by Noor Janjua Zaidi, BSHIM, RHIA
Health Informatics looks at the promise—and the risks—of AI-supported clinical decisions
Imagine this: A digital tool notices that a patient may be at risk, flags a possible medication error, or helps a clinician make a more informed decision within the electronic health record (EHR). That is the promise of artificial intelligence in healthcare. But here is the question: Are healthcare organizations ready to use these tools safely and effectively?
What does the research tell us?
A systematic review published in September 2026 examined 20 studies on healthcare professionals’ adoption of AI-based clinical decision support systems (AI-CDSS). The message was clear: successful adoption depends on more than the technology itself. Staff training, organizational support, and meaningful involvement of healthcare professionals in development and implementation of all matters. (Yushu et al.)
Waldock et al. (2026) reached a similarly cautious conclusion after reviewing 50 studies across 17 medical specialties. AI-based predictive systems showed moderate overall performance, but results varied across clinical settings. In other words, a tool that performs well in one environment may not work equally well in another.
Where could AI make a difference?
In my healthcare experience, accurate patient information, clear communication, and dependable EHR systems are necessary to safe care. From a health informatics perspective, AI could add value by helping teams:
- Reduce administrative burden and streamline clinical workflows
- Identify patient-safety concerns earlier
- Detect possible medication errors or high-risk patterns
- Support—not replace—professional clinical judgment
What could go wrong?
AI can introduce new risks when it relies on inaccurate data, generates too many alerts, or encourages users to depend too heavily on automated recommendations. Before integrating AI into routine care, organizations should ask: Is the data reliable? Are staff prepared? Who monitors the system after launch? How will problems be reported and corrected?
Four priorities for safer AI
- Protect data quality. AI recommendations are only as dependable as the information behind them.
- Train and involve staff. The people using technology should help shape how it is designed and introduced.
- Monitor performance continuously. Organizations must evaluate reliability across real clinical settings, not only during initial testing.
- Keep people at the center. AI should strengthen clinical judgment, patient trust, and equitable outcomes—not to replace human expertise.
Your turn
If AI were introduced in your healthcare setting tomorrow, what would make you trust it—and what safeguards would you want in place? The real measure of success will not be technological efficiency alone. It will be whether AI makes care safer, earns patient trust, and supports fair healthcare outcomes.
References
Waldock, W. J., Guni, A., Darzi, A., & Ashrafian, H. (2026). Performance of predictive AI-based clinical decision support systems across clinical domains: A systematic review and meta-analysis. PLOS Digital Health, 5(3), e0001310. https://doi.org/10.1371/journal.pdig.0001310
Factors influencing healthcare providers’ adoption of artificial intelligence-based clinical decision-support systems: a systematic review using the CFIR framework | Human Resources for Health | Springer Nature Link (2026). Human Resources for Health. https://doi.org/10.1186/s12960-026-01102-x
About the Author
I am Noor Ul Ain Janjua Zaidi, RHIA, and I am currently pursuing a Master of Health Informatics at the University of Cincinnati. My background in Health Information Management and experience in healthcare have inspired a strong passion for using technology, data analytics, and artificial intelligence to improve patient safety, enhance healthcare quality, and support better clinical decision-making. I am also proud to be an active member of OHIMA, serving on the Nominating Committee and AI Task Force, where I enjoy contributing to professional initiatives and exploring the evolving role of artificial intelligence in health information management.
