Colorectal stenting in England: a cross-sectional study of practice

Abstract

Introduction

UK and European guidelines recommend consideration of a self-expandable metallic stent (SEMS) as an alternative to emergency surgery in left-sided colonic obstruction. However, there is no clear consensus on stenting owing to concern for complications and long-term outcomes. Our study is the first to explore SEMS provision across England.

Methods

All colorectal surgery department leads in England were contacted in 2018 and invited to complete an objective multiple choice questionnaire pertaining to service provision of colorectal stenting (including referrals, time, location and specialty).

Results

Of 182 hospitals contacted, 79 responded (24 teaching hospitals, 55 district general hospitals). All hospitals considered stenting, with 92% performing stenting and the remainder referring. The majority (93%) performed fewer than four stenting procedures per month. Most (96%) stented during normal weekday hours, with only 25% stenting out of hours and 23% at weekends. Compared with district general hospitals, a higher proportion of teaching hospitals stented out of hours and at weekends. Stenting was performed in the radiology department (64%), the endoscopy department (44%) and operating theatres (15%), by surgeons (63%), radiologists (60%) and gastroenterologists (48%). A radiologist was present in 66% of cases. Of 14 hospitals that received referrals, 3 had a protocol, 3 returned patients the same day and 4 returned patients for management in the event of failure.

Conclusions

All responding hospitals in England consider the use of SEMS in colonic obstruction. Nevertheless, there is great variation in stenting practices, and challenges in terms of access and expertise. Centralisation and regional referral networks may help maximise availability and expertise but more work is needed to support this.

Introduction

Colorectal cancer (CRC) is the third most common cancer with around 1.4 million new cases and 700,000 deaths per year worldwide, both on the rise.¹ CRC is diagnosed as an emergency presentation in around a quarter of patients despite presenting to primary care with relevant symptoms in just under half of these cases and with red flag symptoms in around a fifth in the preceding years.^2–7

Emergency presentation of CRC includes colonic obstruction, a surgical emergency with high risk of bowel perforation. Major emergency surgery in this scenario features a high rate of complications, high perioperative mortality and poor long-term outcomes.^8–12 Strategies employed to reduce the risk include temporary bowel stoma, intraoperative colonic washout and even subtotal colectomy with or without primary anastomosis. These options, however, have disadvantages, including living with a stoma, electrolyte loss and risk from a second operation.

Initial description in 1994 and the subsequent series on the use of self-expandable metallic stents (SEMS) in obstructing left-sided colonic lesions by Tejero et al offered a means to avoid urgent surgery, thereby enabling the options of planned, lower risk elective surgery or palliative relief.^12–16 In the palliative setting, contemporary studies and meta-analyses have demonstrated no difference in survival, complications or readmissions, and a higher rate of discharge to home and decreased length of hospital stay compared with surgery.^17–21

As a bridge to elective surgery in potentially curable patients, several meta-analyses have demonstrated higher primary anastomosis rates and lower rates of stoma formation.^22–28 Nevertheless, these studies do not comment on the logistical factors such as time from admission to stenting and the degree of perforation or change of decision to surgery while awaiting stenting. Furthermore, there is still debate on the most appropriate use and uptake of this treatment modality owing to the concern regarding stenting complications, tumour dissemination and long-term survival in patients with potentially curable disease.^22–31

In its 2014 guidelines, the European Society of Gastrointestinal Endoscopy (ESGE) did recommend SEMS as a standard treatment to bridge to elective surgery but stated that it can be considered in high operative risk patients and as the preferred treatment in palliative patients.³² Also in 2014, the UK National Institute for Health and Care Excellence (NICE) advised that either emergency surgery or colonic stenting are viable options but cited current evidence as inconclusive.³³

Despite these guidelines, there is a lack of universal consensus among surgeons regarding the use of stenting worldwide. This may be due to factors including individual surgeon or departmental practice, perceptions of colorectal stenting and access to colorectal stenting. The aim of this study was to identify the current application of and access to colorectal stenting in England in relation to the UK and ESGE guidelines.

Methods

The project was registered and approved by the institutional audit department. Hospitals in England with a colorectal surgery department were identified through the National Health Service website. Colorectal surgery leads were contacted via post in March 2018 and invited to complete a questionnaire. The questionnaire contained 12 objective multiple choice questions pertaining to service provision of colorectal stenting including referrals, time, location and specialty (Appendix 1 – available online). Returned responses were collated in July 2018 and anonymised. Data were entered and analysed in Excel^® (Microsoft, Redmond, WA, US) and graphics were generated in Prism^® (GraphPad Software, La Jolla, CA, US).

Results

Of 182 hospitals contacted, 79 (43%) returned the questionnaire. These comprised 55 district general hospitals (70%) and 24 teaching hospitals (30%). All 79 hospitals responded that they considered stenting; 73 (92%) performed stenting at their unit and the remaining 6 (8%) referred patients to another hospital for stenting. Table 1 summarises the responses to the questionnaire.

Table 1.

Questionnaire responses

	All (79)	District general hospital (55)	Teaching hospital (24)
Consider stenting	79/79 (100%)	55/55 (100%)	24/24 (100%)
Perform stenting	73/79 (92%)	50/55 (91%)	23/24 (96%)
Frequency
0–3 per month	68/73 (93%)	47/50 (94%)	21/23 (91%)
4–6 per month	4/73 (5%)	2/50 (4%)	2/23 (9%)
7–9 per month	1/73 (1%)	1/50 (2%)	0/23 (0%)
10+ per month	0/73 (0%)	0/50 (0%)	0/23 (0%)
Setting
Emergency	59/73 (81%)	40/50 (80%)	19/23 (83%)
Elective	55/73 (75%)	36/50 (72%)	19/23 (83%)
During normal hours	70/73 (96%)	49/50 (98%)	21/23 (91%)
Out of hours	18/73 (25%)	10/50 (20%)	8/23 (35%)
Weekdays	70/73 (96%)	49/50 (98%)	21/23 (91%)
Weekends	17/73 (23%)	9/50 (18%)	8/23 (35%)
Referrals
Referred with stenting protocol	3/73 (4%)	3/50 (6%)	0/23 (0%)
Referred with no stenting protocol	11/73 (15%)	2/50 (4%)	9/23 (39%)
Not referred	59/73 (81%)	45/50 (90%)	14/23 (61%)
Referred patients
Patients returned same day	3/14 (21%)	2/5 (40%)	1/9 (11%)
Patients not returned same day	11/14 (79%)	3/5 (60%)	8/9 (89%)
Failed stents managed at stenting unit	10/14 (71%)	4/5 (80%)	6/9 (67%)
Failed stents returned to referring hospital	4/14 (29%)	1/5 (20%)	3/9 (33%)
Stenting location
Endoscopy department	32/73 (44%)	22/50 (44%)	10/23 (43%)
Operating theatre	11/73 (15%)	10/50 (20%)	1/23 (4%)
Radiology department	47/73 (64%)	33/50 (66%)	14/23 (61%)
Stenting doctor
Surgeon	46/73 (63%)	31/50 (62%)	15/23 (65%)
Radiologist	44/73 (60%)	26/50 (52%)	18/23 (78%)
Gastroenterologist	35/73 (48%)	25/50 (50%)	10/23 (43%)
Radiologist
Present	48/73 (66%)	29/50 (58%)	19/23 (83%)
Not present	25/73 (34%)	21/50 (42%)	4/23 (17%)

Frequency

Among hospitals that performed stenting, the vast majority (93%) performed three or fewer stent procedures per month, with four (6%) performing four to six and one (1%) performing seven to nine procedures per month. No hospitals performed more than nine per month.

Timing

Stenting was undertaken during weekday normal hours in 96% of hospitals, with only a quarter of units stenting out of hours (25%) or at weekends (23%). Compared with district general hospitals, a higher proportion of teaching hospitals stented at weekends (35% vs 18%) and out of hours (35% vs 20%). Figure 1 shows the proportion of hospitals stenting in each setting.

Figure 1.

Timing of stenting in teaching and district general hospitals (A); and timing of stenting in hospitals with a radiologist present and those without a radiologist present (B)

Location

Almost two-thirds (64%) of hospitals performed stenting in the radiology department while just under half (44%) of stenting procedures were carried out in the endoscopy department and 15% in operating theatres. No other departments were used. The responses were not mutually exclusive: many hospitals would use more than one location for stenting.

Interventionalist

Stenting was most commonly performed by surgeons (63%), followed by radiologists (60%) and finally, gastroenterologists (48%). The procedure was carried out by more than one specialty in most hospitals. Radiologists were present during stenting procedures more often than not (66%). The presence of a radiologist was not associated with stenting out of hours (Fig 1).

Protocol for referrals

Among the 73 hospitals that stented, 14 (19%) received referrals from other hospitals to perform stenting. Of these, 3 (21%) had protocols for receiving these patients. In only 21% of units that received referrals were patients returned to the referring hospital on the same day, with the majority of hospitals keeping patients for more than a day. In the event of stent failure, most hospitals (71%) managed the patient at the stenting unit as opposed to returning patients to the referring hospital for subsequent management.

Discussion

Since this study was completed, NICE has released guidelines in 2020 on colorectal stenting based on an evidence review.³⁴ It advises that SEMS should be considered in palliative patients because of the high success rate, lower rates of stoma formation and a lack of difference in survival compared with the risks. On the other hand, it states that the evidence for SEMS versus emergency surgery in patients with potentially curable disease is less conclusive and recommends that both are offered.

In the largest study to date, CReST (ColoRectal Endoscopic Stenting Trial) investigated stenting for obstructing cancer in 246 patients, of which 92% were treated with curative intent.²⁸ The trial found no difference in one-year survival, quality of life or length of stay compared with surgery while reducing rates of stoma formation in patients bridged to surgery. Full longer-term results of the CReST study are currently unpublished but will make an important contribution in understanding the oncological outcomes and tumour recurrence.

This study is the first to explore SEMS provision across England. Colorectal stenting was considered by all responding hospitals, implying that SEMS has an appropriate and valuable role to play in managing malignant large bowel obstruction. Notably, our study found a wide variation in stenting practice across the country, highlighting the need for best practice guidance, which would aid access to, uptake of and expertise in colorectal stenting. This issue is echoed by the Getting It Right First Time programme, which aims to reduce variation and replicate best practice across the National Health Service.³⁵ We have identified key areas to address in order to improve access and feasibility, and to reduce length of inpatient stay and costs.

Low rate of procedures performed

The vast majority of hospitals stented fewer than four patients a month, which raises the question of individual and departmental expertise depending on volume. This may have an impact on appropriate patient selection, success and complication rates. There is an abundance of literature investigating the relationship between case volume and outcomes across a variety of operations, with a number of systematic reviews demonstrating volume–outcome associations.³⁶ This association is especially well supported in CRC, with 40 studies collectively demonstrating longer survival and lower rates of perioperative mortality, abdominoperineal excision and anastomotic leaks at high volume centres. This illustrates the impact of experience, volume and subsequent patient selection on improved outcomes.^36–38

Aside from faculty expertise, low volume procedures lack the same infrastructural support as more routine procedures, including support staff, equipment and access to facilities.³⁹ Without such infrastructure, access may be limited to only certain times when support is available. This is especially problematic given that obstructing cancers often present as emergencies.^2–6 Our results show that access to stenting is lower out of hours and at the weekends, especially in district general hospitals. A consequence of this is that during these times, patients may be limited to emergency surgery without the option of stenting.

Entitled ‘Take the Volume Pledge’, a 2015 campaign announced by three US hospitals called for minimum caseload requirements for a certain ten operations.⁴⁰ The campaign sparked controversy, and was met with widespread criticism and poor uptake, with opponents arguing that the regulations would limit access to care and increase travel burden for patients while doing little for improving surgical care in smaller, non-academic and rural hospitals.^41–43 Whether this applies to stenting remains to be seen.

Lack of standardisation regarding resources, personnel and equipment

Further logistical barriers are faced by hospitals in which stenting affects facilities used for other existing services. Fifteen per cent of hospitals in our study perform stenting in operating theatres with consequential impact on emergency operating lists. This hurdle may deter hospitals from offering stenting, especially in hospitals with already overburdened services.

Most hospitals in our study had a radiologist present during the procedure, as recommended in the NICE guidelines,³³ contributing again to the organisational challenge. In contrast, there is evidence that stents can be inserted safely without radiologist guidance.^29,44 Better training and assessing outcomes with or without a radiologist present may remove the requirement for radiological support and would potentially allow improved access to stenting. Our results do not show, however, that hospitals that perform stenting with a radiologist present are less likely to stent out of hours or at weekends.

Lack of referral pathways

A key area of interest of this study was referral practice of colorectal stenting. Large, specialised providers of a procedure yield the advantages of increased experience and expertise as well as a more developed infrastructure to support a higher availability and efficiency of the service.^39,43,45 Nevertheless, as a surgical emergency, patients presenting with malignant obstruction are sent to the nearest hospital regardless of the presence of a colorectal unit and must be referred on. Referral pathways and protocols are therefore central to the management of these patients.

Of the hospitals that did accept referrals, only a minority had a protocol for patient transfer, management and return although patients remained at most units for more than one day, and for further management in the event of failure. The lack of clear consensus towards stenting and the limited hospital caseload may contribute to the difficulty of creating a formal protocol.

Implementation and utilisation of a stenting protocol may help standardise treatment as well as improving access to stenting out of hours and at weekends. Furthermore, this urgent treatment option is delivered by colorectal surgeons, gastroenterologists and radiologists depending on the region. This alleviates the burden from any single specialty and suggests that a network of care could be implemented whereby referrals to another nearby hospital are made where the expertise is available.

In fact, implementation of regionalised networks linking hospitals without expertise to specialised centres has shown to increase access, and improves outcomes for stroke and cardiovascular events.^46–48 One study investigated outcomes following establishment of a regional referral network for left ventricular assist device as a bridge to transplant in patients with postcardiotomy cardiogenic shock.⁴⁸ The authors reported increased access to cardiac replacement options and increased survival in patients presenting to the network. This raises the possibility that referral networks may similarly benefit patients with CRC requiring emergency stenting as a bridge to surgery.

Study limitations

Limitations of this study include those related to cross-sectional surveys, notably response bias and limited view of the temporal dynamics of stenting practices.⁴⁹ Indeed, recent NICE guidance and upcoming results of large trials may impact future practice and provisioning of SEMS in the UK.^28,34 Our study did, however, objectify questions in order to quantify stenting practices and factors contributing to hospitals’ stenting practices (Appendix 1 – available online).

Our response rate of 43% limits the validity of the study through non-response bias and so our results may not be representative of departmental perceptions with regard to choosing stenting. Although this study did not identify any differences in characteristics or geographical location between hospitals that responded and those that did not, there may be differences in stenting practices, attitudes towards stenting and motivations. On the other hand, our response rate of 43% is not significantly different from the average response rate of mailed surveys of physicians (54–57%).^50,51 Nevertheless, our study does capture the logistical challenges encountered. Future studies using national databases to further investigate stenting practices and outcomes in the UK, as well as internationally, would be of great interest, for example identifying the number and proportion of patients treated using stents versus emergency surgery.

Conclusions

Colorectal surgeons in England consider SEMS as an acceptable alternative to major emergency surgery to treat left-sided colonic obstruction in line with NICE and ESGE guidelines. However, there is great variation in stenting practices, out of hours stenting and referral practices due to inherent logical challenges. In light of these issues, future work in formulating best practice guidelines and development of regional referral networks of care may help maximise availability and expertise to hopefully improve outcomes in patients who would benefit from SEMS.