Intraoperative computed tomography-guided pulmonary tumour localization: a thoracic surgeon's learning curve.

OBJECTIVES: With the increasing availability of hybrid operating rooms, single-stage tumour localization and removal under intraoperative computed tomography (CT) guidance is gaining popularity. The objective of this study was to describe the learning curve for this procedure.

METHODS: Over a 15-month period, a single team of thoracic surgeons without experience in intraoperative CT-guided lung tumour localization performed a total of 91 procedures in 89 patients. All these procedures were conducted in a hybrid operating room equipped with cone-beam CT and a laser navigation system. The learning curve was analysed using the cumulative sum method (target success rate 90%), whereas the moving average was used as an indicator of localization time.

RESULTS: The mean lung tumour size on preoperative CT images was 7.81 mm, whereas their mean distance from the pleural surface was 10.16 mm. The localization time (mean 21.19 min) was inversely associated with the surgeon's experience (Pearson's r = -0.6601; P < 0.001). The moving average analysis revealed that localization time stabilized after 32 procedures. There were 6 failures; of these, 2 occurred during lesion localization (as a result of needle puncture-related pneumothorax) and 4 during surgery (caused either by wire dislodgement or dye spillage). The cumulative sum analysis revealed that proficiency was achieved after 38 procedures. The mean localization time and success rates before and after procedure 38 were 32.13 min vs 13.34 min (P < 0.001) and 86.8% vs 98.1% (P = 0.078), respectively.

CONCLUSIONS: The procedural time and success rates of intraoperative CT-guided lung tumour localization were optimized after 38 consecutive procedures.