- Top of page
Aim: To review outcomes of gastroschises originating in North Queensland and repaired in the neonatal centre in Townsville from 1988–2007, and compare these outcomes with published data from other centres.
Methods: A retrospective chart review of outcomes after primary operative repair (POR) with wound closure in fascial layers in the theatre, primary non-operative repair (PNOR) with apposition of the umbilical cord and adhesive dressing in the ward, secondary repair (SR) after a silo, and complex repair (CR) of cases with obstruction, perforation or atresiae. Epidemiological data of babies originating in North Queensland but managed in Brisbane were secured from Mater Mothers' Hospital and the Royal Hospital for Children.
Results: Fifty cases were treated in Townsville: 16 by PNOR, 22 by POR, 6 by SR and 6 by CR. Outcomes of uncomplicated cases treated by PNOR, POR and SR were similar. Outcomes of complicated cases after CR were significantly worse. No significant differences were found between PNOR in Townsvillle and published outcomes after SR from four overseas units. No differences were detected in outcomes of inborn and outborn babies, those delivered by vaginal or caesarean birth or between indigenous and non-indigenous babies. Three (6%) died. Gastroschisis is increasing in North Queensland, particularly in younger mothers. The incidence is 2–3 times higher in indigenous mothers.
Conclusion: No significant differences were found between PNOR and other techniques for uncomplicated cases. PNOR avoids the use of theatre and staff, reducing delay and cost, and the need to transport a sick baby.
- Top of page
Gastroschisis is a congenital defect in the anterior abdominal wall through which abdominal organs protrude into the amniotic fluid, unrestricted by any covering membrane. The full thickness defect is usually to the right of a normally inserted umbilical cord from which it is separated by normal skin.1,2 After birth, protruding bowel is prone to vascular compromise and infection and is a potent source of heat, fluid and electrolyte loss. Replacement should occur as soon as possible but the relative size of the organs may preclude their return to the abdominal cavity without incurring the risk of compartment syndrome because of increased intra-abdominal pressure. Associated abnormalities in the bowel such as matting between the loops, perforations, atresiae and underdevelopment, increase the complexity of early management. Later management may be complicated by obstruction, perforation, necrotising enterocolitis and short bowel.
Surgical management includes primary or secondary repair (SR). In the former, the bowel is returned to the abdominal cavity soon after birth. In the latter, it is covered by a protective silo and returned gradually. Cases complicated by atresia, perforation or obstruction may need resection, anastomosis and the creation of stomata.
This review reports 20 years of experience of surgical management of gastroschisis in the tertiary neonatal centre for North Queensland, Townsville, which serves a widespread population of about half a million people of whom 9.4% are indigenous, and has an annual birth rate of about 10 000 live births. It compares that management with published data from other sources.
The incidence of gastroschisis is rising in North Queensland, especially in younger mothers.
The outcome of treatment of uncomplicated cases by primary, non-operative repair in the ward is similar to that by primary operative repair in the theatre, and to secondary repair with silos.
Gastroschisis complicated by atresiae, obstruction and perforation has the worst outcome.
- Top of page
A retrospective chart review investigated the surgical management and outcome of all cases of gastroschisis treated in the neonatal intensive care unit in Townsville from January 1988 to December 2007. Data were collated concerning the mother (age, residence, month of conception, race, parity), the delivery (date, mode, site), the baby (birth weight, gestational age, associated abnormalities), the mode of surgery, post-surgical management (time on ventilator, peak ventilatory inspiratory pressure, peak concentration of inspired oxygen, urine output), time to start first feeds, time to reach full feeds, time on parenteral nutrition, length of stay), complications (infections proven by positive blood culture, re-operations, necrotising enterocolitis) and mortality.
The modes of surgery were defined as (i) primary operative repair (POR) which involved taking the child to theatre, replacing the abdominal contents and suturing the wound in fascial layers, under general anaesthesia with intubation and ventilation, (ii) primary non-operative repair (PNOR) which occurred in the ward where the bowel was replaced and the defect closed by apposition of the umbilical cord and adhesive dressing in a manner similar to that described by Sandler et al.,3 under pain relief from morphine and with the child intubated and ventilated by neonatal staff, (iii) SR which was employed when the abdominal contents could not be replaced without risk of compartment syndrome, and a protective silo was fitted (initially constructed from a ‘bile bag’ and then supplied by Bentec Medical, Woodland, CA, USA) to allow staged reduction of the bowel over a period of about a week, when the defect was closed by suturing fascial and skin layers under general anaesthesia in the operating theatre. The decision to employ SR was made at the initial assessment and SR was not employed after the failure of an earlier mode of therapy. The final mode of repair was complex repair (CR) which was employed in cases complicated by bowel matting, perforation or atresiae, needing resection, anastomosis and or stomata.
Primary non-operative care was utilised by the surgeon when simple reduction was possible, otherwise SR was employed. In the PNOR groups, pain relief was provided by intravenous morphine and the child was intubated with atropine and suxamethonium, ventilated, and paralysed with pancuronium if extra relaxation was needed.
There was no standard protocol for the administration of intravenous fluids. Most babies were commenced on 80 mls/kg/day with boli of normal saline given in doses of 10 mls/kg for clinical suggestion of hypoperfusion.
In the second decade of the study, reflecting a more conservative and disciplined approach, oral feeds were usually started 2 weeks after surgery at about 2 mls per hour increasing incrementally in uncomplicated cases with the aim of establishing full feeds within the following 2 weeks. Feeds would be stopped for about a week after bilious aspiration. Feeds would not be stopped after aspiration of clear fluids or unaltered milk but would be continued with a ‘reflux tube’ in which the end of the orogastric tube would be connected to a plungerless 10 mL syringe attached to the crib at the level of the baby.
Demographic data of nine babies who originated in North Queensland but were treated in Brisbane, the capital of Queensland, were included in epidemiological analysis but not analysis of management or outcome. That data were secured from The Mater Mother's Hospital and the Royal Hospital for Children. Analysis of perinatal demography of gastroschisis is the substance of another study.
The region of North Queensland comprised the Health Service Districts of Mackay, Townsville, Mt Isa, Cairns and Hinterland, and Torres Strait. No demographic data were available on miscarriages, terminations or stillbirths.
Delivery of babies with gastroschisis was encouraged to occur in the maternity department in Townsville to avoid post-natal transportation. Spontaneous vaginal delivery was employed unless there were obstetric contra-indications.
Indigenous or non-indigenous status was designated by the mother on admission. Indigenous status included Aboriginal and Torres St Islander mothers. Data on birth rates in North Queensland and their stratification for maternal age and indigenous status were supplied by Health Statistics Centre, Queensland Health.
Full feeds was defined as the oral intake of 150 mL/kg/day of milk. Ethical approval was granted by Townsville Health Service District Institutional Ethics Committee.
Data were analysed using Intercooled Stata Version 9.1 (StataCorp, College Station, TX 77845, USA). χ2 and t-test were used where appropriate. P-value < 0.05 was accepted as significant. Data on outcomes in Townsville were compared with published data from international centres.4–7
- Top of page
There were 50 cases of gastroschisis treated in Townsville in the study period, 16 (32%) were treated by PNOR, 22 (44%) by POR, 6 (12%) by SR with a silo and 6 (12%) by CR. There were no statistical differences between PNOR, POR or SR regarding time on a ventilator, ventilatory pressures, maximum concentration of inspired oxygen, urinary output and mean arterial pressure. The time from initiation of feeds to the gaining of full feeds was statistically longer with PNOR (13.5 vs. 7.7 days with POR, P= 0.018), otherwise time to start feeds, time on parenteral nutrition, length of stay and mortality rate were similar. One patient developed necrotising enterocolitis and three needed re-operations for obstruction after PNOR. There were no cases of necrotising enterocolitis but five needed re-operation after POR. After CR, the time to feed, the time to reach full feeds, the duration of total parenteral nutrition (TPN) and the lengths of stay were significantly longer than those in other forms of repair (Table 1).
Table 1. Comparison of outcomes after primary, secondary and complex repair
| ||Primary non-operative||Primary operative||Secondary with silo||Complex|
|Number of patients||16||22||6||6|
|Days on ventilator||3.7||4.7||6.0||2.9|
|Age at 1st feeding||14||16.9||24||37.5|
|Age at full feeding||27.5||24.6||33||65.2|
|Number with central line infections||5||6||1||4|
|Days of TPN||22.3||23.6||29.5||65.2|
|Length of stay||34.4||29.7||53||83.7|
|Mean arterial pressure||48.6||44.6||45||48|
|Max fraction of inspired oxygen FiO2||0.55||0.5||0.44||0.46|
The time to reach full feeds was significantly longer in the second decade of the study (7.3 vs. 15 days. P= 0.04) but time to start feeds, time on parenteral nutrition, length of stay and infection rates were similar.
Twenty-eight (28) cases (56%) were surgically uncomplicated. Regardless of their mode of closure, their times to feed, length on parenteral nutrition, and length of stay were all statistically shorter, though the infection rate was similar to complicated cases (Table 2).
Table 2. Outcome of uncomplicated compared with complicated cases
|Days on TPN||18||43||0.0003|
|Days on ventilator||4||4.8||0.46|
|Age at 1st feeds||12||28||0.0004|
|Age at full feeds||22||44||0.0005|
|Length of stay||28.6||55||0.002|
Fifty-nine cases of gastroschisis were known to have originated in North Queensland in the study period, generating an overall rate of 3.2 cases per 10 000 live births. The incidence increased during the study period, particularly in mothers under 20 years of age. From 1988–1992 to 2003–2007, the overall incidence increased from 0.7 to 4.7 per 10 000 live births. In mothers under 20 years of age it increased from 2.40 to 19.2 per 10 000 live births. The overall rate in indigenous mothers was 2–3 times higher than in non-indigenous mothers (6.6 vs. 2.6 per 10 000 live births. OR 2.5: 95% CI 1.4–4.5, P < 0.001). The average age of indigenous mothers was 21.7 years compared with 22.5 years in non-indigenous mothers which was not statistically different. There was no difference in outcome of indigenous and non-indigenous babies.
Outborn babies were more premature than inborn: (34.1 vs. 36.4 weeks. P < 0.001), but parameters of feeding, length of stay, infection and mortality were the same. No difference was noted in outcome of babies born by spontaneous vaginal delivery compared with caesarean section. In 5 of the 10 outborn babies, diagnosis was made at birth. The others had been diagnosed antenatally but had been unable to be transferred in utero to Townsville because of obstetric considerations. All babies were retrieved by a specialist retrieval team. Their being outborn did not influence the form of management.
Three cases died leading to an overall mortality rate of 6%. None died after PNOR, 2 after POR, and 1 following SR. One died from associated pulmonary hypoplasia and one from withdrawal of life support because of extensive atresia and short bowel. One died from multi-organ failure associated with complex atresia and repeated obstructions.
Compared with published data, little difference was detected between outcome after PNOR in our institution and after SR in several international centres (Table 3).
Table 3. Outcomes after primary, non-operative repair in Townsville, compared with published data of secondary repair in other centres
| ||Primary non-operative (Townsville)||Silo†||Silo‡||Silo§||Silo¶|
|Number of patients||16||26||20||13||21|
|Days on ventilator||3.7||1||3||4||0|
|Age at 1st feeding||14||11||NA||12||12|
|Age at full feeding||27.5||19||25||27||31|
|No with central line infections||5||3||0||2||3|
|Days on TPN||22.3||27||NA||21||30|
|Length of stay||34.4||28.5||29||28||36|
- Top of page
In primary repair of gastroschisis there is the risk of ‘compartment syndrome’ after bowel has been returned to a restricted abdominal cavity. This complication might be revealed by the need for increased ventilatory pressures and concentrations of oxygen,8 and by alterations in cardiac output as monitored by arterial pressure. The rise in intra-abdominal pressure might reduce urinary output, and might impede vascular supply of the bowel, resulting in increased time to initiation and completion of feeds, duration of parenteral nutrition, development of necrotising enterocolitis and other complications, and by an increase in length of stay and mortality.
Because of the risks of compartment syndrome, SR is the preferred management in various centres throughout the world.4 Little difference, however, could be found between parameters of outcome after primary closure of uncomplicated cases in Townsville and secondary closure in other centres4–7 suggesting no increased rate of compartment syndrome. Our findings accord with other recent studies comparing primary and secondary closure.9,10
Comparing POR and PNOR within our unit, no significant differences in outcome were found. The time to reach full feeds was greater after PNOR. Both changes of management (to PNOR and slower progress of feeds) began at the end of the first decade of our study.
Advantages of PNOR in the ward included independence from the operating theatre and staff thus reducing delay and cost, and avoidance of destabilisation from transportation. Neonatal staff were able to stabilise the patient before and after the procedure and to reduce the surgeon's time to an estimated 30 min per procedure.
We have traditionally intubated and ventilated our patients, and provided analgesia with morphine without recourse to inhalational anaesthesia, but, according to Davies et al., intubation and substantial pain relief is unnecessary.11 Conversely, we have found it occasionally useful to paralyse the infant to facilitate bowel replacement.
The determining factor for outcome in our study was the absence of complications such as perforation, atresia and obstruction, as confirmed elsewhere.12,13 Uncomplicated cases progressed more rapidly to full feeds and discharge though their infection rate confirmed by positive blood cultures was similar. Outcome was not affected by mode or place of delivery or by indigenous status.
Not surprisingly, CR of intestinal atresia, perforations and obstructions that required excision, anastomoses and or stomata was followed by longer periods for feeding and lengths of stay in hospital.
We detected no difference in outcome between outborn and inborn cases reflecting, we believe, close cooperation between referring and referral staff to minimise exposure and vascular compromise of protruding bowel. We also detected no difference in outcome between babies delivered vaginally or by caesarean section, in accordance with other data from Australia.14
As in many other countries, the incidence of gastroschisis is increasing in North Queensland, particularly in younger mothers. Though no racial predisposition has been found in other countries, gastroschisis is over twice as common in indigenous mothers in North Queensland, as it has been recorded in the rest of Australia. The reasons for this are not clear but are presumed to reflect a prevailing lifestyle in younger mothers, associated with reduced economic status, exposure to vaso-constricting drugs, change of sexual partners, and peri-conceptual infections.15 Our overall rate in the 20 years of the study was higher than that described nationally (3.2 vs. 2.7 per 10 000) but the rate increased to 4.7 in the last 5 years of the study, considering both indigenous and non-indigenous mothers, and to 9.3 in the former. In indigenous mothers under 20 years of age, in the period 1998–2002 it rose to 34.4 per 10 000, dropping to 11.5 in the last 5 years (Whitehall J, Kandasamy Y, Gill A, Stalewski H, unpublished).16 Ascertainment bias in our study of demographic data was minimised by the isolation of North Queensland which causes most cases to be managed in Townsville, and by availability of data on cases referred to Brisbane.
In conclusion, the incidence of gastroschisis is increasing, especially in younger mothers, and is over twice as common in indigenous mothers. In uncomplicated cases, there is no difference in outcomes after PNOR compared with POR and SR. Its advantages include reduction of delay and cost, and less movement of a sick infant.