Advancing Anesthesia Care and Challenges by Artificial Intelligence: A Prospective Study based on Systematic Review and Meta-Analysis
Background: AI technologies remain essential in modern perioperative care, where they improve clinical decision outcomes and delivery of anesthesia. Health professionals need additional validation testing to determine the practical implementation of these systems for regular medical practice. This...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
ziauddin University
2025-07-01
|
| Series: | Pakistan Journal of Medicine and Dentistry |
| Subjects: | |
| Online Access: | https://ojs.zu.edu.pk/pjmd/article/view/3725 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Background: AI technologies remain essential in modern perioperative care, where they improve clinical decision outcomes and delivery of anesthesia. Health professionals need additional validation testing to determine the practical implementation of these systems for regular medical practice. This study aimed to evaluate the clinical effectiveness and diagnostic accuracy of AI-based tools in anesthesia and perioperative care.
Methods: This PRISMA-2020 based review included studies published till May 2025 on AI use in anesthesia and perioperative care with clinical outcomes. Two reviewers independently extracted data and assessed bias (Cochrane for RCTs, NOS for observational). Meta-analysis was done using RevMan 5.4.1 under a random-effects inverse-variance model. Results were reported as SMDs for clinical effectiveness and ORs for diagnostic accuracy. Heterogeneity (I²), subgroup, and sensitivity analyses were also performed.
Results: Seven research studies (5 as randomized controlled trials, 2 as observational ones) with 1,821 patients were found on inclusion. Anesthesia was implemented with the use of AI in the prediction of the diagnosis, monitoring of sedation, and facilitation of recovery. There were no significant differences in the outcomes of recovery (SMD: -0.36, 95% CI: -1.20 to 0.49). More diagnostic accuracy was achieved under the influence of AI (OR: 2.12; 95% CI: 1.05 to 4.27). The risk of bias was moderate or low.
Discussion: New evidence indicates that AI will transform perioperative care through automated decision support functions and outcome prediction solutions; however, a key limitation is the small number of eligible studies and high variability across clinical settings. Thus, standard evaluation standards and multicenter testing activities are necessary for this potential to become realizable.
|
|---|---|
| ISSN: | 2313-7371 2308-2593 |