Introduction

Early in the laparoscopic era, obesity was considered a relative contraindication to a minimally invasive approach [1]. It is now known that rather than being a detriment, laparoscopy is actually beneficial to obese patients. Compared to open colectomy cases, obese patients undergoing a laparoscopic operation have decreased overall morbidity, fewer wound complications, less post-operative pain, shorter average hospital stays, and decreased mortality [2]. Previous studies have also demonstrated that obese patients [body mass index (BMI >30)] undergoing laparoscopic colectomy have worse perioperative outcomes including longer operative times and increased complications when compared to non-obese (BMI <30) cohorts undergoing identical procedures [3,4,5,6,7,8,9,10]. However, there is little data that specifically compares the challenges and outcomes of obese patients based on the degree of their obesity. We therefore aimed to evaluate the impact of the severity of obesity on perioperative outcomes in patients undergoing laparoscopic colectomy.

Methods

Following institutional review board (IRB) approval, we retrospectively reviewed the data of all patients undergoing laparoscopic colectomy between 1996 and 2013 by three surgeons at the Texas Endosurgery Institute and University Hospitals Cleveland Medical Center. Patients meeting inclusion criteria were identified by the International Classification of Diseases (ICD), 9th revision procedure codes for laparoscopic colectomy, which included 44,204–44,213. Patients were then classified according to BMI as obese (O, BMI 30.0–39.9), morbidly obese (MO, BMI 40.0–49.9), and super obese (SO, BMI >50) in conjunction with previously described World Health Organization (WHO) definitions [11]. Patients in whom colectomy was performed in conjunction with other surgical procedures and those with incomplete records were excluded from analysis. All patients underwent traditional straight laparoscopy and this series included no patients with hand-assist procedures. The primary outcomes of interest included open conversion rate, operative time (minutes), estimated blood loss (mls), post-operative complications, and post-operative length of stay in days, which were compared in O, MO, and SO cohorts. Complications were divided into wound, ileus, anastomotic leak, and systemic (i.e., sepsis, fever, pneumonia, urinary tract infection, etc.). Anastomotic leak was defined as either clinical leak requiring drainage of an abscess or return to the operating room, as well as those confirmed radiographically by CT or contrast enema.

Univariate analysis was performed using independent sample t tests, analysis of variance (ANOVA), or chi-square tests as appropriate. Statistical analysis was performed using PASW Statistics 18.0 (SPSS Inc., Chicago, Illinois) and significance was set at P < 0.05.

Results

We identified 923 patients (50.4% male) that met inclusion criteria. From these, 604 (65.4%), 257 (27.9%), and 62 (6.7%) were classified as O, MO, and SO, respectively. Clinicopathologic characteristics were similar among the three groups with the exception of the underlying diagnosis (Table 1).

Table 1 Clinicopathological features
Full size table

Table 2 lists the primary and secondary outcome measures. SO patients had significantly longer mean operative times (206 min; range, 75–508) compared to both O (163 min; range, 60–360) and MO patients (164 min; range, 60–440, P = 0.01). Similarly, SO patients had higher conversion rates when compared to both the O and MO cohorts (17.7 vs. 6.3 and 7.0%, P = 0.035).

Table 2 Outcomes by severity of obesity
Full size table

While intra-operative complications, including bowel injury, ureter injury, or major vascular injury were similar among the cohorts (O 2.6 vs. MO 3.5 vs. SO 3.2%; P > 0.05); overall post-operative morbidity was higher in the SO group (40.3 vs. O 12.4 vs. MO 16.3%. P = 0.001). This included higher rates of wound infection (P < 0.001) and anastomotic leak including peri-anastomotic abscess (P = 0.027) for SO patients. MO males (p = 0.033) also had higher rates of anastomotic leak when compared to MO females and O patients of either gender. Post-operative ileus longer than 7 days was similar between the groups (P = 0.065). Systemic complications excluding urinary tract infections were similar among all three cohorts (P > 0.05). These complications in O patients included: pulmonary embolus (6), DVT alone (6), pneumonia (5), and myocardial infarction (4); MO patients included: pulmonary embolus (2), MI (1), and DVT (1); and SO patients with pulmonary embolus (1) and MI (2).

Finally, the average length of stay in the O (3.9 days, range 2–12) and MO groups (4.9 days, range 2–12) was significantly less than the SO cohort (6.7 days, range 2–27; P < 0.001).

Discussion

In the present manuscript, we have demonstrated that worse perioperative outcomes are directly correlated with increasing severity of obesity following laparoscopic colectomy. Previous studies have compared the outcomes of laparoscopic colectomy in non-obese patients to obese patients and, not surprisingly, found worse overall outcomes in the obese group, similar to other types of operations for both open and minimally invasive approaches [12, 13]. In contrast, our goal was to distinguish the impact of rising BMI within a cohort where all patients meet obesity criteria. To our knowledge, there have not been any large-scale institutional studies specifically comparing the outcomes of obese patients based on the severity and category of their obesity following laparoscopic colectomy.

Our finding of longer operative times in the SO group is somewhat expected. With increasing BMI, both the abdominal wall and visceral adiposity are increased, adding to the technical complexity of the operation [14]. Rising obesity is associated with a greater sized omentum and loss of domain within the abdomen to create space for proper exposure [15]. Furthermore, the increased adipose tissue in the mesentery often can lead to greater difficulty identifying critical structures such as the ureter and vascular pedicle prior to ligation. While mean operative times for O and MO cohorts were similar, this may be more of a reflection of the experience of the surgeon with operating in these two cohorts, as the rates of obesity (while somewhat steadying) continue to increase to the point that these are more routinely encountered patients [16, 17].

Similarly, it was somewhat expected that our rates of conversion to open increased with rising BMI. Prior studies have shown increased conversion rates to an open procedure when comparing non-obese (~5–15%) to obese (7–25%) patients [9, 18,19,20]. While our results showed a lower conversion rate than previously reported of only 4.8% in patients with a BMI of 30–40 and 7.0% in those with a BMI of 40–50, this is again likely a reflection of the increasing experience with both laparoscopic colectomy in general, as well as its use in obese patients. Yet, once the BMI rises above 50, the conversion rate appears to climb, consistent with Tekkis and associates, who found that conversion rates increase per unit of BMI (odds ratio = 1.7) [21]. It will be interesting, if in the future as experience with minimally invasive colectomy in the SO cohort increases, whether this trend will reverse. Similar to previous reports, higher T stage and previous abdominal surgery remain risk factors for conversion to open procedures [22]. Additionally, there is some data to suggest that hand-assist surgery in the MO patient may be associated with decreased rates of conversion, especially early in the surgeon’s learning curve [23].

Complication rates of 13–30% have been reported in non-obese patients undergoing laparoscopic colectomy [12, 18]. This is comparable to our results for patients with a BMI of 30–40 (13.1%). We found a slight increase in complication rate when patients’ BMI increased to between 40 and 50 (15.9%), and even higher overall complication rate for the SO cohort (41.4%), as well as when stratifying by individual complications such as wound infections, post-operative ileus, intra-abdominal abscess, and anastomotic leaks (0.2 vs. 1.2 vs. 13.8%). Previous authors have similarly demonstrated a higher rate of wound, ileus, and anastomotic leakage with comparing obese patients to non-obese [24,25,26]. Causey and associates have previously demonstrated in a population-based series of Crohn’s patients undergoing bowel resection that with every increase in BMI, the overall complication rate rose by 0.7% (95% CI 0.45–1.04%), which remained consistent with stratification by both open (0.6%) and laparoscopic approaches (1.1%; 95% CI 0.6–1.6%) [11]. However, this series included small intestine, colon, and rectal procedures and was dedicated to patients with Crohn’s disease. Prior studies have also shown significant differences in both gastrointestinal recovery and length of stay between non-obese and obese patients, with non-obese having improved outcomes in each arena [18]. Our results demonstrate this trend continues with the super obese; as post-operative length of stay also showed a similar trend with the SO cohort having even longer hospitalizations.

We acknowledge certain limitations to the present analysis. The mere nature of a retrospective analysis carries a potential source of bias related to the chart review and recording procedures. In addition, while this series was multi-institutional, all of the surgeons were extensively experienced in performing laparoscopic colectomy and the numbers may not reflect the general population as a whole. However, this is likely to be demonstrated only in lowering the overall numbers of complications and conversions, rather than the trend that was identified. Finally, the SO cohort is relatively small compared to the other two groups, composed of only 62 patients. With a larger collection of SO patients, and increasing experience in this group, our results may be altered somewhat. However, despite these limitations, we present the largest series to our knowledge of obese patients undergoing laparoscopic colectomy that have been specifically analyzed by the degree and severity of rising BMI and have demonstrated the negative impact on outcomes—even with experienced surgeons.

Conclusions

In this series of obese patients, we found that the severity of a patient’s obesity is directly correlated with their outcome, with SO patients demonstrating the most dramatic impact when compared to both MO and O patients. While laparoscopic colectomy in all patients requires meticulous perioperative care and sufficient laparoscopic experience, colorectal surgeons must be aware of challenges that the severity of obesity poses.