Stoma reversal in functional bowel disease: managing patient choice

Abstract

Introduction

This study aims to assess the rate of stoma reversal in patients who have undergone stoma formation with permanent intent for functional bowel disorder. We also assessed the incidence of malignancy in defunctioned bowel.

Methods

This is a retrospective study of the outcomes of patients undergoing planned permanent stoma formation for functional bowel disorder over a 10-year period at a single tertiary centre.

Results

Of 92 patients included in the study, 11 (12%) requested and underwent stoma reversal following stoma formation for functional bowel disorder. Of 84 patients with defunctioned bowel left in situ, two (2%) developed bowel malignancy during our study period.

Conclusions

Stoma formation may be necessary for patients with incontinence and constipation when conservative treatments fail. Some patients have very firm views about the need for a stoma, but a significant proportion of patients do not cope with a stoma and request reversal, therefore patient selection and pre-procedure counselling are important. The risk of developing malignant disease in the defunctioned colon is potentially significant, and consideration should be given to appropriate surveillance in this group of patients. Evidence for stoma formation in functional gastrointestinal disorders is lacking; this study reports outcomes in a large cohort of patients over a long period, and highlights areas where further research and practice guidelines are needed. If large numbers of patients are undergoing stoma reversal posing significant mortality and morbidity risks, this suggests that patient selection and preoperative counselling need refinement.

Introduction

Functional gastrointestinal disorders (FGID) can be broadly defined as a collection of abdominal symptoms including pain, bloating, stool irregularity, diarrhoea and constipation, where no biochemical or structural cause can be found for the symptoms.¹

FGID are very common; a recent epidemiology study reports that one in four people in the UK, USA and Canada fulfil the criteria for a diagnosis of functional bowel disorder according to the Rome IV criteria.² Conservative measures of FGID include dietary modifications, pharmacological management and biofeedback therapy.³ Conservative management of FGID is not always successful and in a small percentage of patients with intractable symptoms (less than 1%), stoma formation is performed.^4–6

Sphincteroplasty, sacral nerve stimulation or posterior tibial nerve stimulation, artificial bowel sphincter, muscle transposition and stoma formation are options for managing faecal incontinence.³ Colonic resection, rectal suspension procedures, rectal excision procedures, rectovaginal reinforcement procedures and sacral nerve stimulation are recognised surgical techniques for management of chronic constipation.⁷ Evidence for the effectiveness of stoma formation in FGID however is weak, and there is no firm consensus regarding the indications for surgery in these patients, and very little data on stoma reversal rates. We performed literature searches on Medline Ovid and PubMed which found 16 papers^4–6,8–20 reporting on stoma formation in adults as a management strategy for functional bowel disorders including five review articles,^8–12 ten studies^{4–6,13–19} and one case report.²⁰ Stoma formation appears particularly successful in management of FGID after spinal cord injury^15,16,19 and neurogenic bowel.¹⁷ There were no papers reporting stoma reversal in patients who had undergone stoma formation for functional bowel disorders.

The aim of our study was to investigate what proportion of patients who have stomas formed with permanent intent for the treatment of FGID subsequently undergo stoma reversal. This is an important clinical question, as a high proportion of patients undergoing stoma reversal may indicate that stoma formation was not the most appropriate treatment option, or that patient counselling for stoma formation was inadequate. This study may therefore help to define the role of stoma formation in complex FGID, and to inform changes to practice or further areas of research which could lead to improved outcomes for these patients.

As an important secondary aim, we also investigated the incidence of colorectal malignancy related to defunctioned bowel left in situ; this is clinically relevant because of the difficulties in carrying out surveillance of defunctioned colon and rectum, potentially leading to delayed diagnosis.²¹

Methods

A prospective departmental database was used to identify all patients who underwent stoma formation for functional bowel disease from 2009 to 2019 inclusive. Data were collected retrospectively by reviewing patients’ clinical records (both electronic and paper-based), and nursing documentation to assess their outcomes of stoma formation up to April 2020. Data collected included demographic data, indication for surgery, the type of operation performed (ileostomy or colostomy), complications of surgery, whether the stoma was reversed and the reason for reversal. We subdivided our study group into three subgroups: constipation, incontinence or mixed symptoms of both. We also identified any main contributory factors for symptoms within the categories.

The decision to offer colostomy or ileostomy formation was based on discussions between individual patients and the clinician and was not influenced in any way by this purely observational study. In general, the practice at our unit during the period concerned has been to offer ileostomy for chronic constipation and colostomy for faecal incontinence, however there are exceptions based on the clinical needs of the individual patient. The type of stoma offered was also influenced by previous resections and remaining bowel in situ.

We included patients with FGID only. Patients who had a diagnosed organic pathology that could contribute to their symptoms were excluded. We did not exclude patients with simple diverticular disease, however we did exclude those with evidence of malignancy, active infection, inflammation, stricturing or fistulating disease or those undergoing emergency surgery at the time of proposed stoma formation. We did include patients who had had previous bowel resections as long as they were now deemed free of organic bowel pathology.

Statistical analysis

Categorical data were analysed using Fisher’s exact test (GraphPad software) and continuous data analysis was carried out using Student’s t-test (Microsoft Excel).

Results

Of the initial sample of 97 patients, five were excluded (see Figure 1). Three patients had stoma formation during emergency surgery, one patient did not undergo stoma formation despite considering it and one patient had simple diverticular disease which progressed to complicated diverticulitis.

Figure 1 .

CONSORT diagram representing the patient group followed up

The mean duration of follow-up was 3 years 9 months. Median follow-up was 3 years 6 months (range 2 months to 10 years). Over the course of the study, eight patients were lost to follow-up due to non-attendance at follow-up appointments or moving outside the area (9%). For these eight patients, mean follow-up was 2 years 5 months.

There were missing data for one patient only, including their operation date and other interventions for functional bowel disorder. This was dealt with by including the patient in the analysis for only the known parameters.

Over a 10-year period, 92 patients underwent stoma formation for functional bowel disorder. The age range of the study population was 23–89 years. The mean age was 55 years and the median age was 52 years. Among the study population, 85% were female and 15% were male. Ileostomy formation was undertaken in 47 patients, and 45 underwent colostomy formation.

Neurological disorder (n = 24), previous gastrointestinal resection (n = 16) and pelvic organ prolapse/pelvic floor incompetence (n = 10) were the most common main contributory factors for symptoms of FGID (Table 1). However, there were no physical contributory factors identified in the majority of our group (n = 26).

Table 1 .

Main factors contributing to functional gastrointestinal disorders symptoms in our group

Contributory factor for stoma formation	Constipation	Incontinence	Mixed	Total (n = 92)
Unknown	16	9	1	26
Neurological disorder	10	11	3	24
Previous gastrointestinal resection	4	11	1	16
Pelvic organ prolapse/pelvic floor incompetence	5	4	1	10
Previous pelvic surgery	2	2	2	6
Previous anal cancer chemoradiotherapy	0	3	0	3
Obstetric trauma	0	2	1	3
Previous anal surgery	0	2	0	2
Multiple perianal abscesses	0	0	1	1
Obesity	0	1	0	1

The mean period from initial referral to stoma formation was 41 months (range 2 months to 11 years 4 months). The median was 2 years 8 months. There were missing data for eight patients. Some patients progressed quicker to stoma formation as they had attempted conservative measures without success in other hospital locations prior to referral. Ten patients underwent stoma formation within 6 months of initial referral. Seven of these patients required stoma formation for neurological disorders.

Primary outcome: rate of stoma reversal

From the study population, 11 of 92 patients (12%) underwent stoma reversal (Table 2). We analysed the stoma reversal group further and found that 5 of these 11 patients (45%) then had their stoma re-formed. A second stoma reversal procedure was then performed in four of the five patients who had a second stoma formation procedure. Two of the four had a second stoma reversal and coped without a stoma.

Table 2 .

Analysis of 11 patients who underwent stoma reversal and their outcomes

Patient number	Indication for stoma formation	Type of stoma prior to reversal	Reasons for reversal	Outcome after reversal	Stoma re-formation	Further stoma reversal
1	Incontinence Previous GI resection	Ileostomy	Patient intolerance	Symptoms controlled with loperamide	No	N/A
2	Constipation Unknown	Colostomy	Patient intolerance Parastomal hernia	Symptoms uncontrolled with conservative measures	Yes. Loop ileostomy formation	Yes. Patient intolerance
3	Incontinence Neurological disorder	Colostomy	Patient intolerance	Symptoms controlled with ACE	No	N/A
4	Constipation Pelvic organ prolapse/pelvic floor incompetence	Ileostomy	Patient intolerance	Medical management of constipation but had chronic abdominal pain	No	N/A
5	Constipation Neurological disorder	Ileostomy	Patient intolerance Skin irritation Parastomal hernia Rectal mucus retention Additional ACE helped symptoms	Incontinence	Yes. Ileostomy formation	No. Kept stoma
6	Incontinence Pelvic organ prolapse/pelvic floor incompetence	Colostomy	Patient intolerance Rectal discharge	Symptoms resolved	No	N/A
7	Constipation/chronic pain Previous GI resection	Ileostomy	Patient intolerance	Ongoing chronic pain	Yes. Second ileostomy formation	Yes. Patient intolerance. Had third ileostomy and completion proctocolectomy
8	Incontinence Obstetric trauma	Ileostomy	Patient intolerance Stoma bag leakage	Symptoms resolved	No	N/A
9	Constipation Pelvic organ prolapse/pelvic floor incompetence	Colostomy	Rectal ulcer healed	Biofeedback and rectal irrigation	No	N/A
10	Incontinence Pelvic organ prolapse/pelvic floor incompetence	Colostomy	Patient intolerance	Symptoms returned	Yes. Colostomy twice	Yes. Total stoma (n = 3) Third stoma kept
11	Mixed (constipation predominant) Unknown	Ileostomy	Patient intolerance	Had appendicostomy formation	Yes. Ileostomy with bag leakage	Yes. Continues to cope without stoma

GI = gastrointestinal; N/A = not applicable; ACE = antegrade continence enema

Patient 10 had complex mental health illness and had two colostomies formed and reversed, one prior to referral to our centre. Biofeedback, medical management of functional bowel disorder and a sacral nerve stimulator were attempted before a third and final colostomy formation with antegrade continence enema.

Of the 11 patients (12%) who had stoma reversal (Table 2), the primary reason for stoma reversal was patient intolerance in 10 cases (91%). In the remaining case, stoma reversal was undertaken because the rectal ulcer had healed, and this patient subsequently achieved good control of their constipation with biofeedback and rectal irrigation.

In addition to the primary reasons given for stoma reversal, the following were also deemed contributory factors: skin irritation (n = 2), significant stoma bag leakage (n = 1) parastomal hernia (n = 2) and rectal discharge (n = 1).

There was no statistically significant difference in age or sex between the stoma non-reversal and reversal groups. The mean age for patients who did not have their stoma reversed was 52 years and the mean age for the 11 patients who did have their stoma reversed was 55 years (p = 0.265). Of the 11 patients undergoing stoma reversal, two (18%) were male compared with 12 of 81 patients (15%) not undergoing reversal (p = 0.6723). There was no significant difference in reversal rates between the constipation and incontinence groups (p = 0.5007 Fisher’s exact test) or between ileostomy and colostomy groups (p = 1.00 Fisher’s exact test) (Figure 2).

Figure 2 .

Comparison of stoma reversal between original symptoms and type of stoma formation subgroups

In summary, there was no significant difference between the stoma reversal and non-reversal groups in terms of age, gender, original indication for stoma formation or type of stoma formed.

In the immediate postoperative phase prior to discharge 14 of the 81 patients (17%) in the stoma non-reversal group suffered complications including high-output stoma (n = 5), infection (n = 5), small bowel obstruction (n = 1), ileus (n = 1), stoma stenosis (n = 1) and pulmonary oedema (n = 1). In the stoma reversal group, 2 of the 11 patients (18%) suffered from complications, namely intraoperative small bowel injury (n = 1) and ileus (n = 1). The complication rate was comparable between the stoma non-reversal group (17%) and the stoma reversal group (18%) (see Figure 3).

Figure 3 .

Number of complications and further procedures encountered by our cohort

Secondary outcome: rate of malignancy detected in defunctioned bowel

After initial stoma formation for FGID, 8 of the 92 patients had further bowel resection to remove their remaining defunctioned bowel. This took place in the form of panproctocolectomy for six patients and completion proctectomy for two patients who had had previous colonic resection and anastomosis formation.

In the study population, 84 of 92 patients (86%) had no further bowel resection following stoma formation. Of 84 patients with remaining defunctioned bowel in situ, two (2%) went on to develop colorectal or anal cancer. Neither patient had any bowel resection post stoma formation. The first patient had ileostomy formation for constipation and was diagnosed with adenocarcinoma of the transverse colon 4 years afterwards. The second patient also had ileostomy formation for constipation and developed dyssynergia, bleeding and difficulty with rectal evacuation that led to a diagnosis of anal squamous cell carcinoma 5 years and 9 months after stoma formation.

Discussion

Our study presents long-term follow up data for a large cohort of patients undergoing stoma formation for functional bowel disorders over a 10-year period. Data are reasonably complete for our cohort of 92 patients and our loss to follow-up rate was 9% with a mean follow up of 2 years 5 months in those lost to follow-up.

Stoma formation has been demonstrated as an effective way to manage functional bowel disorder in a small proportion of patients.⁴ Stoma formation complications including necrosis, prolapse, peristomal infection or abscess, retraction, stenosis, parastomal hernia and fistula have been reported as up to 25% with a high incidence of mortality observed in patients who developed stoma-specific complications.²² In stoma reversal, one study reports mortality in 15 of 533 patients (3%) in the first 30 days postoperatively. The surgical complication rate was reported at 20% with the most frequent being wound infection and anastomotic leakage.²³ Our overall complication rates (18%) were similar to these.

Of 92 patients in this study, 81 (88%) did not request stoma reversal, suggesting that stoma formation does have a role to play in the management of FGID that is refractory to less-invasive therapies. This percentage is similar to the 81% of patients who chose to keep their stoma for management FGID in another study.⁴ Another study investigating patient satisfaction with colostomy for faecal incontinence also reported that 84% of patients would ‘probably’ or ‘definitely’ choose to have a stoma again but there were a few patients who had did not adapt to and ‘intensely disliked’ their stoma.⁵

Age, gender, indication for stoma, type of stoma or complications during the immediate postoperative recovery period did not seem to influence stoma reversal in our study population. This could indicate that the optimal type of stoma was chosen in the majority of cases, which was generally ileostomy for constipation and colostomy for incontinence where possible. Alternatively, it could indicate that the type of stoma does not influence the overall success in managing FGID.

No patients (0%) who had stoma formation for symptoms of FGID contributed to by previous pelvic surgery (n = 6), multiple perianal abscesses (n = 1), previous anal cancer chemoradiotherapy (n = 3), previous anal surgery (n = 2) and obesity (n = 1) requested stoma reversal. This may be due to the small numbers of patients in each of these categories. The numbers of patients requesting stoma formation in the remaining groups were 4 of 10 (40%) in the pelvic organ prolapse/pelvic floor incompetence group; 2 of 16 (13%) in the previous GI resection group; 2 of 24 (8%) in the neurological disorder group; and 2 of 26 (8%) in the unknown contributing factor group. The relatively low reversal rate in the neurological disorder group suggests stoma effectiveness in FGID in this group and appears to be in keeping with other studies.^15,16,19 The high rate of reversal (40%) in the pelvic organ prolapse/pelvic floor incompetence group may indicate that conservative measures or other gynaecological interventions should be exhausted prior to embarking on stoma formation in this subcategory.

In the 11 patients who underwent stoma reversal, three (27%) ended up with a permanent stoma suggesting that stoma formation was likely the treatment modality to control their FGID symptoms but their adaptation to having a stoma was poor. It is also possible that preoperative counselling did not adequately prepare these patients for managing a stoma. One patient (9%) opted for reversal after their rectal ulcer healed suggesting stoma formation was potentially therapeutic in this case or at least not required on a permanent basis. Two patients (18%) had suboptimal FGID symptom control after reversal but did not request a further stoma formation. Five patients (45%) achieved symptom control with other treatments or reported resolution of their symptoms after reversal. This suggests that they may not have truly needed stoma in first place, and that conservative therapies could have been further explored.

Arguably there is a small portion of patients at referral who will not engage fully with non-surgical symptom management as they perceive stoma formation as the only solution to their symptoms. These cases are difficult to manage and may indicate an unrealistic patient expectation. They also do not constitute a true indication for stoma formation. Temporary stoma without resection has been proposed^4,9 and may have a role to play in these select cases. Failure of stoma acceptance was a strong contributing factor for stoma reversal in 10 of the 11 cases (91%) and in five cases (45%) this led to multiple stoma formations and reversal which were arguably unnecessary and exposed the patients to increased risk of mortality and morbidity as well as utilising extra time and resources. All 11 patients (12%) who had stoma reversal had attended preoperative counselling with a colorectal stoma nurse specialist. Only one patient (9%) had met another patient with a stoma. Psychological counselling resources are limited and of the nine patients who were offered a referral for psychology counselling, only four (44%) accepted. Improving preoperative counselling and psychological input may better identify those for whom stoma formation may not be appropriate or those who may benefit from stoma formation but require additional individual psychological support and stoma training in the perioperative period.

Additional preoperative counselling and standardised patient selection criteria may help to better control patient expectations and reduce the number of patients requesting a subsequent stoma reversal. This could include a severity score for symptoms, exhaustion of all feasible conservative and less-invasive measures with use of established constipation and incontinence pathways, at least two preoperative counselling sessions with a stoma nurse, a requirement for psychological consultation and a requirement for meeting another patient with a stoma. However, it is acknowledged that meeting another patient with a stoma may be of limited value as selected patients are likely to be those with a satisfactory outcome and quality of life.

In addition, two patients went on to develop cancer in their remaining defunctioned bowel. The main advantage of leaving defunctioned bowel in patients with functional bowel disorder is that it allows stoma reversal at a later date if required. Defunctioned bowel, however, will not produce the same set of symptoms when malignant change has occurred, and common screening tests such as faecal occult blood or FIT are not appropriate.²¹ There are also technical difficulties with administering bowel preparation to perform colonoscopy in patients with an ileostomy. A recent study reported that 50.9% of colorectal consultant surgeons in the UK had encountered malignancy in defunctioned bowel and bowel screening is not offered to 72.2% of age-appropriate patients with defunctioned bowel.²¹ The incidence of colorectal cancer in defunctioned bowel is poorly reported,²¹ however cases of malignancy in this cohort (2%) highlight the need for this population to undergo screening for colorectal cancer, and for clinicians to be vigilant regarding the potential for malignant disease in these patients. There is insufficient evidence to offer resection of defunctioned bowel routinely. Because the number of patients with defunctioned bowel is not very large, it may be reasonable to offer colonoscopy as first line screening for malignancy in this group.²¹

Study limitations

The study limitations include single centre non-randomised cohort and a relatively small population size. We chose to carry out retrospective data collection as prospective data collection would have been impractical. Although we investigated our findings over a 10-year period, we have not carried out a power calculation for optimal sample size. Our sample size is comparable with another cohort reporting outcomes in stoma formation for FGID.¹ We have also reported stoma reversal rates as a primary outcome, although this is likely to have considerable limitations if used as a surrogate marker for effectiveness of treatment. Ideally, validated quality of life and symptom scores would be recorded preoperatively and at fixed points in the patients’ follow-up. However, this was not standard practice in our unit over the period studied, and it is not possible to gather this information retrospectively. A prospective study could allow collection of this information; however, the relatively small numbers of eligible patients would likely make this impractical outside of a multicentre study.

It should also be noted that treatment of complex FGID has evolved over the period studied, and therefore the results may not be fully generalisable to other centres or to current practice. However, we report important data regarding outcomes following stoma formation as a treatment for FGID which we believe are valuable in informing current practice and identifying areas for further study.

Conclusions

Stoma formation is a viable treatment option to improve symptoms of functional bowel disorder which has been refractory to other less-invasive treatment options. This study highlights some of the difficulties inherent in the surgical management of functional gastrointestinal disorders, namely performing surgical procedures on patients with complex and refractory symptoms with a limited evidence base and no clear guidelines. The complication rate from surgery in this study is significant (18%), and the relatively large proportion of patients requesting stoma reversal (12%) and subsequently tolerating FGID or managing FGID by other methods (9%) suggests that this has not been an effective treatment for all cases. This reinforced the need for clear treatment algorithms for refractory FGID, incorporating lifestyle measures, medical therapies, nurse-led interventions (biofeedback, rectal irrigation), physiotherapy and surgical options within a multidisciplinary framework. Over the 10 years covered by this study, the care of patients with FGID has evolved and many centres including our own do now incorporate these measures. However, there is a need for more robust evidence to further inform decision-making in these challenging disorders.

Furthermore, defunctioning stomas make early recognition of malignant transformation in the defunctioned bowel more challenging, therefore a high degree of vigilance is required, and patients should be offered screening for colorectal cancer that is tailored to their needs.