CUSUM Analysis as a Predictive Tool for the Learning Curve in Left Bundle Branch Area Pacing

Figures

• 

  Figure 1. Twelve-lead ECG examples illustrating measurement of native and paced QRS duration and V6 R-wave peak time (V6-RWPT) in two representative patients. Panels (A–C) depict a bradycardia patient treated with a conventional pacemaker (non-CRT). (A): native QRS duration (red vertical line marks QRS onset), measured at 100 ms. (B): paced QRS duration, 100 ms. (C): paced V6-RWPT = 66 ms, measured from QRS onset (red line) to the R-wave peak (green line), consistent with left bundle branch capture. Panels (D–F) depict a patient with a cardiac resynchronization therapy indication. (D): baseline wide QRS (180 ms, left bundle branch block morphology). (E): paced/narrowed QRS after implantation (134 ms). (F): paced V6-RWPT = 77 ms, measured from the red (onset) to the green (R-peak) vertical marker. Together, the shortening of QRS duration and the short V6-RWPT support physiological ventricular activation via the left conduction system.

• 

  Figure 2. Smoothed learning curves of fluoroscopy exposure during left bundle branch pacing implantations. (Upper Panel): fluoroscopy time (seconds) plotted chronologically by implant date. The solid grey line represents the 10-case moving average, while the dashed black line indicates the linear regression trend. A progressive decline is observed across the study period, consistent with improved procedural efficiency. (Lower Panel): radiation dose (mGy) plotted in the same manner. Although inter-case variability was present, the overall trend showed a gradual decrease in exposure with accumulating operator experience. Red asterisks mark CRT implants.

• 

  Figure 3. Fluoroscopy time learning curve for non-CRT procedures (DDD and VVI). (Upper Panel): individual fluoroscopy times (black) plotted sequentially with a 10-case moving average (gray). The dashed horizontal line represents the predefined target of 6 min (360 s). After approximately the 65th case, fluoroscopy times stabilize below the target. (Lower Panel): cumulative sum (CUSUM) analysis for the same cases. The curve rises during the early learning phase and reaches its maximum inflection at case ~65, indicating the transition to procedural proficiency. Thereafter, the CUSUM curve flattens and begins to decline, reflecting sustained achievement of the performance target.

• 

  Figure 4. Fluoroscopy time learning curve for CRT procedures. (Upper Panel): individual fluoroscopy times (black) with a 10-case moving average (gray). The dashed horizontal line represents the predefined target of 9 min (540 s). Despite the limited number of CRT implants (n = 9), most cases remained well below the target, with median values close to 5 min. (Lower Panel): cumulative sum (CUSUM) analysis relative to the 9-min target. The curve reaches its maximum inflection at the second case, suggesting early attainment of acceptable procedural performance. Given the small sample size, these results should be interpreted cautiously.