Cumulative sum analysis of the learning curve for endoscopic resection of juvenile nasopharyngeal angiofibroma

To determine the learning curve with cumulative sum analysis for endoscopic resection of juvenile nasopharyngeal angiofibroma (JNA) and investigate whether the surgeon's expertise is a risk factor for recurrence. We reviewed the medical records of patients with JNA who underwent endoscopic or endoscopic-assisted surgery between 2006 and 2015. We used cumulative sum (Cusum) analysis to plot the learning curve for operation time versus chronological sequence, and verified the Cusum curve by risk-adjusted Cusum (RA-Cusum) analysis. We identified three phases of expertise. The recurrence rate was analyzed using the Kaplan-Meier method and log-rank tests. A multivariable Cox regression analysis was performed to identify the independent risk factors for recurrence. We included 154 JNA patients with a median age of 16 years. The surgeon overcame the learning curve after case 80 with increasing surgical efficiency and competence. The learning curve plotted by Cusum analysis divided the cases into three phases: phase 1, accumulation of initial experience (cases 1-41); phase 2, further accumulation of experience (cases 42-117); and phase 3, mastering the procedure (cases 118-154). Pearson's chi (2) tests showed that tumor stage (P = 0.021), blood loss (P = 0.001), operation time (P < 0.001), and phase (P < 0.001) were associated with recurrence. The log-rank test showed that time to recurrence was significantly shorter in phase 1 than in phases 2 and 3. Blood loss and phase were independently prognostic factors for time to recurrence, with P values of 0.023 and 0.009, respectively. The RA-Cusum analysis identified two inflection points of the curve at case 44 and 83, and verified the results of Cusum analysis. Surgical experience and competence with endoscopic resection affect the recurrence rate in JNA patients. 4.