Introduction

Although primary anastomosis at emergency neonatal laparotomy has been recently advocated [1, 2], enterostomy formation remains the widely accepted approach in neonates where there is intestinal compromise requiring resection [3]. Despite this, there is no clear consensus regarding the optimal site for stomas in relation to the laparotomy wound. Stomas can be created within the laparotomy wound or through a separate incision, allowing complete laparotomy wound closure. This decision is based primarily on surgeon’s preference. It is suggested that stomas through a separate incision offer the advantage of a lower infective complication rate and avoidance of a full laparotomy at the time of closure. Previous studies report a stoma complication rate of up to 42 % [4] but few have correlated this complication rate with the position of the stomas in relation to the laparotomy wound [5, 6]. The aim of this study was to relate the location of neonatal stomas to complications and need for full laparotomy at the time of stoma closure.

Materials and methods

All neonates who underwent formation of an enterostomy at emergency laparotomy between January 1999 and December 2009 in a single institution were identified from a prospectively collected neonatal admissions database. Those who had a stoma formed for anorectal malformation were excluded as these stomas were created through a planned left iliac fossa wound rather than via an emergency laparotomy. Patients with Hirschsprung’s disease were only included if they underwent an emergency laparotomy rather than a planned leveling stoma. A retrospective review of the case notes was performed to identify demographic details, type of stoma and relationship to laparotomy wound, stoma and wound-related complications and clinical details pertaining to the stoma closure. All patients were treated at a single institution.

Patients were separated into two groups: stomas created through the laparotomy wound and stomas created through a separate incision. Wound infection was defined on an intention to treat basis, that is, if antibiotics were commenced due to clinical concerns regarding the wound such as erythema, discharge or dehiscence. End points were wound-related complications, stoma-related complications and need for a full laparotomy at time of closure. Differences between groups were analysed using the Mann–Whitney test for continuous variables, and Chi-squared test or Fisher’s exact test for categorical variables.

Results

During the 11 year study period, 196 neonates underwent enterostomy formation but 70 patients that had a colostomy for anorectal malformation were excluded. Original case notes for a further 20 patients were unobtainable. In total, case notes for 106 patients were reviewed, although 7 patients had stomas formed on 2 separate occasions, giving a total of 113 laparotomies with stoma formation.

Stomas were created within the laparotomy wound in 71 cases, of which 39 had stomas placed side by side in the lateral edge of the wound (Fig. 1). Stomas were created through a separate incision in 42 cases with the majority (35 patients) having stomas placed side by side. An ileostomy was the most common level of stoma in both groups [54/71 through wound (76.1 %) vs. 34/42 separate wound (80.9 %), p = 0.54] shown in Table 1. Necrotising enterocolitis (NEC) was the most frequent indication for laparotomy in both groups [34/71 through wound (47.9 %) vs. 17/42 separate wound (40.5 %), p = 0.44], with other diagnoses shown in Fig. 2. There were no significant differences between the two groups with respect to gestation, birth weight and age or weight at surgery (Table 2).

Fig. 1
figure 1

Site of stoma formation

Full size image
Table 1 Comparison of patients with stomas placed through the laparotomy wound and those with stomas created through a separate incision with respect to the level of the stoma
Full size table
Fig. 2
figure 2

Diagnoses requiring stoma formation

Full size image
Table 2 Patients’ details at birth and time of laparotomy where stomas were created
Full size table

Total complication rates, wound-related complications and stoma-related complications were not significantly different between the two groups (Table 3). The most common stoma-related complication in both groups was prolapse. The group with stomas sited through a separate incision appeared to have an increased likelihood of complications requiring further surgical intervention, although this did not reach statistical significance.

Table 3 Complications in each group
Full size table

Operative data related to enterostomy closure were obtained for 100 surgical procedures as nine patients died prior to closure and four patients with Hirschsprung’s disease underwent pull-through at the time of closure and were excluded from analysis. There were 62 enterostomy closures in patients with stomas created through the wound and 38 closures where stomas had been sited in a separate incision. There was no difference between the groups in age at time of closure (median age 98 days through wound vs. 92 days separate incision, p = 0.9) or body weight (median weight 3.59 kg through wound vs. 3.89 kg via separate incision, p = 0.53). There was no statistically significant difference between the proportion of patients requiring closure for high-output stomas.

A contrast study to assess the distal limb prior to closure was performed 89 patients (the reasons for the remaining 11 % not having one were unclear). Of the 62 patients with stomas brought out through the laparotomy wound, 58 (93.5 %) had a contrast study prior to closure with a stricture detected in 13 patients (22.4 %). Of 38 patients with stomas through a separate incision, 31 (83.8 %) had a contrast prior to closure, with 4 (12.9 %) revealing a stricture. There was no statistically significant difference between the groups with respect to the number having a loopogram prior to closure or the proportion of strictures identified (p = 0.336 and p = 0.398, respectively). All patients who had a stricture demonstrated on contrast went on to have a full laparotomy at closure with re-opening of the original laparotomy wound.

There were a total of 75 patients in which no stricture was suspected or detected on contrast. Of those with stomas through the wound, 37.8 % (17/45) were closed with simple circumexcision but 60 % (27/45) required extension to the full laparotomy wound. Of those with stomas through a separate incision, 56.7 % (17/30) were closed with circumexcision alone, with 20 % (6/30) required reopening of the original laparotomy wound and 23.3 % (7/30) had extension of the stoma incision, resulting in two parallel abdominal wounds.

If all configurations were included, there was a trend towards a higher rate of full laparotomy in the group where stomas had been sited through the wound [43/62 through wound (69.4 %) vs. 20/38 separate incision (52.6 %), p = 0.09]. If stomas had been sited adjacent to each other at the original surgery, there was no difference in the ability to circumexcise the stomas and avoid a full laparotomy at the time of closure [Fig. 3, 19/33 through wound (58 %) vs. 18/31 separate incision (59 %), p = 0.97].

Fig. 3
figure 3

Method of closure (where stomas were formed adjacently at the original laparotomy)

Full size image

Discussion

Despite the enterostomy formation being a common surgical procedure performed in neonates, there are few published series that consider the optimal site or configuration of stoma placement, and the impact this has on complications and subsequent stoma closure. This presented series of neonates with stomas formed at emergency laparotomy did not identify a significant difference in stoma- or wound-related complication rates in neonates with stomas placed through the laparotomy wound compared to those who had their stomas placed through a separate incision.

Other studies have shown a correlation between complications regarding gestational age and birth weight [4] but our groups were well matched in this respect. In a previous study looking at a similar cohort of patients in 1987, Musemeche et al. [5] reported an overall wound-related complication rate of 13 %. This earlier report did not establish any relation in wound-related complications and site of stomas, in accordance with our findings. Our overall wound complication rate was only 4 %, which may reflect under-reporting of minor wound infections, use of prophylactic antibiotics, improvements in standard neonatal care and nutrition, or the beneficial effect of recently introduced specialist nurses in stoma care. Our stoma nurse specialists are involved in the early postoperative stage, in conjunction with the neonatal surgical nurses, to ensure correctly fitted stoma appliances, and our low infection rates are testament to their diligence.

Our study revealed a much lower overall stoma-related complication rate of 14 % compared to published rates of 20–45 % [4, 5]. We are confident that, despite the retrospective nature of this study, we have identified all stoma-related complications as our general and specialist neonatal nursing staff prospectively record any problems in the case notes, and patients were all treated in a single institution. Previous studies have advised caution in the timing of stoma closure [7, 8], but there was no difference in the median age at closure in our groups. No previous studies have correlated stoma complications in the neonate with the position of the stomas in relation to the laparotomy wound, but our impression from informal discussions has been that surgeons feel that stomas sited through a separate “fresh” wound are less susceptible to further complications. Many surgeons believe that the overspill of stoma effluent from stomas sited through the laparotomy wound may lead to wound-related complications, but this was not confirmed by our data. In contrast to a newly formed wound at the end of the procedure, the edges of the laparotomy wound will have been open for the duration of the case and may have been exposed to intra-operative contamination and retraction. This may confer a theoretical benefit to the avoidance of complications such as parastomal hernia or prolapse through increased fibrosis around the exteriorised limbs of the bowel.

At the time of neonatal stoma reversal, there are various options available to the surgeon. These include circumexcision of the stoma avoiding a formal laparotomy (following a satisfactory distal loopogram), re-opening of the previous laparotomy wound or formation of a new laparotomy wound. Reasons for requiring a full laparotomy may include pre-existing knowledge of distal bowel pathology; surgeon’s preference to fully inspect the bowel; or technical difficulties in mobilisation during attempted circumexcision. Advocates of placing the stomas through a separate incision purport that there is a higher chance of achieving closure through circumexcision alone, and our data do suggest a trend towards avoiding a full laparotomy with this configuration, despite not reaching statistical significance. However, in the group with stomas through the original wound, there was a higher rate of distal stricture requiring full laparotomy, and some stomas were sited to incorporate a wide skin bridge that required re-opening of the laparotomy wound to mobilise. In patients with stomas formed through a separate wound, if a full laparotomy is required during closure, the surgical options are to extend the wound around the stomas or reopen the original wound. Both options result in two parallel wounds on the abdominal wall of varying lengths in different dermatomes, which may create challenges in providing effective postoperative analgesia and less satisfactory cosmesis. In our series, where a full laparotomy was required at the time of closure, no patients with stomas originally formed through the wound required an additional separate incision to the original wound but more than half the patients with stomas formed separately had two parallel incisions. In this series, if stomas were sited adjacent to each other, there was no difference in the ability to circumexcise the stomas and re-establish intestinal continuity without a full laparotomy. Again, in the group where stomas were brought out through the original wound, a separate wound was avoided.

The limitations of this study rest mainly in its retrospective, non-randomised nature, preventing robust conclusions from being made. In addition, it is important to accept that the placement of the stomas may have been influenced by the anatomy and pathology at the time of surgery, as well as the surgeon’s personal preference, and this information is not often obvious from the operation note.

We advocate that, if feasible, stomas should be placed side by side through the laparotomy wound. This practice does not increase the risk of postoperative complications compared to forming stomas through a separate incision and offers equal benefit of circumexcision alone at the time of closure, with the option of a full laparotomy without the need for an additional wound.