96, p = 0.0066). Cronbach's alpha was 0.74, indicating acceptable correlation between the items. Conclusion: The Italian version of the ROSIER scale showed satisfactory results, comparable to the original English version, and can be applied by emergency nurses.