The Effect of Overtime Pancreaticoduodenectomy on the Short-term Prognosis of Patients: a Retrospective Cohort Study of 235 Patients

1. Research Background Due to the large number of operations, surgeons sometimes need to work overtime to perform elective surgery. At this time, surgeons need to perform the operation under fatigue or even sleep deprivation. Fatigue and sleep deprivation will affect the cognitive function, leading an increased error rate at work. There have been numerous studies about the effect of overtime surgery on the prognosis of patients. However, the impact of surgery on patients under fatigue and sleep deprivation is still controversial. Halvachizadeh et al. observed higher complication and mortality rates for orthopedic trauma surgery performed after-hours. Boscà et al. believe that there is no poor prognosis in patient undergoing liver transplantation by a fatigued surgeons. Brunschot et al. reported that nighttime kidney transplantation is associated with less pure technical graft failure.

Pancreaticoduodenectomy is widely used to treat pancreatic cancer, bile duct carcinoma, duodenal carcinoma, and ampullary carcinoma. The operation is complicated, and usually lasts more than 5 hours. Postoperative complications such as pancreatic fistula, gastroparesis, abdominal infection, and abdominal hemorrhage are prone to occur. Extensive literature have clarified the risk factors related to complications after pancreaticoduodenectomy. At present, there is no report on the effect of pancreaticoduodenectomy under overtime on the prognosis of patients. Therefore, the study explores the impact of overtime work for pancreaticoduodenectomy on the prognosis of patients. 2. Research method and design This study is a retrospective cohort study. The clinical data of patients who underwent pancreaticoduodenectomy at the Department of Hepatobiliary Surgery, Peking University People's Hospital from January 2017 to December 2019 were retrospectively analyzed. The clinical data of patients are collected through the medical record management system of Peking University People's Hospital; the collected data includes the time of admission, age, gender, BMI, ASA classification, preoperative total bilirubin level, and past concomitant diseases (hypertension, diabetes, coronary heart disease, brain Vascular disease), tumor location, surgeon, operation time, intraoperative blood loss, surgical reconstruction method, anastomosis method, postoperative hospital stay, intra-abdominal infection, gastric emptying disorder, pancreatic fistula, intraoperative death, hemorrhage, and death during hospitalization , Take the number of lymph nodes; The research data is completed by two researchers independently and simultaneously. After completion, the data completed by the two are compared to ensure the accuracy of the data; All patients were scheduled to undergo elective surgery. Patients were stratified by operative start time into the control group (surgery that started between 8:00 and 16:49) and the overtime group (surgery that started between 17:00 and 22:00). The following parameters as possible confounders were compared: patient age, gender, body mass index (BMI), American Society of Anesthesiologists (ASA) grade, medical co-morbidities, preoperative total bilirubin, site of lesion, surgeon, technique of reconstruction, techinique of pancreaticojejunostomy. The following parameters between overtime group and control group were compared, including operative time, blood loss, number of lymph nodes removed, duration of treatment in ICU, and complication rate. 3. Sample size calculation According to the research on the incidence of pancreatic fistula in pancreaticoduodenectomy, it is estimated that the incidence of pancreatic fistula in the normal group is 10%, and the incidence of pancreatic fistula in the fatigue group is 25%. Assuming that the test type I error α is 0.05, the type II error β is 0.20, and the sample size ratio of the normal group and the fatigue group is 3:1. According to the sample size calculation formula, 228 cases in the normal group and 76 cases in the fatigue group were obtained. 4. Statistical analysis Continuous variable was tested by Shapiro-Wilk test to determine whether it is a normal distribution. Continuous variables that were proved to be a normal distribution were reported by mean and standard deviations. Otherwise, continuous variables were reported by median. Categorical variables were reported by frequency or percentage. Continuous, normally distributed variables were compared with the t-test and non-normally distributed variables were compared with the Mann-Whitney test. Chi-square test is used to compare categorical variables. Reverse stepwise multivariable logistic regression was performed to assess the effects of the potential covariates on outcome. Potential confounders were selected based on a p-value less than 0.2 in univariable analysis. P-values less than 0.05 were considered significant. Data was analyzed in Statistical Package for the Social Sciences version 21.0 (SPSS 21.0).