Which, how, and what? Using digital tools to train surgical skills; a systematic review and meta-analysis

Background: Digital tools like digital box trainers and VR seem promising in delivering safe and tailored practice opportunities outside of the surgical clinic, yet understanding their efficacy and limitations is essential. This study investigated Which digital tools are available to train surgical skills, How these tools are used, How effective they are, and What skills they are intended to teach. Methods: Medline, Embase, and Cochrane libraries were systematically reviewed for randomized trials, evaluating digital skill-training tools based on objective outcomes (skills scores and completion time) in surgical residents. Digital tools effectiveness were compared against controls, wet/dry lab training, and other digital tools. Tool and training factors subgroups were analysed, and studies were assessed on their primary outcomes: technical and/or non-technical. Results: The 33 included studies involved 927 residents and six digital tools; digital box trainers, (immersive) virtual reality (VR) trainers, robot surgery trainers, coaching and feedback, and serious games. Digital tools outperformed controls in skill scores (SMD 1.66 [1.06, 2.25], P < 0.00001, I-2 = 83 %) and completion time (SMD -1.05 [-1.72, -0.38], P = 0.0001, I-2 = 71 %). There were no significant differences between digital tools and lab training, between tools, or in other subgroups. Only two studies focussed on non-technical skills. Conclusion: While the efficacy of digital tools in enhancing technical surgical skills is evident - especially for VRtrainers -, there is a lack of evidence regarding non-technical skills, and need to improve methodological robustness of research on new (digital) tools before they are implemented in curricula. Key message: This study provides critical insight into the increasing presence of digital tools in surgical training, demonstrating their usefulness while identifying current challenges, especially regarding methodological robustness and inattention to non-technical skills.