Robotic gynecological surgery is feasible in obese patients, but there remain concerns about the safety of this approach because the positioning required for pelvic surgery can exacerbate obesity-related changes in respiratory physiology. The objective of our study was to evaluate pulmonary and all-cause complication rates in obese women undergoing robotic gynecological surgery and to assess variables that may be associated with complications.
A retrospective chart review was performed on obese patients (body mass index of ≥30 kg/m 2 ) who underwent robotic gynecological surgery at 2 academic institutions between 2006 and 2012. The primary outcome was pulmonary complications and the secondary outcome was all-cause complications. Univariate and multivariate logistic regression analyses were used to determine the associations between patient baseline variables, operative variables, ventilator parameters, and complications.
Of 1032 patients, 146 patients (14%) had any complication, whereas only 33 patients (3%) had a pulmonary complication. Median body mass index was 37 kg/m 2 . Only age was significantly associated with a higher risk of pulmonary complications ( P = .01). Older age, higher estimated blood loss, and longer case length were associated with a higher rate of all-cause complications ( P = .0001, P < .0001, and P = .004, respectively). No other covariates were strongly associated with complications.
The vast majority of obese patients can successfully tolerate robotic gynecological surgery and have overall low complications rates and even lower rates of pulmonary complications. The degree of obesity was not predictive of successful robotic surgery and subsequent complications.
Obesity is a common and growing problem in the United States, currently affecting about one-third (33.8%) of all American adults and 35.5% of women. Obesity causes changes in respiratory physiology, including reduction in functional residual capacity, decreased chest wall compliance, and altered oropharyngeal anatomy. These changes can pose challenges to anesthesiologists and surgeons who care for obese patients in the perioperative period. Obesity has been linked to poorer surgical outcomes, including longer operative times and increased postoperative complications, particularly wound infections.
Historically, hysterectomies and other major gynecological procedures were performed via laparotomy. As technology advanced, many gynecologists have transitioned to minimally invasive surgical techniques such as traditional laparoscopy and robotic-assisted surgery. These approaches have been shown to have a lower rate of surgical complications and length of hospital stay compared with laparotomy.
Despite the advantages of minimally invasive surgery, concerns have been raised about its limitations in obese patients. Surgery in obese patients is more challenging, both for the surgeon and for the anesthesiologist. The pneumoperitoneum and steep Trendelenburg positioning required for minimally invasive pelvic surgery can exacerbate the obesity-related changes in respiratory physiology. This can complicate anesthetic management by creating increased intraabdominal pressure, which can increase peak inspiratory pressure and decrease tidal volume, both of which are associated with perioperative pulmonary complications. When patient respiratory function and ventilator parameters are not carefully monitored and controlled during robotic pelvic surgery, conversion to laparotomy may be required, resulting in additional exposure to the complications associated with laparotomy.
Despite the possible challenges of robotic surgery in obese patients, current data support the feasibility of robotic surgery in this population. Particularly in gynecological oncology patients, robotic surgery has been shown to be feasible and also offers additional benefits of decreased blood loss, fewer transfusion requirements, lower conversion rate to laparotomy, and shorter operative time over traditional laparoscopy.
As obesity becomes increasingly common, the importance of identifying optimal surgical techniques in this population grows. To date, no studies have been published examining ventilatory indices and the rate of pulmonary complications in obese women undergoing robotic gynecological surgery.
The objective of our study was to evaluate the success and complication rates in obese women undergoing robotic gynecological surgery and to determine whether any patient, operative, or ventilator variables are associated with complications, particularly pulmonary complications.
Materials and Methods
A retrospective chart review was performed at the University of Alabama at Birmingham (UAB) and the University of North Carolina (UNC), following institutional review board approval at each institution. At UAB, data collection was performed on all obese patients who underwent robotic gynecological surgery between October 2006 and October 2011. At UNC, data collection was performed on the same patient population undergoing surgery from September 2007 to December 2012.
All robotic surgeries performed by the Gynecologic Oncology faculty at the 2 institutions during the study periods were reviewed. These included both oncological and benign robotic-assisted cases such as total hysterectomies, bilateral salpingo-oophorectomies, pelvic and/or paraaortic lymph node dissection, radical hysterectomies, trachelectomies, and myomectomies.
Outpatient clinic notes, operative notes, intraoperative anesthesia records, and discharge summaries were reviewed to determine patient demographic information, intraoperative respiratory, and ventilatory parameters as well as complications. Preexisting medical comorbidities were recorded, with a focus on pulmonary conditions such as chronic obstructive pulmonary disease (COPD), asthma, sleep apnea, and tobacco use.
Intraoperative details recorded included case length and time in Trendelenburg position. Ventilator and intraoperative respiratory parameters included tidal volume (TV), peak inspiratory pressure (PIP), oxygen saturation (SpO2), and respiratory rate (RR). These ventilator parameters were reported as an average by calculating the mean of the lowest and highest recorded values found in the anesthesia records for each surgery. Hospital and outpatient records were reviewed and all postoperative complications were recorded through 1 month postoperatively.
Both univariate and multivariate logistic regressions were used to evaluate the relative strength of association of selected covariates of interest on complication outcome. The primary outcome was the probability of having a pulmonary complication versus not having a pulmonary complication, and the secondary outcome was the probability of having any complication vs not having a complication.
The covariates of interest included selected patient baseline covariates, operative covariates, and ventilator parameters. Model fit was assessed using an approximation to Bayes factors known as the Schwartz-Bayesian criterion. In univariate modeling, the model fit was assessed by the difference between the null model (the model without the covariate) and the model with the covariate. In multivariate modeling, the relative strength of various multiple covariate models was compared using the Schwartz-Bayesian criterion.
Statistical analyses were performed using both SAS and R statistical software (SAS version 9.3; SAS Institute, Inc, Cary, NC; and R from the R Foundation for Statistical Computing, Vienna, Austria).
One thousand thirty-two patients met inclusion criteria from the 2 institutions, 415 from UAB and 617 from UNC. Table 1 shows the basic patient demographics of our study population. The highest body mass index (BMI) was 71 kg/m 2 and the median BMI was 37 kg/m 2 . Of the study cohort, 353 (34%) had a BMI of 30-35 kg/m 2 , meeting criteria for class I obesity by the World Health Organization criteria. Two hundred ninety-two patients (28%) had a BMI of 35–39 kg/m 2 , meeting criteria for class II obesity, and the remaining 387 patients (37%) met criteria for class III obesity, with a BMI greater than 40 kg/m 2 .
|Patient characteristics||Patients, n (%)|
|BMI, kg/m 2|
|Sleep apnea||41 (4)|
|Tobacco use||178 (17)|
|North Carolina||617 (60)|
|Uterine weight, g|
|Unknown/not applicable||481 (47)|
Two hundred seventy-eight patients had preexisting pulmonary comorbidities prior to surgery: 83 with asthma (8%), 41 with sleep apnea (4%), 17 with COPD (1.6%), and 178 with active tobacco use (17%).
Table 2 shows the operative variables for the patients in our study. The median case length was 150 minutes (range, 67–307 minutes). Only 355 patients (34%) had Trendelenburg time documented. Of these patients, the median Trendelenburg time was 127 minutes. The TV ranged from 207 to 905 mL, with a median TV of 489 mL. The median PIP was 27 cm H 2 O (range, 14–52 cmH 2 O). The median SpO2 was 97.5% (range, 88–100%), and the median RR was 12.5 breaths per minute (range, 6–29).
|Case time, min||25||120||150||187||435|
|Trendelenburg time, min||6||97||127||163||330|
|PIP, cm H 2 O||14||23||27||31||79|
|RR, per min||6||10||13||15||29|
One hundred forty-six patients (14%) had a documented perioperative complication. Table 3 shows the type of complications experienced by our patients. Of all the complications, only 33 patients (3%) had a pulmonary complication. The most common pulmonary complication was desaturation below 90% intra- or postoperatively, affecting 15 patients (1.5%), followed by difficulty with extubation or need for reintubation, affecting 9 patients (1%). Only 2 patients were unable to maintain tidal volumes felt to be adequate by anesthesia during surgery. A time effect was observed over the 6 year study period. The probability of having a pulmonary complication or an all-cause complication decreased over the time of the study ( P = .01 and P = .03, respectively).
|Delayed extubation or reintubation||9|
|Pulmonary edema or pleural effusion||3|
|All-cause complications||146 (14%)|
|Conversion to laparotomy||44|
|Ileus or bowel obstruction||8|
|Bladder or ureteral injury||6|
|Urinary tract infection||6|
|Hypo- or hypertension||3|
Common all-cause complications included 14 patients (1.4%) who required blood transfusion and 13 patients (1.3%) who developed a postoperative wound infection ( Table 3 ). There was 1 death, in an elderly patient with bowel perforation and subsequent sepsis, whose family opted for no resuscitation. The most common all-cause complication was the need for conversion to laparotomy, of which there were 44 patients total (4%). Table 4 shows the indications for all conversions to laparotomy. The most common reasons for conversion were an enlarged uterus or mass that could not be removed in a minimally invasive fashion (17 patients, 39%) and adhesions limiting visibility (10 patients, 23%).
|Reason for conversion||Patients, n (%)||BMI range (median)|
|Enlarged uterus or mass||17 (39)||30–47 (36)|
|Adhesions||10 (23)||30–56 (35)|
|Trendelenburg intolerance||6 (14)||33–69 (50)|
|Advanced disease||3 (7)||30–38 (31)|
|GI or GU injury||3 (7)||32–71 (52)|
|Bleeding||2 (5)||32–51 (43)|
|Body habitus||2 (5)||42–56 (47)|
|Robot malfunction||1 (2)||33 (33)|
Of the patients who were converted for removal of a large uterus or mass, 3 were converted to a minilaparotomy at the end of the procedure for specimen removal, whereas the remainder was converted to laparotomy after initial laparoscopic assessment revealed a specimen felt to be too large to remove through the vagina. Only 6 patients (14% of those needing conversion and 0.6% of total patients) were unable to tolerate Trendelenburg positioning. The reasons for difficulty tolerating the Trendelenburg were difficulty achieving adequate TV (2 patients) and low SpO2 (4 patients).
In univariate analyses, older age, higher estimated blood loss, and longer case time were associated with a higher rate of all-cause complications ( P = .0001, P < .0001, and P = .004, respectively). None of the other variables studied, including BMI, comorbid conditions, length of time in Trendelenburg position, or ventilator parameters, were significantly associated with all-cause complications.
Older age was also associated with a higher rate of pulmonary complications ( P = .01). A history of COPD and a higher average RR during surgery were borderline significant (both at P = .057). None of the other patient variables including BMI, operative variables, or other ventilator parameters were associated with pulmonary complications, except for decreased uterine weight ( P = .02). This likely was due to an independent association between decreased uterine weight and increased age; indeed, the significance of this association disappeared in a multivariate analysis.
In addition to evaluating BMI as a continuous variable, we subdivided patients into clinically important BMI groups of class I and II obesity (BMI, 30–39 kg/m 2 ), class III obesity (BMI, ≥40 kg/m 2 ), and further subdivided class III obesity into those with a BMI of 40–49 kg/m 2 and those with a BMI of 50 kg/m 2 or greater.
When comparing patients with class III obesity with patients with class I and II obesity, we failed to detect a significant difference in pulmonary or all-cause complication rates ( P = .298 and P = .269, respectively). Similarly, when comparing all 3 groups (class I and II obesity, class III obesity with a BMI of 40-49 kg/m 2 , and class III with a BMI of ≥50 kg/m 2 ), no significant difference in pulmonary or all-cause complication rates was detected among the 3 groups ( P = .164 and P = .115, respectively).
When comparing patients with class III obesity and a BMI of 50 kg/m 2 or greater with all other obese patients in our study population (class I, II, and III with a BMI of 40–49 kg/m 2 ), again no significant difference was detected in the pulmonary complication rates ( P = .192); however, patients with class III obesity and a BMI of 50 kg/m 2 or greater had a borderline higher rate of all-cause complications compared with all other obese patients ( P = .066) and were 1.7 times more likely to have a complication of any type (hazard risk, 1.657; 95% confidence interval, 0.97–2.84).
After multivariate analyses, it appears only older age was associated with a higher rate of pulmonary complications, with an odds ratio of 1.04 (95% confidence interval, 1.01–1.08). Figure 1 is a plot comparing patients with and without pulmonary complications by age. The median age of patients with pulmonary complications is higher than the median age of patients without pulmonary complications. In contrast, Figure 2 is a plot demonstrating that median BMI was not significantly different between patients who had pulmonary complications compared with those who did not have pulmonary complications.