In this study, we investigated the surgical outcomes of RAPN according to surgeons’ experience. Even in a high-volume center, one or two specific surgeons (first-generation) commence new surgeries and improve their procedures, and then, subsequent surgeons (second-generation) turn over the surgeries. Second-generation surgeons usually participate in first-generation surgeons’ surgery as an assistant, and they start their surgery according to refined procedures and techniques developed by first-generation surgeons. Therefore, we hypothesized that second-generation surgeons could achieve better surgical outcomes in their early experience period when compared to first-generation surgeons. As shown in our results, we demonstrated that second-generation surgeons tended to be associated with better surgical outcomes, including shorter operation time and better trifecta achievement rate than first-generation surgeons, in the early experience period. This result suggests the importance of an institutional training system to shorten the period of expertise for RAPN.
The importance of institutional experience in RAPN was shown by Zeuschner et al. [3]. They compared the impact of the learning curve of the department, the console surgeon, the bedside assistant, and patient-related factors on the perioperative outcomes of RAPN. The console surgeon’s experience significantly impacted operation time, EBL, complication rate, and length of hospital stay. In contrast, the experience of the department and bedside assistant were significantly associated with more favorable outcomes in terms of the operation time and open conversion rate [3]. Dagenais et al. evaluated the variability in PN outcomes by physician-level discrepancies. They demonstrated that a high proportion of surgeon factors were associated with the length of hospital stay (90%), positive margins (100%), complications (100%), and 30-readmission (90%) in terms of between-surgeon variability. In contrast, a small to moderate proportion of surgeon factor in operative time (20%), estimated blood loss (40%), ischemia time (10%), and excisional volume loss (18%) were observed. As to operative time, unexplained surgeon factors (27%) and unexplained patients factors (54%) were associated with between-surgeon variability, which may include institutional experience or bedside assistance [8]. According to these previous articles, our results of better surgical outcomes in the second-generation surgeons than in first-generation surgeons may have resulted from the accumulation of skills among all participants, including operation staff.
The evaluation of the learning curve for robotic surgery was assessed by several approaches. Meier et al. evaluated the number of repetitions required to reach the expert level using the da Vinci Surgical Skills Simulator™. They showed that robotic surgeons, table-side assistants, and novice surgeons aged 25 years or younger achieved better results than laparoscopic and open surgeons who had no robotic surgery experience and the older novice group [9, 10]. From this study, the importance of the experience of robotic surgery for robotic skills was demonstrated, whether or not a primary surgeon was shown [9, 10].
On the other hand, laparoscopic experience improves the learning curve in real-world robotic RAPN. Pieroprazio et al. examined the transition to RAPN from pure laparoscopic partial nephrectomy (LPN) and investigated the learning curve; they demonstrated that after a learning experience of approximately 25 cases, the transition from LPN to RAPN can be performed without an additional learning curve and can be associated with immediate benefits [2].
Besides urological robotic surgeries, Pernar et al. reviewed the literature on the learning curve in robotic general surgery [11]. Although there are several outcomes of robotic surgeries, time was used to measure the learning curve in all studies. The learning curve of general surgery has focused on the time under robot support; the number of operations required until acquiring surgical proficiency is increasing. The number of cases needed to achieve plateau performance was wide-ranging but overlapping for different kinds of operations: 19–128 cases for colorectal, 8–95 for foregut/bariatric, 20–48 for biliary, and 10–80 for solid organ surgery [11]. Regarding RAPN, WIT and console time tended to be a measure of the learning curve. Mottrie et al. evaluated the impact of the learning curve on perioperative outcomes in patients who underwent RAPN and demonstrated that WIT (< 20 min) and console times were optimized after the first 30 (p < 0.001) and 20 cases (p < 0.001), respectively[12]. In addition, Larcher et al. performed a similar study and demonstrated that WIT showed a steep slope reduction within the first 100 cases, and a plateau was then observed after 150 cases [13].
Regarding our study, although the steep slope reduction of operation time was observed in Fig. 1A, a moderate slope reduction was shown in Figs. 1B, 2A, B, which might be caused by the increasing number of challenging cases with increasing experience.
The learning curve of surgical outcomes other than time has been described in several studies. Mottrie et al. reported that the complication rates remained unchanged over the entire series, concluding that the learning curve for RAPN is short [12]. On the other hand, Larcher et al. described a linear relationship between experience and complication-free course, which did not reach a plateau, even after 300 cases, concluding that the learning curve appears endless with respect to complications [13]. The two studies contained different cohort sizes and different total complication rates; therefore, apparent controversial results might be influenced by several background characteristics.
The present study had several limitations that should be noted. First, the retrospective nature with data collected from a single institution and a population of tertiary care patients are limitations. Second, four surgeons who performed RAPN in this study had adequate laparoscopic experience, which led to relatively good surgical outcomes in their early period of robotic surgery. Given that surgeons without sufficient experience in laparoscopic kidney surgery were included, the results might have been different. Third, comparisons of surgical outcomes between first-generation and second-generation surgeons were performed with univariate analysis, even though there were similarities in patients and tumor background between the two groups. The strength of our study is that it is a relatively rare study investigating the surgical outcomes between surgeons’ generation. In the early period of experience, shorter operation time and higher trifecta achievement were observed with second-generation surgeons, which could stress the importance of an institutional training system.