Impact of First Assistant Surgeon Experience on the Perioperative Outcomes of Laparoscopic Hepatectomies

Introduction

The outcomes of major surgeries are not only influenced by the primary surgeon but a competent surgical team is also of utmost importance. Effective coordination among members of the collective operative team, including surgeons, anesthesiologists, nurses, residents, and auxiliary staff, can contribute to a smooth operative experience and better outcomes. Team familiarity has been associated with better surgical performance, improved safety, reduced operative time, and fewer complications in open abdominal surgeries, inguinal hernia repairs, laparoscopic bariatric surgeries, total knee replacements, reduction mammoplasties, and breast reconstruction procedures.^1–9 Likewise, how experienced the surgical assistant(s) and the collective operative team are has also been found to influence operative efficiency and time, which may have a bearing on outcomes in higher risk patient profiles and procedures.^4,6,7,9–12 In particular, concerns have been raised about performing complex laparoscopic surgeries with assistance from less experienced team members, such as residents, and whether this would negatively impact outcomes. The aforementioned association has been specifically analyzed and described in laparoscopic pancreatoduodenectomies and laparoscopic sigmoidectomies.^9,11 However, to our knowledge, the impact of the first assistant surgeon experience on the perioperative outcomes of laparoscopic hepatectomies has yet to be evaluated.

At our institution, laparoscopic hepatectomies were primarily performed by attendings and aided by two assistants; a first assistant surgeon and a laparoscope holder. The first assistant surgeons may be of various experience levels and can be broadly classified as general surgery residents, fellows who had completed general surgical training and were undergoing fellowship in hepato-pancreato-biliary (HPB) surgery, and attendings who had completed HPB fellowship training. The primary duties of the first assistant surgeon include providing retraction, suctioning, and at times, guiding the laparoscope. This study aims to evaluate the impact of first assistant surgeon experience on the perioperative outcomes of laparoscopic hepatectomies.

Materials and Methods

All consecutive patients who underwent laparoscopic hepatectomies at Singapore General Hospital between 2012 and 2018 were included. These patients were identified from a prospectively maintained surgical database, and data were obtained from the patients' clinical, radiological, and pathological records. Clinical data were collected from a prospective, computerized clinical database (Sunrise Clinical Manager, version 5.8; Eclipsys Corporation, Atlanta, GA) and operative data were obtained from another prospective computerized database (OTM^® 10; IBM, Armonk, NY). This study has been approved by the institutional review board. Before the study period, 100 consecutive cases of minimally invasive hepatectomies were performed between 2006 and 2011. ¹³ These first 100 patients were excluded to minimize the confounding effect of the institutional learning curve in this study.

To study the impact of first assistant surgeon experience, patients were stratified into three cohorts—Group 1 in which assistants were general surgery residents (typically postgraduate year 4 or 5 rotating for 3 months), Group 2 in which assistants were fellows (typically postgraduate year 6–8 rotating for 1–2 years), and Group 3 in which assistants were attendings (surgeons who had completed fellowship training in HPB surgery). Only totally laparoscopic hepatectomies were included in this study and these were further categorized based on technical difficulty according to the Iwate Criteria described by Wakabayashi at the Morioka Consensus Conference in 2014. ¹⁴ Traditional major hepatectomies were classified as resection of three or more contiguous segments. ¹⁵ Laparoscopic major hepatectomies were defined to also include right anterior and posterior sectionectomies.^15–19 Our operative techniques and approaches have been described previously by Goh et al. ²⁰ Postoperative complications were classified according to the Clavien–Dindo grading system and were recorded up to 30 days or during the same hospital stay, including any 30-day readmissions. ²¹ Both 30-day and in-hospital mortalities were recorded.

The augmented inverse probability of treatment weighting (AIPW) approach was used to minimize selection and confounding biases, and adjusted P values associated with average treatment effects are reported. ²² The AIPW methodology combines aspects of covariate adjustment and inverse probability weighting and therefore possesses a doubly robust property against any potentially misspecified propensity score models. Multinomial logistic regression was used to estimate conditional probabilities of assignment to the three groups stratified by the first assistant surgeon's experience, taking into account the following variables: malignant neoplasm, tumor size, multifocal disease, cirrhosis, multiple resections, posterosuperior location, major resection by either traditional or laparoscopic criteria, and Iwate difficulty score. Outcome models were linear and logit for continuous and binary outcomes, respectively. Analyses of baseline characteristics were nominal (i.e., not conditioned on propensity scores) and were performed using the Pearson χ2 test or Mood's equality of medians test for binary and continuous variables, respectively. Statistical analyses were performed using Stata^® (version 13.0, StataCorp, College Station, Tx), and P < .05 was considered to indicate nominal statistical significance.

Results

A total of 385 consecutive patients underwent laparoscopic hepatectomies at the Singapore General Hospital between 2012 and 2018. These comprise 216 patients in Group 1, 56 patients in Group 2, and 113 patients in Group 3. Statistically significant differences were found when comparing proportions of patients who had previous abdominal surgery, previous liver surgery, concomitant surgery not including cholecystectomy, tumor(s) in the posterosuperior location, major resection by laparoscopic criteria, and Iwate difficulty score and level (Table 1).

Group 3 had 21.2% of patients with previous abdominal surgery and 1.8% of patients with previous liver surgery. This was significantly less when compared with Group 1, which had 38.4% (P = .0016) and 8.8% (P = .0133) of patients with previous abdominal and liver surgeries, respectively, and Group 2, which had 41.1% (P = .0068) and 8.9% (P = .0279), respectively. Group 3 also had a smaller proportion of patients who underwent other concomitant surgeries (not including cholecystectomy) at 5.3% compared with 13.4% in Group 1 (P = .0234) and 17.9% in Group 2 (P = .0087). A larger proportion of patients in Group 3 had major resections by laparoscopic criteria (P = .0003) and tumor(s) in a posterosuperior location (P = .0006) when compared with Group 1; 54.9% of patients had tumor(s) in a posterosuperior location and 35.4% had major resections by laparoscopic criteria in Group 3, compared with 35.2% and 17.6%, respectively, in Group 1. These differences were taken into account when developing the propensity score model for the AIPW analysis.

Statistically significant differences were detected when comparing the Iwate surgery difficulty scores between Group 3 and Group 1 (P < .0001) and Group 3 and Group 2 (P = .0423). The mean Iwate scores for Group 1, Group 2, and Group 3 were 5.5, 5.9, and 6.8, respectively, all with a standard deviation of 2.70 (Table 1). When the Iwate scores were further categorized into levels, however, only the comparison between Group 3 and Group 1 demonstrated a statistically significant difference (P = .0072). Group 3 comprised a greater proportion of advanced- and expert-level surgeries than Groups 1 and 2. The percentages of patients who underwent advanced- and expert-level surgeries in Groups 1, 2, and 3 were 33.8%, 32.2%, and 46.0%, respectively.

In terms of intraoperative events, Group 3 was found to have consistently higher operative times as well as more frequent use and longer duration of Pringle's maneuver, when compared with Groups 1 and 2 (Table 2). The median operative times for Groups 1, 2, and 3 were 195, 195, and 290 minutes, respectively. Pringle's maneuver was applied in 26.9%, 33.9%, and 60.2% of patients with a corresponding median duration of 35, 36, and 45 minutes, respectively. These differences were statistically significant when comparing Group 3 with Groups 1 and 2 (P < .05).

Groups 1, 2, and 3 each had an open conversion rate of 6.5%, 10.7%, and 8.9%; median estimated blood loss of 200, 200, and 300 mL; and median hospitalization of 4, 3, and 4 days, respectively (Table 2). Resection margins were close (≤1 mm) in 4.2%, 7.1%, and 3.5% of patients. Major morbidity of greater than Clavien–Dindo grade 2 occurred in 5.6%, 3.6%, and 4.4%, while reoperation was required in 1.9%, 1.8%, and 1.8% of patients. Group 1 had no 30-day mortality and 1 in-hospital mortality; Group 2 had no 30-day mortality and in-hospital mortality; Group 3 had 1 30-day mortality and 2 in-hospital mortalities. None of these perioperative and postoperative outcomes demonstrated statistically significant differences across groups. The reoperation rates, 30-day mortalities, and in-hospital mortalities could not be compared due to low event rates.

Discussion

Laparoscopic hepatectomies are technically demanding procedures with reported learning curves of about 25 cases for minor hepatectomies to 45–75 cases for major hepatectomies.^23–26 Hence, opponents of the approach have frequently cited the steep learning curve required to perform the procedure as a major obstacle to its widespread adoption. Furthermore, questions have also been raised on its applicability to teaching institutions due to the need for well-trained surgical assistants.

The success of a surgical procedure is critically dependent not only on the primary surgeon but also on the seamless coordination and support rendered by the rest of the surgical team. The first assistant surgeon works most closely with the primary surgeon and plays an important role by providing good traction, ensuring optimal view of the dissection plane, and at times, guiding the laparoscope. These tasks contribute to the smooth execution of surgery and are instinctively associated with surgical experience. With all other factors being equal, one would naturally expect the surgery to be faster, smoother, and less eventful when the first assistant surgeon is more experienced. This may also translate to better perioperative and postoperative outcomes. Interestingly, in this study, perioperative and postoperative outcomes were not found to be significantly less favorable when the first assistant surgeon was of a lower experience level.

Finnesgard et al. conducted a similar study, which found that having a more experienced, certified surgical assistant (CSA) led to significantly shorter operative times for laparoscopic pancreatoduodenectomies. ⁹ Kim et al. also found that having a novice assistant resulted in longer operative times for laparoscopic sigmoidectomies, but this difference was negated after an initial learning curve of 10 cases. ¹¹ The first assistant surgeon in this study, however, was at least a fellow and had different responsibilities of modulating the actions of the second assistant who was in charge of operating the laparoscope, establishing pneumoperitoneum, placing initial trocars, minor hemostasis, drain placement, and wound closure. Nevertheless, we were unable to reproduce similar results in our study for laparoscopic hepatectomies despite having a large sample size.

Furthermore, Group 3 had significantly longer operative times than Group 1 (P = .0130) and Group 2 (P = .0053), and this is attributed to greater surgical difficulty, as evidenced by a larger proportion of major resections by laparoscopic criteria (35.4%) and Iwate advanced- and expert-level cases (46.0%). This likely prompted the involvement of a fellow attending as the first assistant surgeon. An earlier analysis of the same cohort published by our group in 2017 showed that as individual surgeons gained more experience, major resections and resections of tumors in the posterosuperior location were increasingly being performed laparoscopically. This corresponded to an increase in operative times and increased use of the Pringle maneuver, which were once again demonstrated in this study. ²⁷ Appropriate selection of first assistant surgeons is important to ensure that their levels of experience are commensurate with the difficulty and complexity of cases.

Besides operative times, Finnesgard et al. also compared intraoperative blood loss, which appeared to be less with a more experienced CSA, but this did not reach statistical significance. Kim et al. also studied lymph node yield, resection margins, and postoperative complications, but found no significant difference in these outcomes. ¹¹ Similarly, in the present study, there was no significant difference in blood loss, resection margins, postoperative complications, open conversion, duration of hospitalization, and postoperative mortality.

There are a few plausible reasons that could explain the results of this study. First, the lack of first assistant surgeon experience could possibly be compensated by the rest of the surgical team. Knowing that the first assistant surgeon is less experienced, the primary surgeon may consciously endeavor to verbalize all instructions more clearly so as to enable better support by the first assistant surgeon. The primary surgeon may also involve more assistants in the surgery for a more even division of tasks. Second, the learning curve for a first assistant surgeon in laparoscopic hepatectomies may be less steep than we expect it to be. Kim et al. reported for laparoscopic sigmoidectomies an assistant learning curve of 10 cases, which can likely be surmounted within 2 weeks in a high-volume center. The assistant learning curve for laparoscopic hepatectomies is not established, but it may not be as steep as we expect it to be. Third, appropriate selection of the surgical assistant based on his or her level of experience could have ensured that the proficiency of assistance rendered was commensurate with the case difficulty and the primary surgeon's experience.

The results should be interpreted in the context of this study's limitations. First, this was a retrospective study confined to a single center's experience, which limits both the study's ability to establish causative relationships and also generalizability of its results. The surgeries were also performed by several different surgeons over 7 years, and we were unable to account for each surgeon's skill and experience at the time of surgery, which could have been possible confounders. On the other hand, this study has a relatively large sample size (n = 385), which far exceeds that of other similar studies, and the robust statistical approach lends considerable strength to the internal validity of its results.

Nevertheless, this study shows that laparoscopic hepatectomies can be performed in teaching hospitals, with the participation of residents, safely and without impacting patient outcomes. These findings are important and support the participation of young residents in complex laparoscopic procedures that constitute an essential part of surgical training and exposure.

Conclusion

Lack of experience in a first assistant surgeon does not adversely impact operative times, intraoperative blood loss, open conversion, resection margins, postoperative complications, duration of hospitalization, and perioperative mortality for laparoscopic hepatectomies. The participation of residents as first assistants in laparoscopic hepatectomies should be encouraged for training and exposure. These cases should be appropriately selected to ensure that their difficulty and complexity are commensurate with the experience of the first assistant surgeons.