Laparoscopic colorectal surgery is associated with a number of postoperative recovery benefits [1–3]. Many techniques have been employed to further minimize surgical trauma including reduced port and single incision laparoscopy. A common technique utilized to minimize incisions when a stoma is required is to enlarge the stoma aperture to extract the specimen, sparing a separate extraction incision. The literature suggests that 46% of patients with a stoma, created either open or laparoscopically, experience postoperative morbidity directly attributable to the stoma. Some studies also suggest that laparoscopic surgery may increase the incidence of parastomal hernia [4]. Many factors affect hernia rates including BMI, wound infection, hand-assisted or conventional laparoscopic approach, and medical comorbidities [5].

The impact of specimen extraction sites on wound complications, including surgical site infections and incisional hernia rates, has been evaluated in a number of studies assessing colorectal surgery procedures [5–7]. However, there are limited data on whether specimen extraction specifically through the stoma site is associated with increased morbidity particularly with stoma site complications including hernia, stenosis, prolapse, or retraction. The aim of this study is to assess whether clinically significant stoma-related complications are influenced by specimen extraction through the stoma site.

Patients and methods

A retrospective comparative review of 738 patients who underwent minimally invasive colorectal surgery with ileostomy creation between January 2008 and December 2014 was performed following institutional review board approval. This single institution study included patients who underwent a minimally invasive colorectal procedure performed by a colorectal-trained surgeon that required a stoma and a specimen to be extracted through the abdominal wall. Patients were excluded if the specimen was extracted through the perineum.

A wound protector was used in all cases at the specimen extraction site. The stoma site aperture creation involved vertical division of the anterior and posterior sheaths with blunt spreading of the rectus muscle layers. After specimen extraction, the anterior rectus sheath was tightened around the stoma using one or more interrupted figure of eight absorbable sutures. The stoma was matured as previously described [8]. The anterior rectus sheath of the separate specimen extraction incisions was also closed using absorbable suture. The SILS procedures all involved SSE. The NSSE extraction sites included midline, Pfannenstiel, umbilical and left lower quadrant (LLQ) sites. All data were obtained from a prospectively maintained institutional colorectal surgery outcome database corroborated by patient chart review as necessary. BMI, age, comorbidities, preoperative steroids, biologics, and chemotherapy use, ASA, LOS, EBL, parastomal complications, and parastomal hernia rate were compared between the two groups. All stoma site complications were obtained from clinical documentation obtained through chart review. If there was no documentation of hernia in the postoperative visits or the stoma nurse documentation, it was assumed no clinically evident hernia was present. CT scans were not routinely performed. The charts were also reviewed for all operations after the surgery to ensure that no clinically significant complications requiring surgery were omitted.

Statistical significance was assumed when the p value (two-sided) was less than 0.05. Categorical variables were reported as frequency (%), and quantitative variables were reported as mean ± standard deviation or as median (25th–75th‰). Categorical variables were analyzed with Chi-square or Fisher’s exact test. Parastomal hernia, separate site hernia, stoma site hernia, and all stomal complications were also described using the Kaplan–Meier method and were compared by log-rank test. Quantitative variables were analyzed with a Wilcoxon rank-sum test. Subsequently, multivariate Cox regression analyses were conducted to assess the associations between each of parastomal hernia, separate site hernia, stoma site hernia, and all stoma site complications, and stoma site extraction, with covariates selected in a stepwise fashion using Akaike’s information criterion (AIC).

Results

738 charts were reviewed which included 333 women (45.1%) and 405 men (54.9%). 23 (3.1%) cases were converted to open and these cases were included in the NSSE group as intention to treat. 139 cases were included in the SSE group (18.8%), while 599 patients had a separate incision NSSE (81.2%). The NSSE sites included 127 midline, 429 Pfannenstiel, 35 umbilical, and 8 left lower quadrant (LLQ) incisions. 736 patients included had a temporary ileostomy created, and only 2 patients in the NSSE group had permanent stomas created. 337 patients had a loop ileostomy created, while 401 patients had an end ileostomy created. 10 surgeries were performed on an emergent basis, while the remainder were performed electively. The method of surgery was based on the surgeon’s comfort. The procedures performed, type of surgery, and diagnoses are summarized in Table 1.

Table 1 Comparison of the type of surgical procedure, diagnosis, and technique between the groups
Full size table

There were no statistically significant differences in BMI, age, comorbidities, or ASA score between the SSE and NSSE groups (Table 2). EBL and operative time (OP) were both significantly decreased in the SSE group (Table 3). The median time to stoma closure interval was 4.7 (3.0–9.0) months for the SSE group (n = 139) and 5.4 (3.1–7.4) months for the NSSE group (n = 597) (p = 0.94). The postoperative outcomes of LOS, ileus, time to stoma function, SBO, anastomotic leak, wound infection rates, and venous thromboembolism rates were similar between the groups (Table 3).

Table 2 Comparison of demographics, preoperative characteristics between the groups
Full size table
Table 3 Comparison of intra- and postoperative outcomes between the groups
Full size table

The parastomal hernia rate for the entire series was 5.3%. The parastomal hernia rate was 10.1% for the SSE group (n = 14) and 4.2% for the NSSE group (n = 25) (p = 0.007). The estimated 2-year parastomal hernia rate for the entire series was 5.0%. The 2-year parastomal hernia rate was 9.8% for the SSE group and 3.9% for the NSSE group (p = 0.002). The parastomal hernia detection time was 4.1 (3.5–5.5) months for the SSE group and 6.9 (4.2–13.9) months for the NSSE group. The stoma prolapse rate was 1.4% for the SSE group (n = 2) and 0.3% for the NSSE group (n = 2). One patient experienced stoma retraction from the SSE group. The stoma stenosis rate was 1.4% for the SSE group (n = 2) and 0.2% for the NSSE group (n = 1). One patient from both the SSE and the NSSE suffered stoma strangulation. The parastomal abscess rate was 1.4% for the SSE group (n = 2) and 0.2% for the NSSE group (n = 1) (Table 4). The incidence of all stomal complications combined was higher in SSE group rather than the NSSE group (Table 5).

Table 4 Descriptive frequencies of postoperative stoma site complications between the groups
Full size table
Table 5 Kaplan–Meier estimated 2-year cumulative rate of stomal complications
Full size table

Prior to ileostomy closure, 21 non-stoma site incisional hernias were noted: 4/139 (2.8%) cases in the SSE group (one LLQ incision hernia, one midline incisional hernia, and two umbilical hernias), and 17/599 (2.8%) cases in NSSE group (7 Pfannenstiel, 6 midline, and 4 umbilical incisional hernias).

The number of patients who required additional surgeries related to stoma site complications was 7/139 (5.0%) for the SSE group and 27/599 (4.5%) for the NSSE group (p = 0.79) (Table 3). 7 patients that had the specimen removed from the stoma site required a second surgery, including: 3 related to stoma site complications, 2 for bleeding, 1 for small bowel obstruction (SBO), and 1 for bowel perforation. 27 patients from the NSSE group required a second operation, including: 3 for stoma site complications, 6 for rectal stump complications, 2 for SBO, 8 for anastomotic leak, 2 for bleeding, 3 for dehiscence, and 3 for hematoma/abscess. The 2-year non-stoma site incisional hernia rate was 4.2% for the SSE group and 3.1% for the NSSE (p = 0.91). The non-stoma site incisional hernia rate for the total series was 3.3% (24/738). The non-stoma site incisional hernia rate was 1.4% (2/139) for the SSE group and 3.7% (22/599) for the NSSE (p = 0.29). The stoma site hernia following closure of the stoma was 2/139 for SSE group and 11/597 for NSSE group (p = 0.75) (Table 4). 11/85 (12.9%) patients who underwent a SILS procedure and had SSE suffered stoma site complications, while 7/54 (12.9%) of patients who had a CL or RAL procedure with SSE had stoma site complications indicating that the SILS port itself does not increase the risk for parastomal hernia (p = 1.0) (Table 4).

Multivariate Cox regression analysis was conducted for factors associated with any stoma-related complication including parastomal hernia, prolapse, retraction, stenosis, strangulation, or abscess. The same analysis was performed for parastomal hernia alone. The following variables were used to construct the model: stoma extraction, steroid use, biologics use, chemotherapy, diagnosis, BMI, transfusion, COPD, age, and gender. SSE (HR: 3.14, 95% CI: 1.74–5.67, p < 0.001), transfusion (HR: 3.26, 95% CI: 1.28–8.28, p = 0.03), and BMI >30 (HR: 4.27, 95% CI: 2.40–7.60, p < 0.001) were found to be independent risk factors associated with all stoma-related complications (Table 6). SSE (HR: 2.93, 95% CI: 1.52–5.66, p = 0.003), BMI >30 (HR: 6.65, 95% CI: 3.47–12.74, p < 0.001) and transfusion (HR: 3.37, 95% CI: 1.19–9.58, p = 0.05) were found to be independent risk factors associated with parastomal hernia (Table 7).

Table 6 Cox regression model for any type of stoma-related complication (parastomal hernia, prolapse, retraction, stenosis, strangulation, or abscess) (n = 723, 47 events)
Full size table
Table 7 Cox regression model for parastomal hernia (n = 723, 39 events)
Full size table

Discussion

Laparoscopy was first described for colorectal operations in the early 1990s [9]. Since that time, the benefits of laparoscopic colorectal surgery are clear and include earlier return of bowel function, decreased postoperative pain, and lower complication rates [10, 11]. As surgeons have become more proficient in laparoscopy and technology has advanced, more complex operations are being performed with less incisions. Remzi et al. described the first single incision laparoscopic surgery (SILS) colectomy in 2008 [12]. Since that time, SILS has become a common platform for colorectal resection, which has resulted in using the stoma site not only as a port site, but also for specimen extraction.

Colorectal operations can include the formation of either an end stoma or a defunctioning stoma to reduce complications of an anastomosis. A large portion of these patients, created open or laparoscopically, experience postoperative morbidity directly attributable to the stoma [1]. Patients suffer from difficulties with stoma pouching, cosmetic concerns, pain, limited mobility, incarceration, retraction, stenosis, and the need for additional surgeries. Many laparoscopic and SILS surgeons are using the stoma site as a port site as well as for specimen extraction in an attempt to minimize incisions. Often the stoma site requires stitches to tighten down the fascia aperture around the bowel. As technology has advanced, it is possible to perform more advanced surgeries with less incisions theoretically decreasing pain, improving cosmetic outcomes, and minimizing wound complications. There are few studies that have examined the effects of enlargement of the stoma site to facilitate specimen extraction on stoma site complications.

Studies regarding parastomal hernias are difficult, as no standardized definition for parastomal hernia exists. Differing definitions of parastomal hernia may explain the wide range of incidences quoted in the literature, from 1 to close to 50% [13]. The use of CT scans is helpful and increases detection of asymptomatic hernias, but still can miss parastomal hernias that reduce in the supine position. This study looked at clinically significant hernias and more importantly stoma site complications that required additional interventions. In this study, prolapse was defined as eversion of the stoma through the abdominal wall, whereas parastomal hernia was defined as a clinically evident bulging around the stoma whether symptomatic or not [1, 14]. Other authors have described parastomal prolapse as a subgroup of parastomal hernia, but it is considered a separate entity in this study [15].

The overall parastomal hernia rate in this series was 5.3%, which is low compared with the reported rates for laparoscopic surgery. In this series, many of the patients had surgery for inflammatory bowel disease or rectal cancer resulting in 99% of the stomas being temporary. All Hartmann’s procedures at this institution were performed open and therefore excluded, and all abdominoperineal resection patients and total proctocolectomy with end ileostomy patients were excluded due to a perineal extraction site. A lower parastomal hernia rate was expected in this study since the average time to parastomal hernia detection was 6 months and the average time to stoma closure was 5 months. The use of clinical exam for diagnosis also falsely lowers the hernia detection rate, though that should be equally biased for both groups.

Patient-specific factors that could play a role in the formation of parastomal hernias including age, gender, diabetes, COPD, steroid use, biologic use, chemotherapy, wound infection, and BMI were evaluated. There were no differences in the patient-related factors between the SSE and NSSE groups (Table 2). We found that both operative time and EBL were both decreased in the SSE group. These findings are consistent with the author’s previous studies looking at SILS compared with conventional laparoscopy (CL) [16, 17]. Since 85/139 patients in the SSE groups were performed using the SILS method, this is likely the reason for the differences noted in the EBL and operative time.

We also found that 8 patients of the 139 from the SSE group experienced stomal prolapse, retraction, stenosis, strangulation, or abscess, while 5 patients of the 599 patients with NSSE suffered these complications. Since the individual complication rates were very low, all stoma site complications, including hernia, were combined and there was a statistically significant increase in complications in the SSE group (p < 0.05). The clinically relevant information in the proposed study is whether the patient suffered consequences of these complications and required an additional operation prior to stoma closure.

The rate of reoperation for stoma site complications was evaluated and was equivalent between the groups. The authors feel that if the patient does not need additional operations prior to stoma closure, the effects of the stoma site complications are temporary and corrected with the stoma closure. Therefore, the statistically significant increase in parastomal hernia rates associated with the SSE group has a much less substantial clinical significance. That does not mean that some patients did suffer adverse effects on quality of life in the interim. In this study, most of patients underwent surgery for IBD or cancer whose stomas were temporary. The increase in parastomal hernias in the patients who had SSE did not increase the need for added operations. The next concern is if the increased parastomal hernia rate led to in an increase in stoma site hernias after closure of the stoma. The follow-up data demonstrate no difference in the rate of hernia at the stoma site after closure of the stoma (Table 4).

The follow-up time for both groups is greater than 2 years, which should be adequate to detect postoperative hernias due to the fact that the majority of stomas were temporary. The NSSE follow-up time is longer than SSE due to the fact that SSE has been performed more recently; therefore, the SSE naturally has a shorter follow-up time. If this was to affect the stoma complications data, it would mean that fewer parastomal hernias were detected than truly exist in the SSE group.

Specific independent risk factors for parastomal hernia have previously been described. These predisposing conditions include respiratory comorbidities, cancer, diabetes mellitus, smoking, and elevated body mass index [1, 18, 19]. Multivariate analyses were performed to evaluate multiple patient-specific factors (Tables 6, 7). Factors that increased the risk of parastomal hernia as well as all stoma complications were SSE, BMI, and transfusion.

Using the stoma aperture as the specimen extraction site does increase the clinically detectable parastomal hernia rate. The increase in parastomal hernia rates in the SSE group is likely due to enlargement of the stoma site for extraction. The possibility exists that the SILS technique could increase the risk of stoma site hernia due to constant tension on the fascia from the SILS port and traction from the instruments. Studies have shown that SILS increases the overall hernia rate [20]. This study shows that the use of the stoma site as a port (SILS) did not increase the risk of stoma site complications versus conventional laparoscopy or robotics (Table 4).

The retrospective nature of this study is a clear weakness. The use of clinical documentation to detect parastomal hernias is also a weakness and leads to a falsely low hernia rates, though this weakness should have affected both the SSE and the NSSE groups equally. All operative reports of the patients were reviewed, so no additional operations were missed in either group. The large sample size is a positive attribute to the study. As well as the fact that few patients were lost to follow-up as the majority had temporary stomas and returned for stoma closure.

Conclusions

Stoma site specimen extraction increases the parastomal hernia rate as well as stoma site complications. This increase did not translate into additional surgeries before or after stoma reversal. The increase in parastomal hernias is likely related to enlargement of the stoma site rather than the SILS technique. Though SSE increases the parastomal hernia rate, there was no need for added intervention or increase in the hernia after stoma closure; therefore, it is reasonable in patients with an acceptable BMI in which the stoma will be temporary. Patients with an elevated BMI or requiring a permanent stoma should have a separate extraction incision created.