Article Text

Download PDFPDF
Outcomes of a uniformly active approach to infants born at 22–24 weeks of gestation
  1. Fanny Söderström,
  2. Erik Normann,
  3. Maria Jonsson,
  4. Johan Ågren
  1. Department of Women's and Children's Health, Uppsala University, Uppsala, Sweden
  1. Correspondence to Dr Fanny Söderström, Department of Women's and Children's Health, Uppsala University, Uppsala, Sweden; fanny.soderstrom{at}


Objective To determine survival and outcomes in infants born at 22–24 weeks of gestation in a centre with a uniformly active approach to management of extremely preterm infants.

Study design Single-centre retrospective cohort study including infants born 2006–2015. Short-term morbidities assessed included retinopathy of prematurity, necrotising enterocolitis, patent ductus arteriosus, intraventricular haemorrhage, periventricular malacia and bronchopulmonary dysplasia. Neurodevelopmental outcomes assessed included cerebral palsy, visual impairment, hearing impairment and developmental delay.

Results Total survival was 64% (143/222), ranging from 52% at 22 weeks to 70% at 24 weeks. Of 133 (93%) children available for follow-up at 2.5 years corrected age, 34% had neurodevelopmental impairment with 11% classified as moderately to severely impaired. Treatment-requiring retinopathy of prematurity, severe bronchopulmonary dysplasia, visual impairment and developmental delay correlated with lower gestational age.

Conclusions A uniformly active approach to all extremely preterm infants results in survival rates that are not distinctly different across the gestational ages of 22–24 weeks and more than 50% survival even in infants at 22 weeks. The majority were unimpaired at 2.5 years, suggesting that such an approach does not result in higher rates of long-term adverse neurological outcome.

  • neonatology
  • neurology

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this topic?

  • With advances in neonatal care, survival rates in extremely preterm infants keep improving.

  • Active management of infants born at the ‘limit of viability’ remains controversial and the attitude towards this patient population vary greatly.

  • There are concerns about poor outcomes, particularly regarding neurodevelopmental disabilities.

What this study adds?

  • Active support to all infants born resulted in survival rates not distinctly different across the gestational ages of 22–24 weeks and over 50% even at 22 weeks.

  • Most survivors had no neurodevelopmental impairment, suggesting that this approach is not futile.


Survival in extremely preterm infants has been steadily improving over the last few decades, but active perinatal management of the most immature infants remains controversial due to concerns regarding poor outcomes, and the attitude towards infants born at the ‘limit of viability’ varies greatly.1–8 While some studies report high mortality and morbidity, others suggest that outcomes can be improved even in infants born at 22 weeks if provided full support, suggesting that attitude in itself might influence outcome.7–15

The aim of this investigation was to report survival, short-term morbidities and neurodevelopmental outcomes in infants born at 22–24 weeks of gestation from a healthcare region with a uniform policy of providing active care to all mother-infant dyads. Short-term outcomes included retinopathy of prematurity (ROP), necrotising enterocolitis (NEC), patent ductus arteriosus, intraventricular haemorrhage (IVH), periventricular malacia (PVL) and bronchopulmonary dysplasia (BPD). Neurodevelopmental outcomes at 2.5 years corrected age included cerebral palsy (CP), visual and hearing impairment and developmental delay.


Study setting

This was a retrospective cohort study from a regional tertiary care centre, Uppsala University Children’s Hospital (UUCH), where all mother-infant dyads at risk for extremely preterm delivery are provided proactive treatment. This includes intrauterine referral when approaching 22 weeks of gestation, provision of tocolytics, antenatal steroids and family counselling. UUCH is the sole tertiary centre for perinatal care in one of Sweden’s six healthcare regions with a population of two million and approximately 23 000 births annually. The region also includes seven secondary care centres with delivery and neonatal care units. UUCH is responsible for all specialised obstetric management and neonatal intensive care at these gestations in the region, and neonatologists and obstetricians work in close collaboration to optimise the care of the mother-infant dyad. All live-born infants are provided active treatment after birth regardless of gestational age, including senior staff in delivery room, intubation with surfactant administration and admission to the neonatal intensive care unit. Should prenatal transport not be possible, a neonatal transport team of a senior neonatologist and a neonatal nurse from UUCH is sent to the secondary care centre, ideally to be present at delivery.


The cohort consisted of infants born during 2006–2015 with a gestational age of 22+0 to 24+6 weeks. By searching the Swedish Neonatal Quality register,16 17 a total of 224 infants born at the hospital or admitted to the neonatal intensive care unit shortly after birth were identified, of which two were excluded due to major congenital malformations.


Gestational age was estimated using ultrasound. NEC was diagnosed by histopathological examination after bowel resection from surgery or autopsy. ROP was classified and staged according to The International Classification of Retinopathy of Prematurity,18 and ROP treatment included laser photocoagulation or injection with antivascular endothelial growth factor, following the Early Treatment of Retinopathy of Prematurity guidelines.19 IVH was diagnosed using ultrasound which was routinely performed on all infants during the first days of life. PVL was diagnosed using ultrasound and/or MRI. BPD was graded according to Jobe et Bancalari with mild BPD defined as oxygen treatment during at least 28 days but not at 36 weeks of gestation, moderate BPD as need for oxygen <30% at 36 weeks, and severe BPD as need for oxygen ≥30% and/or mechanical ventilation or treatment with continuous positive airway pressure.20 Sepsis was defined as clinical signs of infection and a positive blood culture and includes both early-onset and late-onset sepsis. Patient data were collected from medical records including the mothers’ charts for obstetric characteristics and data on intrauterine fetal deaths and stillbirths.


Data are presented as number and percentage, mean and SD or median and IQR. Group comparisons were made using Pearson’s χ² test and Fisher’s exact test for categorical values, and two-tailed Student’s t-test or Mann-Whitney U test for continuous variables. P<0.05 was considered significant. Denominators were adjusted for missing data. Statistical analyses were performed using Stata software, V.16.1 (Stata, College Station, Texas, USA).

Classification of neurodevelopmental impairment

All extremely preterm infants born in Sweden are included in a standardised national follow-up programme aimed at early identification of children with neurodevelopmental impairment. Follow-up visits are initiated at discharge and continue until 5.5 years of age and consist of assessments by paediatricians, physiotherapists and psychologists, which include formal testing such as neurological examination and testing of motor and cognitive function. This study presents the results at 2.5 years of corrected age, when all infants will have been seen by both a physiotherapist and a paediatrician, while formal testing by a trained child psychologist depends on local resources. While the Bayley Scales of Infant Development is often used in follow-up studies, it is not routinely performed as part of the national programme. In this cohort, 46 (35%) were assessed using the Bayley scales. When performed, mild delay was defined as cognitive and/or motor scores −1 SD to −2 SD below mean, and moderate to severe delay as scores more than −2 SD below mean. When not performed, the diagnosis of developmental delay was based on the comprehensive clinical assessment and testing by physicians, physiotherapists and psychologists. All children in Sweden are offered structural follow-up as part of the National child healthcare programme,21 which includes regular health check-ups and basic neurodevelopmental assessment. Classification of neurodevelopmental impairment are based on standard definitions used in this programme, as exemplified in table 1. Follow-up also includes screening for visual and hearing impairment. Children treated for ROP are followed up by ophthalmologists until school-age. Visual impairment was classified as uncorrectable refractive errors, with a visual acuity of <0.3 or complete blindness considered moderate or severe visual impairment. Moderate hearing impairment was defined as hearing loss requiring hearing aid, and severe hearing impairment as hearing loss not correctable with hearing aid or deafness. The severity of CP was defined using the Gross Motor Function Classification System,22 with level I-II considered mild impairment and level III-V moderate-severe impairment.

Table 1

Definitions of developmental delay


Out of the 222 infants born to 193 mothers included, 143 (64%) lived. Survival was 23/44 (52%) at 22 weeks, 56/87 (64%) at 23 weeks and 64/91 (70%) at 24 weeks (p=0.055) (figure 1). There was no difference in survival between inborn and outborn infants (65% vs 62%, p=0.843), and no infant born in a secondary care centre died prior to or during transfer. Fourteen stillbirths among fetuses alive at admission were recorded during this time period, of which two were intrapartum (figure 1). Four died during intrauterine transport from a secondary centre. Neonatal and maternal characteristics are displayed in table 2. Median age at death was 11 days with no differences across gestational ages. NEC was the most common cause of death (18; 23%), followed by BPD (12; 15%), IVH (11; 14%) and sepsis (11; 14%). Two infants died in the delivery room despite active efforts. All infants alive at discharge are alive to date. Seventy (36%) of the deliveries were caesarean, this being significantly more common with increasing gestational age, with no caesarean deliveries in week 22 (table 2). The most common indications for caesarean delivery were breech/transverse fetal lie (20; 29%) or placental abruption/bleeding (18; 26%). In 13 (19%) cases, the indication was maternal only: preeclampsia (n=9) or infection (n=4). Nine (13%) caesarean deliveries were due to fetal distress. Comparing gestational weeks 23 and 24, there were no differences regarding indications for caesarean delivery. Hospital outcomes including length of hospital stay and duration of mechanical ventilation are displayed in table 3. Not all data were retrievable, as some children resided in other healthcare regions at follow-up, and we could not access their medical records. Denominators were adjusted for missing data. The incidence of treatment-requiring ROP and severe BPD increased with decreasing gestational age, as did length of hospital stay and duration of mechanical ventilation (table 3). Follow-up data at 2.5 years corrected age were available in 133 (93%) children of whom 45 (34%) had adverse outcomes (table 4; figure 1). The number of children with developmental delay diagnosis was not different between those assessed with Bayley and those that were not (24% vs 29%, p=0.682). Out of 10 children with CP, 3 were non-ambulant. No infants were blind or deaf but two had need for hearing aid. Visual impairment and developmental delay correlated to lower gestational age (table 4).

Figure 1

Overview of survival and neurodevelopmental outcomes.

Table 2

Neonatal and maternal characteristics of full cohort by gestational week

Table 3

Hospital outcomes in survivors by gestational week

Table 4

Neurodevelopmental outcome at 2.5 years corrected age by gestational week


This study focuses on the very most immature infants, demonstrating survival rates of more than 50% even in infants born at 22 weeks of gestation. While morbidity rates decreased with increasing gestational age, they did not differ substantially across 22–24 weeks of gestation. Providing active care to all infants born regardless of gestational age has been our policy for several decades, with publications from 2004 presenting outcomes of infants born at 23–25 weeks of gestation offered full support after birth.23–25 This approach contrasts to the most common selective one where the mother-infant dyads are provided care depending on the individual preferences of the physicians and parents. At UUCH, parents are actively counselled throughout the process but do not make the decision whether or not to initiate treatment. It is our experience that given information on outcomes such as presented in this study, parents are positive to an active initial approach. The active approach to all infants born provides data on outcomes of the very most immature infants without the inference of perinatal selection,26–32 which we believe to be one key strength of our investigation. One possible source of selection bias not often accounted for is lack of data on stillbirths. In this study, 6% of fetuses alive at admission were declared stillborn, similar to the findings in the EXPRESS study, where the number of stillborn infants in UUCH at 22–24 weeks of gestation (7%) was significantly lower compared with the other centres.26 27 Noteworthy, the outcomes presented in this study are equivalent to or better than those previously presented from cohorts of infants born at 22 weeks including only those selected for active treatment.10–14 We have previously presented data on inborn infants born at 22 weeks of gestation comparing two collaborating centres with similar resources that differ in the approach to the these infants by either providing active care to all infants born, or applying a selective approach.9 That study clearly demonstrated better outcomes at the active centre regarding both survival and incidence of hospital morbidities, in line with the argument that a sceptical attitude towards providing active support at these gestational ages could result in a self-fulfilling prophecy.7 8 15 33 34 Given the outcome data presented in this study, it is our opinion that providing full active support to all infants is justifiable even at 22 weeks of gestation.


The study is limited mainly by its retrospective nature and inconsistencies in neurodevelopmental follow-up. While we consider the clinicians’ assessments sufficient for the purpose of the study, they are not fully objective. We also acknowledge that neurodevelopment at 2.5 years might not be indicative of outcomes later in life, particularly regarding behavioural aspects and difficulties in learning.35 36 There are also other outcomes and aspects of health status besides neurodevelopmental impairment that are not covered in this study. The sample size is small which complicates detailed subgroup comparisons.


A uniformly active approach to all infants born resulted in survival rates not distinctly different across the gestational ages of 22–24 weeks with half of the infants surviving even at 22 weeks. The finding that the majority of the children were unimpaired at 2.5 years suggests that such an approach is worth pursuing, and more detailed studies are urgently needed to optimise the management and further improve outcomes in this highly vulnerable patient population.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statements

Ethics approval

The study was approved by the Regional Ethical Review Board (reference number 2017/115).



  • Contributors FS was responsible for data collection, participated in data analysis and interpretation and drafted the initial manuscript. MJ assisted in data collection and interpretation and reviewed and revised the manuscript. EN assisted in data collection and interpretation and reviewed and revised the manuscript. JA was supervising the project, participated in data analysis and interpretation and drafted the initial manuscript. All authors have reviewed the latest version of the manuscript and approve of the submission.

  • Funding This study was supported by funding from Gillbergska stiftelsen and Födelsefonden.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • Fantoms
    Ben J Stenson