This propensity score matching comparison of emergent open versus laparoscopic colonic resection for IC confirmed that surgical site complications and duration of ventilation support were statistically significantly reduced after the laparoscopic approach. Of note, there was no statistically significant difference found in hospital stay, major complications, or mortality. After multivariate analysis, CCI (OR = 1.29, p = 0.024), APACHE II score (OR = 1.25, p = 0.001), and Favier’s classification (OR = 9.02, p = 0.023) remained statistically significant predictors of mortality (Table 4).

While randomized clinical trials are universally accepted as the gold-standard methodology for measuring the “causal” effects of management on outcomes [25], they are not always possible or adequately powered because they are time-consuming, costly, and may have ethical or practical constraints [26, 27]. Moreover, IC is rare; therefore, the accrual period for a randomized trial would be long and costly. Propensity score matching has gained in popularity in observational studies because potential biases in pretreatment characteristics between treatment groups are minimized [28].

Perioperative mortality can be as high as 60% after surgery and is most likely multifactorial, including degree of parietal involvement, location of ischemia, patient comorbidities, and physiological status [7]. However, after propensity-score matching, there was no statistically significant difference in patient variables thought to be associated with surgical outcomes in our study. Intraoperative findings (Favier’s classification, location of ischemia) were also included in propensity-score matching, even though these variables cannot be known precisely before operation, because they reflect severity of disease and technical difficulty. However, in 2011, Reissfelder et al. presented a risk score that predicts postoperative mortality of patients undergoing surgery for IC; perioperative variables included non-occlusive IC, acute renal failure, extent of bowel ischemia, serum lactate, and duration of catecholamine therapy [29]. As some of these factors were not included in our matching, they could eventually be part of a selection bias for treating patients with IC using an open or laparoscopic approach.

While laparoscopy is widely practiced in elective colorectal surgery, it is much less popular for emergency settings, although its feasibility and safety have been shown [11] and, ideally, should provide a smoother and less complicated postoperative course. Moreover, laparoscopy can be used with diagnostic intent in suspected IC, especially when colonoscopy is deemed dangerous [30] or for post-ischemic stenosis [15] and has the advantage of visualizing the entire colon, but is infrequently indicated for colectomy after acute ischemia and in particular in patients with IC. In the ACS-NSQIP database study, only 125 of 4548 (4.3%) colectomies for IC were performed laparoscopically [31]. Most published studies on the laparoscopic approach in patients with acute colorectal disease have focused on perforated diverticulitis, colorectal anastomosis leakage, or inflammatory colitis [12,13,14, 32,33,34,35,36]. All these studies showed that the laparoscopic approach could decrease the surgical site complication rate in the emergent setting but were not conclusive regarding overall complications, mortality, or hospital stay. This is in line with our results: The surgical site complication rate was lower in the laparoscopic group after propensity score matching. Furthermore, duration of ventilator support was shorter in the laparoscopic group after propensity score matching. Although there was a statistical trend toward shorter hospital stay, larger series are needed to demonstrate the advantages of the laparoscopy approach, and in particular in patients with severe disease or poor physiological status.

Symptoms of IC are often nonspecific, vague, and the diagnosis can be challenging at an early stage. In a review of 364 patients, peritoneal signs were present in only 7.4 percent of patients [1]. Furthermore, it is often difficult to identify symptoms in patients who are unconscious and debilitated, those in intensive care, or who are cognitively impaired, such as those with delirium or dementia. Therefore, IC must be suspected if a patient in an intensive care unit cannot tolerate a normal diet within a couple of days or has signs of infection. As ischemia in IC is initially mucosal, laparoscopic visualization of the serosa may seem normal, but conversely, as transmural ischemia can exist with minimal clinical signs, laparoscopic examination could be an early diagnostic tool for severe IC and lead to earlier surgical intervention.

When we compared the surgical delay between the laparoscopic and open groups, we found that the delay was longer in the open group than in the laparoscopic group (59.6% vs. 43.5%, p = 0.121; after propensity, 67.7% vs. 48.3%, p = 0.123). Although the difference was not statistically significant, this finding might further support the role of laparoscopy as an early diagnostic tool of the acute abdomen (EAES Symposium) [30, 37].

Our overall operation times (open: 200 min, laparoscopy: 180 min, p = 0.693) were comparable to those in another study [18]. However, in contrast to studies that focused on diverticulitis [12, 13], the operation time in the laparoscopy group was not statistically significantly longer than the open group. One reason might be that surgery for IC might have been more complex in the open group compared to laparoscopy group in spite of propensity score matching. Secondly, surgeons take more time to close a laparotomy compared to smaller laparoscopic surgical sites. Thirdly, all surgeries were performed by experienced hands in our institution. Based on our experience, we think laparoscopy should not prolong the operation time in the emergent cases and might even decrease the operation time in experienced hands.

Ischemic colitis is rare. The 2016 Premier Perspectives national inpatient database analysis indicated that only 12/945 (1.5%) laparoscopic colectomies were performed for noninfectious enteritis and colitis [38]. We agree with the authors that emergency laparoscopy has its place in emergency settings such as IC and there is a need to enlarge the indications [38]. Maggiori and Panis found that laparoscopic surgery for severe acute colitis was associated with a similar (or improved) short-term outcome compared with an open approach [39]. Sampietro et al. reported on 145 patients who underwent laparoscopic emergency subtotal colectomy for ulcerative colitis or Crohn’s disease stating that it was safe and feasible for acute severe colitis in IBD [40]. Our colectomy procedure is standardized, and all surgeons in our unit follow it. We believe that standardization of the surgical procedure could avoid mistakes especially when there are anatomical alterations and tissue inflammatory changes in the emergent setting. Finally, a laparoscopic second look is preferable to laparotomy. The timing of the second look is variable but usually should be within 72 h [41, 42].

While the causes that initiate ischemia may be variable, and even multiple, it is widely thought that these patients have vascular anomalies that enhance the onset: These can be constitutional or acquired. Among the former, there is no or a contentious connection between the middle colic and left colic arteries (Griffith’s point) in up to 48% of patients, or between the most distal sigmoid artery and the superior rectal artery (Sudeck’s point) in 5–15% of patients [43]. Previous colectomy or aortic surgery can further modify the vascular supply to the colon [43, 44]. Anastomosis after segmental colectomy for IC under these conditions might need specific maneuvers such as retroileal transmesenteric anastomosis (Toupet technique) or the Deloyers technique [45].

Previous studies have found various laboratory parameters to be associated with mortality, such as LDH > 450 U/l, blood urea nitrogen (BUN) (> 28 mg/dl), Hb < 12 g/dl, and hyponatremia (Na < 136 mEq/l) [46]. In our study, we used the APACHE II score to determine the general physical status of patients with IC. In multivariate regression analysis, APACHE II score (OR: 1.25, 95% CI 1.10–1.41, p = 0.001) was identified as an independent predictive factor for mortality. Furthermore, the APACHE II score model exhibited a high accuracy for the prediction of mortality, with an AUC of 0.86 (77.9–94.1, p < 0.001), comparable to that found by Peixoto et al. where the AUC was 0.89 [46]. In a large series of open colectomy, emergency surgery for IC was associated with high postoperative mortality [22]. In this study, preoperative lactates level, delay to surgery > 12 h, and the occurrence of postoperative acute kidney injury were independent predictors of postoperative mortality. Conversely, the specific cause of IC did not seem to impact postoperative mortality. These authors underlined the key role of prompt diagnosis and surgical intervention in the management of severe IC [22].

After propensity score matching, the conversion rate was 17/31 = 22.5%. The reasons for conversion were bowel distension, total colonic ischemia, diffuse fecal peritonitis, and/or severe adhesions. This is comparable to the literature on diverticular disease [13].

Our study has several limitations. Firstly, the sample size of this single institution case series was small (pre propensity, n = 96; post-propensity, n = 62). Secondly, as the decision to perform the operation laparoscopically was surgeon dependent, there may have been a selection bias even after propensity score matching. Thirdly, because of the retrospective design, it is not sure whether other factors may have influenced the postoperative outcomes, including intraoperative fecal spillage and bowel distension. Moreover, surgeon judgment intervened in the evaluation of the extent and degree of bowel ischemia to decide how much bowel should be resected and whether to perform primary anastomosis. In the future, intra-operative assessment by ICG might be an interesting avenue to explore [47]. Last, only known potential risk factors were matched. All confounding factors, those measured as well as those that are not measured, can only be eliminated by adequately conducted randomization. However, ischemic colitis is a rare disease, and therefore, a properly conducted and adequately powered randomized study would be difficult to perform. Finally, the inclusion time period of this study was long (2011–2020), and there have been many advances in surgical techniques and equipment, intensive care which may have impacted the surgical outcomes (such as operating time, postoperative complication rate, and hospital stay).