Learning curve for oesophageal cancer surgery

Abstract

Abstract Background Surgical training and experience are frequently claimed to influence early and late outcome measures. The aim of this study was to examine any improvement in an individual surgeon's performance in one operation over a period of 7 years from initial appointment to date. Methods Patients undergoing Ivor Lewis subtotal oesophagectomy performed by a single surgeon between April 1990 and December 1996 were identified from a prospectively compiled oesophageal cancer database. Operating time (abdominal, thoracic and ‘one-lung time’), blood loss, transfusion requirements (intraoperative and total), extent of lymphadenectomy (number of lymph nodes sampled), intensive treatment unit (ITU) stay, hospital stay, postoperative morbidity and mortality, pathological stage, grade and survival were recorded. Results The records of 150 patients were identified for analysis. The cohort was split into five groups, each of 30 patients operated on consecutively. Each of the groups was comparable for age, sex, smoking history, preoperative haemoglobin and creatinine levels, weight loss, American Society of Anesthesiologists' grade, and histological stage and grade of disease. Analysis of the variables pertaining to operation revealed a significant improvement with time including reduced single-lung operating time (P = 0·01), reduced blood loss (P = 0·03), reduced transfusion requirement (P < 0·0001), reduced ITU stay (P < 0·0001), reduced inpatient stay (P < 0·0001) and an increased yield of lymph nodes (P < 0·0001). Conclusion This study showed a continuing improvement in a surgeon's performance over a 7-year period. With the current trend to shorter training periods there is a case for continuing supervision of the ‘fully trained’ surgeon within highly specialist units.