All patients of the simple MI group (n = 18) were initially given a contrast enema under fluoroscopy. All four patients without CF and 3/14 CF patients (21%) were successfully treated with 1:1 diluted Gastrografin^® enemas. In 10 patients, Gastrografin^® enemas were unsuccessful, as was iohexol in another patient. None of the patients suffered from complications after the enema administration.

The 11 patients (all with CF) in whom contrast enema had failed underwent surgical intervention Table 3. RPA was performed in two patients, both had several complications as outlined in Table 3. Both required additional surgery because of ASBO: enterostomy in one patient and adhesiolysis in the other. The former suffered from 14 episodes of septicemia and died after 380 days of admission. RES was performed in five patients, one of them requiring RPA because of ASBO. Four patients were treated with PSI and irrigation with 1:1 Gastrografin^® enema, two of whom required enterostomy and additional adhesiolysis because of peritonitis following perforation. No differences existed between the treatment groups with regard to complications, time to complete enteral feeding, length of resection, initial hospital stay, or mortality rate.

Fifteen patients of the complex MI group, all without signs of perforation or calcifications, were initially given 1:1 Gastrografin^® enemas for diagnostic purposes. In one patient extravasation of the enema fluid was seen as an indication of perforation. All 23 patients underwent surgery; complications after surgery are outlined in Table 4. RPA was performed in 14 patients, three of whom required additional surgery: suture of perforation and enterostomy in two patients and adhesiolysis because of ASBO in one. Additionally, one patient suffered from a total of 11 episodes of septicemia and died after 431 days of admission. RES was performed in nine patients, two of whom required additional surgery: adhesiolysis for ASBO in one and exploratory laparatomy persisting signs of intestinal obstruction in the other. All patients of this group survived.

Patients with simple MI were initially treated with contrast enema. In our study, 1:1 diluted Gastrografin^® was safe and effective in resolving the obstruction in 39% of the patients with simple MI, which is in line with the published success rates ranging 5–83% [¹⁰–¹⁷]. The fact that all four patients without CF in this group were successfully treated may well reflect the fact that meconium in these patients is less viscous. Though a recent report suggests that repeated enemas might increase the success rate of the conservative treatment [¹⁸], we chose to reduce the risk of complications by not repeating an unsuccessful attempt.

In mice, Gastrografin^® was shown to be more efficacious for the relief of constipation than agents such as saline, Tween-80, Dnase and N-acetyl cysteine [¹¹]. Apart from saline, all agents reduced the viscosity of human meconium immediately after administration; of these, N-acetylcysteine reduced viscosity with more than 99%, but only after an incubation period of 6 h. The use of hyperosmolar agents such as Gastrografin^® is, however, not without risk. Perforation, hypovolemia, electrolyte loss, and hepatotoxicity have been reported [¹⁴, ¹⁸–²⁰] the risk of complications can be diminished by taking care of precautions such as diluting the solution with equal amounts of water, limitation of administration pressure, and the provision of intravenous access to maintain optimal fluid administration. Furthermore, contraindications including signs indicative of perforation should be identified with the help of X-ray studies. None of our simple MI patients suffered from complications, whereas in one complex MI patient leakage of contrast into the abdominal cavity occurred, apparently due to the pre-existing perforation sealed by the intestinal loops. Overall, the minimal invasive nature, the relatively high success rates and the low complication rates make Gastrografin^® enema the first choice in simple MI.

In case of failed contrast enema or complications in simple MI, resection with temporary double-barreled enterostomy seems to be the preferable approach. Severe complications related to the surgical procedure itself, such as peritonitis due to anastomotic leakage and perforation due to frequent manipulation, were more often seen in RPA and PSI. Decreased viability of severely dilated bowel in combination with the previous injection of hypertonic solution and manipulation during surgery aimed at advancing the meconium might underlie the greater complication rate following surgery without enterostomy. Moreover, the enterostomy in RES has the advantage that enemas can be applied after surgery in order to evacuate any remaining meconium, so that manual manipulation during surgery can be kept at minimum. The disadvantages of enterostomy, such as enterostomy care, the need for an extra surgical procedure, prolonged hospital stay, and the risk of high-output diarrhea do not outweigh the obvious reduced risk of anastomotic leakage and perforation.

Recent literature does not provide clear guidelines concerning the surgical treatment of simple MI. The Bishop–Koop enterostomy, introduced in 1957 consisting of an end-to-side anastomosis combined with a chimney- like enterostomy, has been widely used [¹³, ²¹, ²²]. It has the advantage of easy extraperitoneal, bed-side closure of the enterostomy. However, anastomosis is made, in the intraperitoneal cavity, leading to a higher risk of anastomotic leakage. For this reason, we prefer the double-barreled enterostomy. Although PSI is reported to result in reduced admission duration and a relatively low complication rate [¹³, ¹⁶, ²⁰], our data suggest that PSI is not without risk. The injection of hypertonic solution combined with squeezing and milking of the neonatal gut increases the risk of perforation, as is the case with suturing the ileotomy in tissue with impaired vascularisation. RPA, finally, has fallen out of grace for simple MI [²³], as the primary anastomosis of compromised intestine might be hazardous. Recent techniques include variants of enterotomy with irrigation and T-tube ileostomy [²⁴, ²⁵], but information on these techniques is limited.

In the complex MI group, surgical intervention is mandatory. In this group as well, resection with enterostomy in our hands was the safest treatment. None of the patients developed peritonitis, as compared to 21% of the RPA group. The reoperation rates, apart from closure of enterostomy, in RES and RPA were similar. Although RES creates the need of additional surgery to close the enterostomy, this is a safe, elective procedure. Complication rates between patients with and without CF were similar.

Conclusive evidence for the superiority of one or another procedure in complex MI has not been published. RPA has been used infrequently, mostly with atresia. Rescorla and Grosfeld [¹³] did not mention the complications while reporting on RPA in 13 out of 27 complex MI patients with atresia, volvulus or perforation. Escobar et al. [¹⁸] suggest that RPA should only be used in case of atresia, RES being preferable in complex MI patients. However, most studies report no significant differences between RPA and RES apart from duration of hospital stay, or focus only on one surgical procedure [¹³, ²⁴, ²⁵]. Jawaheer et al. [²⁶] report complications requiring additional surgery in 4 out of 13 patients (31%) after RPA, which is in line with our experience (complications in 21%). They consider RPA as a safe option for all complex MI patients apart from those with unstable condition and impaired perfusion of intestinal margins, the advantages being reduced hospital stay, avoidance of stoma-related morbidity and second laparotomy for stoma closure [²⁶]. However, impaired intestinal perfusion, is often present in complex MI patients, while the manipulation of neonatal intestine for the evacuation of viscous meconium might further impair blood supply and intestinal integrity. As it is hard to objectively assess the viability and condition of the intestine, we consider intra-abdominal anastomosis in these circumstances as hazardous. Fashioning of a double-barreled enterostomy requires less manipulation of the neonatal intestine and enables the application of enemas after surgery, while elective enterostomy closure is a safe procedure.