AI in Surgical Safety: Seeing What Humans Can’t

Surgical environments are complex, characterized by rapid decision-making, high stakes, and a reliance on seamless team coordination. Despite extensive training and well-defined protocols, even the most experienced teams encounter preventable errors and near-misses. These lapses can stem from unpredictable variables, communication breakdowns, or workflow inefficiencies, and they occur across institutions and specialties. 

Current methods of quality improvement fail to meet the urgency or scale of the problem. Intraoperative data are often manually recorded, introducing inaccuracy, bias, and incompleteness. Valuable insights are lost when many events go undocumented or are too resource-intensive to analyze. Risk-adjusted outcome data, such as those obtained through the National Surgical Quality Improvement Program,¹ offer only partial snapshots. These reports are expensive, delayed by months, and limited to a fraction of procedures. 

AI Can Elevate Surgical Safety Without Assigning Blame 

The use of AI in healthcare² has emerged as a powerful tool to support surgical teams. AI-enabled technologies can capture and analyze full perioperative workflows, offering objective insights into performance, safety, and communication. Rather than focusing on individual errors, AI helps teams recognize patterns and systemic vulnerabilities that may compromise patient care. This shift toward systems-based analysis enables a safer, more adaptive surgical environment. 

What Does AI “See” in the OR?

AI-enabled technologies are capable of processing multimodal data from the OR to deliver a comprehensive, real-time understanding of surgical performance. These technologies observe how teams communicate and collaborate during procedures, providing insight into behavioral dynamics and safety compliance.

Several key metrics include adherence to surgical safety checklists, timing of critical steps, identification of workflow anomalies, clinical and operational outliers, and tone and sequence of communication. Rather than serving as a surveillance tool, AI functions as an impartial observer—offering data to support learning and system improvement.

The Black Box Platform™: Continuous Monitoring Meets Predictive Insight

Inspired by aviation’s flight data recorders, the Black Box Platform acts as a surgical procedure recorder. It captures audiovisual, environmental, and physiological data—the entire clinical experience—and transforms raw input into actionable insights for safety, quality, and operational performance.

The platform detects deviations from protocols, identifies delays and coordination gaps, and highlights safety events that would otherwise go unnoticed. It operates passively, allowing the surgical team to remain focused on patient care without distraction. By design, it supports honest reflection on care delivery and drives systemic improvement—without assigning blame.

What are Real-World Examples of AI in Surgical Safety?

The Black Box Platform has been in use in ORs around the world for over a decade, and its impact is supported by a growing body of peer-reviewed research. The following studies illustrate how AI-driven insights are translating surgical safety from theory into practice.

• Using the OR Black Box® to Determine Quality Improvement Outcomes:³ This study examined the effectiveness of implementing simulation training to improve surgical safety checklist performance, utilizing the OR Black Box for assessment. The findings revealed that surgeon-led Timeouts and Debriefs may inadvertently silence other team members, while interprofessional simulation discussions created rare opportunities for collaborative dialogue that improved shared understanding of institutional policies and enhanced team cohesion. The results demonstrated significant improvements in debrief-related metrics for surgical teams that participated in the simulation training compared to those who did not.

• Impact of Team Performance on the Surgical Safety Checklist and Patient Outcomes:⁴ This study evaluated the association between surgical safety checklist compliance and patient outcomes. The results showed that higher overall checklist compliance was associated with lower mortality rates, shorter hospital stays, and reduced ICU admissions. Most notably, better performance during timeouts and debriefings was linked to improved patient outcomes including 30-day readmissions and 30-day mortality rates. The researchers concluded that AI can be instrumental in measuring and improving our ability to understand and address patient safety issues in real-time during surgical procedures.

• Remote Assessment of Real-World Surgical Safety Checklist Performance:⁵ This study evaluated surgical safety checklist performance across 7 North American academic medical centers and focused on checklist compliance, team engagement, and the quality of checklist content review. The analysis revealed widespread variability in compliance, timing, and execution quality across centers. The study found that a time-out was performed during most surgical procedures (98.4%), whereas a debrief was performed during 62.3% of procedures. Furthermore, when a team introduction was performed, it was associated with more checklist prompts completed, a higher engagement score, and a higher percentage of team members who ceased other activities during time-out. These insights would be nearly impossible to gather using traditional documentation methods.

• Using the OR Black Box® to Assess Surgical Team Member Adaptation Under Uncertainty:⁶ This study analyzed how different surgical team members contribute to teamwork and adapt their skills during uncertain situations, particularly when dealing with intraoperative adverse events (IAEs). The findings revealed distinct patterns of adaptation among team members during IAEs. Nurses significantly increased their backup behavior (recognized when assistance was needed and asked other team members for support), while surgeons and medical trainees demonstrated a substantial increase in psychological safety-related skills (openly described errors committed, verbalized a personal or team safety concern, and provided a message focused on problem-solving and avoided assigning blame to the operator). Furthermore, all roles showed a decrease in situation assessment skills during IAEs (communicating the status of activities or responsibilities to the team, instructing or informing the team about the operative plan, and bringing the team together to problem solve). These results highlight the unique contributions and adaptations of each role in the surgical team during critical moments.

• Analyzing Interprofessional Teamwork in the Operating Room:⁷ This prospective observational study evaluated different methods for assessing surgical teamwork and patient safety, comparing two traditional assessment tools, the Non-Technical Skills Assessment Tool (NOTECHS) and Team Emergency Assessment Measure (TEAM) with a modified-Systems Engineering Initiative for Patient Safety (SEIPS) model. The study found that while NOTECHS and TEAM assessments showed consistently high teamwork scores, the SEIPS approach revealed a more nuanced picture, capturing both optimal and suboptimal teamwork behaviors as well as team resilience. The researchers concluded that while traditional tools were useful for summative assessments, the SEIPS model provided deeper insights into teamwork processes and contextual factors, suggesting it could be particularly valuable for healthcare organizations seeking to develop more effective teamwork interventions.

Together, these studies highlight how AI-enabled technologies like OR Black Box, powered by the Black Box Platform, offer unprecedented visibility into the intraoperative environment. Rather than relying on retrospective reports or anecdotal recall, healthcare teams can access objective, data-rich insights that inform safety strategies and improve real-world outcomes.

Beyond validation, these findings point to tangible benefits for surgical teams—benefits that are already reshaping perioperative practice.

Proven Benefits for Perioperative Teams

Perioperative teams benefit from the objective, data-backed coaching of AI-enabled technologies rather than relying on anecdotal feedback. These insights can inform skill development, optimize team coordination, and support credentialing decisions.

• Enhanced Safety Checklist Performance and Patient Outcomes. Research demonstrates that AI-driven monitoring leads to measurable improvements in surgical safety checklist compliance. The objective measurement capabilities allow teams to identify specific areas where compliance falters and implement targeted improvements.

• Improved Team Communication and Psychological Safety. AI analysis reveals communication patterns that would otherwise remain invisible, helping teams understand how different roles contribute during critical moments. The technology has been shown to foster psychological safety in healthcare⁸ by creating opportunities for interprofessional dialogue and enabling team members to openly discuss errors and safety concerns without fear of blame. This shift from individual fault-finding to systems-based learning encourages staff to speak up and participate more fully in safety protocols.

• Role-Specific Adaptation and Skill Development. The ability to analyze how different team members adapt during adverse events provides unprecedented insights into professional development. These learnings enable targeted coaching that leverages each role's natural strengths while addressing specific areas for improvement. 

• Operational Excellence and Workflow Optimization. AI-enabled monitoring captures workflow anomalies, timing delays, and coordination gaps that traditional methods miss. The technology identifies clinical and operational outliers in real-time, enabling proactive interventions rather than reactive responses.

• Systems-Based Learning and Continuous Improvement. The ability to detect and address risk factors early has been linked to a reduction in adverse events. Rather than relying on delayed, incomplete manual documentation, teams access comprehensive, objective data that captures the full perioperative experience. This continuous monitoring approach transforms quality improvement from episodic reviews to ongoing, evidence-based enhancement of surgical practice.

See AI in Action

AI is not meant to replace human judgment. Instead, it serves as a force multiplier for safety and performance in surgery. By revealing the unseen patterns and blind spots in perioperative care, AI enables teams to protect patients and improve outcomes with precision. Are you ready to see how data-driven insights can elevate safety in your OR? Book a demo of the Black Box Platform today.

Recommended Reading

1. American College of Surgeons. (2025). ACS National Surgical Quality Improvement Program. https://www.facs.org/quality-programs/data-and-registries/acs-nsqip/

2. Surgical Safety Technologies. (2025, June 17). AI in Healthcare: Navigating Nurses’ Top Concerns [blog post]. https://www.surgicalsafety.com/blog/nursing-concerns-ai-in-healthcare

3. Campbell, K.K., Abreu, A.A., Zeh, H.J., et., al. (2024). Using OR Black Box Technology to Determine Quality Improvement Outcomes for In-situ Timeout and Debrief Simulation. Annals of Surg;10, 1097. https://journals.lww.com/annalsofsurgery/abstract/9900/using_or_black_box_technology_to_determine_quality.976.aspx

4. Al Abbas, A.I., Meier, J., Daniel, W., et., al. (2024). Impact of team performance on the surgical safety checklist on patient outcomes: An operating room black box analysis. Surg Endosc;38, 5613–5622. https://link.springer.com/article/10.1007/s00464-024-11064-7

5. Riley, M.S., Etheridge, J., Palter, V., et., al. (2024). Remote Assessment of Real-World Surgical Safety Checklist Performance Using the OR Black Box: A Multi-Institutional Evaluation. Journal of the American College of Surgeons;238(2), 206-215. https://journals.lww.com/journalacs/abstract/2024/02000/remote_assessment_of_real_world_surgical_safety.7.aspx

6. Incze, T., Pinkney, S.J., Li, C., et., al. (2024). Using the Operating Room Black Box to Assess Surgical Team Member Adaptation Under Uncertainty: An Observational Study. Annals of Surg;280(1), 75-81. https://journals.lww.com/annalsofsurgery/fulltext/2024/07000/using_the_operating_room_black_box_to_assess.13.aspx

7. Boet, S., Burns, J.K., Brehaut, J., et., al. (2023). Analyzing interprofessional teamwork in the operating room: An exploratory observational study using conventional and alternative approaches. J Interprof Care;37(5), 715-724. https://pubmed.ncbi.nlm.nih.gov/36739535/

8. Surgical Safety Technologies. (2025, May 8). Psychological Safety in Healthcare Drives High-Performance Teams - and AI Should Support It [blog post]. https://www.surgicalsafety.com/blog/psychological-safety-in-healthcare-ai-should-support-it