Colonic stenting: is the bridge to surgery worth its cost? A cost-effectiveness analysis at a single Asian institution.

Mechanical large bowel obstruction is a common surgical condition encountered in the acute hospital setting, and the most common etiology is colorectal malignancy [ 1 ]. Colorectal cancer is the third most common cancer diagnosed annually, with an annual incidence ranging between 25 and 50 in 100,000 [ 2 ]. It is estimated that up to a third of patients newly diagnosed with colon cancer—especially left-sided colon cancers—present with symptoms of large bowel obstruction [ 3 ]. Furthermore, 8% to 13% of colon cancer patients develop large bowel obstruction, which can lead to bowel ischemia and perforation if left untreated [ 4 ]. For patients presenting with acute left-sided colonic obstruction, the European Society for Gastrointestinal Endoscopy (ESGE) in 2020 recommended stenting as a valid alternative to emergency surgery [ 5 ]. Endoscopic colonic stenting aims to convert an emergency operation to a semi-elective procedure to mitigate the surgical and anesthetic risks of emergency surgery, as well as to improve the rates of laparoscopic resection [ 6 , 7 ]. Furthermore, stenting also seeks to avoid stoma creation during the definitive resection, thereby mitigating the psychosocial implications of having a stoma, as well as reducing the need and cost for subsequent stoma reversal, with its associated hospitalization [ 7 , 8 ]. As a bridge to surgery, successful colonic stenting increases the rates of laparoscopic surgery increases and primary anastomosis, while reducing postoperative complications [ 3 , 9 , 10 ]. Despite these positive results for colonic stenting, there remains doubt regarding the safety and efficacy of colonic stenting as a bridge to surgery due to its high reintervention rates [ 11 , 12 ]. A randomized controlled trial in 2008 had a disproportionately high number of perforations in the stenting group (6 out of 11 adverse events, vs. 1 adverse event in the surgical group), which led to the trial being stopped prematurely [ 13 ]. There have also been debates regarding oncological outcomes, as perforation may occur during stent placement, resulting in tumor spread and hence increasing the risk of local recurrence and reducing the overall survival rate [ 9 , 14 – 16 ]. In contrast, other long-term studies have reported that stenting as a bridge to surgery was associated with low recurrence rates, low complication rates, and overall acceptable oncological outcomes [ 7 , 9 , 17 – 19 ]. Aside from its psychosocial and oncological effects, colorectal cancer also represents an economic burden to countries in terms of healthcare costs and productivity losses [ 20 , 21 ]. Hospitalization costs represent a significant proportion of the economic burden due to prolonged hospitalization stays, with an average of 10 to 11 days estimated in an international retrospective analysis [ 22 , 23 ]. By avoiding emergency upfront open resection and stoma creation, allowing time to optimize the patient for laparoscopic surgery with its associated reduced duration of hospitalization, endoscopic colonic stenting can decrease both patient and healthcare costs. However, limited data have been reported on the cost-effectiveness of colonic stenting compared to emergency surgery in Asia; instead, stenting has generally been evaluated in either a European or American setting [ 12 , 24 ]. As data on healthcare costs are variable depending on the healthcare system, we aimed to evaluate the efficacy of colonic stenting as a viable bridge to elective surgery in Asia, as well as its cost benefits over emergency surgery to address malignant large bowel obstruction.

The study was approved by the SingHealth Centralised Institutional Review Board (No. 2020/2525). Written consents were obtained from patients participating in the study. We conducted a review of a prospectively collected colorectal cancer database of all endoscopically stented presumed-malignant large bowel obstructions at presentation at Sengkang General Hospital in Singapore between April 2019 and October 2022. We excluded patients who presented with metastatic unresectable disease or eventually declined surgery in favor of expectant management in the final analysis. Endoscopic, perioperative, and financial details for these patients were collected and analyzed. Outcomes were defined as technical and clinical success, as well as successful surgical resection. Technical success was defined as successful stent deployment across the colonic stricture with immediate establishment of luminal patency, evident by the endoscopic observation of the passage of proximal feculent material through the stent [ 25 – 27 ]. Clinical success was defined as a successful scheduled surgical outcome of definitive resection and primary anastomosis [ 25 – 27 ]. Clinical failure included technical failure, unplanned return to the emergency room, and failure of stenting in decompression requiring subtotal or total colectomy. Histopathological reporting of colon cancer was graded in accordance with the American Joint Committee on Cancer (AJCC) staging. Thirty-day morbidity from the stent and subsequent surgical procedure was graded according to the Clavien-Dindo classification, and mortality data were analyzed. Data on the surgical outcomes of patients who underwent emergency surgery for left-sided colonic obstruction were also collected. We performed a cost-effectiveness analysis on colonic stenting for acute colonic obstruction based on hospital financial data modelling and compared this against the modelled financial data of those who had upfront emergency surgery for malignant large bowel obstruction, including stoma closure. The costs of adjuvant treatment and its associated stoma-related complications were not included in the cost analysis. Colonic stenting procedures were either performed in the endoscopic suite with the patient under sedation or in the emergency operating theatre with the patient under monitored sedation with fluoroscopic guidance. All procedures were performed by 2 accredited endoscopists, with at least 1 specialist-accredited consultant colorectal surgeon. Following flexible sigmoidoscopy with an EVIS EXERA III GIF-1TH190 endoscope (Olympus) to visualize the offending stricturing lesion, a 450 Jagwire Guidewire (Boston Scientific) supported with a tandem catheter (Boston Scientific) was advanced through the lumen of the stricture under direct vision and fluoroscopic guidance. Contrast was then injected to estimate the approximate length of the lesion to aid in selecting the length of the stent. A Wallflex Colonic Stent (22 mm diameter with varying length [60/90/120 mm] and pliability [regular/soft]; Boston Scientific) was passed through the narrowed lumen after exchanging the stent over the guidewire. Deployment took place under fluoroscopic and endoscopic guidance. Technical success was evaluated at the time of endoscopy, and clinical success was achieved when the planned interval surgical resection was performed. In addition to clinical parameters, all patients had an abdominal x-ray performed within 24 hours after the procedure to ensure that the deployed stent remained in a satisfactory position with adequate expansion after the procedure. After successful stent insertion, elective surgery was planned for within 2 weeks as recommended by the ESGE guidelines [ 5 ]. Categorical data, such as patient demographics, were presented as proportions or percentages while continuous data were presented as median values with interquartile ranges (IQRs). Cost-effectiveness analysis was performed by obtaining the mean costs between a patient who would have undergone emergency surgery and those who would have undergone colonic stenting as provided by our finance office. The incremental cost-effectiveness ratio was calculated. For those who underwent upfront emergency surgery, cost modelling considered surgical costs, including the index operation with its related hospitalization costs (including high-dependency stay), as well as stomal consumables, stoma nursing support, and the associated stoma closure with its related hospitalization costs. For those who underwent colonic stenting, cost modelling included the endoscopic procedural costs and the costs of the stent, definitive surgery, and all its associated hospitalization.

During the 42-month study period, 40 patients presented with presumed-malignant large bowel obstruction and underwent endoscopic stenting at our institution. Of these, 29 patients had colonic stenting performed as an eventual bridge to definitive surgery, as shown in Fig. 1 . The other 11 patients did not undergo definitive resection due to metastatic disease and comorbidities. Additionally, 1 patient is still on neoadjuvant chemotherapy for systemic control of disease prior to definitive resection. Table 1 demonstrates the demographics and disease factors of patients undergoing colonic stenting. The median age of patients presenting with acute left-sided colonic obstruction was 66 years (IQR, 57–71 years), and most of them were classified as American Society of Anesthesiologists (ASA) physical status class II at baseline. They were symptomatic from the obstruction for a median length of 2 days (IQR, 3–7 days). Based on computed tomography, the offending lesion was most frequently located in the sigmoid colon (19 patients, 66%), followed by the descending colon (8 patients, 28%). The median length of the stented lesion was 5.5 cm. Twenty-six (90%) out of the 29 patients who underwent colonic stenting achieved technical success, while 24 patients (83%) achieved clinical success and underwent elective resection. Of the patients who did not achieve technical success, 2 patients (7%) had perforation and none had stent migration. Both patients who had immediate perforation on the table during colonic stent insertion presented with a relatively long duration of symptoms, 7 and 11 days, respectively. In the other patient who did not achieve technical success, stenting failed as the wire was unable to be passed through the obstructing lesion and the decision was made for emergency surgery. Two other patients did not achieve clinical success despite initial technical success, resulting in a total of 5 patients (17%) not achieving clinical success. One patient’s surgery was delayed due to COVID-19 infection, and he presented with delayed perforation 17 days after colonic stenting, while another experienced re-obstruction despite technical success, requiring emergency surgery. The majority of patients with colonic stents had 6-cm-long stents inserted (23 patients, 79%), followed by 9-cm-long stents (5 patients, 17%). The last patient had two 6-cm stents deployed in an overlapping technique to traverse a long stricture. Table 2 summarizes the stenting and surgical outcomes of patients who underwent colonic stenting, while Table 3 summarizes the difference in surgical outcomes between colonic stenting and emergency surgery for left-sided colonic obstruction. In patients who achieved technical and clinical success, the median time to surgery was 13 days (IQR, 9–17 days). Twenty-one patients (72%) underwent laparoscopic surgery. Eight patients (28%) required open surgery. Colonic stenting had failed in 5 of these patients, and the other 3 had undergone successful colonic stenting but required open surgery due to adhesions from prior operations (2 patients) or bleeding (1 patient). Of the 29 patients, 2 presented with benign causes of intestinal obstruction—1 with endometriosis resulting in a left-sided colonic fibrotic stricture, and the other with benign fissuring ulcers secondary to segmental colitis associated with diverticular disease. All 27 patients who presented with a malignant cause of intestinal obstruction had primary colonic cancer. A majority (17 patients, 63%) presented with stage II disease; all had at least T3 disease and above, with invasion of the muscularis propria, according to the AJCC staging system [ 28 ]. Twenty-four patients (83%) had primary anastomosis without stomas, while the other 5 patients (17%) required stoma creation. Of these 5 patients with stomas, colonic stenting had failed in 3, while the other 2 required a stoma in view of low anterior resection. The 30-day morbidity rate from surgery was 14% (4 patients), while the 30-day mortality rate was 0% (0 patients). Of note, no anastomotic leakage was encountered. Hence out of the 24 patients who achieved clinical success, 21 successfully underwent laparoscopic surgery, and 22 did not require a stoma creation. In comparison, all patients who underwent upfront emergency surgery required open surgery (63 patients, 100%), while for colonic stenting, only 8 patients (28%) required open surgery. Patients who underwent emergency surgery were also more likely to require a stoma (67%) than those who underwent colonic stenting (17%). A large majority (24 patients, 83%) stayed for 1 day (IQR, 1–1 day) in the high-dependency unit after definitive resection of their colonic lesion, and a median of 7 days (IQR, 5–13.5 days) in the general ward. Table 4 shows the median postresection hospital stay. Most patients who underwent colonic stenting (21 patients, 73%) underwent elective resection without stoma creation, while the patients who underwent emergency surgery were most likely to undergo resection with a stoma (42 patients, 67%), with a further 62% of these patients (26 patients) undergoing stoma reversal. An example of the cost breakdown of colonic stenting as compared to emergency surgery is demonstrated in Table 5 , and the overall costs of colonic stenting as compared to emergent surgery based on surgical outcomes are summarized in Table 6 . Based on these projections, cost-effectiveness and estimated savings were calculated. The expected cumulative cost of management with colonic stenting from acute management to definitive treatment calculated was $32,900. This included the initial hospital stay after colonic stenting, as well as the subsequent elective admission for definitive surgery, with a median of a 1-day stay in the high-dependency unit and a total of 10 days of general ward stay based on our hospital’s financial records. The expected cost of emergency surgery, including eventual stoma reversal, was $40,700, which included the costs of stoma supplies, the 2 procedures, a 1-day stay in the high-dependency unit, and a median of 12 days in the general ward. In total, this resulted in an estimated healthcare cost-savings of $7,800 per person, which were mainly due to the avoidance of upfront emergency surgery with stoma creation. The incremental cost-effectiveness ratio was 0.81, demonstrating that colonic stenting carries significant healthcare cost-savings for patients. The overall difference in the duration of hospitalization for management with colonic stenting over upfront emergency surgery was 2 days, favoring stenting. Fig. 2 demonstrates the outcomes and costs of patients who underwent colonic stenting as compared to emergency surgery.

Our overall complication rate from colonic stenting was 10%. These patients presented with early perforation (2 patients, 7%) or delayed perforation (1 patient, 3%) during endoscopic colonic stenting, and none (0%) presented with stent migration. The overall technical success rate was 90% (26 patients), and the clinical success rate was 83% (24 patients). Reasons for failure included perforation (3 patients, 10%), re-obstruction (1 patient, 3%), and technical difficulty (1 patient, 3%). The predictors of outcomes for colonic stenting include preprocedural factors (e.g., the duration of symptoms), procedural factors (e.g., tumor location, tumor length, and stent diameter), and postprocedural factors (e.g., length of time to surgery) [ 29 – 31 ]. Longer symptom duration is a known risk factor for technical failure, and the patients who presented with immediate perforation on table had symptoms for longer than the median duration (7 and 11 days, respectively) [ 26 ]. A meta-analysis in 2021 found overall rates of technical (92%) and clinical (82%) success that were comparable to our technical (90%) and clinical (83%) success rates [ 31 ]. The ESGE guidelines recommend that colonic stenting be performed by experienced endoscopists, noting that studies have shown an increase in technical success after performing 20 endoscopic stenting procedures [ 5 ]. At our institution, colonic stenting was carried out only by experienced endoscopists, with at least 1 other colorectal surgeon on standby. This is likely one of the factors accounting for our high technical and clinical success rates. Our institutional data demonstrate that colonic stenting is safe, with comparable technical and clinical success rates to those of other institutions. With proper patient selection and available specialists, colonic stenting is a safe and effective bridge to surgery. Patients with unsuccessful stent procedures had a higher rate of open procedures, which could be attributed to the failure to successfully decompress the bowel, making laparoscopic work unfeasible. When patients present with obstruction and are not optimized for emergency surgery, they often have longer postoperative recovery, with poor surgical outcomes (e.g., respiratory complications and readmissions) and an elevated risk of surgical complications [ 32 , 33 ]. By stenting an obstructed lesion, we convert these patients to elective surgery and hence enable preoperative optimization [ 6 , 33 ]. This can be demonstrated by the short stay in the high-dependency unit, as well as the shorter hospitalization stay (about 3 days overall). Our overall complication rate was also acceptable, at approximately 14%, with 10% of patients having Clavien-Dindo III to IV complications. Emergency surgery with resection has been reported to show a high rate of anastomotic leak, at approximately 15% [ 34 ]; notably, however, our study did not have any patients who presented with anastomotic leak. Fast recovery from surgery is also essential, as patients present with obstruction, which is a poor prognostic factor; hence, adjuvant chemotherapy would be recommended [ 35 ]. Aside from preoperative optimization, an elective laparoscopic resection also avoids wound complications from an open midline incision, shortens the hospital stay, and hence allows the earlier initiation of postoperative adjuvant chemotherapy [ 36 ]. A large proportion of our patients who underwent colonic stenting as a bridge to surgery achieved clinical success and were able to proceed for definitive resection and subsequent chemotherapy. Our analysis found that patients who presented with acute left-sided obstruction and proceeded for emergency resection could expect an initial hospital bill ranging from US $39,100 to $46,100. This corresponds with findings in the literature, where a systematic review [ 37 ] reported the initial management costs for colorectal cancer could vary US $7,893 to $60,289, while another systematic review [ 38 ] described a similar range of USD $19,929 to $67,195. As our patients presented with obstructed colorectal cancer with a need for longer postoperative high-dependency monitoring, it may be expected that their healthcare bill would be on the higher end of the spectrum. Our study found that, as a bridge to surgery, colonic stenting offered a cheaper alternative, with an initial hospital bill of US $6,500. Overall, the hospital bill after colonic stenting and definitive resection is expected to range from US $28,400 to $52,900. The cumulative costs are US $32,900 owing to our low complication rate. This translates to a difference of approximately US $7,800 per patient. Furthermore, colonic stenting also reduces both the temporary and permanent stoma rates and mitigates the psychosocial effect of stoma creation [ 25 , 39 ]. Furthermore, the incremental cost-effectiveness ratio was found to be 0.81, favoring colonic stenting as compared to emergency surgery and suggesting that healthcare resources should be directed towards stenting where possible. These findings are concordant with various studies in Italy and Canada, which reported that colonic stenting is a cost-effective and safe alternative to emergency surgery [ 39 , 40 ]. To the best of our knowledge, no cost analysis on colonic stenting has been performed in Asia prior to our study, and our paper is the first to demonstrate that colonic stenting in an Asian institution is a safe and cost-effective procedure that reduces healthcare costs and burden. The healthcare costs of intestinal obstruction and emergency surgery translate into a heavy strain on patients, their families, and the healthcare system. With these savings, physicians should consider colonic stenting where appropriate in the management of an obstructed large bowel with left-sided pathology. There are some limitations to this study, including selection bias as patients who were suitable for colonic stenting generally had better prognostic factors, such as no bowel ischemia or perforation. Furthermore, our single institution has a relatively small number of patients who underwent colonic stenting. More data from a larger, multicenter multinational cohort study would be required prior to making definitive conclusions with regards to the cost-efficacy of colonic stenting. Due to the acute presentation of our patients with large bowel obstruction, a randomized controlled trial may not be feasible due to the inherent risks associated with both procedures and ethical considerations. Hence, we were unable to make a comparison between stenting and 1-stage surgery where stoma formation can be avoided through colonic washout intraprocedurally. In patients with left-sided colonic obstruction, stenting as a bridge to surgery is a safe, clinically effective, and cost-effective means to treat acute colonic obstruction with high success rates and low complication rates, while avoiding open emergency operation. In the future, more data will be needed to determine the optimal timing from stenting to resection, as well as the utility of colonic stenting prior to neoadjuvant chemotherapy.