Reliability and validity of a tool to assess oncology nurses' experiences with prognosis-related communication.

PURPOSE/OBJECTIVES To establish the reliability and validity of a previously developed survey measuring nurses' experiences with prognosis-related communication. DESIGN Psychometric testing of survey. SETTING Mailed survey of Oncology Nursing Society (ONS) members. SAMPLE 392 ONS members. METHODS Reliability was analyzed using Cronbach's alpha. Total scale and subscale characteristics were evaluated through inter-item correlation matrices, average inter-item correlations, corrected item-to-total correlations, and Cronbach's alpha coefficients if the items were removed. Construct validity was assessed using exploratory factor analysis and contrasted group comparisons. MAIN RESEARCH VARIABLES Measures of attitudes toward prognosis-related communication, demographic variables. FINDINGS A three-factor structure emerged with acceptable reliability and validity. Contrasted group comparisons revealed differences in prognosis-related communication by nurses' years of experience with patients with cancer, level of education, and extent of education about prognosis-related communication. CONCLUSIONS The final three-factor instrument, Prognosis-Related Communication in Oncology Nursing, was found to have acceptable reliability and validity. IMPLICATIONS FOR NURSING The final instrument can serve as a tool to measure nurses' experiences with prognosis-related communication. Such measurements may guide interventions that aim to improve the process of prognostic disclosure and elucidate the role of the nurse in the process.