Skip to main content

Evaluation of the Learning Curve Threshold in Robot-Assisted Lung Cancer Surgery: A Nationwide Population-Based Study.

Cancers2025-01-08PubMed
Total: 71.5Innovation: 7Impact: 8Rigor: 7Citation: 7

Summary

This nationwide analysis shows that hospitals typically need around 110 robot-assisted lung cancer surgeries (range 94–174) to reach a learning curve associated with fewer severe complications. Institutions not reaching this threshold had significantly higher rates of severe events including acute respiratory distress syndrome, challenging the commonly cited 25-procedure benchmark.

Key Findings

  • Learning-curve thresholds ranged from 94 to 174 procedures (median 110), not 25.
  • Severe complications, including ARDS, were significantly more frequent in hospitals that did not validate the threshold.
  • Robotic procedures increased from 195 (2019) to 1567 (2022), totaling 3706 surgeries.
  • 24.7% of patients experienced Clavien-Dindo > II postoperative complications.

Clinical Implications

Hospitals should reconsider minimum case volumes before independently offering robot-assisted lung cancer surgery and implement quality monitoring to mitigate severe complications, including ARDS. Credentialing standards may need to be raised to approximately 100–170 cases.

Why It Matters

Redefining learning-curve thresholds has immediate implications for credentialing, centralization, and patient safety, including ARDS prevention. The nationwide scope and methodologically robust approach increase its policy and practice relevance.

Limitations

  • Observational design with potential residual confounding
  • Reliance on administrative coding (misclassification risk)
  • Generalizability may be limited outside France

Future Directions

Prospective evaluation of training and credentialing thresholds, linkage with patient-level clinical data, and external validation in other healthcare systems.

Study Information

Study Type
Cohort
Research Domain
Prevention
Evidence Level
III - Observational population-based cohort analysis using administrative data.
Study Design
OTHER