Article intro - Delphi study on Surgical AI

 

Collins et al. published in npj Digital Surgery: "Guidance on AI-enhanced surgical practice: a Delphi consensus on ontology, data, implementation and evaluation". Full reference: Justin W. Collins, Marco Montesi, Danail Stoyanov, Carla M. Pugh, Marcio C. Moschovas, Denny Yu, Shady Saikali, Jeffrey S. Levy, Zafer Tandogdu, Filippo Filicori, Dimitrios Stefanidis, Swaroop S. Vedula, Martin A. Martino & Gretchen P. Jackson, npj Digital Surgery volume 1, Article number: 17 (2026). 

Abstract

AI-enhanced surgical care lacks standardized frameworks for implementation, evaluation, and governance, limiting safe and scalable adoption. This study integrates a systematic review, expert consensus meeting, and accelerated Delphi process involving 50 expert stakeholders to address these gaps. Consensus was reached with high reliability in over 150 elements defining standards in ontology, data, implementation, and evaluation. The resulting guidance emphasizes rigorous study design, real-world evidence, and lifecycle oversight, providing a structured, consensus-based framework to support clinicians, leaders, and policymakers in responsibly implementing and evaluating AI in surgical practice. Across all stakeholders, success depends on maintaining a closed-loop system in which evaluation continuously informs both data strategy and implementation. Operational adoption may be approached in phases, beginning with institutional readiness and governance, followed by pilot deployments, scaled integration, and ultimately the development of continuous learning systems. This consensus on guidance is intended to serve as an operational blueprint, enabling coordinated, multidisciplinary implementation that ensures AI enhances surgical performance while maintaining safety, trust, and clinical accountability. 

Source: Digital Surgery