Electromagnetic navigation bronchoscopy-guided preoperative lung nodule localization in video-assisted thoracic surgery (VATS): a learning curve analysis.

Abstract

Background: Electromagnetic navigation bronchoscopy (ENB) has been widely used to mark small peripheral pulmonary nodules (PPNs) in video-assisted thoracic surgery (VATS) resection. This technique offers the advantages of a high accuracy and fewer complications. However, few studies have analyzed the learning curve of ENB-guided preoperative localization. We aimed to describe the learning curve and factors influencing ENB-guided thoracoscopic pulmonary nodule resection.

Methods: This study included 300 consecutive patients with PPNs who underwent ENB-guided localization by the same endoscopist in our department between November 2019 and December 2021. The cumulative sum (CUSUM) method was used to analyze the learning curve of ENB-guided localization and the learning curve in different lobes, while logistic regression was used to analyze the risk factors affecting ENB operative time (OT).

Results: In 184 patients with 300 nodules, three learning phases were identified through turning points of the learning curve: Phase I (the 16^th nodule), Phase II (the 17^th to the 107^th nodule), and Phase III (the 107^th to the 300^th nodule). No significant difference was found in the success rate of ENB-guided localization in each phase of the learning curve (100%, 96.7%, and 97.9%, P=0.78). The distance from the localization to the pleura in Phase I was statistically significantly shorter than that in Phase II and Phase III (0.6±0.4 vs. 1.1±0.6 vs. 1.0±0.5 cm, P=0.001 and P=0.003). Furthermore, the learning curves for nodules in different lobes were different. The learning curve for the upper lobe nodules was divided into two phases; the learning curve for the middle lobe disclosed more negative values; and the learning curve for the lower lobe nodules displayed no obvious pattern. Significant differences were found in nodule location, distance from the localization to the pleura and learning curve phase (P=0.003, P<0.001, P=0.02). The independent factors for OT included gender, smoking history, nodule type, distance from localization to the pleura, and learning curve phase.

Conclusions: ENB OT at the 107^th nodule leveled off and showed a downward trend. Different lobes have different learning curves, the middle lobe is the easiest lobe to learn with ENB and can be used as the first lobe of choice for beginners. The learning curve can objectively evaluate the accuracy of ENB location and help endoscopists identify areas for improvement.