
Dr. Khaled Aboeldahab
AI in Critical Care: Evidence-Based Guides for ICU and Long-Term Care Teams
AI in critical care means using machine learning and large language models to help ICU teams spot deterioration earlier, reduce routine work and plan recovery, with a clinician always making the final decision. This hub gathers my evidence-based guides on what these tools actually do, what the published studies show, and how to use them safely.
I am a physician in critical and long-term care. I wrote these pages for intensivists, ICU nurses, hospital leaders and families who want straight answers rather than vendor claims. Every guide cites peer-reviewed sources, separates proven uses from experimental ones, and ends with practical steps.
Start here: the core guide
AI in ICU Management: A Practical, Evidence-Based Guide
The pillar guide. It covers how AI is used in the ICU today, what the evidence on sepsis prediction really shows, why most ICU AI never reaches the bedside, and a checklist for starting safely. Also available in Arabic.
Guides by topic
Generative AI and large language models
Using ChatGPT and LLMs Safely at the ICU Bedside (read the guide)
Where tools like ChatGPT help (documentation, summaries, education, a second look at a differential) and where they fail (hallucinations, ignoring guidelines, bias). Includes a bedside safety checklist.
Recovery, rehabilitation and long-term care
The Role of AI in ICU Rehabilitation Therapy (read the guide)
What early mobilisation trials show, and how sensors and prediction models may help rehabilitation teams inside the ICU.
AI in Long-Term Care and ICU Rehabilitation (read the guide)
What happens after the ICU: post-intensive care syndrome, prolonged ventilation, falls, pressure injuries and delirium, and what AI can and cannot predict in long-term care.
Coming next
-
AI sepsis alerts: what works and what doesn't. Alert burden, lead time and how to read a vendor's AUROC.
-
AI for ventilator weaning and sedation. Extubation-readiness prediction and patient–ventilator asynchrony detection.
-
How to evaluate an ICU AI tool before you buy it. Local validation, silent running and monthly audit.
-
AI governance for hospital ICUs. Who owns the model, who audits it, and who switches it off.
How I choose what to write about
Each guide answers a question that clinicians or families actually ask. I keep three rules:
-
Evidence first. Each claim links to a peer-reviewed study, guideline or systematic review. When the evidence is weak or mixed, I say so.
-
The clinician decides. AI can flag risk or draft text. It does not examine the patient, talk to the family, or carry responsibility.
-
Fit the unit. A model that worked in one hospital may not work in another. Local validation comes before any alert reaches the bedside.
Frequently asked questions
What is AI in critical care?
It is the use of machine learning, deep learning and language models on ICU data such as vital signs, labs, ventilator settings and notes. Most tools predict risk or summarise information, and a clinician still decides.
Is AI already used in ICUs?
Some tools are, mainly early-warning and sepsis alerts and documentation aids. A 2025 systematic review in JAMA Network Open found that very few ICU AI studies reach clinical integration, so most uses remain experimental. See the pillar guide for the figures.
Will AI replace intensivists?
No. The evidence describes AI as support for clinical judgement, not a replacement for it.
References for this hub
-
Berkhout WEM, et al. Operationalization of Artificial Intelligence Applications in the Intensive Care Unit: A Systematic Review. JAMA Netw Open. 2025;8(7):e2522866. https://doi.org/10.1001/jamanetworkopen.2025.22866
-
Shi T, et al. Large Language Models in Critical Care Medicine: Scoping Review. JMIR Med Inform. 2025;13:e76326. https://doi.org/10.2196/76326
About the author
Dr. Khaled Aboeldahab, physician in critical and long-term care. MBBS (1993); Master's Degree in Critical Care Medicine; MBA in Hospital Management; Professional Diploma in Infectious Diseases (RCI). Member of ESICM, AMDA – The Society for Post-Acute and Long-Term Care Medicine, and the American College of Chest Physicians. Formerly held senior leadership roles in critical care, long-term care and rehabilitation. Featured in a Wolters Kluwer expert insight (2024).
More about the author · LinkedIn
Last updated: 19 November 2026. For education only; not individual medical advice.