RESUMO
BACKGROUND: Despite the fact that clinicians face uncertainty in their decisions, there is no comprehensive framework to measure it in medical practices which is the knowledge gap especially for Iran. Therefore, this study aimed to evaluate the reliability and validity of a Persian questionnaire which is designed to measure different determining aspects of uncertainty from clinical physicians' perspectives in Iran. METHODS: Clinical Uncertainty Measurement Questionnaire (CUMQ) has been derived from a mixed method study since March 2019 to January 2021. To exclude raw items of the questionnaire, the literature was reviewed and in-depthinterviews were implemented with 24 residents,specialists and sub-specialists in all major clinical fields which resulted in the first theoretical uncertainty in clinical decision making framework. CUMQ content validity has been evaluated using content validity index (CVI) and content validity ratio (CVR). The structural validity of the questionnaire was assessed using confirmatory factor analysis and factor loading and t-value for each indicator of uncertainty is reported. Moreover, to analyze the research model we used the Partial Least Squares (PLS) technique using the SmartPLS software. Convergent (using Average Variance Extracted (AVEs) for each latent variable) and discriminant validity (using the criteria of Fornell and Larckerand cross loading) of the model was also evaluated. After that, the quality of the model was evaluated adjustment through predictive validity (Q2) and effect size (f2). In addition, the reliability was also assessed using Cronbach's alpha and composite reliability. RESULTS: The CVR and CVI ranged from 0. 80 to 1. 00 which illustrates high content validity. Out of 30 items, 24 items had acceptable factor loading and remained in the questionnaire which have been categorized as five main clinical uncertainty dimensions; general determinants, individual determinants of the physician, individual determinants of patient, dynamics of medical sciences, diagnostic and instrumental limitations. The value of composite reliability and Cronbach's alpha for all dimensions were above the threshold value of 0. 7 and the reliability has been confirmed. As AVE values were greater than 0. 5, convergent validity is confirmed. The result of Fornell-Larcker and cross-loadings also indicated that discriminant validity is well established. CONCLUSION: This CUMQ is as avalid and reliable instrument and a suitable tool to measure clinical uncertainty in the Iranian Medical community. However, the reliability of this questionnaire can be studied in other languages and in other countries.
