Postoperative Care, Perioperative Risks, and the Continuum of Care in Thoracic Surgery: A Comprehensive AI-Augmented Review of Evidence.

Abstract

Background: Thoracic surgery remains essential for lung cancer treatment, yet postoperative pulmonary complications (PPCs) affect up to 50% of patients, driving up mortality and prolonging hospital stays. As the surgical population ages and frailty increases, optimising perioperative care is critical to improving outcomes.

Methods: This narrative review utilised an AI-augmented drafting and verification workflow. Initial drafts were generated using large language models (Antigravity/Google Gemini 3 Pro, Anthropic Claude), with systematic evidence verification carried out with the Consensus.app. The search focused on high-quality trials, meta-analyses, and guidelines published between 2016 and 2026. All clinical interpretations remained under human control.

Results and discussion: Preoperative frailty and nutritional screening identify high-risk patients for targeted optimisation. Prehabilitation significantly reduces PPCs and length of hospital stay. Intraoperative strategies, including individualised PEEP titration and erector spinae plane blocks, offer safer alternatives to traditional methods. Recent evidence supports high-flow nasal oxygen over non-invasive ventilation for most patients and highlights the efficacy of digital drainage systems and restrictive fluid management. Furthermore, procalcitonin-guided antibiotic stewardship reduces exposure, while the "Failure to Rescue" framework, bolstered by continuous wearable monitoring, is essential for preventing ICU readmissions.

Conclusion: Improving thoracic surgery outcomes requires shifting from reactive management to proactive, individualised care. Integrating risk stratification, prehabilitation, modern ventilation, and AI-assisted monitoring are fundamental to advancing precision medicine in thoracic surgery.