Objective
We sought to examine outcomes in an expanding robotic surgery (RS) program.
Study Design
In all, 1000 women underwent RS from May 2006 through December 2009. We analyzed patient characteristics and outcomes. A total of 377 women undergoing RS for endometrial cancer staging (ECS) were compared with the historical data of 131 undergoing open ECS.
Results
For the entire RS cohort of 1000, the conversion rate was 2.9%. Body mass index increased over 3 time intervals: T1 = 26.2, T2 = 29.5, T3 = 30.1 (T1:T2, P = .01; T1:T3, P = .0001; T2:T3, P = .037). Increasing body mass index was not associated with increased major complications: T1 = 8.7%, T2 = 4.3%, T3 = 5.7%. In the ECS cohort, as compared with open ECS, women undergoing RS had lower blood loss (46.9 vs 197.6 mL, P < .0001), shorter hospitalization (1.4 vs 5.3 days, P < .0001), fewer major complications (6.4% vs 20.6%, P < .0001), with higher lymph node counts (15.5 vs 13.1, P = .007).
Conclusion
RS is associated with favorable morbidity and conversion rates in an unselected cohort. Compared to laparotomy, robotic ECS results in improved outcomes.
Minimally invasive surgery (MIS) is now accepted as an integral part of gynecologic surgery. While the reproducibility and feasibility of MIS in gynecology has motivated surgeons to expand the use of laparoscopy in their practices, the implementation of laparoscopy for more advanced procedures has been precluded by the technical and mechanical limitations of the equipment. Accordingly, these limitations have driven the technological evolution of MIS tools. Robotic-assisted MIS effectively surmounts the technical obstacles posed by traditional laparoscopy by incorporating 3-dimensional visualization, articulated instrumentation, motion scaling, enhanced stability of the operative image and instrumentation, and optimization of ergonomics and autonomy.
We have previously reported clinical outcome data arising from the first year of incorporation of robotic surgery (RS) into our gynecologic oncology practice. Analysis of our initial experience suggests that compared with comparable open cases performed the year prior to, and during the first year of our robotic experience, RS offered advantages that include decreased estimated blood loss (EBL), faster recovery, and a trend toward reduced major complications (MC). Equally important is the fact that we were able to dramatically expand our MIS program utilizing this technology.
Given these encouraging findings, but recognizing that careful patient selection had been critical to our success, we looked forward while posing several questions. Would our complication rate stay low as we expanded a robotics approach to women with more complex clinical problems, as well as those unselected for body mass index (BMI) or previous number of abdominal surgeries? Additionally, would our conversion rate remain low when a higher proportion of obese women were included in the robotics cohort? What would the impact of incorporating resident and fellow robotics training have on outcomes? Finally, would our efficiency improve?
We now report the outcomes of the first 1000 women who consecutively underwent RS upon referral to our gynecologic oncology practice.
Materials and Methods
One thousand women underwent RS from May 2006 through December 2009. Initially, the first and second authors completed the credentialing requirements in May 2006, and within 4 months 2 additional gynecologic oncologists completed training. In May 2009 a fifth attending gynecologic oncologist joined our practice and began performing robotic cases. Resident involvement began with the 17th case. In November 2006, we began training our gynecologic oncology fellows on the console (case 57). All cases were performed with the da Vinci surgical system (Intuitive Surgical, Sunnyvale, CA). Our RS technique has been previously described.
In the first year, data collection and analysis were performed prospectively for a quality assurance project per request of Swedish Cancer Institute Cancer Committee. Institutional review board approval was obtained upon completion of the quality assurance project. A database in compliance with the Health Insurance Portability and Accountability Act of 1996 was created for all patients who underwent RS during the study period. We prospectively collected data including age, BMI, number of previous abdominal surgeries, EBL, operative times (OT), pathologic findings such as lymph node count (LN), major and minor surgical complications, and length of stay (LOS). Calculated OT for RS was from placement of the uterine manipulator to skin closure, or incision to skin closure for those cases in which a uterine manipulator was not utilized. Postoperative complications were included if occurring up to 6 weeks postoperatively. Vaginal cuff separation and/or dehiscence were included regardless of the number of weeks from the surgical procedure. LOS included the time from arrival in the operating room to the time of discharge. Conversion to laparotomy was defined as any laparotomy required to complete the planned robotic procedure including mini-laparotomy for organ removal even if the entire procedure was performed robotically up to the point of specimen delivery. For patients with biopsy-proven endometrial carcinoma, the performance of and extent of lymphadenectomy was at the discretion of the attending surgeon.
Data analysis was performed using software (Stata, version 11.0; StataCorp, College Station, TX). Descriptive statistics were initially performed, followed by data analysis to determine whether there was a parametric vs nonparametric distribution. In most cases, a nonparametric distribution was observed; therefore, the Kruskal-Wallis 2-sample rank sum test was utilized to determine whether there was a statistically significant difference between groups. When the distribution appeared to be parametric, a χ 2 test or Student t test was used (a 2-tailed test was utilized, with alpha level <0.05 considered significant).
Results
First consecutive 1000 robotic cases
The most common indications for surgery in our robotics cohort were endometrial carcinoma (n = 377) and adnexal mass (n = 194). The remaining indications were genetic predisposition to ovarian and tubal cancer (n = 145), cervical carcinoma in situ or adenocarcinoma in situ or benign uterine indication (n = 85), cervical or endometrial cancer amenable to radical hysterectomy and staging (n = 65), ovarian or fallopian tube cancer (n = 59), locally advanced cervical cancer (n = 17), complex atypical endometrial hyperplasia (n = 27), vaginal vault prolapse (n = 19), vaginal intraepithelial neoplasia (n = 3), and other (n = 8).
Surgical procedures included simple hysterectomy with or without adnexectomy (n = 461), hysterectomy with bilateral salpingo-oophorectomy and pelvic with or without periaortic lymphadenectomy (n = 277), adnexectomy (n = 95), radical hysterectomy with lymphadenectomy (n = 65), staging pelvic and periaortic lymphadenectomy for either incompletely staged endometrial/cervical cancer or surgical staging of cervical cancer (n = 50), resection and staging of ovarian or fallopian tube carcinoma (n = 23), sacrocolpopexy (n = 12), sacrocolpopexy with hysterectomy (n = 7), myomectomy (n = 5), hysterectomy/bilateral salpingo-oophorectomy with upper vaginectomy (n = 2), upper vaginectomy (n = 1), oophoropexy prior to radiation (n = 1), and trachelectomy with lymphadenectomy (n = 1).
The number of MIS cases performed in our practice has continued to grow each year since the inception of our robotics program. This trend is illustrated in Figure 1 . As shown, in the year preceding the introduction of robotics in our practice, only 9% of all cases performed at our primary institution were approached laparoscopically. By the third year of our robotics experience, the proportion of MIS cases had grown to 36%.
Within the entire RS cohort the total complication rate was 9.9%. The MC rate was 5.7%, while minor complications occurred in 4.2% of patients. Details of the MC are listed in Table 1 . The most common MC were infectious in 1.2%, ileus in 0.7%, and cuff dehiscence in 0.6%. Infectious complications included vaginal cuff abscess requiring percutaneous drainage and antibiotics (n = 6), vaginal cuff cellulitis requiring intravenous antibiotics (n = 2), vaginal cuff cellulitis requiring only oral antibiotics (n = 2), postoperative acute cholecystitis (n = 1), and fatal sepsis arising from bacterial endocarditis in a patient with a mechanical valve (n = 1).
| Complication | n | Percent |
|---|---|---|
| Major vessel injury | 3 | 0.3 |
| Intraoperative hemorrhage | 1 | 0.1 |
| Postoperative anemia requiring transfusion | 5 | 0.5 |
| Deep venous thrombosis or pulmonary embolus | 5 | 0.5 |
| Ureteral stricture/injury | 2 | 0.2 |
| Ileus | 7 | 0.7 |
| Infection | 12 | 1.2 |
| Small bowel obstruction | 4 | 0.4 |
| Cuff dehiscence or separation | 6 | 0.6 |
| Enterotomy | 1 | 0.1 |
| Postoperative vault prolapse | 1 | 0.1 |
| Chylous ascites | 1 | 0.1 |
| Cystotomy | 2 | 0.2 |
| Transient postoperative arrhythmia//labile hypertension/fluid overload | 7 | 0.7 |
| Total | 57 | 5.7 |
We compared trends in BMI and preexisting medical comorbidities in the early, middle, and final thirds of the study time interval (T1-T3). BMI increased significantly with increasing robotic experience: T1 = 26.2, T2 = 29.5, T3 = 30.1 (T1 vs T2, P = .01; T1 vs T3, P = .0001; T2 vs T3, P = .037). Despite a significant and steady rise in BMI, a concomitant rise in MC did not occur ( Table 2 ). The MC rate remained low throughout each subsequent time interval: T1 = 8.7%, T2 = 4.3%, T3 = 5.7%. The conversion to laparotomy rate in the entire robotics cohort was low at 2.9%. Comparing the conversion rates by time interval, despite a steady rise in BMI, an escalation in conversions was not realized: T1 = 4.0%, T2 = 2.5%, T3 = 2.8% ( Table 3 ). Conversion rates were examined in each of 3 BMI categories. Conversion rates remained low in all BMI categories: 2.7% in normal-weight and overweight women, 2.1% in obese women with BMI between 30-40, and 6.1% in morbidly obese women with BMI >40. The most prevalent reasons for converting to laparotomy were adhesions precluding acceptable exposure (1.2%), unanticipated carcinomatosis requiring laparotomy for optimal debulking (0.7%), and inability to deliver an intact specimen vaginally (0.4%). Rare indications for conversion included extensive subcutaneous emphysema that occurred in 2 cases, and major vessel injury, poorly controlled bleeding, ureteral injury, and robotic malfunction each occurring in 1 instance in the study interval.
| Year | Major complications, n | Major comorbidities, n | Mean BMI (range) | P value for BMI comparisons |
|---|---|---|---|---|
| 2006 through 2007 T1 | 13/149 (8.7%) | 35/149 (23.5%) | 26.5 (17.4–49.4) | |
| 2007 through 2008 T2 | 14/323 (4.3%) | 102/323 (31.6%) | 29.5 (14.6–69.2) | .01 a |
| 2008 through 2009 T3 | 30/528 (5.7%) | 239/529 (45.2%) | 30.1 (15.9–70.1) | .037 b .0001 c |
| Time interval | Conversions, n | Conversions | Mean BMI (range) | P value for BMI comparisons |
|---|---|---|---|---|
| All | 29/1000 | 2.9% | ||
| 2006 through 2007 T1 | 6/149 | 4.0% | 26.5 (17.4–49.4) | |
| 2007 through 2008 T2 | 8/323 | 2.5% | 29.5 (14.6–69.2) | .01 a |
| 2008 through 2009 T3 | 15/528 | 2.8% | 30.1 (15.9–70.1) | .037 b .0001 c |
There were 832 of the 1000 robotic cases in which colpotomy and vaginal cuff closure was a component of the case. Vaginal cuff dehiscence or separation occurred in 6 of these 832 women for an overall cuff dehiscence rate of 0.72%. There was a steady decline in vaginal cuff dehiscence with increasing experience: T1 = 2.6%, T2 = 0.72%, T3 = 0.22% despite a significant rise in BMI with each subsequent time period ( Table 4 ).
The impact of fellow and resident involvement was evaluated. Trainees participated in 460 of the first 1000 robotic procedures in our primary institution. As shown in Table 5 , there were no significant differences in total complications, MC, or minor complications, or conversions to laparotomy when residents and fellows participated compared to cases performed exclusively by attending gynecologic oncologists. Recognizing that gynecologic oncology fellows were more likely to perform a more substantial proportion of, if not the entire procedure as compared to resident physicians, we compared outcomes in cases with resident vs fellow participation. No differences were observed in total complications, MC, or minor complications, or conversion to laparotomy ( Table 6 ).
| Variable | Fellow/resident, n | Attending only, n | P value |
|---|---|---|---|
| Total cases | 460 | 540 | |
| Total complications | 36 (7.8%) | 63 (11.7%) | .2715 |
| Major complications | 20 (4.3%) | 37 (6.9%) | .0889 |
| Minor complications | 16 (3.5%) | 26 (4.8%) | .2941 |
| Conversions to laparotomy | 18 (3.9%) | 11 (2.0%) | .0782 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
