Introduction

The creation of a stoma for temporary or permanent deviation of the bowel is a common procedure in colorectal surgery. When used as a temporary solution, the intent is to reverse the stoma and close the aperture in the abdominal wall. This is usually performed a minimum of 6 months after the index surgical procedure in order to let the tissue and anastomoses heal properly. Nevertheless, many of these patients never undergo the reversal procedure. A permanent stoma may be an integral part of the index surgical procedure itself, or an active strategy decided preoperatively in selected groups of patients (i.e., elderly or frail), or a measure determined by situations related to the cancer or comorbidity.

Living with a stoma implies coping with changes in body appearance and functional ability, which demands several lifestyle adjustments [1,2,3]. Unfortunately, stomal complications are not unusual and may add to an already strained situation for the patient. Complications range from minor such as skin irritation and leakage, dehydration from high output stomas, and cosmetically poor results with difficulties finding clothes that fit, to more serious problems such as prolapse, bowel obstruction, and parastomal hernia causing recurrent episodes of abdominal pain and the risk of bowel strangulation. All these factors lead to impaired quality of life.

A parastomal hernia is an incisional hernia defined as a herniation of abdominal contents through the trephine in the abdominal wall alongside the enterostomy [4, 5]. The risk for developing a parastomal hernia has been found to be high. Some, in fact, would describe this complication as a more or less inevitable long-term outcome. In previous studies, the cumulative incidence of parastomal hernia varies between 5 and 50% [4, 6,7,8]. The diversity of results in the literature is partly due to differences in defining parastomal hernia. Distinguishing criteria between stomal hernia and prolapse are also lacking. Some studies define parastomal hernia based solely on clinical signs such as bulging adjacent to the stoma or palpable defect in the abdominal wall, while others complement this with a radiological definition [9,10,11,12,13]. Cohort size and follow-up period also differ considerably. It is known from studies on ventral hernia that the rate of herniation increases with time, which is probably also true for parastomal hernia. All in all, this makes it difficult to compare and estimate the true rate of parastomal hernia from previously published studies. Whatever the criteria used, clinically the most important hernias are those that cause complications, symptoms, or problems to the patient that warrant surgical repair [9].

Among risk factors mentioned in the literature are those related to the surgical technique such as the diameter of the aperture in the abdominal wall and the location of the stoma, through the rectus abdominis muscle or lateral to it [4, 14]. There are also patient-related risk factors, including old age, chronic respiratory disorder, corticosteroid use, obesity, wound infection, and malnutrition [15,16,17,18].

In this study on a large population-based cohort of adult patients with permanent stoma and with long-term follow-up, our aim was to assess the cumulative incidence of parastomal hernia requiring surgical repair or causing symptoms in patients who have undergone colorectal cancer surgery, and to identify risk factors that could predispose to the development of this type of hernia.

Materials and methods

Study design

Data were retrieved from the Swedish Colorectal Cancer Registry (SCRCR), originally the Swedish Rectal Cancer Registry (SRCR) founded in 1995, and since 2007 including the Swedish Colon Cancer Registry (SCCR). It is compulsory for every healthcare provider in Sweden to report patients diagnosed with cancer to the Swedish Cancer Registry (SCR) [19]. Together with the SCR, the SCRCR approaches 100% coverage or all patients diagnosed with colorectal cancer in Sweden [20,21,22]. The other database used was the National Patient Register (NPR) [23] to which all Swedish healthcare providers report. Since 1987, this registry has included all in-patient care, and from the year 2001 even outpatient visits to both private and public healthcare provider except visits to primary healthcare physicians, and has been shown to have a high reliability [23, 24]. The SCRCR and the NPR were cross-matched using the patient’s Swedish personal identity number: a 10-digit number unique to each resident [25]. The personal identity number makes it possible to follow every individual over time.

Inclusion and exclusions criteria

In this study, all patients who underwent surgery for colorectal cancer in Sweden between January 1, 2007 and September 3, 2013 were identified in the SCRCR. From this cohort, procedures that included a permanent stoma were selected. Data on gender, age, BMI, tumor staging based on TNM-classification [26], preoperative radiochemotherapy, and index surgical procedures were also obtained from the SCRCR.

Information on development of parastomal hernia during the postoperative period, identified by International Classification of Diseases codes (ICD-10 codes K43, K45, or K46 or procedure codes JAR10-81), was obtained from the NPR until November 7, 2014, which was the limit of the follow-up period. These specific ICD codes enabled us to include all known parastomal hernias causing symptoms great enough for patients to bring it to the attention of their physician. Whereas minor asymptomatic hernias would not have been detected, the vast majority of stomal hernias found at clinical examination or diagnosed by radiological imaging, as well as hernias requiring surgical treatment were registered in the NPR during the period of the study.

The NPR also provided all diagnoses from admissions and visits prior to the index surgical procedure. These were scrutinized for relevant comorbidity including cardiovascular disease, connective tissue disorders, liver cirrhosis, kidney failure, diabetes, chronic obstructive lung disease, and chronic inflammatory conditions, i.e. conditions mentioned in the literature as possible risk factors for developing a hernia.

Statistical methods

Cox proportional hazard analysis was performed to identify the impact of each risk factor and to estimate the cumulative incidence of parastomal hernia. The analyses were performed with IBM SPSS Statistics for Windows, Version 22.0.

Results

A total of 39,984 patients were registered in the SCRCR during the period 2007 until 2013. Of these, 7649 received a permanent stoma. Patients with data lacking on any variables in the multivariate analysis were excluded and the final study group consisted of 6,329 patients (Fig. 1). Baseline characteristics of the total cohort are shown in Table 1. There was a predominance of men receiving a permanent stoma at the index surgical procedure (4430 male patients (57.9%) compared to 3219 female patients (42.1%)). The mean age was 71.7 years. The follow-up time was equivalent to the registry period 2007–2013, with data from the NPR up to November 7, 2014. Altogether 3,276 patients (42,8% of the total cohort) died during the period of study. In Cox proportional hazard analysis, death was treated as a censored event.

Fig. 1
figure 1

Flow chart of cohort assembly

Table 1 Baseline characteristics

Some 65% of the patients had advanced stage tumors. Tumors infiltrating muscularis propria, i.e., T2, constituted almost 18% of the cohort. Categories T0-T1, corresponding to the least invasive tumors, accounted for 6.8% of the patients receiving a permanent stoma. In approximately 10% of the patients, the T-category was not known. Almost half of the patients had no spread to nearby lymph nodes and 72.5% had no distant spread.

Abdominoperineal resection (APR) was the most common surgical procedure in the cohort (43.9%). Anterior resection (AR) used for carcinoma situated in the upper or middle part of rectum where an oncologic safe distal margin could be achieved, leaving the rectal sphincter intact and allow for an anastomosis, was combined with a permanent stoma in 2.1% of the cases. Some 30.6% of the patients underwent resection of the sigmoid colon, in most cases registered as Hartmann’s procedure with a permanent stoma. Of the patients operated with resection of other parts of the colon (ascending, transverse, or descending colon), or when total colectomy was performed, 9.6% of cases had a permanent stoma. Of all surgical procedures resulting in a permanent stoma, 12.5% were performed as an emergency procedure.

In the statistical analyses, obesity (BMI ≥ 30) was found to be the only independent risk factor for developing a parastomal hernia. Multivariate Cox proportional hazard analysis showed a hazard ratio of 1.49, 95% CI 1.02–2.17, for BMI ≥ 30 (Table 2). None of the other potential risk factors, including TNM-categories, were statistically significant in predicting development of a stomal hernia. There was a slightly elevated hazard ratio of 1.36 (95% CI 0.96–1.91) for preoperative radiotherapy.

Table 2 Univariate and multivariate Cox proportional hazard analysis of risk for stomal hernia. The analyses were based on patients with complete data for all covariates included (N = 6329)

The cumulative incidence in this population-based cohort of patients diagnosed or surgically treated for parastomal hernia after a follow-up period of 5 years was 7.7%, 95% CI 6.1–9.2% (Fig. 2).

Fig. 2
figure 2

Cumulative incidence of parastomal hernia by BMI

Discussion

In the present study, the cumulative incidence of stomal hernia requiring surgery or causing symptoms was 7.7% over a 5-year follow-up period. Although this figure is lower than reported from previous studies using more sensitive methods and other criteria for determining the presence of stomal hernia [9, 27,28,29], the outcome in the present study focused on clinically relevant hernias.

Previous studies have shown obesity to be a risk factor for parastomal hernia [29, 30]. This should be considered when creating a stoma in obese patients. To some extent, the risk for future stomal hernia may be minimized by appropriate choice of stomal site, aperture diameter, and perhaps by reinforcing the stoma with mesh [31, 32].

As shown in Fig. 2, the incidence of stomal hernia requiring repair does not level off. This probably depicts the natural course, with pressure and distention of the tissues around the stoma. In the early period, tissues around the stoma usually retain their tensile strength. However, over time, the fascia becomes distended and loses its tensile strength resulting in distention of the tissues that progresses as time goes by [6].

The present study has some limitations. The primary endpoint was defined by the diagnoses and interventions registered in the NPR or SCRCR by the surgeons responsible for the patients. Although both registers are population-based and have national coverage, stomal hernias are only registered if the surgeon finds it relevant. This probably explains the relatively low incidence of stomal hernia seen in the present study. It is possible that there were some patients with a large stomal hernia causing problem who were not considered candidates for surgery due to comorbidity or because of anatomical conditions related to the hernia or previous abdominal surgery. Furthermore, no clear distinction between stomal hernia and stoma prolapse was made.

The SCRCR does not include data on measures taken to prevent the development of stomal hernia, and thus, we know nothing of what was done intraoperatively to reduce this risk. Type of stoma is not registered more specifically than stating whether it is intended for protective and temporary or permanent use, nor is there any information on the placement of the stoma site or the size of the aperture in the abdominal wall in the SCRCR. Also in this context, mesh reinforcement is of particular concern. Most of the procedures, however, were performed before mesh reinforcement became routine. Over the last decade, a large number of studies on mesh reinforcement have been published, with conflicting results [31,32,33,34,35,36]. In a Cochrane review based on 10 studies, the risk for stomal hernia was found to be halved if the stoma is reinforced with a mesh [37].

According to the European Hernia Society guidelines on stomal hernia prevention, the evidence regarding measures to prevent stomal hernias is insufficient [38]; thus, patient characteristics and specific preference should be taken into account when creating a stoma. Reinforcement with a mesh may be an option but should only be performed after assessing the specific circumstances in each case, including expected survival, the likelihood of stoma reversal, the risks associated with reoperation, and the risk of developing a stomal hernia. Whereas prevention of a stomal hernia may be high on the priority list of some patients, the side effects and risks associated with a permanent mesh may outweigh the potential benefit. Based on the results of the present study, obesity is one of the most important risk factors that must be considered when deciding on mesh reinforcement.

Conclusions

This population-based study showed that the cumulative incidence of clinically relevant parastomal hernia causing symptoms or requiring surgery after a follow-up period of 5 years was at least 7.7%. The only risk factor associated with stomal hernia was BMI ≥ 30. Our findings also suggest that the cumulative incidence increased with BMI over time, with higher risk for patients with a BMI greater than the median. In our opinion, this risk factor should be taken into consideration when deciding on measures to prevent stomal hernia, such as the use of mesh reinforcement.