Share this article on:

Outcome at 3 Years of Age in a Population-Based Cohort of Extremely Preterm Infants

De Groote, Isabel PhD1; Vanhaesebrouck, Piet MD, PhD2; Bruneel, Els MD3; Dom, Lina MD4; Durein, Isabelle PT Master5; Hasaerts, Danielle MD6; Laroche, Sabine MD7; Oostra, Ann MD8; Ortibus, Els MD9; Roeyers, Herbert PhD7; van Mol, Christine MD10

doi: 10.1097/01.AOG.0000284447.43442.55
Original Articles

OBJECTIVE: To assess health and neurodevelopmental outcome at 3 years of age in neonatal intensive care unit (NICU)–surviving children who were born at 26 or fewer weeks of gestation in a geographically defined region of Belgium from 1999 through 2000.

METHODS: The study included a clinical examination and a standardized neurologic and developmental assessment. Disabilities were defined by international criteria. In 97% (92 of 95) of the children, accurate information on the presence of overall disability could be collected.

RESULTS: Thirty-six percent (95% confidence interval [CI] 25–47%) of the formally assessed children (28 of 77) had deficient neuromotor development, with 5% of them showing severe sensory-communicative impairment. Mean (±standard deviation) scores on the Mental Developmental Index and Psychomotor Developmental Index were 81.2 (18.8) and 73.2 (17.8), respectively. Seventy percent (95% CI 60–80%) had a mental (Mental Developmental Index) or psychomotor (Psychomotor Developmental Index) impairment or both, assessed to be more than 1 standard deviation below the population mean. Mental and psychomotor outcome did not differ significantly when compared according to either gestational age, gender, or multiple birth (all P>.05).

When either minor central dysfunction or cerebral palsy was not taken into account, normal mental development was recorded in 62% of the subjects. The cumulative of poor outcome (ie, disability- or prematurity-related death) among the 95 infants discharged alive was estimated to be 58% (95% CI 48–68%), representing 25 (26%) mildly-to-moderately disabled and 28 (29%) severely disabled toddlers, including two infants whose postdischarge deaths were directly related to prematurity.

CONCLUSION: The average developmental outcome is poor in children born as extremely preterm infants. Finding early predictors of adverse outcome is a major challenge.


Health and neurodevelopmental outcome at 3 years of age is poor in children born as extremely preterm infants.

From the 1Research Group Developmental Disabilities, Ghent University, Ghent, Belgium; 2Department of Neonatology, Ghent University Hospital, Ghent, Belgium; 3Department of Neonatology, Algemeen Ziekenhuis Zuid-Oost Limburg, Genk, Belgium; 4Department of Child Neurology, Koningin Paola Kinderziekenhuis, Antwerp, Belgium; 5Center for Developmental Disorders Brugge, Brugge, Belgium; 6Department of Child Neurology, Brussels University Hospital, Brussels, Belgium; 7Department of Neonatology, Antwerp University Hospital, Antwerp, Belgium; 8Center for Developmental Disorders Ghent, Ghent, Belgium; 9Department of Pediatric Neurology, University Hospital Gasthuisberg, Leuven, Belgium; and 10Department of Neonatology, Algemeen Ziekenhuis St-Augustinus, Antwerp, Belgium.

* For members of the EPIBEL Study Group of the Flemish perinatal centers, see the Appendix.

Supported by the Foundation Marguerite Marie Delacroix and the Belgian Ministry of Health.

The authors thank the follow-up teams who were very helpful in collecting data in the centers and the children and families who participated in the study.

Corresponding author: Piet Vanhaesebrouck, Department of Neonatology, Ghent University Hospital, De Pintelaan 185 B-9000 Gent, Belgium; e-mail:

Financial Disclosure The authors have no potential conflicts of interest to disclose.

Mortality of extremely preterm infants has decreased significantly over recent years. However, concerns about the long-term development in these vulnerable infants cast shadows over the better survival rate.1–3 At present, the importance of prospective, population-based cohort studies is recognized, and the results are useful in the decision of either continuation or withdrawal of intensive care for infants with borderline viability. Because neonatal intensive care has been substantially extended and diversified in the 1990s, reports on infant cohorts studied before the surfactant era are hardly informative at present. Although the probability of disability is very high in extremely premature infants, it varies substantially according to the definitions used in the data reported.1–3 Cerebral palsy and neurosensory handicap such as blindness, deafness, and cognitive impairment are frequent consequences. The national EPICure study, while approaching the matter functionally, defined disability as the need of physical assistance to perform daily activities (ie, inability walking without assistance, blindness). It showed that 49% of the surviving children with gestational age of 25 weeks or less showed disability at the corrected median age of 30 months. Among them, 23% met the criteria for severe disability.4 Population-based reports about the outcome in such children at an older age indicate that major neurodevelopmental impairments persist5 and that the number of minor problems increases into school age.6–7

A population-based study, called the EPIBEL (Extremely Preterm Infants in Belgium) study, of infants born at 26 weeks or less of gestation was initiated in all 19 perinatal centers of Belgium.8 The overall survival rate of liveborns was 54% (175 of 322). Of these 175 survivors, 63% did show one or more major complications at the time of discharge from the neonatal intensive care unit (NICU), eg, serious neuromorbidity, chronic lung disease, and treated retinopathy of prematurity. Hence, the risk for subsequent abnormal development had been expected to be very high. This study was designed to examine the outcome of the subjects at the corrected age of 3 years and to compare the results with the ones obtained in other population-based studies.

Back to Top | Article Outline


The infants assessed in the follow-up study (ie, alive at the onset of labor) were delivered in the perinatal centers of Flanders from January 1, 1999, until January 1, 2001. Gestational age ranged from 22 to 26 completed weeks (more details available in Vanhaesebrouck et al8) (Table 1). All infants (n=95; ie, 54% of the survivors of the original EPIBEL group) born in Flanders, the Dutch-speaking region in Belgium, were invited for formal assessment. The parents were invited by letter to have a multidisciplinary assessment of their 3-year-old child at the follow-up center associated with the NICU where the newborn had been admitted. In practice, the children were formally assessed at the median corrected age of 36 months (range 30–42 months). Seventy-five percent of the children were tested between the ages of 34 and 38 months. In each center, assessment was performed by a pediatrician or pediatric neurologist, physiotherapist, and psychologist. If the latter was not available locally, the first author (I.D.G.) performed the developmental assessment (n=28). The local staff of examiners knew the status of the patients because they had already examined the majority of the children near their first or second birthdays.

Table 1

Table 1

The assessment comprised a detailed clinical examination and full developmental evaluation. The clinical evaluation included collecting the recent medical history and a global health and anthropometric assessment, as well as a standardized neurologic and sensory examination (see Appendix).9 The Dutch edition of the second version of the Bayley Scales of Infant Development (BSID-II-Nl)10 was used to assess mental and psychomotor development. The BSID-II-Nl is standardized on a mean score of 100 and a standard deviation of 15 points. In the BSID-II-Nl, a Mental Developmental Index or a Psychomotor Developmental Index score of 85 or greater is considered normal. The indexes of 70–84, 55–69, and less than 55 yield the assessments of mild, moderate, or severe impairment, respectively. The study was approved by the local ethics committee, and the parents or tutor of the subjects gave written informed consent.

The “overall disability outcome” is based on the degree of neurologic (ie, neuromotor and sensory-communicative functioning and nonfebrile seizures) and developmental (ie, Mental Developmental Index and Psychomotor Developmental Index) functioning. According to recommendations by others,4 but contrary to the classification of the Bayley Scales of Infant Development-II, the Mental Developmental Index and Psychomotor Developmental Index scores of 70 or greater were considered as normal in defining the “overall disability outcome.” The synthesized categories were “no overall disability” (ie, no significant developmental delay [Mental Developmental Index and Psychomotor Developmental Index 70 or greater] and no impairment detected in the neuromotor or sensory-communicative domain) and “severe disability” (ie, one or more severe impairments). The remaining was classified as “mild-to-moderate disability.”

Functioning in the neuromotor, sensory-communicative, and developmental domain was categorized according to the degree of impairment and labeled “none,” “mild to moderate,” and “severe” based on criteria used in previous international studies.4,11 Criteria determining the level of impairment are presented in Table 2, with the ones for severe impairment in the subdomains shown in bold type. Mental and psychomotor functioning, expressed as Mental Developmental Index and Psychomotor Developmental Index, respectively, were taken into account for discerning the levels of developmental impairment.

Table 2

Table 2

The classification of type and location of cerebral palsy was based on describing function, tone, and reflexes in each limb. In addition, it comprised the results of the neurologic examination (see Appendix). Minor central motor dysfunction was defined as poor motor coordination without clear evidence of any neurologic abnormality, such as cerebral palsy. Cerebral palsy and minor motor dysfunction (ie, neurologic classification) were not taken into account in the “overall disability outcome” (ie, functional classification), although admittedly both classifications overlap substantially.

Data were entered into a standardized database (FileMaker Pro 6.0, FileMaker Inc, Santa Clara, CA) by the attending physician and sent to either I.D.G. or P.V. Data were encoded for analysis with SPSS 12.0 (SPSS Inc, Chicago, IL), checked twice for accuracy, and outlying results verified. In case of missing or ambiguous data, the attending physician was contacted to discuss the data. Results on ordinal variables were compared by using Kruskal-Wallis (three groups) or Mann-Whitney U (two groups) analyses. Differences in categorical variables were analyzed with χ2. Differences were considered significant when P<.05.

Back to Top | Article Outline


There were 251 births recorded in this regional cohort. Ninety (36%) of them were not admitted to any NICU, and 91% of these infants were reported to have died after the onset of labor and 9% in the delivery room after intentional resuscitation efforts. Most of the intrapartum deaths (83%) occurred before 25 weeks of gestation, and 41% were late terminations of pregnancy, mainly because of the detection of a major congenital or hereditary fetal anomaly (see Vanhaesebrouck et al8 for more details). Among the remaining 161 infants, 66 infants (41%) died in the NICU, and 95 (59%) were discharged alive (Table 1). Mean (±standard deviation) birth weight was 789 (±170) g, and the mean gestational age was 25.4 (±0.7) weeks.

Of the 95 survivors, three infants died before the study of formal follow-up assessment had started. Two of them died of clinical consequences directly related to extreme prematurity (eg, chronic lung disease and tracheostomy), and the third infant succumbed after an accidental fall. A fourth child died of a fatal burning casualty shortly after the follow-up study had started, but the data on this child were included in its results. The parents of three children could not be reached, and 12 parents declined cooperation for various reasons. Hence, of 92 surviving infants, 77 three-year-old children (84%) could be formally assessed. For 12 of the 15 children who did not participate physically, valuable information was collected by contacting the local pediatrician (n=4) or developmental center (n=6) and recording neuromotor and physical development of the child. In two children, the local neonatologist performed the neurodevelopmental assessment at 12 (n=1) or at 24 months (n=1) of age. In all instances, the assessor completed the standardized electronic file on health outcome and neurologic development and transmitted the available mental and psychomotor development test results. If not available (n=5), overall estimates of the mental and psychomotor development were provided. Among these 12 children, five had no overall disability, four were mildly disabled, two were moderately disabled, and one child was severely disabled. In three cases out of the total of 92, no information could be collected (Table 1). Gestational age was 26 weeks in those three children; two of them had chronic lung disease, one showed cystic periventricular leukomalacia, and one was treated for threshold retinopathy of prematurity during NICU stay.

In general, further results pertain to the formally assessed children (n=77). A number of clearly disclosed analyses were performed on the broader denominator of children (n=89), which also included the 12 children examined by qualified health care assessors locally.

Sufficient data were available on global health and specific chronic medical treatment and support during the last 12 months before the formal assessment in 87 of the 92 long-term survivors, 47 of whom (54%) had one or more somatic difficulties. Recurrent upper (25%) and/or lower airway disease (23%) were most frequently encountered with chronic aerosol treatment in 18% of them. Chronic intestinal disorders were present in 10%, with two toddlers still dependent on gastrostomy feeding. Shunt for hydrocephalus was present in five children (6%).

The mean (±standard deviation) body weight at 36 months of age was 1.25 (±1.48) standard deviation below the mean of the specific Flemish population norms (data available for 73 children).12,13 Average (±standard deviation) head circumference (n=66) was 0.80 (±1.30) standard deviation lower, and stature (n=72) –0.76 (±1.23) standard deviation shorter than the corresponding figures in age-matched controls. Differences from the population scores were less pronounced at 36 months compared with the corresponding anthropometric data at 2–6 months (both data series available in at least 52 children). This catch-up phenomenon was most prominent for stature (from –1.65 [±1.55] to –0.76 [±1.23] standard deviation below the population mean (P=.012).

Data on neuromotor development (Table 2) of the formally assessed toddlers (n=77) are subdivided according to specific function. Severe neuromotor impairment (boldface type) was due to either dysfunction in the lower limb (no independent walking) and/or upper limb performance (incapable to dress and feed self). The sole child with gestational age of 23 weeks had a normal outcome in all neuromotor subdomains. Three children (4%, 95% confidence interval [CI] 0–8%) had severe neuromotor, and 25 (32%, 95% CI 22–42%) children had mild-to-moderate neuromotor impairment. The distribution of degrees of neuromotor impairment did not differ significantly when correlated with gestational age, gender, or plurality (data not shown).

Cerebral palsy was present in 19 (25%, 95% CI 15–35%) of the 77 formally assessed children and in 20 of 89 children in the broader study group (22%, 95% CI 13–31%). The type of cerebral palsy mainly found was pyramidal (spastic) (Table 3). In the group of children with cerebral palsy, impairment in the sensory-communicative as well as the developmental (mental or psychomotor) domain was common (47% and 42%, respectively). The majority (15 or 79%, 95% CI 61–97%) of children with cerebral palsy had also delayed mental development as formally revealed by the Bayley assessment (severe delay in eight children, moderate in three, and mild in four). An additional 13 of the 77 formally assessed children (17%, 95% CI 9–35%) had also minor central motor dysfunction.

Table 3

Table 3

Seventy-five percent of the children had no sensory-communicative impairment, 5% had severe impairment, and 19% had mild-to-moderate impairment (Table 2). There were no deaf children among these 77 subjects, but one deaf child (ie, profound hearing loss that could not be corrected with hearing aids) belonged to the “broader” study group (n=89). Two children of the formally assessed group had no useful vision, but none in the additional sample of patients. Twenty-one percent of the children had strabismus, and 7% of them needed strabismus surgery. Nine percent wore spectacles. Abnormal eye movement was noticed in 5% (n=4), deficient fixation in 6% (n=5), and inadequate tracking in 4% (n=3). In four children (5%), there was no recognizable speech at 36 months, despite the fact that hearing was unaffected in three of the latter.

Mental (Mental Developmental Index) and psychomotor (Psychomotor Developmental Index) Bayley development indexes were available for 62 children (80%). Formal testing could not be performed in five children (6%) because of their severe sensory impairment and/or severe global disability. In three (4%) children, the Brunet-Lézine scale14 was administered because of their different mother tongue, and in seven (9%) children the first version of the Bayley Scales of Infant Development15 was still in use at the follow-up center. The results obtained in the formally assessed patient group are presented in Table 4 and subdivided according to gestational age, gender, and plurality. Data pertaining to the only child of 23 weeks of gestation were added to the data for children with a gestational age of 24 weeks. This child had a normal Mental Developmental Index and Psychomotor Developmental Index.

Table 4

Table 4

The average (±standard deviation) Mental Developmental Index was 81.2 (±18.8) and the mean Psychomotor Developmental Index was 73.2 (±17.8). Overall, 70% (95% CI 60–80%) had a Mental Developmental Index and/or Psychomotor Developmental Index more than 1 standard deviation below the population mean. Only one in three (30%, 95% CI 20–40%) and less than half of the children (44%, 95% CI 33–55%) showed normal psychomotor (Psychomotor Developmental Index 85 or greater) and mental (Mental Developmental Index 85 or greater) development, respectively. Mental and psychomotor outcome did not differ significantly when compared according to either gestational age, gender, or multiple birth (all P>.05).

When only children without minor central motor dysfunction and without cerebral palsy were taken into account (n=45), 62% of the children had a normal Mental Developmental Index, and the proportion of severe mental delay did drop from 18% to 7%. The subgroup of children with a normal Psychomotor Developmental Index (85 or greater) increased from 30% to 38%, and that of children with severe psychomotor delay (Psychomotor Developmental Index less than 55) decreased from 27% to 16%.

Results on the overall disability outcome are presented in Table 5 and subdivided according to either number of NICU admissions (n=161) or number of infants discharged alive (n=95) as denominator. The degree of overall disability was similar, irrespective of gestational age (categorized in groups of 24 or less, 25, and 26 completed weeks, P>.05), gender, and plurality (data not shown). In Figure 1, the distributions of the various types of overall outcome at 3 years are shown, weighted either against the primary denominator (column A), the number of NICU admissions (B), the total number of live births (C), or the complete cohort of 251 recorded extreme preterm births in Flanders during the study time (D).

Table 5

Table 5



An “all-in” calculation of poor outcome (ie, disability or prematurity-related death) among the 95 infants discharged alive can be estimated to be 58% (95% CI 48–68%) (n=55), representing 25 (26%, 95% CI 17–25%) mildly-to-moderately disabled children and 28 (29%, 95% CI 20–38%) severely disabled toddlers, as well as two postdischarge deaths directly related to prematurity. As proposed by Tin et al,16 it was anticipated that, with high probability, 3 year olds with extreme preterm birth would be evenly distributed over the three types of outcome discerned. Therefore, in this adding-up the three children lost to follow-up were subdivided, one in each outcome fraction.

The Venn diagram (Fig. 2) pertains to the 51 (66%) children with formally documented overall disability subdivided according to the specific domain of impairment (neuromotor, sensory-communicative, Mental Developmental Index, and Psychomotor Developmental Index less than 70). The most frequent type (17%, 95% CI 9–25%) of overall disability was an encompassed impairment on each one of the four domains (13 of 77). In the majority of these 13 children, at least one severe impairment was observed. The combination of neurologic, mental, and psychomotor impairment (16%) and the disability limited to the mental and psychomotor developmental domains (14%) formed almost equally large groups.



Back to Top | Article Outline


The survival rate of liveborn infants with a gestational age of 26 weeks or less was 56% at the time of discharge from hospital in this population-based study. Outborn infants were excluded from the study because the outcome of these infants is substantially worse, often because of suboptimal circumstances and care.17,18 Formal recording of outcome and morbidity was done at 3 years of age, regarding global health on the one hand and different domains of neurologic and developmental functioning on the other.

In more than half of the surviving extremely preterm infants, significant health problems required regular medical attention, pointing to the long-term consequences of neonatal respiratory and gastrointestinal complications.8 In addition, extreme prematurity is complicated by sustained poor somatic growth persisting into preschool age.

Remarkably, about two thirds (63%) of the toddlers had no neuromotor deficit or dysfunction upon formal neurologic examination only. However, psychomotor developmental assessment (Psychomotor Developmental Index) by application of the Bayley Scales found normal psychomotor development in only 30% of the children. Hence, at age 3 years, delay of psychomotor development was a common finding, even when clear neuromotor dysfunction could not be documented. Our data confirm the findings by Wood et al4 that sole consideration of neuromotor deficits overestimates motor ability in preterm children.

More than half (56%) of the formally assessed children had delayed mental development, defined as a Mental Developmental Index below 85. In about one out of five children (18%), this delay was severe, with Mental Developmental Index below 3 standard deviations (ie, less than 55) of the population mean. However, omitting all children with either minor central motor dysfunction or cerebral palsy showed normal mental development (Mental Developmental Index 85 or greater) in 62% of the residual group. Such exercise points to comorbidity of central motor deficits and dysfunctions. To a lesser degree, it also illustrates the lack of specificity of some developmental tests such as the mental scale of the Bayley assessment that, while intending to estimate mental development, also includes items on language, personal, and social aspects. Moreover, performance on the mental scale can also be influenced by specific motor abilities.10 Children with psychomotor delay may therefore be at a disadvantage when mental abilities are judged by a developmental test. Fortunately, intelligence tests designed for older preschoolers are less flawed by the inherent interaction of mental and motor ability. Hence, potentially abnormal physical, neuromotor, and/or cognitive development can be assessed more accurately at a later age in preterm children, both with and without central motor deficit.

Only 40% of the 3-year-old children who survived to discharge had no disability. Severe disability was identified in 28% of them. A mental and psychomotor developmental index of 70 or greater was considered “normal” in the overall disability outcome assessment according to internationally recommended practice, based on the fact that the predictive value of subnormal mental and psychomotor performance in extremely preterm infants for subsequent school age development is poor.19 However, if children with a suboptimal score (ie, Mental Developmental Index and/or Psychomotor Developmental Index 70–84) on the Bayley Scales of Infant Development had been included in the “mild-to-moderate disability” group, the overall rate of children free of disability would have decreased from 40% to 26%. Toddlers with a Mental Developmental Index or Psychomotor Developmental Index of 70–84 are at considerable risk for developing subsequent problems. Hence, here is a second reason for follow-up assessment at later ages to yield more precise and trustworthy data on the outcome in extremely preterm infants.

Regarding gestational age, the results suggest that this study group of extremely preterm children is fairly homogeneous in outcome. This is in line with the EPICure study, which reported that, according to differences in gestational age, no significantly different types of outcome at 30 months were observed.4 However, this is at variance with the hospital discharge results recorded in this study, which indicate that short-term outcome improved markedly between 24 and 26 weeks of gestation.8 This supports the idea that outcome at discharge is different from that at longer term. Moreover, the decision whether to start or withhold active curative care cannot solely be based on gestational age. Not in keeping with other studies, we did not find a higher rate of disability in males.3,4,20

Rates of disabilities reported in the literature vary strongly, due to differences in definitions, in patients as well as in time period and age selected. Hence, comparison of study results is challenging.1,20 To partly resolve this problem, we have adopted the definitions used in recent cohort studies and have presented various denominators to calculate several disability rates. The overall rate of neuromotor impairment was 36% (95% CI 25–47%), but this increased to 43% when children with a gestational age of 26 weeks were excluded. It amply surpassed the 24% found in the EPICure study. However, the rate of severe neuromotor impairment was only half as large in our study (5% compared with 10%). Ten percent of the children in the EPICure study did not walk without assistance, compared with only 2% in this cohort. This illustrates once more the importance of selection criteria. The international criteria for severe neuromotor impairment were developed for 2-year-old children.11 The average age of the toddlers in our study was about 6 months more than that of the EPICure children. Hence, they had more time to reach the set milestones.

The rate of cerebral palsy (25%, 95% CI 15–35%) found in our study is higher than in comparable multicenter studies of extremely preterm infants, but still within the margin of error.21 Apparently this higher rate is due to a higher proportion of mild cerebral palsy. In the event of equivocal or mild motor problems, cerebral palsy may not be recognized clinically and diagnosed as a Developmental Coordination Disorder. However, when results of neuroimaging (ie, white matter disease, enlarged cerebral ventricles) were taken into account, a mild form of cerebral palsy could be diagnosed in a substantial number of those children. Continuation of follow-up efforts in our study group should answer the question whether or not the mild cerebral palsy evolves into a clinical diagnosis of Developmental Coordination Disorder.21 Sorting out these hypotheses is important because reports are available about children who “outgrow” their cerebral palsy, diagnosed in early childhood. Similarly, in some children cerebral palsy can be diagnosed only later in development.22–24

The rate of severe (5%) and moderate-to-mild (19%) impairment in the sensory-communicative domain are within the range of that encountered in other multicenter studies, reporting a 1–5% severe hearing and/or visual impairment.4,20 Not correcting for gestational age would have resulted in higher overall disability rates. This practice seems justified because the EPICure study revealed that the rate of severe disability at 6 years of chronological age was similar to that at 30 months of corrected age.6 However, the outcome of some children categorized as “mildly-to-moderately” disabled at the age of 30 months evolved to a normal outcome at 6 years of age, pointing to the low predictivity of mild-to-moderate disabilities for subsequent preschool outcome.

In conclusion, in this first Flemish long-term follow-up study, almost two out of three extremely preterm children (58%) discharged alive from the NICU have an overall disability, of whom more than half (32%) are categorized as severe. The high rate of neurodevelopmental deficits in early childhood confirms the extreme vulnerability of this population, although, remarkably, at 3 years of age no disability was detectable in the various domains of neurodevelopment in 42% of the children. As normal neurologic development becomes more complex and refined with age, efforts to further monitor this at-risk population becomes very relevant. As in many others, the results in this study point to the paramount importance of extending the stay in utero until there is substantial evidence that the extremely preterm fetus is seriously compromised. Finding early predictors of adverse outcome8 and determining best perinatal practices are real challenges to improve survival without long-term morbidity in extremely preterm infants.3

Back to Top | Article Outline


1. Hack M, Fanaroff AA. Outcomes of children of extremely low birthweight and gestational age in the 1990s. Semin Neonatol 2000;5:89–106.
2. Watts JL, Saigal S. Outcome of extreme prematurity: as information increases so do the dilemmas. Arch Dis Child Fetal Neonata Ed 2006;91:F221–5.
3. Fanaroff AA, Stoll BJ, Wright LL, Carlo WA, Ehrenkranz RA, Stark AR, et al. Trends in neonatal morbidity and mortality for very low birthweight infants. Am J Obstet Gynecol 2007;196:147.e1–8.
4. Wood NS, Marlow N, Costeloe K, Gibson AT, Wilkinson AR. Neurologic and developmental disability after extremely preterm birth. N Engl J Med 2000;343:378–84.
5. Doyle LW, Anderson PJ, and the Victorian Infant Collaborative Study Group. Improved neurosensory outcome at 8 years of age of extremely low birthweight children born in Victoria over three distinct eras. Arch Dis Child Fetal Neonata Ed 2005;90:F484–8.
6. Marlow N, Wolke D, Bracewell MA, Samara M. Neurologic and developmental disability at six years of age after extremely preterm birth. N Engl J Med 2005;352:9–19.
7. Mikkola K, Ritari N, Tommiska V, Salokorpi T, Lehtonen L, Tammela O, et al. Neurodevelopmental outcome at 5 years of age of a national cohort of extremely low birth weight infants who were born in 1996–1997. Pediatrics 2005;116:1391–400.
8. Vanhaesebrouck P, Allegaert K, Bottu J, Debauche C, Devlieger H, Docx M, et al. The EPIBEL study: outcomes to discharge from hospital for extremely preterm infants in Belgium. Pediatrics 2004;114:663–75.
9. Evans P, Johnson A, Mutch L, Alberman E. A standard form for recording clinical findings in children with a motor deficit of central origin. Dev Med Child Neurol 1989;31:119–27.
10. van der Meulen BF, Ruiter SA, Spelberg LHC, Smrkovský M. Bayley Scales of Infant Development (BSID) – II. Nederlandse versie. Lisse, The Netherlands: Swets Test Publishers; 2002.
11. Johnson A. Follow-up studies: a case for standard minimum data set. Arch Dis Child Fetal Neonatal Ed 1997;76:F61–3.
12. Devlieger H, Martens G, Bekaert A, Eeckels R. Standaarden van geboortegewicht-voor-zwangerschapsduur voor de Vlaamse boreling. Tijdschr. voor Geneeskunde 2000;56:1–14.
13. Laboratorium Antropogenetica Vrije Universiteit Brussel en Dienst Jeugdgezondheidszorg Katholieke Universiteit Leuven. Groeicurven Vlaanderen 2004. Available at: Retrieved July 30, 2007.
14. Josse D. Le Brunet-Lézine révisé. CPA, Issy les Moulinaux, 1997.
15. van der Meulen BF, Smrkovský M. Bayley Ontwikkelingsschalen (BOS 2-30). Lisse, The Netherlands: Swets & Zeitlinger; 1983.
16. Tin W, Fritz S, Wariyar U, Hey E. Outcome of very preterm birth: children reviewed with ease at 2 years differ from those followed up with difficulty. Arch Dis Child Fetal Neonatal Ed 1998;79:F83–7.
17. Chien LY, Whyte K, Aziz P, Thiessen D, Matthew SK, Lee, SK, and the Canadian Neonatal Network. Improved outcome of preterm infants when delivered in tertiary care centers. Obstet Gynecol 2001;98:247–52.
18. Yu VY. Is neonatal intensive care justified in all preterm infants? Croat Med J 2005;46:744–50.
19. Hack M, Taylor G, Drotar D, Schluchter M, Cartar L, Wilson-Costello D, et al. Poor predictive validity of the Bayley Scales of Infant Development for cognitive function of extremely low birth weight children at school age. Pediatrics 2005;116:333–41.
20. Msall ME. Neurodevelopmental surveillance in the first 2 years after extremely preterm birth: evidence, challenges and guidelines. Early Hum Dev 2006;82:157–66.
21. Hadders-Algra M. Developmental coordination disorder: is clumsy motor behaviour caused by a lesion of the brain at early age? Neural Plast 2003;10:39–50.
22. Ford GW, Kitchen WH, Doyle LW, Rickards AL, Kelly E. Changing diagnosis of cerebral palsy in very low birthweight children. Am J Perinatol 1990;7:178–81.
23. Nelson KB, Ellenberg JH. Children who “outgrew” cerebral palsy. Pediatrics 1982;69:529–36.
24. Ross G, Lipper EG, Auld PA. Consistency and change in the development of premature infants weighting less than 1,501 grams at birth. Pediatrics 1982;76:885–91.
Back to Top | Article Outline


The following items recorded in a standardized file were used in the analyses: perinatal center number, year of admission, patient number, date of birth/admission to neonatal intensive care unit/discharge; gender; gestational age; multiple birth; died after discharge (yes-no, date, cause), weight, length, head circumference (at birth, discharge, 2–6 months and 2–3 years corrected age); somatic data (respiratory problems upper airway, respiratory problems lower airway, cardiac problems, “failure to thrive,” gastrointestinal problems, epilepsy partial simple, epilepsy complex, epilepsy generalized, orthopedic problems, genetic problems, other, unknown); date of examination; abnormal unwanted movements (at rest: none, short and jerky, slow and writhing, tremor, flexor/extensor spasms; with excitement or goal-directed movement: none, short and jerky, slow and writhing, tremor, flexor/extensor spasms, incoordination not secondary to increased tone, abnormal posture/grimacing resulting from voluntary movements elsewhere, other); tone (for each limb separately: within normal range, increased, decreased, varying); distribution of involvement (right/left asymmetry of tone or function, which side worse, upper/lower limbs more affected); head and neck, sitting, gait, hand use (see Table 2 for criteria); central motor deficit (no, minor central motor dysfunction, central motor deficit suspected, central motor deficit present; location: monoparesis, hemiparesis, diparesis, triparesis, quadriparesis; degree: mild, moderate, severe, and type: see Table 3 for criteria); visual impairment (yes-no-uncertain, spectacles, usual vision: see Table 2 for criteria, squint, operation for squint; abnormal eye movements, fixation problems, tracking problems); hearing impairment (yes-no-uncertain, hearing aids, usual hearing: see Table 2 for criteria); communication difficulties (yes-no-uncertain: usual communication: see Table 2 for criteria); seizures (yes-no-uncertain, type: generalized, petit mal, focal, myoclonic, akinetic, other, unknown; age); Bayley Scales of Infant Development-II-Nl: mental and psychomotor scales (age, raw score, Mental Developmental Index (MDI), Psychomotor Developmental Index (PDI), testing age).

All members of the EPIBEL Study Group of the eight perinatal centers in Flanders, Belgium, are neonatologists who contributed data to the follow-up study and whose invaluable help we gratefully acknowledge: Brussels University Hospital: Adel Bougatef, Filip Cools; Algemeen Ziekenhuis Middelheim (Antwerp): Hilde Van de Broek, Martine Docx; Algemeen Ziekenhuis St-Augustinus (Antwerp): Christine Vandeputte; Algemeen Ziekenhuis St-Jan Brugge: Dominiek Lecoutere, Paul Van Laer; Algemeen Ziekenhuis Zuid-Oost Limburg (Genk): Claire Theyskens; Antwerp University Hospital: Patrick Van Reempts, Bart Van Overmeire; University Hospital Gasthuisberg Leuven: Hugo Devlieger, Karel Allegaert; University Hospital Ghent: Koenraad Smets.

© 2007 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.