Attitudes and perceptions toward artificial intelligence among practicing rheumatologists in Iraq: a cross-sectional web-based survey
Background Artificial intelligence is becoming a bigger part of healthcare, particularly in rheumatology, where clinicians must integrate complex clinical, laboratory, and imaging data. Physicians’ perceptions of AI may substantially influence the extent to which it is implemented responsibly in routine clinical practice. Objective This study aimed to assess the attitudes of practicing rheumatologists in Iraq toward artificial intelligence and to examine their association with demographic and professional charac…

In brief
A web-based cross-sectional survey ran from October 2025 to January 2026 and gathered 162 complete responses from practicing rheumatologists in Iraq. Most respondents were aged 35 to 44, male, and specialists. About 62% said they were familiar with AI and 69% had already used it in rheumatology practice.
Despite that use, attitudes were mostly poor or moderate on a composite scale, which suggests cautious engagement rather than enthusiasm. This matters because physician perceptions shape how responsibly AI is built into routine care. The design is a self-reported snapshot, so it cannot show how attitudes change or whether they affect patient outcomes.
Main points
- The survey collected 162 complete responses from practicing rheumatologists in Iraq over four months.
- 56.8% had a poor attitude toward AI, 32.1% moderate, and 11.1% good.
- 61.7% reported familiarity with AI, and 69.1% had already used AI in rheumatology practice.
- Familiarity with AI (p = 0.003) and involvement in research (p = 0.009) were significantly associated with more favorable attitudes.
The problem
More than half of surveyed Iraqi rheumatologists scored as having a poor attitude toward AI, even though most had already used it, which may hinder responsible clinical adoption.
The rundown
The questionnaire was adapted from earlier studies, refined through expert review and pilot testing, and covered demographics, Likert-scale attitude statements, and general perceptions of AI. Composite scores were sorted into poor, moderate, and good categories using modified Bloom's cutoff criteria, and analysis used SPSS and R.
Respondents were mostly specialists (67.9%), male (56.8%), and aged 35 to 44 (42.0%). Bivariate analysis tested links between attitude and demographic and professional characteristics, and only familiarity with AI and research involvement reached significance.
The authors conclude that targeted digital literacy and specialty-specific education could support effective AI adoption in clinical practice, since greater familiarity and research exposure went with more favorable attitudes.
Sources
- Peer-reviewedFrontiers in Medicine2026-10-16
ace
The debate