The Learning Curve Of Da Vinci Robot-assisted Mediastinal Tumor Resection

Objective Da Vinci robot-assisted surgery is evolving as an ideal surgical approach in the field of Mediastinal Tumor resection.We aimed to evaluate the learning curve of da Vinci robot-assisted mediastinal tumor resection(DRMTR).Methods A series of 50 consecutive DRMTR procedures were performed by Wang Shumin,the director of thoracic surgery in Shenyang General Hospital between March 2011,when Da Vinci surgery robot was used firstly,and September 2012.All data were collected in a database and analyzed retrospectively.The learning curve was evaluated using the cumulative sum(CUSUM)analysis,and the curve-fitting of CUSUM was designed through Statistical Package for Social Science(SPSS)software.Results Women comprised 54% and men 46% of this 50 cases.These patients had a mean age of 46.9years.This series included 39 cases of anterior mediastinal tumor,1 case of middle mediastinal tumor and 10 cases of posterior mediastinal tumor.All 50 operations were performed successfully.The tumors of 50 cases were completely resected,no residual tumor tissues,no patient death occurred or was converted to thoracotomy.The mean operative time(OT)was 124.6 min(range,41-335 min).According to the scatter diagram of OT,we can see the OT is on the decline with the increase of case number as a whole.The CUSUM learning curve was best modeled as a third-order polynomial(parabola)with equation CUSUM in minutes equal to 0.046 ×case number3-4.681×case number2 + 127.508×case number-237.940,which had a high R2 value of 0.868.The fitting curve crossed the top at the 19 th case,and as a boundary,the learning curve was divided into two phases: phase A(the learning phase)and phase B(the mastey phase),and all cases were divided into two groups.Compared with two groups,no differences were statistically significant between the two groups with respect to gender,age,body mass index(BMI),American society of Anesthesiologists(ASA)score,the length of medical history,the length and location of the tumor(P>0.05).In terms of OT,themean OT was(124.6±66.8)min,and the OT in group A(mean 171.0±72.6 min)was definitely longer than that in group B(mean 96.2±44.0 min,the difference was statistically significant(P(27)0.01).In the aspect of the intraoperative blood loss(IBL),the mean of IBL was(26.2±22.0)ml,there was significant difference between two groups(P(27)0.01),and the IBL in group A(mean 40.0±27.3 ml)was obviously more than that in group B(mean 17.7±12.2 ml).About the postoperative chest tube drainage(PCD),the mean time of PCD was(4.5±2.4)d,the time of group A(mean 6.1±1.2 d)was significantly longer than that of group B(mean 3.5±2.4 d)(P(27)0.01).In terms of the rate of postoperative complications(RPC),the RPC in all operation was 6%,the RPC in group A was 15.8%,and that in group B was 0%,the difference was statistically significant(P(27)0.05).The mean of postoperative hospital stay(PHS)was(6.5±2.9)d,the comparison showed no significant differences in two groups(P>0.05).Conclusions The two phases identified with CUSUM analysis of OT represented characteristic stages of the learning curve of DRMTR,and the learning curve consisted of two distinct phases: phase A and phase B,which represented the learning phase and mastery phase,respectively.Through the data,for an experienced surgeon in thoracoscopic mediastinal tumor resection,the surgeon needs 19 cases experience to cross the learning curve approximately.This study of the learning curve of DRMTR is only for reference about the development of da Vinci robot system in surgery in our country.