Background The survival rate for children born with gestational ages 22–27 weeks is increasing, and this may be associated with higher rates of disability. The aims of this study were to determine the outcomes at age eight for a regional cohort of children born at 22–27 weeks during 1997, and to compare their rates of disability with a cohort of the same gestational age born in 1991–1992.
Methods Consecutive children with gestational ages in the range 22–27 weeks born in the state of Victoria, Australia, in 1997 and matched term controls were assessed at 8 years. Outcomes included blindness, deafness, cerebral palsy (CP) and intellectual impairment and disabilities caused by these impairments. These outcomes were compared with a cohort of 22–27-week and term children born in 1991–1992 in the same region.
Results Follow-up rates for the 1997 cohort at 8 years of age were 95% (144/151) for 22–27 weeks survivors and 89% (173/195) for controls. Rates of disability were substantially higher in the preterm cohort than the controls. The 1997 and 1991–1992 preterm cohorts had similar rates of moderate or severe disability (19%), however the rate of mild impairment was greater in 1997 (40% vs 24%). Rates of disability were almost identical in control groups. Intellectual impairment and CP were the major reasons for the higher rates of disability.
Conclusions The high prevalence of adverse neurodevelopmental outcome in children born at 22–27 weeks compared with term controls at school age persists, and may even be increasing over time.
Statistics from Altmetric.com
Extremely preterm birth is an important public health issue as increasing numbers of these vulnerable children are offered neonatal intensive care and their survival rates continue to increase over time.1 The biggest increase in survival rates has been seen for the most immature infants <25 weeks' gestational age.1 The risk of long-term adverse outcomes, however, is also known to increase as gestational age at birth decreases.2 Moreover, when followed to school age, very preterm children continue to experience high rates of adverse developmental outcomes, with approximately one in two children having academic or behavioural difficulties.3 In addition to increases in the survival rate of the most susceptible survivors at <25–26 weeks, secular trends in the use of therapies known to improve long-term outcome, such as caffeine for apnoea of prematurity,4 mean that very preterm cohorts from different eras must be assessed to continually monitor the risks versus benefits of neonatal intensive care.1
What is already known on this topic
The survival rate to school age, of children born at 22–27 weeks of gestation, continues to rise.
These children have very high rates of adverse neurodevelopmental outcomes, compared with term peers.
What this study adds
As the survival rate has increased, moderate to severe disabilities in children born at 22–27 weeks' gestation have remained stable but mild disabilities are more prevalent.
The increase in mild disabilities is primarily driven by an increase in the diagnosis of mild cognitive impairment.
The aims of this study were to determine the rates of neurodevelopmental disability according to gestation, at age eight, of children born with gestational ages 22–27 weeks in the state of Victoria, Australia, in 1997, and to compare results with similarly recruited children of the same gestation born in 1991–1992 from the same geographic region. It was hypothesised that rates of disability at school age would increase as gestation decreased, but that overall disability rates would remain stable over time.
Preterm births in 1997
The preterm cohort born between 1 January 1997 and 31 December 1997 comprised 151 survivors to 8 years of age out of 223 consecutive live births at 22–27 completed weeks of gestation, in the state of Victoria, Australia, a survival rate of 67.7%. The term cohort comprised 199 randomly selected children of gestational age >36 weeks born in the 3 level III perinatal centres in Victoria, all of whom survived to 8 years of age. Controls were born on the expected date of birth of a preterm child, matched for gender, and the mother's country of birth (English speaking or not) and health insurance status (private health insurance or not). All children were enrolled in the newborn period in a prospective longitudinal study of growth and development throughout childhood; results at 2 years of age have been reported.1 Written informed consent was obtained from parents of term children. Follow-up was considered routine clinical care for the 22–27 weeks children. The Research and Ethics Committees of the Royal Women's Hospital, Mercy Hospital for Women and Monash Medical Centre, Melbourne, approved these follow-up studies.
Survivors were assessed at 8 years of age by paediatricians and psychologists blinded to perinatal details, predominantly in specialised follow-up clinics, although a few were tested at school or home if they could not attend the clinics. Age was corrected for prematurity, to be consistent with assessments of earlier cohorts and to avoid bias in the cognitive test scores.5
Impairments that were diagnosed at follow-up included cerebral palsy (CP), blindness, deafness and intellectual impairment (IQ <−1 SD relative to the mean and SD for term controls). General intelligence was assessed using the Wechsler Intelligence Scale for Children, fourth edition (WISC-IV).6 The preterm children were compared with the term controls, rather than with the test norms, in assigning disability criteria, in order to study the preterm children in the context of their typically developing peers from the same geographic area. A small number of children were not able to complete all the subtests of the WISC-IV, primarily due to neurosensory impairments. For children who were unable to complete visual–motor subscales of the WISC-IV due to CP-related motor impairment or visual impairment (n=4), their verbal comprehension index was used as an estimate of IQ. For the child who was unable to complete language-based subscales of the WISC-IV due to significant hearing impairment (n=1), the perceptual reasoning index score was used as an estimate of IQ. Two children were unable to complete any of the WISC-IV subtests due to severe disability. These children were assigned an IQ standard score of −4 SD. Two children did not complete all subscales of the WISC-IV due to a lack of compliance and their IQ score was calculated based on the completed subscales. Analyses were carried out including and excluding the children with incomplete assessments.
Severe disability comprised severe CP (child not walking), blindness (visual acuity <6/60 in the better eye), or an IQ <−3 SD; moderate disability comprised moderate CP (walking with considerable difficulty, with or without appliances), deafness (requiring hearing aids or worse), or an IQ from −3 SD to <−2 SD; mild disability comprised mild CP (walking with minimal limitation), or an IQ from −2 SD to <−1 SD.7
Preterm births in 1991–1992
The preterm cohort born in 1991–1992 comprised 225 survivors of 438 consecutive live births at 22–27 completed weeks of gestation, a survival rate of 51.4%. The follow-up rate at age eight was 93% (210 children). The term cohort comprised 265 randomly selected children of gestational age >36 weeks born in the 3 level III perinatal centres in Victoria, 262 of whom survived to 8 years of age. Their neurological outcomes at 2,1 8 9 57 10 11 and 87 10 11 years of age have been reported, but largely by birth weight, rather than gestational age. Impairments and disabilities were coded identically for the 1991–1992 and the 1997 cohorts, including using the mean and SD for IQ scores (WISC-III for 1991–1992 cohorts) of controls to calculate cut-off scores for intellectual impairment.
Treatments available for the 1991–1992 and the 1997 preterm cohorts were similar, including the availability of exogenous surfactant, but the biggest difference was that more infants <28 weeks' gestational age were offered intensive care in 1997 than in 1991–1992, particularly those <24 weeks.1
Data were analysed by SPSS for Windows, V.17.0 (SPSS, Chicago, Illinois, USA).12 Between group differences for dichotomous outcome data were analysed by χ2 analysis or by Fisher exact test if cell sizes were small, or by χ2 for linear trend and for continuous data by t test, computing mean differences and 95% CIs. Logistic regression was used to examine the effect of birth in the early 1990s compared with birth in 1997 on disability status, while adjusting for gestational age.
The overall rates of disabilities were much higher in 22–27 weeks children than in controls (χ2trend=61.0, p<0.001; table 2). Fewer than half (41%) of the preterm children were free of any disability at age 8, compared with the majority (87%) of the term children (χ2=73.7, degrees of freedom (df)=1, p<0.001); 19% of the preterm group had a moderate or severe disability, compared with only 3% of the controls (χ2=21.7, df=1, p<0.001). Within the preterm cohort alone there was no statistically significant change in disabilities with gestational age (χ2=17.1, df=15, p=0.31).
Compared with the term children, the preterm children had higher rates of deafness and blindness (table 3), but as the overall rates were low the differences were not statistically significant (p=0.093 by Fisher exact test for both comparisons). However, substantially more preterm children had CP compared with controls (χ2=20.3, df=1, p<0.001), and more preterm children had intellectual impairment of increasing severity than controls (=48.3, p<0.001). Overall 53.5% of preterm children had some intellectual impairment relative to the rate of 12.9% in the controls. In contrast, if intellectual impairment was categorised on a mean of 100 and SD of 15, only 23.9% of 22–27 weeks children would have been classified as having intellectual impairment, compared with 5.9% of controls. The 22–27 weeks children had lower IQ scores than controls, with a mean difference of 13.2 (95% CI 10.0 to 16.4; p<0.001). Within the preterm group alone, there were no statistically significant changes in blindness, deafness or CP with gestational age; however, mean IQ scores decreased significantly with decreasing gestation (table 3).
Comparison of the 1997 and 1991–1992 eras
The rates of disability for the term children were similar over the two eras (table 4; χ2=13.0, df=3 p=0.72). In contrast, the overall rate of disability was higher in 22–27 weeks children born in 1997 compared with 22–27 weeks children born in 1991–1992 (table 4; χ2=12.2, df=3 p=0.007), which was due to an increase in children with mild disability, as rates of moderate and severe disability combined were the same in both eras (19.1%). Rates of mild CP were similar for both eras (4.9% in 1997; 6.7% in 1991–1992), and the increase in mild disability in 1997 was due to more children identified as mildly intellectually impaired (IQ <−1 SD to −2 SD was 39% in 1997, and 23% in 1991–1992).
The overall increase in disability rates in 1997 was not due to increasing survival of the most preterm infants but was due to an increase in mild disability rates in the more mature children (25–27 weeks), who make up the largest proportion of the preterm cohorts (figure 1). Increasing gestation was associated with reduced odds of moderate/severe disability compared with zero/mild disability (OR per week increase 0.72, 95% CI 0.57 to 0.90, p=0.005) and with any disability compared with zero disability (OR per week increase 0.82, 95% CI 0.67 to 0.99, p=0.04). Birth era only had an effect when children with any disability were compared with zero disability, and birth in 1997 was associated with increased odds of any disability (OR 1.86, 95% CI 1.20 to 2.87, p=0.005).
As expected, the 22–27 weeks children born in 1997 had much higher rates of disability than the term controls. However, contrary to expectations, rates of overall disability did not vary significantly with gestation within the preterm cohort, although mean IQ decreased significantly with decreasing gestation. When rates of neurodevelopmental disability for the 1997 and 1991–1992 preterm cohorts were compared, similar rates of moderate to severe disability were found but an increased rate of mild disability was observed in the 1997 cohort. Surprisingly, the increase in disability overall in the 1997 cohort occurred predominantly in the more mature (25–27 weeks) survivors, and not those <25 weeks.
There is no doubt that 22–27 weeks survivors have more neurosensory disability than term controls. In a recent review 20% to 35% of survivors <28 weeks' gestational age aged from 18 months to 11 years from geographical cohorts had substantial problems with mobility, vision, hearing or cognition,13 similar to moderate and severe disability combined, as defined in the current study. However, the majority of the studies in this review had assessed the children before school age. The current study is the only one to our knowledge where 22–27 weeks cohorts from the same geographical region across different eras have been assessed at school age.
The finding that rates of disability in the 1997 preterm cohort were not related to gestational age was unexpected, and contrasts with data from other geographic cohorts.13 Examining individual areas of impairment, however, yielded more information than overall disability rates. The inverse relationship of IQ scores and gestation was clear, and has been previously documented.2 CP rates have previously been shown to increase with decreasing gestation.14 The small numbers of children with CP in the 1997 cohort reduced the power to find a significant relationship in the current study. Likewise, rates of deafness and blindness were so low overall that statistical significance would not be expected.
The odds of having any disability were higher for 22–27 weeks children born in 1997 than in 1991–1992 due to an increase in mild disabilities. This trend was specific to the preterm population over time, as this increase was not seen in the term control groups between the two eras. This increase in mild disabilities was driven primarily by an increase in mild cognitive impairment in the 22–27 weeks children in 1997. In 1997 IQ scores were measured using the WISC-IV and in 1991–1992 by the WISC-III. Subtle differences between the two tests may be one possible explanation for the increase in mild cognitive impairment in the 1997 preterm cohort. IQ scores on the WISC-IV tend to be about five points lower than on the WISC-III.15 Further, there are a number of subtest differences between the WISC-III and the WISC-IV, with the more recent edition having a greater focus on working memory and conceptual and perceptual reasoning. These higher order cognitive abilities have been previously shown to be areas of concern for very preterm children.16 Consistent with the possibility that the WISC-IV may be more sensitive to the areas of vulnerability of 22–27 weeks children is the observation that rates of cognitive delay at age 2 years between the 1997 and 1991–1992 preterm cohorts were similar (23% in 1991–1992; 24% in 1997%).1 However, it is also possible that the increase in mild cognitive impairment in the preterm children in the 1997 cohort represents a true increase, and highlights the need to follow-up successive cohorts from the same geographic region and also to expand the focus of follow-up from overall measures of disability to a more detailed assessment of cognitive and functional outcomes.
The current study has several strengths. The 22–27 weeks cohorts were geographically determined and therefore not subject to referral bias. The follow-up rates were high (>90% for the preterm cohorts). The children were assessed at school age, by which time neurosensory impairments can be diagnosed with certainty and cognitive abilities can be assessed reliably. The controls were recruited prospectively and were matched for variables such as gender and sociodemographic factors that are known to influence long-term outcomes.17 The cognitive scores of the 22–27 weeks children were assessed relative to the controls to maintain consistency with previous eras. Had we only used test norms of mean 100 and SD 15 we would have substantially underestimated the difficulties that the Victorian 22–27 weeks children face, compared with their peers. Using controls to categorise cognitive scores for very preterm survivors has been used in other geographic cohorts, with similar results.18 The study weaknesses include the slightly lower retention rate, at age eight, of the term controls, although overall our retention rate was higher than most other previous studies following-up geographic cohorts. Also, the disability criteria used in the current study do not include other areas of difficulty that occur commonly in the very preterm population, including behaviour, attention and learning difficulties3; these outcomes will be the subject of future reports. We acknowledge that examining a limited number of outcomes underestimates the true frequency of all difficulties in this high-risk population.
In conclusion, the high prevalence of adverse neurodevelopmental outcome in children born at 22–27 weeks' gestation compared with term controls at school age persists, and may even be increasing over time. The impact of strategies recently proven to improve developmental outcome in early childhood, such as caffeine after birth,4 antenatal magnesium sulfate,19 and dietary supplementation with docosahexanoic acid,20 on outcomes of geographical cohorts at school age is eagerly anticipated in the hope that the rates of disability can be reduced in these highest-risk children. Other strategies to improve neurodevelopmental outcomes are also urgently required and regular developmental surveillance into school age is necessary for these high-risk children.
Participants Convenor: Lex W Doyle (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia; Department of Obstetrics and Gynaecology, The University of Melbourne, Melbourne, Australia; Department of Paediatrics, The University of Melbourne, Melbourne, Australia; Murdoch Children's Research Institute, Victoria, Australia). Collaborators (in alphabetical order): Peter Anderson (Department of Psychology, The University of Melbourne, Melbourne, Australia; Murdoch Children's Research Institute, Victoria, Australia); Catherine Callanan (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia); Elizabeth Carse (Monash Medical Centre, Melbourne, Australia); Margaret P Charlton, (Monash Medical Centre, Melbourne, Australia); Noni Davis (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia); Cinzia R De Luca (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia; Department of Obstetrics and Gynaecology, The University of Melbourne, Melbourne, Australia); Julianne Duff, (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia); Marie Hayes (Monash Medical Centre, Melbourne, Australia); Esther Hutchinson (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia; Elaine Kelly (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia; Mercy Hospital for Women, Melbourne, Australia); Marion McDonald (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia); Gillian Opie (Mercy Hospital for Women, Melbourne, Australia); Gehan Roberts (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia; Department of Paediatrics, The University of Melbourne, Melbourne, Australia; Murdoch Children's Research Institute, Victoria, Australia); Linh Ung (The Premature Infant Follow-up Program at the Royal Women's Hospital, Melbourne, Australia); Andrew Watkins (Mercy Hospital for Women, Melbourne, Australia); Amanda Williamson (Mercy Hospital for Women, Melbourne, Australia); Heather Woods (Mercy Hospital for Women, Melbourne, Australia).
Funding This study was funded by the Victorian Government, who had no involvement in design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.
Competing Interests None.
Ethics approval This study was conducted with the approval of the Research and Ethics Committees of the Royal Women's Hospital, Mercy Hospital for Women and Monash Medical Centre, Melbourne.
Patient consent Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.
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.