Recruitment was terminated early because of staffing limitations affecting delivery of the intervention, expiration of study funding, and slower-than-anticipated recruitment, which made completion of the planned sample size infeasible.

Optimizing enteral nutrition (EN) is challenging in extremely preterm infants due to feeding intolerance that relates to the functional gastrointestinal immaturity. Early feeding is a safe way to promote postnatal gastrointestinal maturation and, when compared with delayed enteral feeding, provide benefit, such as reduced time to full enteral feedings (TFF) and number of parenteral nutrition (PN) days. Failure to develop oral feeding competence often leads to growth failure, longer hospital stays, dependence on PN and its complications, and influences long-term growth and developmental outcomes. Feeding with human breast milk has a protective effect against necrotizing enterocolitis (NEC) compared with formula, whereas feeding intolerance is one of the early signs of NEC. Delayed passage of meconium is a risk factor for feeding intolerance in preterm very low birth weight neonates and specific meconium microbiota characteristics have been linked to increased risk of NEC. This randomized controlled trial (RCT) aims at evaluating the effect of regular intestinal lavage using normal saline on the TFF and severe complications such as NEC and sepsis, in extremely preterm infants. Investigators aim also to follow children´s neurological development until 5,5 years of age. The study will include one intervention group of 100 subjects that will receive regular rectal washout with normal saline and equal number of control subjects, treated according to current routine. The trial is preliminarily estimated to last between year 2018 and 2022. Investigators will monitor closely for possible adverse events. The results are going to be published in reviewed medical journal.

The study will be designed and conducted as a two-arm parallel, open-label, randomized controlled trial (RCT) at a tertiary-care hospital.

No masking will be applied, in order not to expose the control group to sham interventions of no benefit or potential harm.

A blinded research coordinator will randomly assign eligible infants using a computer software-based randomization system (default) or sealed, numbered, and opaque envelopes (back-up), with a 1:1 allocation ratio, as soon as possible and not later than 24 hours from birth.

Sample size calculation is based on the incidence of NEC, as the rarest of the events that will be studied, which is currently about 15% among extremely preterm infants at the investigators' unit, as mentioned above. For study power of at least 70%, confidence level of 5%, and an expected decrease of NEC incidence from 15% to 8%, each study arm should include about 100 participants. Each year, the investigators' unit treats about 70 infants born in GA 22+0 - 26+6. Supposing a participation rate of 70%, it is estimated that the recruitment period should last between 3 and 4 years, preliminarily starting during 2018. However, a recruitment period of maximum 5 years will be allowed.

Current feeding regimen and feeding intolerance guidelines will be applied for both study arms.

Intervention The following intervention will be applied to the intervention group: specially trained pediatric surgeon will administer 10ml/kg pre-warmed (37oC) normal saline via a single-use rectal tube of size 6FR twice per day, aiming at a depth of maximum 10 cm/kg, starting after randomization and not later than 24 hours of age, and continued until full enteral nutrition of 170ml/kg/day is achieved or NEC diagnosis (Bell stage II or more) is established, which one comes first. However, the intervention will be applied at a maximum of 2 weeks from birth. Besides, the intervention will be withheld in case of suspicion of infection or NEC but will be resumed if antibiotics are not introduced and discontinuation of feeds does not exceed 48 hours. Early stopping criteria include infection / sepsis or NEC suspicion necessitating prolonged discontinuation of feeds > 48 hours and / or antibiotics, circulatory instability or adverse events. The intervention will only be applied at the NICU of University Children's Hospital in Uppsala and will be discontinued if the infant is transferred to another hospital.

Standard departmental guidelines will be followed regarding ventilation, choice of PN, invasive monitoring, and management of sepsis and/or necrotizing enterocolitis (NEC) for both the intervention and the control groups.

Comparison group

The following guidelines are currently applied at the investigators' unit for extremely preterm infants that do not defecate adequately:

If the child has not passed meconium or has already passed meconium / feces but has not defecated for 3-4 days:

• If there are no symptoms: continued EN, follow-up of clinical status and defecation pattern. If the child has not defecated for another 1-2 days (and remains asymptomatic), consider the following treatment strategy to facilitate defecation:

  1. Tactile perineal stimulation is performed by a nurse, using a room temperature damp compress
  2. Cautious rectal stimulation is performed by a nurse, using a rectal catheter size 6FR; depth 1 - 2 cm
  3. Step 1 & 2 is repeated depending on effect and defecation pattern
  4. If Step 1 & 2 does not result in defecation and the clinical status is unchanged, enema with 4mL / kg sodium chloride 9mg / mL is given
  5. In case of no effect, a pediatric surgeon is consulted, and consideration is given to continued treatment; (1) watchful waiting alternatively repeated enema (2) rectal washout with 10 mL / kg sodium chloride 9mg / mL (3) further radiological and laboratory investigations

• In case of symptoms (cardiorespiratory instability, impaired abdominal status, vomiting, bilious residuals, etc.): EN is withheld, pediatric surgeon is consulted and radiological / laboratory investigations as well as early treatment with antibiotics (tazobactam / piperacillin + gentamicin) is considered.

Follow-up All extremely preterm infants hospitalized in the investigators' unit are discharged home at about GA 35+0. Infants regionally belonging to another hospital are usually transferred to the "home hospital unit" not earlier than GA 28+0. After hospital discharge, all extremely preterm infants are routinely followed-up according to a nationwide schedule, until the age of 5,5 years. Data mostly concerning children's growth as well as neurological and psychomotor development are collected, and interventions are made as needed.

Study participants will be followed-up until full term corrected age (GA 40 +0), and medical information including possible treatments as well as diagnosis of NEC or other gastrointestinal complication, sepsis, PDA, BPD, ROP, IVH, hyperbilirubinemia, growth parameters (weight, length, head circumference) and mortality will be collected by the investigators from the treating physicians and electronic medical records. All the clinical, laboratory and radiological findings of infants treated at the investigators' unit are routinely and continuously registered in detail, with the use of medical computer software, during the whole hospitalization period, and these data will be readily accessible to the investigators. Besides, data on children's growth and neurological development will be collected from the electronic medical records, until the age of 5,5 years. The above-mentioned data of study participants that have been transferred to the home or other hospital unit will be collected by the investigators though communication with the treating physician at the respective hospital.

Documentation All data relevant to the study collected by the investigators will be registered and de-identified in electronic database with password-protected access allowed only to the investigators.

Refuse to enrollment and subject dropouts Refusals to enroll the study and dropouts of the study and the respective reasons will be documented and reported in the results. Effort will be made to replace each of these subjects with another infant of the same gestational age.

Completion of the study The study will be completed when the number of 100 subjects per study arm is reached (a maximum of 130 subjects per study arm will be allowed), estimated to be achieved by the year 2022 (and not later than year 2023).

Statistical analysis Numerical data with a gaussian distribution will be expressed as means (standard deviation), otherwise as medians (ranges). Anderson-Darling test will be used to assess the normality of the data. Numerical values will be compared between intervention and control groups using t-tests or the Mann-Whitney U test. Categorical variables will be compared using χ2 or Fisher's exact tests. Times to achieve full enteral feeding will be compared by the log rank test. Multiple Cox regression analysis will be used to adjust for covariates, such as gestational age and birth weight. IBM SPPS software (current version at the time of publication) will be used for all calculations and a p-value of <0,05 will be considered significant.

Interim analysis will be conducted when 35 infants have been recruited in each study arm (estimated about 1,5 years from study start). Another interim analysis will be conducted when 65 infants have been recruited in each study arm (estimated about 3 years from study start). Unplanned interim analysis will also be conducted in case of subjectively estimated too obvious difference between the groups regarding the primary outcomes, at any time during the study. The study will be discontinued as soon as an interim analysis shows statistically significant results regarding the primary outcomes or severe adverse events.

Ethical considerations The study protocol will be submitted for approval to the Regional Ethical Review Board in Uppsala and subsequently registered at the international registry of clinical trials, ClinicalTrials.gov. The parents of the infants will be informed about the study by specially trained personnel, both verbally and in writing, before the child is born or, whenever that is not possible, as soon as possible after birth. Upon agreement to participate, written informed consent will be obtained from the parents, before enrollment in the study. Additionally, the parents will be clearly informed about the possibility to leave the study any time they wish.

None of the studies published so far reported any adverse events by induced meconium evacuation, with the exception of a higher proportion of NEC (although not statistically significant), vomiting, nausea and bradycardia with the use of oral gastrografin, and a non-significant trend towards increased risk of NEC with the use of glycerine enemas or suppositories compared with no treatment. However, in the RCT study described herein, both health personnel involved in the participating infants' daily care and the investigators are going to closely monitor for possible adverse events. Possible adverse events could be: rectal bleeding (including occult rectal bleeding-induced anemia), fluid loss due to overstimulation of immature intestines of preterm infants or the opposite effect, namely intestinal absorption of given saline, leading to hypervolemia and electrolyte disturbances, mainly hyponatremia. A safety monitoring committee will monitor adverse events with predefined stopping rules. Early stopping criteria will be applied in case of medical instability, as mentioned above. Furthermore, the whole study will be discontinued in case of too obvious positive effect, making it unethical to deprive the control group of the intervention, or in case of severe adverse events.

The inclusion of study participants in other interventional studies could intervene with the outcomes intended to be investigated in the study herein, rendering it difficult to retrieve definitive conclusions. Thus, study participants will not be included in other interventional studies at the same time.

The intervention will be discontinued if the infant is transferred to another hospital, as mentioned before. The decision to transfer a study subject to home or other hospital unit will only be based on current criteria and will not be affected by the participation in the study. Study participation is not expected to delay the transfer to home hospital and prolong the stay at our NICU in Uppsala, unless a severe adverse event should occur.

Reporting and dissemination Following the study, data will be analysed by the research group and the results will be published in reviewed medical journals and possibly presented in a neonatology conference, hopefully during year 2023. It is difficult to provide a precise time schedule. Patient data collection is estimated to last about 4 years. Data analysis thereafter is time consuming and may last one more year until the whole project is completed and ready for publication. All the principal investigators will have free access to raw data and the right to publication.

Inclusion Criteria:

• born at Akademiska hospital in Uppsala
• gestational age between 22 weeks + 0 days and 26 weeks +6 days
• written informed consent obtained from both guardians before enrollment in the study.

Exclusion Criteria:

• major dysmorphic features consistent with chromosomal abnormality
• major congenital anomalies, such as gastrointestinal disorders
• circulatory instability during the first hours of life