Psychometric evaluation of an instrument measuring artificial intelligence utilization in decision-making domains of healthcare organizations.

The decision-making process in healthcare services encounters numerous challenges. Artificial intelligence (AI) has significantly contributed to enhancing healthcare decision-making. There is a lack of validated instruments available in the literature to measure AI utilization across various healthcare domains. This study aimed to validate an instrument designed to assess the level of AI utilization across various healthcare domains within healthcare organizations. This study was conducted in Iran during the 2024-2025 period, utilizing a methodological design for the development of the study instrument. Initially, the authors formulated and constructed items for a preliminary questionnaire based on a previously published study within the relevant context. To ensure the instrument's validity and reliability, a comprehensive evaluation was performed using multiple methods, including assessments of face validity, content validity, construct validity, and reliability analysis. The final version of the study instrument consisted of 12 items. The instrument demonstrated excellent validity and reliability. The average factor loading across the instrument's items was 0.8, and the principal component accounted for 65.31% of the total variance. Additionally, both Cronbach's alpha and the intraclass correlation coefficient (ICC) values were 0.95, indicating high internal consistency and reliability. Furthermore, the findings indicated that the level of AI utilization in Iran was predominantly low across most assessed items. The study presented a validated instrument for assessing AI implementation across healthcare decision-making domains. Further research is needed to develop specialized instruments for each decision-making domain to enhance data comprehensiveness.