Improve the Survival Without Morbidity of Extremely Preterm Infants (PREMEX)

Extremely preterm infants, those born between 22 and 26 weeks of gestation, are at very high risk of neonatal death, severe morbidity, and disabilities. These children's survival basically depends on the obstetric-pediatric team's willingness to provide active care. In France, practices vary strongly according to the child's place of birth. The rates of active antenatal care (reflects the intention to manage the child actively at birth), standardized by gestational age, vary from 22% to 61% between regions. One of the consequences of this heterogeneity is that the survival rate in France of these extremely premature infants is much lower than that of countries offering similar levels of care. Among the children live-born in France, survival rates at hospital discharge were 0% at 22 weeks, 1% at 23, 31% at 24, 59% at 25, and 75% at 26 weeks. Survival in the USA, Great Britain, Japan, Australia, and Sweden ranges from 10 to 50% higher than in France.

The great variability of antenatal practices for the management of extremely preterm infants reflects the disarray of medical teams. In the absence of consensus and written decision-making processes, this uncertainty culminates in management decisions based at best on local habits and at worst on the individual habits of the clinician present. These habits vary with the physicians' beliefs or convictions, and their level of knowledge and experience of extremely preterm births. This variability of practices between hospitals presents a problem of equity.

The hypothesis is that standardization of the management of extremely preterm infants, from the moment that the pregnant woman is hospitalized until the child's birth should improve survival without severe morbidity in this group of children at very high risk

The intervention proposed is a new organization of care, based on the EXPRIM (EXtrem PRematurity Innovative Management) protocol (MC Lamau, et al, PMID: 34059380), involving early, standardized, and multidisciplinary management of women hospitalized for a risk of extremely preterm birth and their children. It will take place in each perinatal network for all pregnant women hospitalized between 22 and 26 weeks with a risk of preterm delivery.

Setting up the protocol requires taking into account the parents' time and timing issues, and its potential for change, to plan the implementation of the protocol, especially the degree of emergency of the situation and the probability of imminent delivery.

Complex intervention, named the EXPRIM protocol, aimed at standardizing the organization of care. It is based on the following principles:

• A collective obstetric-pediatric prognostic assessment, in a non-emergency setting, no longer based only on gestational age, and ensuring better team consistency in terms of the attitude to propose to parents;
• A consensus decision by the multidisciplinary team about the obstetric and neonatal management proposed-either active or palliative care-at the conclusion of this prognostic evaluation
• An interview with the parents to:
• inform them about the obstetric situation and the child's prognosis
• answer their questions
• propose either active or palliative care to them as management
• and then ask their opinion.
• The administration of corticosteroid therapy, independently of the decision about management (active or palliative care), starting at 23 weeks will be of broad indication;

If the hospitalization takes place in a level-1 or 2 hospital: after general information about preterm delivery, a proposal to transfer the woman to the perinatal network's level-3 hospital, starting at 23 weeks

The choice of a stepped-wedge cluster randomization design allows all perinatal networks to implement the intervention.

After the random drawing of clusters for the intervention, the teams will be trained in the EXPRIM protocol and then will undergo a 3-month transition period (not considered either a control or intervention period) to learn the protocol thoroughly. The data collected during this period will not be used in the analysis. In summary, the 3 stages before recruitment during the intervention are:

1. allocation of clusters
2. training of perinatal teams
3. transition period After random drawing, 4 perinatal networks will be allocated to the intervention in 5 waves every 3 months, with 3 months of recruitment for the last wave. Twenty perinatal networks will be randomized and will participate in the trial with a total recruitment period of 21 months (including the 3-month transition periods). Overall, the duration of the study at each center will be 49 months. All perinatal networks will have a period with and without the intervention and will participate in the comparison accordingly. After 15 months all the networks will have begun the intervention.

No specific follow-up visit for the purposes of the study is planned. The hospital staff physicians will perform follow-up as part of the standard usual care for women and their newborns.

The follow-up collected for this study will take place:

• At D4 post-delivery (or before, if the woman leaves the hospital before then): A questionnaire about the parents' experience of the information delivered and the decisions made will be given to and collected from the parents. There are 2 "parental experience" questionnaires: one destined to parents of newborn alive and admitted in NICU and one destined to parents of stillborn or deceased child at birth. For parents of a stillborn or deceased child at birth, this questionnaire may be given during hospitalization in maternity (D0-D4), or after returning home or during the postnatal visit.
• At Day 28, post-delivery: A "Neonatal individualized developmental care questionnaire about the parents' experience of care for their child will be given to and collected from the mother and the co-parent.
• At the child's discharge from the hospital, or if he or she dies in the hospital:
• Collection of clinical data (principal endpoint) from data in the medical file. If the patient was transferred to another hospital, the PREMEX investigator and/or the CRT will obtain the follow-up data to enter on the e-CRF.
• Data to measure practices and adherence to the intervention will be collected
• The neonatologist-investigator will ask the parents at the child's discharge from hospitalization for written consent to continue their participation in this study for the follow-up at age 2.
• When the child reaches the corrected age of 2 years:
• Nearly all the surviving children will be receiving care within structured networks for the follow-up of vulnerable children, responsible for optimizing their medical care and follow-up. Accordingly, a short questionnaire will be completed by the physician caring for the child at the corrected age of 2 years. The data collected will concern motor and sensory development, in particular, cerebral palsy, blindness, and deafness. The questionnaire will be sent to the parents, who will ask the physician to complete it at the medical visit planned at this age. The physician will send the completed questionnaire to the project coordinators.
• Information about the child's development will also be collected with a questionnaire including a standardized assessment scale, the PARCA-R questionnaire (PArent Report of Children's Abilities-Revised). Parents will be invited to complete the PRO (Patient Reported Outcomes, the PARCA-R questionnaire in our case) on an electronic platform-Patient Reported Outcomes-and will receive specific access codes.

At the time of inclusion, the contact data of the parents taking part in the study (e-mail address, telephone number and postal address) were collected on contact forms by the clinical teams of each maternity unit and kept in the medical files.

As part of the 2-year follow-up of these children, the coordinating midwife and/or CRT centralized will collect this contact data from the maternity units via a secure.