Mamidanna Ravikrishna, Ni Zhifang, Anderson Oliver, Spiegelhalter Sir David, Bottle Alex, Aylin Paul, Faiz Omar, Hanna George B
*Department of Surgery and Cancer, Queen Elizabeth the Queen Mother Wing, St Mary's Hospital, Imperial College London, London, UK †Statistical Laboratory, Centre for Mathematical Sciences, Cambridge University, Cambridge, UK ‡Dr Foster Unit at Imperial College, Department of Primary Care and Public Health, Imperial College London, London, UK §Surgical Epidemiology, Trials and Outcome Centre (SETOC), St Mark's Hospital and Academic Institute, Harrow, Middlesex, UK.
Ann Surg. 2016 Apr;263(4):727-32. doi: 10.1097/SLA.0000000000001490.
The aim of the study was to assess whether there is a proficiency curve-like relationship between surgeon volume and operative mortality and determine the minimum surgeon volume for optimum operative mortality.
The inverse relationship between hospital volume and operative mortality is well-established for esophageal, gastric, and pancreatic cancer. The recommended minimum surgeon volumes are however uncertain.
We retrieved data on esophagectomies, gastrectomies, and pancreatectomies for cancer from the NHS Hospital Episodes Statistics database from April 2000 to March 2010. We defined mortality as in-hospital death within 30 days of surgery. We determined whether there was a proficiency curve relationship by inspecting surgeon volume-mortality graphs after adjusting for patient age, sex, socioeconomic, and comorbidity indices. We then statistically determined the minimum surgeon volume that produced a mortality rate insignificantly different from the optimum of the curve.
Sixteen thousand five hundred seventy-two esophagectomies, 12,622 gastrectomies, and 9116 pancreatectomies were examined. Surgeon volume ranged from 2 to 29 esophagectomies, from 1 to 14 gastrectomies, and from 2 to 31 pancreatectomies per surgeon per year. We demonstrated a proficiency relationship between surgeon volume and mortality in esophageal, gastric, and pancreatic cancer surgery. Each additional case of esophagectomy, gastrectomy, and pancreatectomy would reduce 30-day mortality odds by 3.4%, 7.2%, and 4.1%, respectively. However, as surgeon volume increased, mortality rate continued to improve. Therefore, we were unable to recommend minimum surgeon volume.
Mortality after resections for esophageal, gastric, and pancreatic cancer falls as surgeon volume rises up to 30 cases. Within this range, we did not demonstrate any statistical threshold that could be recommended as a minimum volume target.
本研究旨在评估外科医生手术量与手术死亡率之间是否存在类似熟练曲线的关系，并确定实现最佳手术死亡率的最低外科医生手术量。
医院手术量与手术死亡率之间的负相关关系在食管癌、胃癌和胰腺癌手术中已得到充分证实。然而，推荐的最低外科医生手术量尚不确定。
我们从英国国家医疗服务体系（NHS）医院事件统计数据库中检索了2000年4月至2010年3月期间食管癌切除术、胃癌切除术和胰腺癌切除术治疗癌症的数据。我们将死亡率定义为手术后30天内的院内死亡。在调整患者年龄、性别、社会经济和合并症指数后，通过检查外科医生手术量-死亡率图表，我们确定是否存在熟练曲线关系。然后，我们通过统计学方法确定能产生与曲线最佳值无显著差异的死亡率的最低外科医生手术量。
共检查了16572例食管癌切除术、12622例胃癌切除术和9116例胰腺癌切除术。每位外科医生每年的食管癌手术量为2至29例，胃癌手术量为1至14例，胰腺癌手术量为2至31例。我们证明了外科医生手术量与食管癌、胃癌和胰腺癌手术死亡率之间存在熟练关系。每增加一例食管癌切除术、胃癌切除术和胰腺癌切除术，30天死亡几率分别降低3.4%、7.2%和4.1%。然而，随着外科医生手术量的增加，死亡率继续下降。因此，我们无法推荐最低外科医生手术量。
食管癌、胃癌和胰腺癌切除术后的死亡率随着外科医生手术量增加至30例而下降。在此范围内，我们未发现任何可推荐为最低手术量目标的统计学阈值。
