Minimally invasive surgery for diverticulitis: a single-center retrospective study in the United States focusing on splenic flexure mobilization in elective sigmoid colectomy

Abstract

Purpose

Splenic flexure mobilization (SFM) is occasionally utilized during sigmoidectomy to facilitate a tension-free colorectal anastomosis. Present literature regarding its use, safety, and efficacy predominantly evaluates traditional laparoscopic vs. open techniques, thus not adequately representing the current minimally invasive surgical landscape.

Methods

This retrospective cohort analysis evaluated SFM during sigmoidectomy for diverticular disease via traditional laparoscopic, single-incision laparoscopic, and robotic techniques at a United States academic institution from 2019 to 2022. Primary outcomes were the rate of SFM, independent predictors, and short-term outcomes.

Results

A total of 117 patients underwent sigmoidectomy for diverticulitis by six colorectal surgeons, of whom 27 underwent SFM. SFM was associated with higher rates of complicated diverticulitis (85.2% vs. 46.7%, p < 0.001), including abscess (55.6% vs. 33.3%, p = 0.037) and fistulae (59.3% vs. 17.8%, p < 0.001). Presence of a fistula was an independent predictor of SFM by logistic regression (odds ratio [OR], 5.886; p < 0.001). Mobilization was associated with more concomitant surgical procedures (66.7% vs. 28.9%, p < 0.001). There was a significant association between SFM and conversion to open approach (14.8% vs. 3.3%, p = 0.049), longer length of stay (4.2 ± 2.8 days vs. 3.6 ± 4.3 days, p = 0.041), and increased operative time (241.7 ± 76.4 minutes vs. 199.2 ± 75.7 minutes). There was no difference in 30-day morbidity/mortality.

Conclusion

Routine SFM can be considered in minimally invasive sigmoidectomy for complicated diverticulitis without significant impact on 30-day morbidity/mortality regardless of surgical approach and may be anticipated preoperatively, particularly when a fistula is present.

INTRODUCTION

Despite the increasing prevalence of diverticular disease in Western society, there remains a lack of consensus guidelines regarding definitive indications and optimal timeline for surgical intervention. Previous studies have found surgical intervention to be advantageous in more complicated cases, particularly involving fistula or obstruction [1–3]. The sequela of chronic inflammation, such as distorted planes and foreshortened mesentery, often make the critical step of creating a tension-free colorectal anastomosis during resection more challenging [4,5]. Although there are multiple intraoperative techniques that can be utilized to help achieve this goal, splenic flexure mobilization (SFM) is often one of the first employed.

SFM has previously been shown to be safe and effective during laparoscopic resection of left-sided diverticulitis [6]. Nevertheless, it is currently debated whether routine SFM is warranted for left-sided colectomies for both benign and malignant indications, or if a more selective approach should be taken due to the potential risks of morbidity [7,8]. SFM can be technically challenging due to the need for additional meticulous dissection and careful identification of important vasculature and surrounding organ structures. This is especially true in the reoperative field, setting of extensive active or chronic inflammation, in patients with elevated body mass index (BMI), or when being performed via a minimally invasive approach [4]. SFM has been more extensively studied in the setting of left-sided colectomy for malignant disease, when it is often necessary to allow for adequate pelvic dissection and lymph node harvest, and low pelvic anastomosis [8].

The present research available surrounding SFM specifically for benign disease is sparse. A meta-analysis evaluating outcomes of SFM vs. no-SFM performed by Nowakowski et al. [7] evaluated studies involving traditional laparoscopic (TL) and open sigmoid and rectal resections for both benign and malignant indications. Only two of the 10 papers included in their final analysis included benign conditions in their study populations [7]. A more recent meta-analysis by Rondelli et al. [9] evaluated outcomes of SFM vs. no-SFM for benign and malignant indications, and included TL and open approaches only for rectal resections. The breadth of these study populations should make surgeons cautious in applying these results to the management of individual patients with diverticulitis.

Minimally invasive surgical (MIS) techniques in colorectal surgery are rapidly evolving. The present data comparing outcomes of SFM for benign colorectal disease across various MIS approaches ([TL vs. single-incision laparoscopic [SILS] vs. robotic-assisted laparoscopic [RALS]) is virtually non-existent, particularly those involving robotic-assisted cases. Furthermore, those studies that are available do not control for the indication or type of surgery performed (i.e., sigmoidectomy vs. left hemicolectomy vs. rectal resection), which are important factors when discussing clinical outcomes and their implications. Additionally, there are no studies reported evaluating predictive factors of SFM for these various MIS approaches. Evaluating for differences across the various MIS approaches, rather than treating them as one entity, is important to illuminate any potential impact the inherent ergonomic and technical differences a particular technique may have on the need for SFM. Any identifiable significant differences could pique interest in developing trials that would have an impact on surgical decision-making and, therefore, affect the capability of surgeons to deliver the most effective, evidence-based surgical care for patients. This study aims to evaluate for patient or disease factors that were predictive of the need for SFM during various approaches to MIS sigmoid resection for diverticulitis.

METHODS

Study design, patients, and data collection

A retrospective cohort analysis of patients who underwent minimally invasive, elective sigmoid resection for diverticulitis at a single, academic, quaternary referral center in the United States from 2019 to 2022 was performed. Patients who underwent minimally invasive surgery for left-sided diverticulitis (classified under K57 in the International Classification of Diseases, 10th Revision) with one of six fellowship-trained and American Board of Colon and Rectal Surgery-certified attending colorectal surgeons were identified. The electronic medical record was extensively reviewed for relevant clinical data and was extracted into an institutional database in a de-identified format.

Patients were included if they underwent TL, SILS, or RALS sigmoid resection between 2019 and 2022. It is important to note that all surgeons involved in this analysis perform SFM selectively as part of their typical practice pattern. Patients were excluded if they were less than 18 years old, pregnant, incarcerated, had urgent or emergent surgery, underwent low anterior resection, left hemicolectomy, Hartmann’s reversal, or an open procedure as their planned operation to minimize variation due to baseline extent of dissection and resection.

Outcome measures

Our primary endpoint was the identification of any patient or disease factors that were predictive of the need for SFM during elective MIS sigmoid resection for diverticulitis. Secondary endpoints included: comparison of MIS approach, indication for surgery, operative time, concomitant intraoperative procedures, conversion to open, estimated blood loss (EBL), postoperative hospital length of stay (LOS), time to return of bowel function, 30-day morbidity and mortality. All postoperative complications were graded using the Clavien-Dindo classification. Complicated diverticulitis cases were defined as diverticular disease associated with abscess, fistula, stricture, obstruction, or perforation identified pre- or intraoperatively. All other cases were deemed uncomplicated.

Statistical analysis

Continuous variables are presented as means with standard deviations. Categorical variables are presented as percentages of the total. Comparative analyses of the SFM vs. no-SFM cohorts were made using either the independent samples t-test, Mann-Whitney U-test for continuous variables, and the chi-square test or Fisher exact test for categorical variables. Our threshold for statistical significance was a p-value of less than 0.05. Multivariable binary logistic regression analysis was performed to assess predictors of intraoperative SFM. All statistical analyses were performed using IBM SPSS version 28.0 (IBM Corp.) with the assistance of a university-employed statistician.

RESULTS

Our retrospective chart review identified 117 patients who underwent elective, MIS sigmoid resection for diverticular disease during the study timeframe. Of these, 27 (23.1%) underwent SFM and 90 (76.9%) did not.

Patient demographics and preoperative comorbidities

There was no statistically significant difference in age (62.5 years vs. 58.8 years, p = 0.066), BMI (28.4 kg/m² vs. 29.8 kg/m², p = 0.535), incidence of previous abdominal surgeries (p = 0.412), sex (p = 0.391), or race distribution (p = 0.591) between the SFM and no-SFM groups (Table 1).

Table 1.

Patient demographics and preoperative comorbidities

Variable	Splenic flexure mobilization	No splenic flexure mobilization	p-value
Demographics
No. of patients	27 (23.1)	90 (76.9)
Age (yr)	62.5 ± 14.2	58.8 ± 11.0	0.066
Sex			0.391
Female	14 (51.9)	55 (61.1)
Male	13 (48.1)	35 (38.9)
Prior abdominal surgery	20 (74.1)	58 (64.4)	0.412
Race			0.591
Caucasian	21 (77.8)	72 (80.0)
African American	2 (6.7)	10 (11.1)
Others	4 (15.5)	8 (8.9)
Body mass index (kg/m2)	28.4 ± 3.6	29.8 ± 6.2	0.535
ASA PS classification ≥III	19 (70.4)	55 (61.1)	0.381
Comorbidities
Current smoker	2 (7.4)	12 (13.3)	0.517
Congestive heart failure	0 (0)	1 (1.1)	>0.999
Coronary artery disease	3 (11.1)	7 (7.8)	0.695
Chronic kidney disease	2 (7.4)	8 (8.9)	>0.999
Dialysis	2 (7.4)	2 (2.2)	0.227
Obstructive sleep apnea	3 (11.1)	17 (18.9)	0.560
Chronic obstructive pulmonary disease	0 (0)	6 (6.7)	0.334
Steroid use	0 (0)	7 (7.9)	0.198
Bleeding disorder	0 (0)	0 (0)	NA

Values are presented as number (%) or mean ± standard deviation.

ASA, American Society of Anesthesia; PS, physical status; NA, not applicable.

There were no statistically significant differences in preoperative comorbidities, including smoking status (p = 0.517), congestive heart failure (p > 0.999), coronary artery disease (p = 0.695), chronic kidney disease (p > 0.999), dialysis (p = 0.227), obstructive sleep apnea (p = 0.560), chronic obstructive pulmonary disease (p = 0.334), steroid use (p = 0.198), and bleeding disorder (Table 1).

Approach, indications, and intraoperative events

The three minimally invasive approaches utilized for sigmoidectomy during the study period at our institution were TL, SILS, and RALS. The most common approach was SILS (47.9%), followed by RALS (36.8%) and TL (15.4%). This distribution of operative approach was similar in both the SFM and the no-SFM cohorts. There was no statistically significant difference in conversion rate for the various MIS approaches within the SFM cohort.

On univariate analysis, patients who underwent SFM were significantly more likely to have complicated diverticulitis (85.2% vs. 46.7%, p < 0.001). More specifically, they were more likely to have an abscess (55.6% vs. 33.3%, p = 0.037) or fistula (59.3% vs. 17.8%, p < 0.001). Patients undergoing SFM were significantly more likely to have had a colovaginal fistula specifically (p = 0.043), but there was no significant association between SFM and fistulas to other organs. There was no significant difference in the proportion of patients with stricture or obstruction between the SFM and no-SFM cohorts (3.7% vs. 3.3%, p = 0.934) (Table 2).

Table 2.

Approach, indications, and intraoperative events

Variable	Splenic flexure mobilization (n = 27)	No splenic flexure mobilization (n = 90)	p-value
Approach			0.885
Laparoscopic (n = 18)	4 (14.8)	14 (15.6)
Single-incision laparoscopy (n = 56)	12 (44.4)	44 (48.9)
Robotic (n = 43)	11 (40.7)	32 (35.6)
Operating room timea) (min)	241.7 ± 76.4	199.2 ± 75.7	0.009*
Indication
Complicated diverticulitis (n = 65)	23 (85.2)	42 (46.7)	<0.001*
Uncomplicated, recurrent (n = 52)	4 (14.8)	48 (53.3)
Intraoperative findings and events
Abscess (n = 45)	15 (55.6)	30 (33.3)	0.037*
Fistula (n = 32)	16 (59.3)	16 (17.8)	<0.001*
Colovesical	6 (37.5)	12 (75.0)	0.049*
Colocutaneous	1 (6.3)	0 (0)	>0.999
Colovaginal	5 (31.3)	0 (0)	0.043*
Colocolonic	1 (6.3)	0 (0)	>0.999
Enterocolovesical	1 (6.3)	0 (0)	>0.999
Not specified	2 (12.5)	4 (25.0)	0.654
Stricture or obstruction (n = 4)	1 (3.7)	3 (3.3)	0.934
Concomitant procedures
Extensive lysis of adhesions	8 (29.6)	9 (10.0)	0.025*
Ureteral stent	2 (7.4)	6 (6.7)	>0.999
Enterorrhaphy	2 (7.4)	1 (1.1)	0.133
Small bowel resection	3 (11.1)	3 (3.3)	0.135
Fistula takedown	13 (48.1)	15 (16.7)	0.002*
Bladder repair	3 (11.1)	7 (7.8)	0.695
Ureteral repair/reconstruction	2 (7.4)	3 (3.3)	0.326
Gynecologic procedure	2 (7.4)	3 (3.3)	0.326
Splenectomy	0 (0)	0 (0)	-
Ostomy creation	1 (3.7)	4 (4.4)	0.867
Conversion to open (n = 7)	4 (14.8)	3 (3.3)	0.049*
Traditional laparoscopy	1 (25.0)	2 (66.7)	0.486
Robotic-assisted laparoscopy	1 (25.0)	0 (0)	>0.999
Single-incision laparoscopic	2 (50.0)	1 (33.3)	>0.999
Intraoperative bleeding requiring transfusion	0 (0)	0 (0)	-
Estimated blood loss (mL)	131.3 ± 177.5	53.5 ± 73.5	0.028*
Wound class			0.032*
Clean	0 (0)	0 (0)
Clean-contaminated	2 (7.4)	29 (32.2)
Contaminated	8 (29.6)	16 (17.8)
Dirty	17 (63.0)	45 (50.0)

Values are presented as number (%) or mean ± standard deviation.

^a)Measured from the time of incision to the end of anesthesia in minutes.

*Statistically significant.

The SFM cohort had significantly longer operative time (241.7 minutes vs. 199.2 minutes, p = 0.009), were more likely to be converted to open (14.8% vs. 3.3%, p = 0.049), and had a higher EBL (131.3 mL vs. 53.5 mL, p = 0.028) compared to the no-SFM cohort. Operative approach did not appear to contribute to the need for conversion to open in those cases requiring SFM. Of those SFM cases that were converted to open, SILS was the most common approach (n = 2), followed by TL and RALS (n = 1). This distribution was not statistically significant, and the exceedingly small sample size warrants cautious interpretation. Intraoperatively, patients who underwent SFM were significantly more likely to undergo concomitant fistula takedown (48.1% vs. 16.7%, p = 0.002) and require extensive adhesiolysis (29.6% vs. 10.0%, p = 0.025). The rates of other concomitant procedures, including ostomy creation (p = 0.867), ureteral stent placement (p >0.999) or ureteral repair (p = 0.326), enterorrhaphy (p = 0.133), small bowel resection (p = 0.135), bladder repair (p = 0.695), gynecologic procedure (p = 0.326), and splenectomy were statistically similar between the two cohorts (Table 2).

Wound class was also significantly different between the cohorts (p = 0.032), with SFM patients having a greater proportion of wounds classified as dirty (63.0% vs. 50.0%) or contaminated (29.6% vs. 17.8%) and fewer wounds classified as clean-contaminated (7.4% vs. 32.2%) (Table 2).

Postoperative morbidity and mortality

LOS was longer in patients who underwent SFM (4.2 days vs. 3.6 days, p = 0.041). Further, a greater proportion of the SFM group also experienced a “prolonged LOS,” defined as ≥75th percentile LOS of the entire cohort (33.3% vs. 14.4%, p = 0.028). There were no statistically significant differences in postoperative morbidities in the SFM vs. no-SFM cohort, including anastomotic leak (p > 0.999), surgical site infection (SSI; p = 0.680), wound dehiscence requiring return to operating room, ileus (p = 0.350), urinary retention (p > 0.999), urinary tract infection (UTI) (p > 0.999), stroke (p = 0.231), and bleeding requiring transfusion (p > 0.999). There were zero cases of acute renal failure, deep venous thrombosis, pulmonary embolism, or myocardial infarction in either group (Table 3). Both cohorts experienced a statistically similar time to return of bowel function (1.8 ± 0.9 days vs. 2.2 ± 2.6 days in SFM vs. no-SFM, respectively; p = 0.978). The SFM and no-SFM cohorts experienced similar 30-day return to the operating room and 30-day readmission rates (p > 0.999 and p > 0.999, respectively). There were no mortalities within 30 days of operation in either cohort (Table 3).

Table 3.

Postoperative complication rates

	Splenic flexure mobilization (n = 27)	No splenic flexure mobilization (n = 90)	p-value
Time to return of bowel function (day)	1.8 ± 0.9	2.2 ± 2.6	0.978
Length of stay (day)	4.2 ± 2.8	3.6 ± 4.3	0.041*
Prolonged length of staya)	9 (33.3)	13 (14.4)	0.028*
Postoperative complications
Anastomotic leak	0 (0)	1 (1.1)	>0.999
Surgical site infectionb)	1 (3.7)	7 (7.8)	0.680
Ileus	3 (11.1)	4 (4.4)	0.350
Urinary retention	1 (3.7)	3 (3.3)	>0.999
Urinary tract infection	1 (3.7)	5 (5.6)	>0.999
Stroke	1 (3.7)	0 (0)	0.231
Bleeding requiring transfusion	1 (3.7)	5 (5.6)	>0.999
Clavien-Dindo classification			0.641
0	21 (77.8)	73 (81.1)
I–II	4 (14.8)	8 (8.9)
III–V	2 (7.4)	9 (10.0)	>0.999
Return to OR	0 (0)	2 (2.2)	>0.999
Readmission within 30 days	3 (11.1)	9 (10.0)	>0.999
30-Day mortality	0 (0)	0 (0)	-

Values are presented as mean ± standard deviation or number (%).

OR, operating room.

^a)Prolonged length of stay was defined as >75th percentile for the entire cohort. ^b)Superficial, deep, organ/space.

*Statistically significant.

Predictors of splenic flexure mobilization

Multivariable binary logistic regression analysis demonstrated the presence of a fistula as a significant independent predictor of need for SFM during elective minimally invasive sigmoid resection for diverticulitis (odds ratio [OR], 5.89; 95% confidence interval, 2.14–16.17; p < 0.001). Age, sex, approach, and the presence of an abscess or stricture were not significant predictors of need for SFM (Table 4).

Table 4.

Predictors of minimally invasive splenic flexure mobilization on multivariable analysis

Variable	Odds ratio (95% CI)	p-value
Age	1.022 (0.978–1.067)	0.334
Sex
Male	Reference
Female	0.488 (0.169–1.411)	0.185
Body mass index	0.948 (0.855–1.052)	0.315
Approach
Laparoscopic single incision	1.042 (0.238–4.559)	0.957
Robotic	1.207 (0.266–5.476)	0.808
Laparoscopic	Reference
Abscess	1.862 (0.690–5.024)	0.220
Fistula	5.886 (2.142–16.173)	<0.001*
Stricture	1.506 (0.101–22.458)	0.766

CI, confidence interval.

*Statistically significant.

DISCUSSION

The role of routine SFM during left-sided colon resections has been debated amongst surgeons for decades. Now, with the widely accepted practice of RALS and the growing popularity of the SILS technique, there is even more demand for updated data regarding the indication-specific utilization, safety, and efficacy of SFM in minimally invasive abdominopelvic surgery. Our retrospective cohort analysis, performed at a single, quaternary referral, academic institution in the United States, sought to investigate SFM during elective, minimally invasive sigmoid resection for diverticulitis. All surgeons involved in this analysis perform SFM selectively as part of their typical practice pattern, making this an ideal population with which to investigate patient or procedure-specific predictors of need for SFM during sigmoid resection.

Overall, we found SFM significantly more likely to be performed in complicated diverticulitis, specifically in the presence of abscess or fistula in univariate analysis. There was no significant difference in SFM in the presence of stricture or obstruction. On multivariate analysis, we found the presence of a fistula to be an independent predictor of the need for concomitant SFM. This result implies that the transmural inflammation resulting in perforation and fistula development is more likely to affect the mesenteric pliability and thus require SFM in order to create a tension-free anastomosis. Additionally, there is likely a greater area of organ involvement and more caudad location, specifically in pelvic fistulae and abscesses, which would necessitate a larger resection area with a more distal transection point and, therefore, increase the likelihood that SFM would be necessary to complete a tension-free colorectal anastomosis. There have been very few studies involving predictive models for SFM, and the studies that are available are not adequate representations of the breadth of today’s MIS landscape. For example, Pettke et al. [8] compared SFM in only TL vs. open resections, and furthermore, included both benign and malignant left-sided pathologies. Their significant finding was that SFM was significantly more likely to occur during the TL approach (86% vs. 40%, respectively). Schlussel et al. [10] also investigated SFM in only TL vs. open colectomies, exclusively in diverticulitis cases, and also found significantly increased odds of performing an SFM if the TL approach was utilized (OR, 2.4; p = 0.02). They also found a significant association between SFM and disease located in the descending rather than sigmoid colon (82% vs. 12%, p < 0.01). However, neither study included SILS or RALS techniques in their analysis. Our SFM analysis is the first to-date that includes all of these minimally invasive approaches, including SILS and RALS, while specifically controlling for surgical indication and type of surgery performed.

In concordance with previous studies, the SFM cohort was associated with significantly longer operative times [7,9,10]. The reasoning for this is likely multifactorial, as not only is it an additional operative step, but it is often done after initial resection reveals a deficit in adequate colon length for colorectal anastomosis. The previous literature reporting this finding compared TL to open procedures, whereas we demonstrated this same finding when comparing various MIS techniques. Variance in operative time based on approach to SFM (anterior, inferomedial, lateral) [11] has not been evaluated to our knowledge and similarly was not evaluated in our study. Additionally, our observed increased operative time may also be secondary to the fact that our study found SFM significantly more likely in complicated diverticulitis cases, as discussed above. Our study is one of the first to not only include information regarding uncomplicated vs. complicated disease, but also to stratify complicated disease by anatomic findings. We found our SFM cohort was associated with significantly more concomitant intraoperative procedures, specifically fistula takedown and extensive adhesiolysis, as well as higher rates of conversion to open surgery. Overall, these findings in concert highlight the challenges of dissection in chronically and severely inflamed anatomic planes. The degree of added complexity and difficulty of dissection encountered in the surgical management of complicated diverticulitis, rather than the SFM technique itself, is likely the primary reason for higher rates of conversion to open approach found in our SFM cohort and could also explain why studies evaluating SFM outcomes in cancer resections do not demonstrate similar higher conversion rates [7]. Additionally, it is worth noting that true colovaginal fistulas are rare in the absence of previous hysterectomy, but were a relatively frequent finding in our series. This would also imply a reoperative nature of the pelvis, which also increases the complexity of the overall operation. Interestingly, in concordance with previously reported studies that evaluated SFM vs. no-SFM in TL vs. open resections for all indications, we found no significant differences in the rates of postoperative complications, anastomotic leak, SSI, return to OR, or 30-day readmission rates between the SFM and no-SFM cohorts in MIS [12,13]. However, our finding is discordant with the only other study analyzing left-sided resection specifically for diverticulitis, which demonstrated an increased risk of minor morbidity (defined by researchers as superficial or deep SSI, pneumonia, unplanned intubation, UTI, or deep vein thrombosis) in patients who underwent SFM [10]. This finding is of particular importance as it suggests that while SFM can add time and complexity to a procedure, short-term outcomes appear to be equivalent. This could provide reassurance to surgeons when deciding whether SFM is necessary.

Interestingly, we found that the type of minimally invasive approach was not a significant independent predictor of need for SFM. This is a rather unexpected finding, given that a study using the American College of Surgeons NSQIP (National Surgical Quality Improvement Program) database by Al-Temimi et al. [14] demonstrated SFM was significantly more common in laparoscopic compared with robotic colectomies for diverticulitis. This divergence may be due to the fact that we limited our cohort to sigmoid resections only, whereas their study looked at all left-sided resections and their results may have been impacted by disease location rather than the technique itself, as disease location in the descending colon was significantly more likely to result in SFM in another previous study (OR, 7.4; p < 0.01) [10]. Disease location and extent certainly influence the length of bowel resected and the extent of mobilization necessary for creation of a colorectal anastomosis, thus having implications on the need for SFM. Overall length of the colon, as influenced by several factors including ethnicity, can also impact the need for SFM to achieve a tension-free anastomosis. Unfortunately, this data is not available in our institutional database and therefore was not able to be evaluated in the current study. This would make for an interesting future analysis and should be considered going forward. Additionally, all cases included in our analysis were performed by experienced colorectal surgeons who regularly practice both advanced laparoscopy and robotic surgery, whereas the aforementioned study included general surgeons, in addition to colorectal surgeons, with likely heterogeneous degrees of comfort with MIS colorectal resections. This could imply that as surgeons become more experienced with these approaches, the decision to perform SFM becomes more dependent on patient and disease-specific factors rather than the limitations or capabilities of the specific surgical approach.

Further research is needed with prospective trials involving larger populations to replicate and validate our results. Our results should be interpreted with caution due to the limitations discussed below, but if validated, these findings could legitimize the role of routine SFM in minimally invasive sigmoid resections for complicated diverticulitis when a fistula is present and, arguably, for cases when an abscess is encountered or extensive adhesiolysis is required. Additionally, knowing patients with a fistula are likely to require SFM may encourage surgeons to perform this maneuver earlier in the procedure during initial mobilization of the left colon, rather than after extensive and difficult dissection of the chronically inflamed sigmoid colon. As SFM can be a technically challenging endeavor, performance earlier in the procedure may be of benefit from a psychological standpoint to the surgeon. This could, in turn, potentially impact not only clinical outcomes but also potentially decrease operative times.

Significant limitations to our study include its retrospective nature and the potential for selection bias. The decision to undergo a TL vs. SILS or RALS approach was surgeon-dependent and potentially influenced by operative time constraints and equipment availability. Additionally, the decision whether to perform SFM was also surgeon-dependent and not-standardized, therefore, also potentially contributing to selection bias. Nevertheless, there were no significant differences in patient demographics or preoperative comorbidities between the SFM and no-SFM cohorts. An additional limitation includes the single-center setting, which may limit the power and generalizability of our findings. Future multicenter prospective studies could provide more definitive evidence on the role and safety of MIS SFM in diverticulitis. Further, our small sample size of cases requiring SFM limits the statistical power and precision of our analysis, as evidenced by our wide confidence interval.

Overall, routine SFM can be considered in patients undergoing minimally invasive sigmoidectomy for complicated diverticulitis when a fistula is present, and potentially even when an abscess is present, without concern for significant impact on 30-day morbidity/mortality. This decision should be independent of the MIS approach utilized.