Comparison of the Learning Curves of Ultrasound-Guided In-Plane Needle Placement Among Four Different Puncture Modes: A Randomized, Crossover, Simulation Study

The effects of different positional relationships between the probe, needle, and puncture model on in-plane puncture performance have not been fully evaluated. In this simulation study, we used a 4-period crossover design to compare the learning curves of ultrasound-guided in-plane needle placement among 4 different puncture modes by novices. Forty residents were randomly assigned to receive training in one of 4 puncture modes according to the placement of the puncture model and the orientation of the probe to the operator: horizontal phantom-parallel probe (HP), horizontal phantom-vertical probe (HV), vertical phantom-parallel probe (VP), and vertical phantom-vertical probe (VV). They were allowed 10 trials on each mode and then received the other 3 trainings following the predefined sequences based on a Williams design. Puncture time was recorded from needle entry until successful in-plane puncture under ultrasound guidance. Linear and generalized linear models indicated significant effects of puncture mode and trial number on puncture time (P < .001 for all models). The mean (standard deviation [SD]) puncture times for 10 trials were 44 (44) s for HP, 37 (34) s for HV, 80 (57) s for VP, and 46 (48) s for VV. HV had the shortest puncture time, while VP had the longest. No significant difference was observed in puncture time between VV and HP modes (P = .330). Within each mode, puncture time significantly decreased from the first to the tenth trial (P = .001 for HP, P < .001 for HV, P < .001 for VP, and P = .002 for VV). VP showed the steepest learning curve; however, even after 10 trials, its puncture time remained significantly higher than that of the other 3 modes (P < .001 for all comparisons). Ultrasound-guided in-plane puncture difficulty follows the order VP > HP = VV > HV.