Artificial intelligence-based predictive analytics tools consistently outperformed traditional methods in preoperative settings, achieving AUROCs ≥0.85 for mortality and complication predictions and improving operational efficiency.
Systematic Review (n=14)
Do artificial intelligence-based predictive analytics tools improve preoperative surgical decision-making and workflow efficiency compared to traditional methods?
AI-based predictive analytics enhance preoperative decision-making and operational efficiency, though direct clinical outcome improvements require further study.
Artificial intelligence (AI)-based predictive analytics tools hold promise for enhancing preoperative surgical decision-making, but their prospective clinical and operational impacts remain underexplored. This systematic review evaluates studies assessing AI tools in preoperative settings, focusing on their influence on clinical decisions and workflow efficiency. A comprehensive search of PubMed and Google Scholar, supplemented by citation tracking, was conducted up to December 2024, using MeSH terms and keywords related to AI, preoperative care, and prospective study designs. Inclusion criteria targeted prospective clinical trials, implementation studies, or high-fidelity simulations of AI tools for risk stratification, resource planning, or procedural decisions in surgical specialties. Two reviewers screened records and extracted data on study design, AI tool application, and outcomes (e.g., predictive accuracy, decision impact). Of the 950 screened records, 14 studies (randomized trials, implementations, and simulations) met the inclusion criteria. AI tools consistently outperformed traditional methods, achieving AUROCs ≥0.85 for mortality and complication predictions. Randomized trials showed improved clinician agreement (e.g., weighted kappa increases of 0.13–0.25) and operational efficiency (e.g., 11.2% fewer underpredicted surgical cases). However, direct clinical outcome improvements remain understudied. AI-based predictive analytics enhance preoperative decision-making and operational efficiency but require further multi-center trials to confirm clinical benefits. Seamless workflow integration and bias monitoring are critical for adoption.
Sharifzadeh et al. (Thu,) conducted a systematic review in Preoperative surgical decision-making (n=14). Artificial intelligence-based predictive analytics tools vs. Traditional risk assessment methods and standard clinical judgment was evaluated on Clinical decisions, workflow efficiency, and predictive accuracy. Artificial intelligence-based predictive analytics tools consistently outperformed traditional methods in preoperative settings, achieving AUROCs ≥0.85 for mortality and complication predictions and improving operational efficiency.