The article describes a study that aimed to characterise themes, perioperative processes, system factors and outcomes in complaints about surgical care using a scalable large language model approach. The researchers focused on how patient and family complaints could be systematically analysed to reveal patterns in surgical safety and quality.
The study applied a large language model to narrative complaint data to identify recurring themes across the perioperative pathway, including preoperative communication, intraoperative care, and postoperative follow up. By using this method, the team sought to move beyond manual review of complaints and to create a framework that can process large volumes of unstructured text consistently and efficiently.
The authors discuss how insights from the model could help health systems better understand system factors contributing to dissatisfaction and harm, and how these insights might inform quality improvement initiatives in surgical services. The work is presented as an example of how large language models can support learning from patient complaints at scale, with the goal of improving both patient experience and surgical outcomes.
