Learning curve analysis across three surgical eras in Ivor Lewis esophagectomy
摘要
Comprehensive learning curve analyses across the evolution from open to robot-assisted esophagectomy remain limited. This retrospective single-center study analyzed 376 consecutive Ivor Lewis esophagectomies (2012–2023) across three eras: open (n = 127), hybrid-open (n = 52), and robot-assisted minimally invasive esophagectomy (RAMIE, n = 197). Learning curves were assessed using CUSUM analysis, and patient-specific factors were evaluated to guide case selection. Open esophagectomy reached plateau at case 65 with subsequent outcome deterioration. Hybrid-open achieved early plateau (case 15) but showed fluctuating metrics. RAMIE demonstrated steady improvement with CUSUM-derived stabilization achieved at cases 45–50 (hybrid-RAMIE) and 70 (total RAMIE), showing significant reductions in operative times, hospital stay, and ICU stay (p < 0.001). Higher BMI, advanced nodal disease, and comorbidity burden negatively impacted outcomes. In our center, CUSUM-derived stabilization was observed after approximately 45–50 cases for hybrid-RAMIE and about 70 cases for total RAMIE. Strategic case selection—prioritizing lower BMI, limited nodal burden, and minimal comorbidities may accelerate early stabilization of performance. High-volume centers may benefit from total-robotic approaches, while lower-volume programs may favor hybrid pathways.