|| Checking for direct PDF access through Ovid
To define the learning curve required to improve the peri-operative outcomes of patients operated for recurrent and locally advancedprimary rectal cancer.Consecutive patients undergoing exenterative surgery for colorectal pelvic cancer, by one surgical team, between 2006 and 2009, were included. The patients were divided into quartiles (Q1-Q4) according to the date of surgery. Peri-operative outcomes wereprospectively assessed. A risk-adjusted Cumulative Sum (CUSUM) model was used to evaluate the learning curve.Forty-four patients (29 males; median age 61 (29-78); 11primary cancers) were included in the study. Fifteen patients underwent abdominosacral resection and 26 other pelvic exenterative operations. Pelvic and/or perineal plastic reconstructionprocedures (n = 17) increased significantly from two patients in Q1 to 7 in Q4 (P = 0.058). Consequently, intraoperative time remained unchanged (P = 0.774). The median intra-operative blood loss was reduced from 1.5 (Q1) to 0.6 (Q4) liters (P = 0.548). Complete resection (R0) was achieved in 30 patients. Median hospital stay declined from 22 (Q1) to 15 (Q4) days (P = 0.008). Major complications (re-operations) were significantly reduced from six in Q1 to one in Q4 (p = 0.011). Risk-adjusted CUSUM analysis demonstrated an improvement in the major complications after 15 operations.Fifteen cases are required to gain the desirable training and experience to enhance perioperative outcomes.