Sex of the child

Low quality evidence from two studies found no association between sex of the child and feeding problems. One study (n=1151) examined the association between sex and no independent feeding at 18–22 months corrected age among children born with birth weight <1000 g (Vohr 2000). Another study (n=584) found no association between sex of the child and feeding difficulties at 2 years (corrected age) among moderate to late children born preterm (32–36 weeks) (Johnson 2016).

Small for gestational age

Low quality evidence from two studies show somewhat mixed results. One low quality study (n=1151) examined the association between being preterm and small for gestational age and no independent feeding at 18–22 months corrected age among children born with birth weight <1000 g (Vohr 2000). No significant association was found. Another low quality study (n=584) found a borderline significant increased risk of feeding difficulties at 2 years of corrected age among children born small for gestational age at 32–36 weeks of gestation (Johnson 2016).

Ethnicity

Low quality evidence from one study (n=1151) examined the association between the ethnicity or race of the preterm child and no independent feeding at 18–22 months corrected age (Vohr 2000). No significant association was found.

Brain abnormalities

Low quality evidence from one study (n=1151) among children born with birth weight <1000g found an increased odds of lack of independent feeding at 18–22 months corrected age with neonatal intraventricular haemorrhage (IVH) grade III-IV (Vohr 2000).

Sepsis

Low quality evidence from one study (n=1151) among children born with birth weight <1000g found no association between neonatal culture-proven sepsis (neither early-onset nor late-onset) and lack of independent feeding at 18–22 months of corrected age (Vohr 2000).

Retinopathy of prematurity (ROP)

No evidence was identified on the relationship between ROP and functional problems with feeding.

Necrotising enterocolitis (NEC)

Low quality evidence from one study (n=1151) among children born with birth weight <1000g found no association between NEC and lack of independent feeding at 18–22 months of corrected age (Vohr 2000).

Antenatal exposure to steroids

Low quality evidence from one study (n=1151) among children born with birth weight <1000g found no association between antenatal exposure to steroids and lack of independent feeding at 18–22 months of corrected age (Vohr 2000).

Postnatal exposure to steroids

Low quality evidence from one study (n=1151) among children born with birth weight <1000g found no association between postnatal exposure to steroids and lack of independent feeding at 18–22 months of corrected age (Vohr 2000).

Bronchopulmonary dysplasia (BPD)

Low quality evidence from one study (n=1151) showed an increased odds of lack of independent feeding at 18–22 months of corrected age with bronchopulmonary dysplasia at 36 weeks among children born with birth weight <1000 g (Vohr 2000).

Socioeconomic status

Low quality evidence from one study (n=584) found no association between socioeconomic status and feeding difficulties at 2 years (corrected age) among children born at 32–36 weeks of gestation (Johnson 2016).

Maternal substance abuse

No evidence was identified.

Multiple pregnancy

No evidence was identified.

Chorioamnionitis

No evidence was identified.

Neglect

No evidence was identified.

Maternal age

No evidence was identified.

Maternal mental health disorder

No evidence was identified.

Sex of the child

Low quality evidence from two studies (n=246 and n=1151) found no associations between the sex of the child and motor delay (PDI <70 and lack of independent walking) among preterm babies (born at <25 weeks of gestation or with birth weight of 401–1000 g), assessed at 18–22 months of corrected age (Shankaran 2004; Vohr 2000).

Small for gestational age (SGA)

Low quality evidence from one study (n=1151) found no association between being born SGA and PDI score of <70 and lack of independent walking at 18–22 months of corrected age among children born with birth weight 401–1000 g (Vohr 2000).

Ethnicity

Low quality evidence from two studies (n=246; n=1151) on the relationship between ethnicity/race and motor delay among children born preterm show no association among preterm children (born at <25 weeks of gestation or with birth weight of 401–1000 g), on PDI <70 (Shankaran 2004; Vohr 2000) and lack of independent walking (Vohr 2000) between black and non-black children (Shankaran 2004) and between white and non-white children (Vohr 2000) assessed at 18–22 months of corrected age with BSID.

Brain abnormalities

Low to moderate quality evidence from four studies (sample sizes ranging from 246 to 6161) was available on the relationship between neonatal brain lesions among children born preterm (born at <28 weeks of gestation or with birth weight <1000 g) and motor delay at 18–24 months corrected age (Adams-Chapman 2008; O’Shea 2008; Shankaran 2004; Vohr 2000). All studies found increased odds of PDI <70 with different types of brain lesions (intraventricular haemorrhage [IVH], IVH grade III-IV, IVH III with shunt, IVH IV with shunt, periventricular leukomalacia [PVL], cystic PVL, early PVL, periventricular haemorrhagic infarction). One study (n=1151) also found an association with IVH or PVL grade III-IV and lack of independent walking (Vohr 2000). One publication (n=246) found no association between intracranial haemorrhage (ICH) grade III-IV and PDI <70 (Shankaran 2004).

Sepsis

Low to high quality evidence from four studies (sample sizes ranging from 1151 to 6314) on the relationship between neonatal sepsis and motor delay show mixed results (Martin 2010; Stoll 2004; Vohr 2005; Vohr 2000). High quality evidence from a study (n=1155) found no association between culture-proven late-onset neonatal sepsis and abnormal PDI at 2 years of age (Martin 2010). Moderate quality evidence from another study found an increased odds of abnormal PDI score at 18–22 months corrected age among preterm children (with birth weight 1000 g or less) that had had neonatal culture-proven sepsis with antibiotic therapy for more than five days, that had had neonatal sepsis with NEC, and that had had neonatal meningitis with or without sepsis (Stoll 2004). Low to moderate quality evidence from two publications of the same study project examining cohorts born at different times (n=3785 and n=1151) found no association between sepsis and abnormal PDI score at 18–22 months corrected age (Vohr 2005; Vohr 2000). The latter also did not fund an association between sepsis and lack of independent walking.

Retinopathy of prematurity (ROP)

Moderate quality evidence from one study on the association between different severities of ROP (vs no ROP) and abnormal PDI score (either <55 or 55–69) show mixed findings (Allred 2014). The evidence shows a general tendency of increased odds of abnormal PDI score for all severities of ROP, however, not all of them reached statistical significance. ROP stage 3+, however, showed significantly increased odds of PDI <55 and PDI 55–69. The children were born earlier than 28 weeks of gestation and they were assessed at 24 months of age.

Necrotising enterocolitis (NEC)

Low to high quality evidence from four studies (sample sizes ranging from 865 to 2948) on the association between necrotising enterocolitis (NEC) and psychomotor development (assessed by BSID) show somewhat mixed results (Hintz 2005; Martin 2010: Shah 2012; Vohr 2000). High quality evidence from one study (n=1155) and moderate quality evidence from another study (n=2948) showed a significant increase in the odds of an abnormal PDI for preterm infants (23 to 27+6 weeks of gestation or birth weight of 401–1000 g) who had NEC requiring surgery but not for ones with medically managed NEC (Hintz 2005; Martin 2010). Moderate quality evidence from one study (n=865) showed an increased odds of abnormal PDI score with NEC grade II or higher and low quality evidence from another study (n=1151) showed an increased odds of abnormal PDI score with NEC (unspecified) (Shah 2004; Vohr 2000). The same low quality publication also reported that there was no association between NEC and lack of independent walking (Vohr 2000). All outcomes were assessed at around 2 years of age.

Antenatal exposure to steroids

Low to moderate quality evidence from five studies on the association between antenatal steroid exposure and motor delay (assessed by BSID) show mixed results (Carlo 2011; Laughon 2009; Shankaran 2004; Vohr 2005; Vohr 2000). Moderate quality evidence from two studies (n=4924; n=3785) found reduced odds of PDI score <70 at 18–22 months of corrected age among preterm children (born 22–32 weeks of gestation) with exposure to antenatal steroids (Carlo 2011; Vohr 2005). The first study also performed stratified analysis for each week of gestation (from 22 to 25 weeks), the findings are mixed but largely did not reach statistical significance. Low quality evidence from two other studies (n=246; n=1151) found no association between antenatal steroids and PDI <70 at 18–22 months of corrected age among extremely preterm children (<25 weeks of gestation or with birth weight 401–1000 g) (Shankaran 2004; Vohr 2000). The latter publication also found no association on lack of independent walking. Moderate quality evidence from one study (n=915) found an increased odds of PDI score <55 among preterm children (born <28 weeks of gestation) at 24 months of age (Laughon 2009).

Postnatal exposure to steroids

Low to moderate quality evidence from two studies (=3785 and n=1151, respectively) on the relationship between postnatal exposure to steroids and motor delay found an increased odds of PDI score <70 (Vohr 2005; Vohr 2000). The latter publication also found an increased odds of lack of independent walking. The children were born at 22–32 weeks of gestation or with birth weight 401–1000 g and assessed at 18–22 months of corrected age.

Bronchopulmonary dysplasia (BPD)

Low to moderate quality evidence from four studies (sample sizes ranging from 246 to 3785) on the association between bronchopulmonary dysplasia (BPD, need of additional oxygen at 36 weeks) and motor delay show mixed results (Laughon 2009; Shankaran 2004; Vohr 2005; Vohr 2000). Moderate quality evidence from one study (n=915) found no association with PDI score of <55 when looking at BPD without mechanical ventilation and a near-significant association when looking at BPD with mechanical ventilation among children born <28 weeks of gestation and assessed at 24 months of age (Laughon 2009). Low to moderate quality evidence from two publications from one large study project (n=3785 and n=1151, respectively) found an increased odds of PDI <70 at 18–22 months of age with BPD among children were born at 22–32 weeks of gestation or with birth weight 401–1000 g (Vohr 2005; Vohr 2000). The latter publication also found an association with lack of independent walking. Low quality evidence from one study (n=246) found no association among children born <25 weeks of gestation and assessed at 18–22 months of corrected age (Shankaran 2004).

Socioeconomic status

Low quality evidence from one study (n=246) found no association between socioeconomic status (household income <$20000/year vs >=€20000) and PDI <70 (assessed by BSID) among children born at <25 weeks of gestation and assessed at 18–22 months of corrected age (Shankaran 2004).

Maternal substance abuse

Low quality evidence from one study (n=82) found a significant association between maternal cocaine use and abnormal psychomotor developmental index score (BSID) at three years of age among children born with birth weight <1500 g (Singer 2001).

Multiple pregnancy

No evidence was identified.

Chorioamnionitis

No evidence was identified.

Neglect

No evidence was identified.

Maternal age

No evidence was identified.

Maternal mental health disorder

No evidence was identified.

Sex of the child

Moderate quality evidence from two studies (n=638; n=834) showed increased odds of developmental delay (identified using screening tools) for male preterm children as compared to females (Johnson 2015a; Kerstjens 2013). Developmental delay were assessed by ASQ in the first publication; and moderate to severe cognitive impairment was assessed by PARCA-R screening tool in the second publication. These children were born at 32 to 36 weeks and were assessed at 2 years of corrected age in the first study and at 43 to 49 months of age in the second study.

Small for gestational age (SGA)

Moderate quality evidence from one study (n=834) showed an increase in the risk of developmental delay (assessed by ASQ) for SGA preterm children, when compared to those preterm children born appropriate for gestational age (Kerstjens 2013). The children were assessed at between 43 and 49 months of age, and were born at 32 to 35 weeks.

Ethnicity

Moderate quality evidence from one study (n=1403) found an increased odds of moderate to severe cognitive impairment (assessed by PARCA-R) among non-white children compared with white children (born at 32–36 weeks of gestation) assessed at 2 years of corrected age even after adjusting for socioeconomic status (Johnson 2015).

Brain abnormalities

No evidence was identified.

Sepsis

Moderate quality evidence from one study (n=832) found no association between neonatal sepsis (defined as clinical symptoms and at least one positive blood culture) and developmental delay (ASQ total problems <2SD) among children born at 32–35 weeks of gestation and assessed at 43–49 months of age (Kerstjens 2012a).

Retinopathy of prematurity (ROP)

No evidence was identified.

Necrotising enterocolitis (NEC)

No evidence was identified.

Antenatal exposure to steroids

Moderate quality evidence from one study (n=834) found no association between antenatal exposure to steroids and developmental delay (ASQ total problems <2SD) among children born at 32–35 weeks of gestation and assessed at 43–49 months of age (Kerstjens 2013).

Postnatal exposure to steroids

No evidence was identified.

Bronchopulmonary dysplasia (BPD)

No evidence was identified.

Socioeconomic status

Moderate quality evidence from one study (n=1403) on the association between socioeconomic status and moderate to severe cognitive impairment show that lower socioeconomic status was associated with increased odds of cognitive impairment (Johnson 2015a). This study included children born at 32–36 weeks of gestation and they were assessed at 2 years of corrected age using PARCA-R screening tool.

Maternal substance abuse

No evidence was identified.

Multiple pregnancy

Moderate quality evidence from one study (n=834) shows an association between multiple pregnancy and developmental delay (ASQ total problems <2Sd) among children born at 32–35 weeks of gestation and assessed at 43–49 months of age (Kerstjens 2013).

Chorioamnionitis

No evidence was identified.

Neglect

No evidence was identified.

Maternal age

Moderate quality evidence from one study (n=834) found no association between maternal age under 20 years and developmental delay (ASQ total problems <2SD) among children born at 32–35 weeks of gestation and assessed at 43–49 months of age (Kerstjens 2013).

Maternal mental health disorder

Moderate quality evidence from one study (n=834) found no association between maternal mental illness and developmental delay (ASQ total problems <2SD) among children born at 32–35 weeks of gestation and assessed at 43–49 months of age (Kerstjens 2013).

Sex of the child

Low to moderate quality evidence from two studies (three publications, sample sizes ranging from 625 to 1228) shows no association between child’s sex and behavioural problems among children born preterm (Delobel-Ayoub 2009; Delobel-Ayoub 2006; Johnson 2015b). The first study assessed children born <33 weeks of gestation at 3 and 5 years of age with the SDQ (Delobel-Ayoub 2006; Delobel-Ayoub 2009) and the second study assessed moderate to late preterm (32–36 weeks) children at two years corrected age on delayed socioemotional competence (assessed with BITSEA) (Johnson 2015b).

Small for gestational age (SGA)

Moderate quality evidence from two studies (n=1228; n=1277) showed no difference in total behavioural difficulties for SGA preterm infants as compared to those who were appropriate for gestational age. Children were assessed at 3 to 5 years of age and were born at 22–32 weeks. However, one of these studies did observe an increase in the risk of inattention-hyperactivity symptoms for SGA preterm infants born at 29–32 weeks (Delobel-Ayoub 2006; Guellec 2011). In addition, low quality evidence from one study (n=625) found no association between being born SGA and delayed socioemotional competence (assessed with BITSEA) at 2 years (corrected age) among moderate to late children born preterm (Johnson 2015b).

Ethnicity

Low quality evidence from one study (n=625) show an association between being non-white and delayed socioemotional competence (assessed with BITSEA) at 2 years (corrected age) among moderate to late children born preterm (Johnson 2015b).

Brain abnormalities

Moderate quality evidence from one study show an increase in the risk of behavioural difficulties (assessed with the SDQ) for preterm infants with major cerebral lesions when assessed at the age of 3 years (Delobel-Ayoub 2006). The children were born at 22–32 weeks, and 1228 children were included. The same study (different publication, n=1102) conducted further follow-up at 5 years of age and found no association between brain lesions (level of severity not considered) and behavioural problems (Delobel-Ayoub 2009).

Sepsis

No evidence was identified.

Retinopathy of prematurity (ROP)

No evidence was identified.

Necrotising enterocolitis (NEC)

No evidence was identified.

Antenatal exposure to steroids

Low quality evidence from one study (n=625) show no association between exposure to antenatal steroids and delayed socioemotional competence (assessed with BITSEA) at 2 years (corrected age) among moderate to late children born preterm (Johnson 2015b).

Postnatal exposure to steroids

No evidence was identified.

Bronchopulmonary dysplasia (BPD)

Moderate quality evidence from one study (n=1228) did not show any difference in the risk of behavioural problems for preterm infants who had bronchopulmonary dysplasia, as compared to those who did not (Delobel-Ayoub 2006). The children were born at 22–32 weeks and followed up at 3 years of age.

Socioeconomic status

Low to moderate quality evidence from three studies show mixed results on behavioural outcomes in relation to socioeconomic status. Moderate quality evidence from one study (n=1102) show no association between socioeconomic status and behavioural problems (assessed with the SDQ) in very preterm (22–32 weeks) at 5 years (Delobel-Ayoub 2009). Moderate quality evidence from another study (n=1458) show an increase in the odds of behavioural problems and internalising problems (assessed with the CBCL) for children born to families with lower socioeconomic status (Potijk 2015). Increased odds of externalising problems was borderline significant. This study included children born between 32 and 41 weeks of gestation and followed up at 4 years. Low quality evidence from a third study (n=625) found an association between lower socioeconomic status and delayed socioemotional competence (assessed with BITSEA) at 2 years of age (corrected) among moderate to late children born preterm (Johnson 2015b).

Maternal substance abuse

Low quality evidence from one study (n=625) show an association between recreational use of drugs during pregnancy and delayed socioemotional competence (assessed with BITSEA) at 2 years (corrected age) among moderate to late children born preterm (Johnson 2015b).

Multiple pregnancy

Low quality evidence from one study (n=625) show no association between multiple pregnancy and delayed socioemotional competence (assessed with BITSEA) at 2 years (corrected age) among moderate to late children born preterm (Johnson 2015b).

Chorioamnionitis

No evidence was identified.

Neglect

No evidence was identified.

Maternal age

Moderate quality evidence from one study (n=1228) show an increase in the risk of behavioural problems (assessed by SDQ) for preterm infants (born at 22–32 weeks gestation and followed up at 3 years of age) born to mothers less than 25 years (compared with mothers 25–34 years) (Delobel-Ayoub 2006). Maternal age of 35 years or more was not associated with behavioural problems in this study. When the children were followed up at 5 years of age (n=1102), the increased odds of behavioural problems of preterm children of mothers younger than 25 years at the time of birth remained borderline significant (Delobel-Ayoub 2009). The association between maternal age 35 years or older and behavioural problems also became borderline significant (borderline reduced odds of behavioural problems compared with maternal age of 25–34 years).

Maternal mental health disorder

Moderate quality evidence from one study (n=1102) show an increase in the risk of behavioural problems (assessed by the SDQ) at 5 years of age for preterm children (born at 22–32 weeks) born to mothers with poorer self-reported mental health (Delobel-Ayoub 2009).

Sex of the child

Low quality evidence from one study (n=219) show male children born at <26 weeks of gestation to be more likely to be provided special educational needs support at 11 years compared to their female peers (Johnson 2011).

Small for gestational age (SGA)

Low quality evidence from one study (n=1439) that examined the association between being born SGA and having school difficulties (defined as needing special schooling or having low grades, reported by parents) among children born preterm at eight years of age was available (Guellec 2011). No association was found among children born SGA born at 24–28 weeks of gestation but an increased odds of school difficulties was found among children born SGA born at 29–32 weeks of gestation.

Ethnicity

Low quality evidence from one study (n=219) show no association between maternal ethnicity and special educational needs among extremely children born preterm at 11 years of age (born at <26 gestational weeks) (Johnson 2011).

Brain abnormalities

Low quality evidence from one study (n=219) show a significant association between abnormal cerebral ultrasound scan and special educational needs at 11 years among children born at <26 gestational weeks (Johnson 2011).

Sepsis

No evidence was identified.

Retinopathy of prematurity

No evidence was identified.

Necrotising enterocolitis (NEC)

Low quality evidence from one study (n=219) show no association between necrotising enterocolitis and special educational needs among extremely children born preterm at 11 years of age (born at <26 gestational weeks) (Johnson 2011).

Antenatal exposure to steroids

Low quality evidence from one study (n=219) show no association between any exposure to antenatal steroids and special educational needs among extremely children born preterm at 11 years of age (born at <26 gestational weeks) (Johnson 2011).

Postnatal exposure to steroids

Low quality evidence from one study (n=219) show no association between any exposure to postnatal steroid for chronic lung disease and special educational needs among extremely children born preterm at 11 years of age (born at <26 gestational weeks) (Johnson 2011).

Bronchopulmonary dysplasia (BPD)

No evidence was identified.

Socioeconomic status

Low quality evidence from one study (n=219) show no association between socioeconomic status and special educational needs among extremely children born preterm at 11 years of age (born at <26 gestational weeks) (Johnson 2011).

Maternal substance abuse

No evidence was identified.

Multiple pregnancy

No evidence was identified.

Chorioamnionitis

Low quality evidence from one study (n=219) show no association between chorioamnionitis (suspected or proven) and special educational needs among extremely children born preterm at 11 years of age (born at <26 gestational weeks) (Johnson 2011).

Neglect

No evidence was identified.

Maternal age

Low quality evidence from one study (n=219) show no association between maternal age and special educational needs among extremely children born preterm at 11 years of age (born at <26 gestational weeks) (Johnson 2011).

Maternal mental health disorder

No evidence was identified.

Sex of the child

Low to moderate quality evidence from 6 studies (sample sizes ranging from 187 to 53078) showed mixed results on the association between sex of the child born preterm and CP.

Moderate quality evidence from 1 study (n=208) showed that there was no significant risk of cerebral palsy in male infants (versus female) assessed at 18–22 months corrected age born at ≥22 weeks gestational age. Low quality evidence from 1 study (n=246) found no association between male sex and risk of CP among children born at <25 weeks of gestation and assessed at 18–22 months corrected age. Moderate quality evidence from 1 study (n=187) showed that there was no significant risk of cerebral palsy for male children (versus female) at follow-up of 2 years. Moderate quality evidence from 1 study (n=252) showed that there was no significant risk of cerebral palsy in males (versus females) born <28 weeks gestational age at follow-up of 5 years. Moderate quality evidence from 1 study (n=2457) showed that there was no increase in risk of cerebral palsy in male children born 30–34 weeks gestational age assessed at 5 years of age. Low quality evidence from 1 study (n=53078) showed that there was a significant increase in the risk of cerebral palsy in males (versus females) who were born at <32 weeks gestational age and assessed at 7 years of age. In the same study, no significant association was found between being male and CP among children born at 32–33 weeks of gestation.

Small for gestational age (SGA)

Moderate quality evidence from 5 studies (sample sizes ranging from 187 to 53078) showed mixed results on the association between being born SGA and CP.

Moderate quality evidence from 1 study (n=2971) showed a significant increase in the risk of moderate or severe cerebral palsy for children who were small for gestational age (SGA, versus not SGA) during a follow-up period of 18–22 months corrected age. Moderate quality evidence from 1 study (n=2846) showed that there was no increase in the risk of cerebral palsy in children born SGA (versus appropriate for gestational age) at 24–28 weeks or 29–32 weeks gestational age. Moderate quality evidence from 1 study (n=187) showed that there was no association between being born SGA (versus appropriate for gestational age) and CP among children born preterm at 2 years. Moderate quality evidence from 1 study (n=53078) showed that the risk of cerebral palsy in children born at <32 weeks of gestation who were SGA (versus appropriate for gestational age) was lowered. Among children born at 32–33 weeks, there was no association with SGA and CP, however, among children born at 34–36 weeks, there was an increased risk of CP among preterms born SGA.

Ethnicity

High quality evidence from 1 study (n=375) showed that there was a lowered risk of CP among children of African American origin (versus not African American) among children born between 23 and 32 weeks gestational age followed up at 6 years of age.

Brain abnormalities

Moderate to high quality evidence from 10 studies (sample sizes ranging from 187 to 6161) largely showed increased risk in CP in children exposed to IVH grade III-IV, severe PIVH, PVL, IVH/shunt, IVH grade III-IV and/or grade II-IV, parenchymal pathology and/or ventriculomegaly, IVH grade III or echodensities or ventricular dilation, cystic PVL or intraparentchymal, intracranial haemorrhage compared with those unexposed to those risk factors. Children in these 11 studies were born at different gestational ages and assessed at age 18 months, 24 months, 18 to 22 months corrected age, 2 years, 30 months, 5 years, 6 years, and 7 years. Only 1 study (n=246) found no significant association between IVH grade III-IVH and CP when children were assessed at 18–22 months corrected age (moderate quality).

Sepsis

Moderate quality evidence from 5 studies (sample sizes ranging from 208 to 6347) showed mixed findings with regard to the association between sepsis and CP.

Two studies showed that preterm children exposed to sepsis were at an increased risk for CP in comparison with those unexposed when assessed at age 18–22 months corrected (moderate quality evidence). However, another 3 studies showed no significant association between the two when preterm children were assessed at age 18 to 22 months corrected, 18 months, and 7 years (moderate quality evidence).

Retinopathy of prematurity (ROP)

Moderate quality evidence from 2 studies (n=1085; n=283) showed no significant association between ROP and the risk of CP when children were assessed at age 24 months and 30 months. The same non-significant association was found when ROP of different severities (such as ROP threshold, ROP pre-threshold) and the various forms of CP (for example CP quadriparesis, CP diparesis, and CP hemiparesis) were assessed in 1 of the studies (moderate quality evidence).

Necrotising enterocolitis (NEC)

Low to high quality evidence from 6 studies (sample sizes ranging from 252 to 2948) showed mixed findings regarding the risk of CP in relation to NEC. Four studies found no significant association between NEC and CP when children born preterm were assessed at 18–22 months corrected age, age 18 months, and age 5 and 6 years. However, significantly increased risk in CP among those exposed to NEC compared with those unexposed was found in 2 studies when children were assessed at 18 to 22 months corrected age and 5 years, respectively.

Antenatal exposure to steroids

Low to moderate quality evidence from 10 studies (sample sizes ranging from 193 to 6347) reported mixed findings regarding the association between antenatal steroids and CP. Seven studies found no significant association between those exposed to antenatal steroids and CP compared with those unexposed and when children were assessed at age 24 months, 18 to 22 months corrected age; 18 months; 30 months, 5 years, and 7 years. However, moderate to low quality evidence from three studies (sample size ranged from 193 to 1924) showed a significantly reduced risk in CP associated with antenatal steroids when children born at 27.3 (mean) weeks’ GA were assessed at age 18 months and 5 years, respectively; and children born at 22–25 weeks’ GA were assessed at age 18–22 months corrected.

Postnatal exposure to steroids

Moderate quality evidence from 6 studies (sample sizes ranging from 280 to 6347) reported mixed findings. Three studies (n=280; n=672; n=3785) found a significantly increased risk in CP among those exposed to postnatal steroids compared with those unexposed when children were assessed at age 18–22 months corrected, 24 months, and 5 years. However, nonsignificant association between postnatal steroids and CP was reported in another three studies (n=1472; n=1812; n=283) when children were assessed at age 18–22 months corrected age, 30 months corrected age, and 5 years.

Bronchopulmonary dysplasia (BPD)

Moderate quality evidence from 4 studies (sample sizes ranging from 246 to 3785) reported mixed findings on the risk of CP in relation to BPD at 36 weeks. No association was found in 4 studies when children born at 22–32 weeks’ GA, <28 weeks GA were assessed at age 18–22 months corrected, 24 months corrected, and 5 years. However, in 1 study, when BPD with mechanical ventilation was assessed, no significant association was found between it and CP when children born at <28 weeks GA were assessed at age 24 months corrected.

Socioeconomic status

Evidence from 3 studies (n=641) showed no impact of socioeconomic status on the risk of cerebral palsy (Shankaran 2004; Tommiska 2003; Toome 2013). The quality of evidence from these studies ranged from low to high.

Maternal substance abuse

No evidence was identified.

Multiple pregnancy

High quality evidence from 1 study (n=208) showed that multiple pregnancy did not significantly affect the risk of cerebral palsy in a group of extremely low birth weight infants assessed at 18 months corrected age. High quality evidence from another study (n=187) showed no significant effect of multiple pregnancy on the risk of cerebral palsy at 2 years (corrected age). Moderate quality evidence from 1 study (n=1461) reported no significant change in the risk of cerebral palsy for multiple pregnancy (as compared to singletons) born at 30–34 weeks. Further analysis of the same cohort included preterm infants from 24–32 weeks (n=812). This also showed no significant change in the risk of cerebral palsy for multiple pregnancy or with maternal age (moderate quality evidence). Low quality evidence from 1 study (n=53078) reported no association between multiple pregnancy and cerebral palsy.

Chorioamnionitis

High quality evidence from 1 study (n=2235) showed no significant impact of chorioamnionitis on the risk of cerebral palsy in a group of very preterm babies (<27 weeks’ gestation) at 18–22 months of corrected age. Moderate quality evidence from 1 study (n=283) did not find an association between chorioamnionitis and CP among children born before 26 weeks of gestation and assessed at 30 months corrected age. Low quality evidence from 1 study (n=2202) showed no association between histological chorioamnionitis and cerebral palsy in children born before 34 weeks of gestation at 3 years of age (uncorrected).

Neglect

No evidence was identified.

Maternal age

High quality evidence from 1 study (n=208) showed that maternal age did not affect the risk of cerebral palsy in a group of extremely low birth weight infants assessed at 18 months corrected age. High quality evidence from another study (n=187) showed no significant effect of maternal age on the risk of cerebral palsy at 2 years (corrected age) among children born before 32 weeks of gestation. Low quality evidence from 1 study (n=53078) showed no association between maternal age and CP among children born preterm.

Maternal mental health disorder

No evidence was identified.

Sex of the child

Low quality evidence from 1 study (n=560) showed that an increase in the risk of DCD in male children (versus female) born before 28 weeks of gestation and assessed at 8 to 9 years age.

Sex of the child

Low to moderate quality evidence from 8 studies (sample sizes ranging from 187 to 14147) showed somewhat mixed findings on the association between the sex of the preterm child and intellectual disability.

Moderate quality evidence from 1 study (n=963) showed that there was no association between male sex and cognitive impairment (MDI <70) in children born before 27 weeks of gestation and assessed at 18–22 months corrected age. High quality evidence from 1 study (n=246) showed that there was no increased risk of cognitive impairment (MDI<70) in male children born before 25 weeks of gestation (versus females) at 18–22 months corrected age. Moderate quality evidence from 1 study (n=14147) showed that there was a significant increase in risk of intellectual disability in male children (versus female) with birth weight of 401–1000 grams (mean gestational age 25.5 weeks) at 18–22 months corrected age. Low quality evidence from 1 study (n=1151) did not find an association between male sex (versus female) and cognitive impairment in children born before 27 weeks of gestation and assessed at 18–22 months corrected age. High quality evidence from 1 study (n=187) showed no significant increase in the risk of cognitive impairment in male children (versus female) born at a mean 28.8 weeks gestational age and assessed at 2 years (corrected age). Moderate quality evidence from 1 study (n=1506) showed that there was a significant increase in the risk of cognitive impairment in male children (versus female) born before 28 weeks of gestation and assessed at 2 years (corrected age). Moderate quality evidence from 1 study (n=1503) found no association between male sex (versus female) and mild or severe cognitive impairment in children born between 24 to 32 weeks gestational age and assessed at 5 years of age. Moderate quality evidence from 1 study (n=252) found no association between male sex (versus female) and cognitive impairment in children born before 28 weeks of gestation assessed at 5 years of age.

Small for gestational age (SGA)

Moderate quality evidence from 5 studies (sample sizes ranging from 187 to 2846) showed somewhat mixed results on the association between being born SGA and intellectual disability among children born preterm.

Moderate quality evidence from 1 study (n=963) found a significant increase in risk of cognitive impairment (MDI <70) in children born before 27 weeks of gestation who were SGA (versus appropriate for gestational age) at 18–22 months corrected age. Low quality evidence from 1 study (n=1151) did not find an association between SGA (versus appropriate for gestational age) and cognitive impairment in children born before 27 weeks of gestation and assessed at 18–22 months corrected age. High quality evidence from 1 study (n=187) showed that three was no significant increase in the risk of cognitive impairment in children born SGA born at a mean 28.8 weeks gestational age and assessed at 2 years (corrected age). Moderate quality evidence from 1 study (n=1503) found an increased risk of severe cognitive impairment in children born SGA (versus appropriate for gestational age) between 24 to 32 weeks gestational age and assessed at 5 years of age. Moderate quality evidence from 1 study (n=2846) showed that there was no increased risk of cognitive impairment at 5 years in children born SGA at 24–28 weeks gestational age, however, there was a significant increase in the risk of impairment at 29–32 weeks gestational age.

Ethnicity

Low to moderate quality evidence from 4 studies (sample sizes ranging from 246 to 3790) showed mixed findings on the association between ethnicity and intellectual disability in children born preterm.

Low quality evidence from 1 study (n=246) showed that there was no increased risk of cognitive impairment (MDI<70) in children of black ethnicity (versus non-black) born before 25 weeks of gestation assessed at 18–22 months corrected age. Moderate quality evidence from 1 study (n=3790) showed no significant increase in the risk of cognitive impairment in children of non-white race (versus white) at 18–22 months corrected age. Low quality evidence from 1 study (n=1151) did not find an association between black ethnicity (versus non-black) and cognitive impairment in children born before 27 weeks of gestation and assessed at 18–22 months corrected age. However, moderate quality evidence from 1 study (n=1506) showed that there was a significant increase in the risk of cognitive impairment in children of non-white ethnicity (versus white) born before 28 weeks of gestation and assessed at 2 years (corrected age).

Brain abnomalities

Low to moderate quality evidence from 11 studies (sample sizes ranging from 187 to 6161) largely showed an increased risk in intellectual disability defined in different ways across studies associated with PVL, IVH and infarct. Children in those studies were assessed at age 18 to 22 months corrected, 24 months corrected, 2 years, and 5 years. However, non-significant association was found in 2 studies when children were assessed at age 18–22 months corrected and 5 years.

Sepsis

Moderate quality evidence from 6 studies (sample sizes ranging from 1472 to 6314) reported mixed findings. Three studies found a significantly increased risk in intellectual disabilities associated with sepsis when children were assessed at age 18 to 22 months corrected age. However, another three studies (sample size ranging from 963 to 3785) reported non-significant association between the two when children assessed also at age 18–22 months corrected.

Retinopathy of prematurity (ROP)

Moderate quality evidence from 1 study (n=1085) showed mixed results when different degrees of ROP and intellectual disability of different levels were assessed among children aged 24 months. ROP stage 3 showed an increased risk associated with MDI <55 (Bayley). However, when MDI 56–69 was assessed as the outcome, the significantly increased risk associated with ROP was found for ROP zone 1, ROP threshold, and ROP pre-threshold.

Necrotising enterocolitis (NEC)

Moderate quality evidence from 9 studies (sample sizes ranging from 193 to 6314) reported mixed findings regarding the association between NEC and intellectual disability defined in different methods. Six studies showed an increased risk in MDI < 70 associated with NEC (e.g., NEC surgery, NEC perforation) when children were assessed at age 18 to 22 months corrected, 2 years, 5 years. However, another 3 studies showed non-significant association between the two when children were assessed at age 18 to 22 months corrected, 5 years.

Antenatal exposure to steroids

Low to moderate quality evidence from 10 studies (sample sizes ranging from 193 to 4924) showed largely non-significant association between antenatal steroids and intellectual disability measured in different ways when children were assessed at age 18–22 months corrected and 5 years. In 1 study (n=193), antenatal steroids were found to be associated with an IQ score <70 when children were assessed at age 5 years.

Postnatal exposure to steroids

Moderate quality evidence from 4 studies (sample sizes ranging from 2901 to 3705) showed mixed results regarding the association between postnatal steroids and intellectual disability. Three studies found an increased risk in MDI < 70 associated with postnatal steroids when children were assessed at age 18 to 22 months corrected. However, 1 study (n=2901) found no significant association between it and severe cognitive deficiency assessed by Kaufman Assessment Battery for Children (K-ABC) scale when children at 5 years were assessed.

Bronchopulmonary dysplasia (BPD)

Low to moderate quality evidence from 7 studies (sample sizes ranging from 193 to 3785) reported mixed findings. Four studies found a significantly increased risk in intellectual disability associated with BPD at 36 weeks when children were assessed at age 18 to 22 months corrected. However, 3 studies found no significant associations between BPD with or without mechanical ventilation and intellectual disability when children were assessed at age 18 to 22 months corrected, and at age 24 months.

Socioeconomic status

Low quality evidence from 1 study (n=246) showed no association between low socioeconomic status (household income <$20K) and cognitive impairment at 18–22 months corrected age among children born before 25 weeks of gestation. High quality evidence from another study (n=187) showed no significant effect of socioeconomic status on the risk of cognitive impairment at 2 years (corrected age) among children born before 32 weeks of gestation. Moderate quality evidence from 1 study (n=1503) found a significant increase in the risk of mild and severe intellectual disability for preterm infants (24–32 weeks) of families with lower socioeconomic status. Further analysis of the same study (n=1461) also showed a significant increase in moderate/severe cognitive deficiency for moderately preterm infants (30–34 weeks) born to families of lower socioeconomic status.

Maternal substance abuse

Low quality evidence from 1 study (n=82) found that maternal use of cocaine significantly increased the risk of intellectual disability among children born preterm at 3 years of age.

Multiple pregnancy

Moderate to high quality evidence from 2 studies (n=643) showed no significant effect of multiple pregnancy on the risk of cognitive impairment at 2 and 2.5 years of age among children born before 27 weeks and before 32 weeks of gestation.

Chorioamnionitis

High quality evidence from 1 study (n=2235) showed a significant increase in the risk of cognitive impairment at 2 years of age for preterm infants with chorioamnionitis that was diagnosed both clinically and histopathologically. Moderate quality evidence from 1 study (n=456) showed no significant effect of chorioamnionitis on cognitive function at 2.5 years among children born before 27 weeks of gestation. Low quality evidence from another study (n=2202) showed no association between histological chorioamnionitis and cognitive function in children born before 34 weeks of gestation at 3 years of age (uncorrected).

Neglect

No evidence was identified.

Maternal age

Moderate quality evidence from 1 study (n=3790) showed no association between maternal age and cognitive impairment at 18–22 months corrected age among children born before 27 weeks of gestation. High quality evidence from 1 study (n=187) showed no significant effect of maternal age on the risk of cognitive impairment at 2 years (corrected age) among children born before 32 weeks of gestation.

Maternal mental health disorder

No evidence was identified.

Brain abnormalities

Moderate quality evidence from 1 study (n=161) showed an increased risk in low attainment in mathematicws associated with IVH of all grades when children born preterm were assessed at age 5 years.

Sepsis

No evidence was identified.

Retinopathy of prematurity (ROP)

No evidence was identified.

Necrotising enterocolitis (NEC)

No evidence was identified.

Antenatal exposure to steroids

No evidence was identified.

Postnatal exposure to steroids

No evidence was identified.

Bronchopulmonary dysplasia (BPD)

Moderate quality evidence from 1 study (n=161) showed an increased risk in low attainment in mathematics associated with BPD at 36 weeks when children born preterm were assessed at age 5 years.

Sex of the child

High quality evidence from 1 study (n=187) showed that there was a significant increase in the risk of language impairment in male children (compared to female) born at a mean gestationa age of 28.8 weeks at 2 years of age.

Small for gestational age

No evidence was identified.

Ethnicity

Moderate quality evidence from 1 study (n=3790) showed no association between being of non-white ethnic background and language impairment at 18–22 months’ corrected age in children born preterm when compared to children born preterm of white ethnicity.

Brain abnormalities

Moderate quality evidence from 2 studies (n= 1472; n=187) showed an increased risk in speech and language disorders associated with severe PIVH and IVH grade III/IV or PVL grade II-IV when children born pre-term were assessed at age 18–22 months corrected age and 2 years.

Sepsis

Moderate quality evidence from 1 study (n=1472) found an increased risk in speech and language disorders associated with sepsis when children born pre-term were assessed at age 18–22 months corrected age.

Retinopathy of prematurity (ROP)

No evidence was identified.

Necrotising enterocolitis (NEC)

No evidence was identified.

Antenatal exposure to steroids

Moderate quality evidence from 2 studies (n= 1472; n=1934) found no significant association between antenatal steroids and language disorders when children born pre-term were assessed at age 18–22 months corrected age.

Postnatal exposure to steroids

No evidence was identified.

Bronchopulmonary dysplasia (BPD)

No evidence was identified.

Socioeconomis status

High quality evidence from 1 study (n=187) showed no significant effect of socioeconomic status on the risk of language impairment at 2 years (corrected age) among children born before 32 weeks of gestation.

Maternal substance abuse

No evidence was identified.

Multiple pregnancy

High quality evidence from 1 study (n=187) showed no significant effect of multiple pregnancy on the risk of language impairment at 2 years (corrected age) among children born before 32 weeks of gestation.

Chorioamnionitis

No evidence was identified.

Neglect

No evidence was identified.

Maternal age

Moderate quality evidence from 1 study (n=3790) showed no significant effect of maternal age on the risk of language impairment at 18–22 months corrected age among children born before 27 weeks of gestation. High quality evidence from another study (n=187) showed no significant effect of maternal age on the risk of language impairment at 2 years (corrected age) among children born before 32 weeks of gestation.

Maternal mental health disorder

No evidence was identified.

Brain abnormalities

No evidence was identified.

Sepsis

No evidence was identified.

Retinopathy of prematurity (ROP)

No evidence was identified.

Necrotising enterocolitis (NEC)

Moderate quality evidence from 1 study (n=307) showed an increased risk in any psychiatric disorder associated with NEC when children born preterm were assessed at age 11 years.

Antenatal exposure to steroids

No evidence was identified.

Postnatal exposure to steroids

No evidence was identified.

Bronchopulmonary dysplasia (BPD)

No evidence was identified.

Sex of the child

Low quality evidence from 2 studies (n=1078; n=85535) showed an increased risk of ASD in male preterm children compared to female preterm children.

Low quality evidence from 1 study (n=1078) showed that there was a significant increase in the risk of infantile autism among male children born preterm/extremely low birth weight at 8–11 years follow-up compared to their female peers. Low quality evidence from 1 study (n=85535) showed that there was a significant increase in the risk of autism spectrum disorder in males born preterm (compared to females) when asked from parents if the doctor had ever told that their child born preterm aged 2 to 17 years had ASD.

Small for gestational age (SGA)

High quality evidence from 2 studies (n=235198; n=21717) showed mixed findings on the association between being born SGA and ASD.

High quality evidence from 1 study (n=235198) showed that there was a significant increase in the risk of ASD diagnosis in children born preterm who were born small for gestational age compared to children born preterm appropriate for gestational age. High quality evidence from 1 study (n=21717) showed no association between being born SGA and autism among children born preterm (at 23–31 weeks’, 32–33 weeks’, and 34–36 weeks’ gestation) at 11 years of age.

Ethnicity

Low quality evidence from 1 study (n=95535) showed mixed results regarding association between ethnicity and ASD in children born preterm. No association was found in Hispanic or non-Hispanic mixed race children compared to non-Hispanic white children. A reduced risk of ASD was reported among non-Hispanic black children compared to non-Hispanic white children. The study measured ASD by asking parents of children born preterm if the doctor had ever told that their child born preterm aged 2 to 17 years had ASD.

Brain abnormalities

Moderate quality evidence from 1 study (n=3807) showed an increased risk in autism associated with IVH grade III-IV when children born preterm were assessed at age 2 to 11 years, However no significant association between cystic PVL and autism was found in the same study.

Sepsis

Moderate quality evidence from 1 study (n=3807) showed no significant association between sepsis and autism when children born preterm were assessed at age 2 to 11 years.

Retinopathy of prematurity

No evidence was identified.

Necrotising enterocolitis (NEC)

No evidence was identified.

Antenatal exposure to steroids

No evidence was identified.

Postnatal exposure to steroids

No evidence was identified.

Bronchopulmonary dysplasia (BPD)

Moderate quality evidence from 1 study (n=1078) showed no significant association between BPD at 36 weeks and autism when children born preterm were assessed at age 8 to 11 years.

Sex of the child

Low quality evidence from 1 study (n=85535) showed an increase in the risk of ADHD among male children born preterm (compared to female) when asked from parents if the doctor had ever told that their child born preterm aged 2 to 17 years had ADHD. The same study reported a reduced risk of ADHD, as reported by parents, among children born preterm of Hispanic and non-Hispanic black ethnicity compared to children born preterm of non-Hispanic white ethnicity.

Small for gestational age

No evidence was identified.

Ethnicity

No evidence was identified.

Sex of the child

No evidence was identified.

Small for gestational age (SGA)

Moderate quality evidence from 1 study (n=297) showed a significant increase in the risk of blindness (<20/200 vision bilaterally) among children born at 23–26 weeks’ gestation who were born SGA compared to children of the same gestation age who were born appropriate for gestational age.

Ethnicity

No evidence was identified.

Brain abnormalities

Moderate quality evidence from 1 study (n=6161) showed an increased risk in blindness associated with IVH grade III/shunt when children born preterm were assessed at age 18–22 months corrected.

Sepsis

Moderate quality evidence from 1 study (n=6161) showed an increased risk in blindness associated with sepsis, meningitis with our without sepsis when children born preterm were assessed at age 18–22 months corrected.

Retinopathy of prematurity (ROP)

Moderate quality evidence from 1 study (n=193) showed an increased risk in blindness associated with ROP when children born preterm were assessed at age 5 years.

Necrotising enterocolitis (NEC)

No evidence was identified.

Antenatal exposure to steroids

Moderate quality evidence from 1 study (n=6161) showed no significant association between antenatal steroids and blindness when children born preterm were assessed at age 18–22 months corrected.

Postnatal exposure to steroids

No evidence was identified.

Bronchopulmonary dysplasia (BPD)

No evidence was identified.

Socioeconomic status

No evidence was identified.

Maternal substance abuse

No evidence was identified.

Multiple pregnancy

No evidence was identified.

Chorioamnionitis

Low quality evidence 1 study (n=2202) showed no association between histological chorioamnionitis and visual impairment in children born before 34 weeks of gestation at 3 years of age (uncorrected).

Neglect

No evidence was identified.

Maternal age

No evidence was identified.

Maternal mental health disorder

No evidence was identified.

Sex of the child

No evidence was identified.

Small for gestational age (SGA)

Moderate quality evidence from 1 study (n=2971) showed no association between being born SGA and hearing loss among children born at 23 to 26 weeks’ gestation.

Ethnicity

No evidence was identified.

Brain abnormalities

Moderate quality evidence from 1 study (n=6161) showed no significant association between IVH grade III/shunt and deafness when children born preterm were assessed at age 18–22 months corrected.

Sepsis

Moderate quality evidence from 1 study (n=6314) showed an increased risk in deafness associated with sepsis when children born preterm were assessed at age 18–22 months corrected. However, the same study showed no significant association between meningitis with our without sepsis and deafness.

Retinopathy of prematurity (ROP)

No evidence was identified.

Necrotising enterocolitis (NEC)

No evidence was identified.

Antenatal exposure to steroids

Moderate quality evidence from 1 study (n=4924) showed no significant association between antenatal steroids and deafness when children born preterm were assessed at age 18–22 months corrected.

Postnatal exposure to steroids

No evidence was identified.

Bronchopulmonary dysplasia (BPD)

No evidence was identified.

Socioeconomic status

No evidence was identified.

Maternal substance abuse

No evidence was identified.

Multiple pregnancy

No evidence was identified.

Chorioamnionitis

Low quality evidence 1 study (n=2202) showed no association between histological chorioamnionitis and severe hearing impairment in children born before 34 weeks of gestation at 3 years of age (uncorrected).

Neglect

No evidence was identified.

Maternal age

No evidence was identified.

Maternal mental health disorder

No evidence was identified.

Sex of the child

Low to high quality evidence from 4 studies (sample sizes ranging from 246 to 3041) showed mixed findings on the association between the sex of the child and composite neurodevelopmental or neurosensory outcome in children born preterm.

High quality evidence from 1 study (n=1473) showed an increased risk of moderate to severe functional disability (1 or more of the following: developmental delay, cerebral palsy, bilateral blindness, or bilateral deafness) among males (compared to females) born before 29 weeks’ gestation and assessed at 2–3 years corrected age.

Moderate quality evidence from 1 study (n=3041) showed an increased risk of neurodevelopmental disability (1 or more of the following: mental developmental index score or physomotor developmental index score < 70, moderate or severe cerebral palsy, bilateral blindness, or deafness) among males (compared to females) born at a mean gestational age of 25.8 weeks and assessed at 18 to 22 months corrected age.

Moderate quality evidence from 1 study (n=373) showed no association between the sex of the child and major neurosensory disability (1 or more of the following: cerebral palsy, blindness, or complete deafness) at 2 years in children born at 22–27 weeks’ gestation.

Low quality evidence from 1 study (n=246) showed no association between the sex of the child and neurodevelopmental impairment (1 or more of the following: cerebral palsy, mental developmental index score or psychomotor developmental index score < 70, bilateral blindness, or hearing impaired with amplification) at 18 to 22 months corrected age in children born before 25 weeks’ gestation.

Small for gestational age (SGA)

Moderate to high quality evidence from 3 studies (sample sizes ranging from 187 to 1473) showed mixed results on the association between being born SGA and composite neurodevelopmental or neurosensory outcome in children born preterm.

High quality evidence from 1 study (n=1473) showed an increased risk of moderate to severe functional disability (1 or more of the following: developmental delay, cerebral palsy, bilateral blindness, or bilateral deafness) among SGA children (compared to children born appropriate to gestational age) born before 29 weeks’ gestation and assessed at 2–3 years corrected age. Moderate quality evidence from 1 study (n=373) showed no association between being born SGA and major neurosensory disability (1 or more of the following: cerebral palsy, blindness, or complete deafness) at 2 years in children born at 22–27 weeks’ gestation.

Ethnicity

Low to moderate quality evidence from 2 studies (n=246; 2=3041) showed mixed findings on the association between ethnicity and composite neurodevelopmental outcome in children born preterm.

Moderate quality evidence from 1 study (n=3041) showed an increased risk of neurodevelopmental disability (1 or more of the following: mental developmental index score or physomotor developmental index score < 70, moderate or severe cerebral palsy, bilateral blindness, or deafness) among children of non-white ethnicity (compared to children of white ethnicity) born at a mean gestational age of 25.8 weeks and assessed at 18 to 22 months corrected age.

Low quality evidence from 1 study (n=246) showed no association between ethnicity and neurodevelopmental impairment (1 or more of the following: cerebral palsy, mental developmental index score or psychomotor developmental index score < 70, bilateral blindness, or hearing impaired with amplification) at 18 to 22 months corrected age in children born before 25 weeks’ gestation.

Brain abnormalities

Moderate quality evidence from 11 studies (sample sizes ranging from 166 to 6161) showed largely increased risk in neurodevelopmental impairment or neurosensory impairment associated with IVH grade III, IVH grade IV, IVH grade III-IV, severe PIVH, cystic PVL, IVH III/shunt, severe cerebral lesions when children born preterm were assessed at age 18–22 months corrected, 22–30 months, 2 years, and 2–3 corrected year.

Sepsis

Moderate quality evidence from 6 studies (sample sizes ranging from 166 to 6314) reported mixed findings. Three studies showed an increased risk in neurodevelopmental/neurosensory impairment associated with sepsis when children were assessed at 18–22 months corrected age. However, 3 studies found no significant difference between those exposed to sepsis and those who were not when children were assessed at 18–22 months corrected age and 2 years.

Retinopathy of prematurity (ROP)

Moderate quality evidence from 3 studies (sample sizes ranging from 79 to 1472) showed a borderline increased or increased risk in neurodevelopmental impairment and or neurosensory impairment associated with ROP when children born preterm were assessed at age 2 years, 2 to 3 corrected year, and 7 to 10 years.

Necrotising enterocolitis (NEC)

Moderate quality evidence from 7 studies reported mixed findings regarding the relationship between NEC and composite outcomes either measured as neurodevelopmental impairment or neurosensory impairment. Five studies showed an increased risk in neurodevelopmental impairment or neurosensory impairment when children were assessed age 18 to 22 months corrected, 2 years, and 7 to 10 years, however 3 studies showed no significant associations when children were assessed at age 18–22 months corrected, 2 years, and 2–3 corrected years.

Antenatal exposure to steroids

Low to moderate quality evidence from 8 studies (sample size ranging from 246 to 4924) showed no significant association between antenatal steroids and composite outcomes either measured as neurodevelopmental impairment or neurosensory impairment. This was the same when children were assessed at age 18–22 months corrected, 2 years, 2.5 corrected years, and 2–3 years,

Postnatal exposure to steroids

Moderate quality evidence from 6 studies (sample sizes ranging from 166 to 3041) reported mixed findings regarding the relationship between postnatal steroids and neurodevelopmental impairment or neurosensory impairment. Four studies showed an increased risk in the composite outcomes associated with postnatal steroids when children were assessed at age 18–22 months corrected and 2 years. However 2 studies found no significant association between the two when children were assessed at age 18–22 months corrected and 2 years as well.

Bronchopulmonary dysplasia (BPD)

Low to moderate quality evidence from 4 studies (sample sizes ranging from 246 to 3785) reported mixed findings. Three studies found no significant association between BPD and neurodevelopment impairment or neurosensory impairment when children born preterm were assessed at age 18–22 months corrected, and 2 years. However, a significantly increased risk in neurodevelopmental impairment associated with BPD was found in 1 study when children born at 22–32 weeks’ GA were assessed at age 18–22 months corrected.

Socioeconomic status

Low quality evidence from 1 study (n=246) showed no significant association between low socioeconomic status (household income <$20K) and composite neurodevelopmental impairment outcome at 18–22 months corrected age among children born before 25 weeks of gestation. High quality evidence from 1 study (n=187) also showed no significant risk of neurodevelopmental impairment at 2 years (corrected age) among children born before 32 weeks of gestation from low income households (versus non-low income).

Maternal substance abuse

No evidence was identified.

Multiple pregnancy

Moderate quality evidence from 2 studies (n=829) showed no significant effect of multiple pregnancy on the risk of neurosensory impairment (1 or more of the following: CP, moderate/severe visual, or hearing impairment) at 2 and 2.5 years corrected age among children born between 22–27 weeks of gestation. High quality evidence from 1 study (n=187) also showed no significant risk of neurodevelopmental impairment (1 or more of the following: intellectual disability, cerebral palsy, hearing impairment, or visual impairment) at 2 years (corrected age) for multiple births as compared to singletons among children born before 32 weeks of gestation.

Chorioamnionitis

High quality evidence from 1 study (n=2235) showed no impact of histological chorioamnionitis on the risk of a composite outcome measure of neurodevelopmental impairment (including CP, deafness, blindness and cognitive delay) at 18–22 months corrected age among children born before 27 weeks of gestation. This study also showed that infants with both clinical and histological chorioamnionitis also had no increase in the risk of neurodevelopmental impairment. Moderate quality evidence from 1 study (n=456) showed no significant effect of chorioamnionitis (including prolonged and premature rupture of membranes) on the risk of a neurosensory impairment (1 or more of the following: CP, moderate/severe visual impairment, or hearing impairment). However, moderate quality evidence from 1 study (n=373) showed a significant increase in the risk of major neurosensory disability (1 or more of the following: CP, blindness, or deafness) at 2 years of age in children born between 22–27 weeks of gestation with chorioamnionitis compared to those without.

Neglect

No evidence was identified.

Maternal age

High quality evidence from 1 study (n=187) showed no effect of maternal age on the risk of neurodevelopmental impairment at 2 years (corrected age) among children born before 32 weeks of gestation.

Maternal mental health disorder

No evidence was identified.

Receptive communication

Low quality evidence from one study (n=394) showed that among children born at <27 weeks of gestation the prevalence of receptive communication problems (Bayley, mild −1SD to −2SD) was 24.9% (95%CI 20.7 to 30.0%) at 2.5 years age. In the same study, the prevalence of moderate receptive communication problems (Bayley −2SD to −3SD) was 9.1% (95%CI 6.5 to 12.4%). The prevalence of moderate to severe (Bayley −3SD) receptive communication was 5.8% (95%CI 3.7 to 8.6%) (Mansson 2014).

Low quality evidence from one study (n=1506) showed that among children born at <28 weeks of gestation the prevalence of receptive communication problems (OWLS <=−2SD) was 19% (95% CI 16.5 to 21.8) when assessed at 10 years age (Joseph 2016b).

Expressive communication

Low quality evidence from one study (n=394) showed that among children born at <27 weeks of gestation the prevalence of expressive communication problems (Bayley, mild −1SD to −2SD) was 31.3% (95%CI 26.7 to 36.1%) at 2.5 years age (Mansson 2014). In the same study, prevalence of moderate expressive communication (Bayley moderate −2SD to −3SD) problems was 8.1% (95%CI5.6 to 11.3%), and for moderate to severe expressive communication problems (Bayley −3SD), the prevalence was 6.4% (95%CI 4.2 to 9.3%) (Mansson 2014).

Low quality evidence from one study (n=1506) showed that among children born at <28 weeks of gestation the prevalence of expressive communication problems (OWLS <=−2SD) was 19% (95% CI 16.5 to 21.8) when assessed at 10 years age (Joseph 2016b).

Total behavioural problems

Low to moderate quality evidence from two separate studies (n=1645 to 2382) showed that among children born at 24–28 weeks and 24–27weeks of gestation the prevalence of total behavioural difficulties (SDQ, 10^th percentile) ranged from 24.1% (95%CI 19.2 to 29.6%) 22.2% (95%CI 17.1 to 28.1%) at 3 years age and 5 years age respectively (Delobel-Ayoub 2006; Foix-Helias 2008).

Low quality evidence from one study (n=224) showed that among children born at <26 weeks of gestation the prevalence of total behavioural difficulties (SDQ, 10^th percentile, parent reported) was 38.5% (95%CI 32.0 to 45.2%) and 34.6% (95%CI 29.2 to 41.5%) (teacher-reported) at 6 years age (Samara 2008).

Low quality evidence from one study (n=2901) showed that among children born at 24–28 weeks of gestation the prevalence of total behavioural difficulties (SDQ, 10^th percentile, parent reported) was 27.8% (95%CI 23.0 to 32.9%) at 8 years age (Larroque 2011). In another moderate quality study (n=189), among children born at <28 weeks of gestation, the prevalence of total behavioural difficulties was 18% (95%CI 12.8 to 24.2%) (Hutchinson 2013).

Low quality evidence from one study (n=568) showed that among children born at <28 weeks gestation the prevalence of total behavioural problems (at risk, BASC) was 15.0% (95%CI 11.0 to 19.7%) whereas in the same population those who had clinically significant behavioural problems (BASC) the prevalence was 7.0% (95%CI 4.2 to 10.6%) at 8 years corrected age (Anderson 2003).

Moderate quality evidence from one study (n=154) showed that among children born at <27 weeks of gestation the prevalence of total difficulties (SDQ score 17 to 40) was 28.0% (95%CI 18.2 to 39.6%) at 7–9 years age (Stahlman 2009).

Moderate quality evidence from one study (n=169) showed that among children born at <26 weeks of gestation the prevalence of total behavioural problems (CBCL, 90^th percentile, parent reported) was 28.9% (95%CI 19.5 to 39.9%) and 24.1% (95%CI 15.4 to 34.7%) (teacher-reported CBCL, 90^th percentile) at 11 years age (Farooqi 2007).

Low quality evidence from one study (n=2855) showed that among children born SGA at 24–28 weeks of gestation the prevalence of total behavioural difficulties (SDQ, 10^th percentile) was 33.3% (95%CI 14.6 to 57%) compared with those children born MGA (27.3% (95%CI 13.3 to 45.5%)) at 5 years age. For those children born AGA the prevalence was 23.7% (95%CI 19.3 to 28.5%) (Guellec 2011).

ADHD symptoms

Low quality evidence from one study (n=201) showed that among children born at <28 weeks of gestation the prevalence of ADHD symptoms (CADS-P, inattentive symptoms, T score >60) was 32.1% (95%CI 20.3 to 46%) at 8 years corrected age (Anderson 2011). In the same study, the prevalence of ADHD symptoms (hyperactivity-impulsive symptoms, T score >60) and ADHD index (CADS-P, T score >60) was 41.8% (95%CI 28.7 to 55.9%) and 43% (95%CI 30.3 to 57.7%) respectively (Anderson 2011).

Low quality evidence from one study (n=298) showed that among adolescents born at <28 weeks of gestation the prevalence of ADHD (DSM-IV, <−1.5 SD, parent reported) was 17.6% (95%CI 12.5 to 23.7%) at 17 years age whereas the prevalence of ADHD reported by adolescents themselves was 5.2% (95%CI 2.5 to 9.4%) at 17 years age (Wilson-Ching 2013).

ASD symptoms

Low quality evidence from one study (n=307) showed that among children born at <26 weeks of gestation the prevalence of positive ASD screen (SCQ ≥15, parent reported) was 15.8% (95%CI 10.9 to 22.0%) at 11 years age (Johnson 2010).

Moderate quality evidence from one study (n=1198) showed that among children born at <27 weeks of gestation the prevalence of positive ASD screen (SCQ ≥11, parent reported) was 12.4% (95% CI 10.2 to 14.8%) at 10 years age (Joseph 2016a).

Attention/hyperactivity symptoms

Low quality evidence from one study (n=2855) showed that among children born SGA at 24–28 weeks of gestation the prevalence of inattention/hyperactivity (SDQ 10^th percentile) was 19% (95%CI 5.5 to 42%) compared with those children born MGA (21.2% (95%CI 9 to 38.9%)). For those children born AGA the prevalence was 21.7% (95%CI 17.5 to 26.4%)) (Guellec 2011).

Moderate quality evidence from one study (n=826) showed that among children born at <28 weeks of gestation the prevalence of attention problems (CBCL 93^rd percentile) was 10.7% (95%CI 8.6 to 13.0%) at 24 months corrected age (Downey 2016).

Total behavioural problems

Low quality evidence from one study (n=2382) showed that among children born at 29–32 weeks of gestation the prevalence of total behavioural difficulties (SDQ, 10^th percentile) was 18.2% (15.8 to 20.9%) at 3 years (Delobel-Ayoub 2006). At 5 years age (moderate quality evidence, n=1645), the prevalence of total behavioural problems (SDQ, 10^th percentile) in children born at 28–32 weeks gestation was 21% (95%CI 18.9 to 23.2%) (Foix-Helias 2008).

Low quality evidence from one study (n=2901) showed that among children born at 29–32 weeks of gestation, the prevalence of total behavioural difficulties (SDQ, 10^th percentile) was 18.9% (95%CI 16.6 to 21.4%) at 8 years age (Larroque 2011).

Total behavioural problems

Low quality evidence from one study (n=235) showed that among children born at <32 weeks of gestation the prevalence of total behavioural problems (CBCL, >98^th percentile) was 8.9% (95%CI 4.9 to 14.4%) at 2 years (corrected age) (Stoelhorst 2003a).

Low quality evidence from one study (n=2382) showed that among children born at <33 weeks of gestation the prevalence of total behavioural problems (SDQ, 10^th percentile) was 20% (95%CI 17.7 to 22.3%) at 3 years age (Delobel-Ayoub 2006).

Low quality evidence from one study (n=1504) showed that among children born at 25–31 weeks of gestation the prevalence of emerging total behavioural problems (CBCL, >85^th percentile) was 5.2% (95%CI 3.3 to 7.9%) at 4 and 5 years age (Hornman 2016). In the same study, the prevalence of resolving and persistent total behavioural problems (CBCL, 85^th percentile) was 5.5% (95%CI 3.5 to 8.2%) and 8.2% (95%CI 5.7 to 11.4%) respectively (Hornman 2016).

Low to moderate quality evidence from two separate studies (sample size ranging from 566 to 924) showed that among children born at <32 weeks of gestation the prevalence of total behavioural problems (CBCL, >=55, parent reported) was 13.8% (95%CI 10.6 to 17.5%) and 3.4% (95%CI 2.1 to 5.2%) when measured on FTF for emotional and behavioural problems at 5 years age respectively (de Kleine 2003; Rautava 2010). In another study (n=1645) among children born at 24–32 weeks of gestation the prevalence of total behavioural problems (SDQ, 10^th percentile) was 21.2% (95%CI 19.2 to 23.2%) at 5 years age (Foix-Helias 2008).

Low quality evidence from one study (n=2901) showed that among children born at 24–32 weeks of gestation the prevalence of total behavioural difficulties (SDQ, 10^th percentile) was 21.1% (95%CI 18.9 to 23.3%) at 8 years age (Larroque 2011).

Low quality evidence from one study (n=2855) showed that among children born SGA at 29–32 weeks of gestation the prevalence of total behavioural difficulties (SDQ 10^th percentile) was 19.1% (95%CI 12.4 to 27.5%) compared to those children born MGA (26.5% (95%CI 18.8 to 35.2%)) at 5 years age. For those children born AGA the prevalence was 19.4% (95%CI 17 to 21.9%)) (Guellec 2011).

Attention/hyperactivity symptoms

Low quality evidence from one study (n=2855) showed that among children born SGA at 29–32 weeks of gestation the prevalence of inattention/hyperactivity was 23.5% (95%CI 16 to 32.3%) compared with those children born MGA (15.7% (95%CI 9.7 to 23.4%)) at 5 years age. For those children born AGA the prevalence was 15% (95%CI 12.9 to 17.3%)) (Guellec 2011).

Total behavioural problems

Low quality evidence from one study (n=625) showed that among children born at 32–36 weeks of gestation the prevalence of behavioural problems (BITSEA, >25^th percentile) was 21% (95%CI 17.8 to 24.4%) at 2 years (corrected age) (Johnson 2015). In the same study, the prevalence of delayed social incompetence (BITSEA <15^th percentile) was 26.4% (95%CI 23 to 30%). For children who had behavioural problems or delayed social competence (BITSEA), or both, the prevalence was 37.3% (95%CI 33.5 to 41.2%) and 10.1% (95%CI 7.8 to 12.7%) respectively (Johnson 2015).

Moderate quality evidence from one study (n=916) showed that among children born at 32–35 weeks of gestation the prevalence of total behavioural problems (CBCL, 90^th percentile) was 7.9% (95%CI 6.2 to 9.8%) at 4 years age (Potijk 2012).

Low quality evidence from one study (n=1504) showed that among children born at 32–35 weeks of gestation the prevalence of emerging total behavioural problems (CBCL, >85^th percentile) was 3.7% (95%CI 2.4 to 5.4%) at 4 and 5 years age (Hornman 2016). In the same study, the prevalence of resolving or persistent total behavioural problems (CBCL, >85^th percentile) was 8.7% (95%CI 6.7 to 11.2%) and 6.6% (95%CI 4.8 to 8.8%) respectively (Hornman 2016).

Total behavioural problems by week of gestation

Low quality evidence from one study (sample 2382) showed that among children born at 24–28 weeks of gestation the prevalence of total behavioural difficulties (SDQ, 10^th percentile) was 24.1% (95%CI 19.2 to 29.6%) at 3 years age (Delobel-Ayoub 2006). In the same study, the prevalence decreased to 16.9% (95%CI 13 to 21.3%) among children born at 29–30 weeks of gestation whereas there was an increase in prevalence of 19% (95%CI 15.9 to 22.4%) among children born at 31–32 weeks of gestation (Delobel-Ayoub 2006).

A similar pattern was observed in another low quality study (n=2901) showed that among children born at 24–28 weeks the prevalence of total behavioural difficulties (SDQ, 10^th percentile) was 27.8% (95%CI 23 to 32.9%) at 8 years age (Larroque 2011). In the same study, the prevalence decreased to 17.2% (95%CI 13.5 to 21.4%) among children born at 29–30 weeks of gestation, whereas there was an increase in prevalence of 19.9% (95%CI 16.9 to 23.1%) among children born at 31–32 weeks of gestation (Larroque 2011).

ASD symptoms by week of gestation

Low quality evidence from one study (n=2035) showed that there was an increase in prevalence of positive autism screening (M-CHAT) with decreasing gestational age, ranging from 54.8% (95%CI 36 to 72.7%) at 23 weeks of gestation to 38.1% (95%CI 31.7 to 44.7%) at 26 weeks of gestation (assessed at 2 years age) (Moore 2012).

Low quality evidence from one study (n=1130–2035) showed that among children born at <27 weeks of gestation the prevalence of autism (positive screen, M-CHAT) was 41% (95%CI 37 to 45.7%) at 2 years age (Moore 2012) compared to the prevalence of those children born at 32–33 or 34–36 weeks of gestation (9.3% (95%CI 4.1 to 17.5%) and 15.3% (95%CI 12.4 to 18.6%)) respectively (Guy 2015).

Total externalising behavioural problems by gestational group

Moderate quality evidence from one study (n=169) showed that among children born at <26 weeks of gestation the prevalence of externalising problems (CBCL, 90^th percentile) was 9.6% (95%CI 4.3 to 18.1%) at 11 years age (Farooqi 2007). In the same study, the prevalence of externalising problems (TRF, 90^th percentile) was 18.1% (95%CI 10.5 to 28.1%).

Moderate quality evidence from one study (n=916) showed that among children born at 32–35 weeks of gestation the prevalence of externalising problems (CBCL, 84^th percentile) was 9.5% (95%CI 7.7 to 11.6%) at 4 years age (Potijk 2012).

Low quality evidence from one study (n=1054) showed that among children born at 25–31 weeks and 32–35 weeks of gestation the prevalence of emerging externalising problems (CBCL, >85^th percentile) was 5.2% (95%CI 3.3 to 7.9%) and 5.4% (95%CI 3.8 to 7.4%) respectively at 4 and 5 years age (Hornman 2016). In the same study, the prevalence for resolving externalising problems at 25–31 and 32–35 weeks of gestation was 5.2% (95%CI 3.3 to 7.9%) and 8.4% (95%CI 6.4 to 10.3%) respectively. The prevalence of persistent externalising problems (CBCL, >85^th percentile) at 25–31 and 32–25 weeks of gestation was 8.2% (95%CI 5.7 to 11.4%) and 8.4% (95%CI 6.4 to 10.8%) respectively at 4 and 5 years age (Hornman 2016).

Total internalising behavioural problems by gestational group

Moderate quality evidence from one study (n=169) showed that among children born at <26 weeks of gestation the prevalence of internalising problems (CBCL, 90^th percentile) was 32.5% (95%CI 22.7 to 43.7%) at 11 years age (Farooqi 2007). In the same study, the prevalence of internalising problems (TRF, 90^th percentile) was 25.3% (95%CI 16.4 to 36%).

Moderate quality evidence from one study (n=916) showed that among children born at 32–35 weeks of gestation the prevalence of internalising problems (CBCL, 84^th percentile) was 9.7% (95%CI 7.9 to 11.8%) at 4 years age (Potijk 2012).

Low quality evidence from one study (n=1054) showed that among children born at 25–31 and 32–35 weeks of gestation the prevalence of emerging internalising problems (CBCL, >85^th percentile) was 8% (95%CI 5.5 to 11.1%) and 6.7% (95%CI 4.9 to 8.9%) at 4 and 5 years age respectively (Hornman 2016). In the same study, the prevalence of resolving internalising problems (CBCL, >85^th percentile) at 25–31 and 32–35 weeks of gestation was 7.2% (95%CI 4.9 to 10.2%) and 7.5% (95%CI 5.6 to 9.8%) respectively. The prevalence of persistent internalising problems at 25–31 and 32–35 weeks gestation was 11.7% (95%CI 8.7 to 15.3%) and 10.1% (95%CI 7.9 to 12.7%) respectively (Hornman 2016).

Attention/hyperactivity problems

Moderate quality evidence from one study (n=169 to 916) showed that among children born at <26 weeks of gestation the prevalence of attention problems (CBCL, 90^th percentile) was 30.1% (95%CI 20.5 to 41.2%) and 24.1% (95%CI 15.4 to 34.7%) using the TRF (90^th percentile) at 11 years age (Farooqi 2007). In another moderate quality study (n=916 the prevalence of attention problems (CBCL, >97^th percentile) was 4.15% (95%CI 3 to 5.7%) among children born at 32–35 weeks of gestation, assessed at 4 years age (Potijk 2012).

Low quality evidence from one study (n=1643) showed that among children born at <34 or 34–36 weeks of gestation the prevalence attention problems (failure to pay attention when crossing street (CBCL) was 22.8% (95%CI 18.5 to 27.5%) and 20.6% (95%CI 18.4 to 22.9%) respectively (Higa-Diez 2016). In the same study the prevalence of adverse outcomes for all attention problems (CBLC) was 9.4% (95%CI 5.6 to 14.6%) and 5.6% (95%CI 4 to 7.6%) among those born at <34 or 34–36 weeks of gestation, assessed at 8 years age.

Low quality evidence from two studies (sample size ranging from 201 to 224) showed a trend of higher prevalence of attention problems (using different tools) among children born at <26 or 28 weeks of gestation (range 30.1% (95% CI 23.3 to 37.5%) to 54% (95%CI 47 to 60.8%)) assessed at 6 and 8 years age respectively (Samara 2008; Anderson 2011).

Low quality evidence from one study (n=1643) showed that among children born at <34 or 34–36 weeks of gestation the prevalence of interrupting people (CBCL) was 41.9% (95%CI 36.7 to 47.2%) and 40.3% (95%CI 37.6 to 43.1%) respectively (Higa-Diez 2016). In the same study, the prevalence of inability to wait turn (CBCL) was 12.6% (95%CI 9.4 to 16.6%) and 9,1% (95%CI 7.6 to 10.8%) respectively.

Adolescents (n=298) born at <28 weeks of gestation had a lower prevalence of hyperactive or inattentive (CADS <−1.5SD) problems, ranging from 14.5% (95%CI 9.9 to 20.1%) (Wilson-Ching 2013). In the same study, the prevalence of shifting attention (CNT, <−1.5SD) or divided attention (Telephone search wile counting/Test of Everyday Attention <−1.5SD) was 41.1% (95%CI 34.4 to 48.2%) and 15.3% (95%CI 10.6 to 21.1%) respectively (Wilson-Ching 2013).

Moderate to low quality evidence from four studies (sample size ranging from 224 to 2901) showed a trend of high prevalence of hyperactivity problems (SDQ, >90^th percentile) among those born at low gestational age of <26 weeks (48% (95%CI 41.3 to 54.8%)) (Samara 2008) compared to a lower prevalence among those born at higher gestational age of 24–32 weeks (17.2% (95%CI 15.3 to 19.3%) (Larroque 2011).

Low quality evidence from two separate studies (sample size ranging from 2382 to 2901) showed that among children born at 24–28 weeks of gestation the prevalence of hyperactivity (SDQ, 10th percentile) was 24.1% (95%CI 19.2 to 29.6%) and 18.5% (95%CI 14.5 to 23.1%) at 3 years and 8 years age respectively (Delobel-Ayoub 2006; Larroque 2011). In the same two studies, the prevalence of hyperactivity ranged from 17.1% (95%CI 13.3 to 21.6%) to 15.1%(95%CI 11.6 to 19.1%) at 29–31 weeks of gestation, assessed at 3 and 8 years age respectively (Delobel-Ayoub 2006; Larroque 2011). The prevalence ranged from 18.5% (95%CI 14.5 to 23.1) to 17% (95%CI 15 to 20.9%) at 31–32 weeks of gestation, assessed at 3 and 8 years (Delobel-Ayoub 2006; Larroque 2011).

Conduct problems

Low to moderate quality evidence from four separate studies (sample size ranging from 224 to 2901) showed a general trend of decreasing prevalence of conduct problems (SDQ, 10^th percentile) with increasing gestational age, ranging from 36.2% (95%CI 29.9 to 42.9%) (<26 weeks gestational age) (Samara 2008) to 9.4 % (95%CI 8.0 to 11.1%) (24–32 weeks gestational age) (Larroque 2011).

Low quality evidence from one study (n=2901) showed that the prevalence of conduct problems (SDQ, 10^th percentile) decreased with increasing gestational age group from 16.1% (11.9 to 21%) at 24–28 weeks of gestation to 15% (95%CI 12.2 to 18.1%) at 31–32 weeks of gestation (assessed at 3 years age) (Delobel-Ayoub 2006). At 8 years, there was no clear trend of prevalence with gestational age group (Larroque 2011).

Emotional problems

Low quality evidence from two separate studies (n=2901) showed that among children born at 24–28 weeks of gestation the prevalence of emotional symptoms (SDQ, 10^th percentile) was 17.2% (95%CI 12.9 to 22.2%) and 20.3% (95%CI 16.1 to 25%) at 3 years and 8 years respectively (Delobel-Ayoub 2006; Larroque 2011). In the same two studies, the prevalence of emotional problems among children born at 29–30 weeks of gestation was 14.1% (95%CI 10.6 to 18.3%) and 14.3% (95%CI 10.9 to 18.2%) at 3 and 8 years age respectively. Prevalence of emotional problems among those born at 31–32 weeks of gestation was 15% (95%CI 12.2 to 18.1%) and 17.2% (95%CI 14.4 to 20.3%) at 3 and 8 years age (Delobel-Ayoub 2006; Larroque 2011).

Moderate quality evidence from one study (n=916) showed that among children born at 32–35 weeks of gestation the prevalence of emotionally reactive problems (CBCL, >97^th percentile) was 3.7% (95%CI 2.6 to 5.2%) (Potijk 2012). In other studies, the prevalence of emotional problems was higher among those born at lower gestational age of <26 weeks of gestation (29.9% (95%CI 23.8 to 36.5%) (Samara 2008).

Peer and prosocial problems

Low quality evidence from one study (n=224) showed that among children born at <26 weeks of gestation the prevalence of peer problems (SDQ, >90^th percentile) was 36% (95%CI 29.7 to 42.7%, parent reported) and 50% (95%CI 43.5 to 57.4%, teacher reported) respectively (Samara 2008). The prevalence of peer problems (SDQ, >90^th percentile) was lower with varying gestational age groups, ranging from 17.4% (95%CI 15.4 to 19.5%) at 24–32 weeks of gestation (Larroque 2011) to 20% (95%CI 17.7 to 22.6%) in those born at 22–32 weeks of gestation (Delobel-Ayoub 2009).

Low quality evidence from two separate studies (sample size ranging from 2382 to 2901) showed a trend of decreasing prevalence of peer problems (SDQ, 10^th percentile) with increasing gestational age, ranging from 17.9% (95%CI 13.5 to 22.9) among those born at 24–28 weeks of gestation to 12% (95%CI 9.5 to 14.9%) among those born at 31–32 weeks of gestation (Delobel-Ayoub 2006; Larroque 2011). A similar trend was observed in another low quality study (sample size 2382) with prevalence ranging from 19.4% (95%CI 15.3 to 24.1%) among those born at 24–28 weeks of gestation to 15.4% (95%CI 12.8 to 18.4%) among those born at 31–32 weeks of gestation (Larroque 2011).

Low quality evidence from one study (n=2382) showed a trend of decreasing prevalence of prosocial behaviour (SDQ, 10^th percentile) with increasing gestational age, ranging from 20.1% (95%CI 15.5 to 25.3%) among those born at 24–28 weeks of gestation to 13% (95%CI 10.4 to 16%) among those born at 31–32 weeks of gestation (Delobel-Ayoub 2006), assessed at 3 years age.

Special education needs (overall and individual problems)

Low quality evidence from one study (n=152757) showed that among children born at 24–27 weeks of gestation the prevalence of SEN was 29.5% (95%CI 25.4 to 33.8%) at 5–18 years age (Mackay 2010).

Low quality evidence from one study (n= 237894) showed that among children born at 24–27 weeks of gestation the prevalence of sensory SEN was 3% (95%CI 1.6 to 4.9%), physical or motor SEN was 6.1% (95%CI 4.1 to 8.7%), language SEN was 0.63% (95%CI 0.13 to 1.83%), intellectual SEN was 14.1% (95%CI 11.1 to 17.6%), specific learning difficulties SEN was 2.1% (95%CI 1.0 to 3.8%), ASD SEN was 1.1% (95%CI 0.3 to 2.4%), and social, emotional behavioural SEN was 1.3% (95%CI 0.5 to 2.7%) at 5–18 years (Mackay 2013).

School difficulties (low grade, repetition of grade, adaption difficulties)

Moderate quality evidence from one study (n=169) showed that among children born at <26 weeks of gestation, the prevalence of school difficulties (repetition of grade and/or use of SEN resources) was 59.3% (95%CI 48.2 to 69.8%) at 11 years age (Farooqi 2007). In the same study, the prevalence of grade repetition was 15.7% (95%CI 8.6 to 25.3%).

Low quality evidence from one study (n=2382) showed that among children born at <26 weeks of gestation the prevalence of school adaption difficulties (parent reported) was 33% (95%CI 36.7 to 39.8%) compared to a prevalence (teacher reported) of 39.2% (95%CI 32.6 to 46.2%) at 6 years age (Samara 2008).

Low quality evidence from one study (n=2855) showed that among children born at 24–28 weeks of gestation and were small for gestational age, the prevalence of school difficulties (special schooling or low grades, parent reported) was 35.3% (95%CI 14.2 to 61.7%) at 8 years age compared to those who were born MGA (prevalence 44.8% (95%CI 26.5 to 64.3%) (Guellec 2011).

Identified special education needs

Low quality evidence from one study (n=219) showed that among children born at <26 weeks of gestation the overall prevalence of identified SEN (teacher reported) was 62.3% (95%CI 55.5 to 68.8%) at 11 years age. In the same study, the prevalence of SEN identified in mainstream schools only (teacher reported) was 56.5% (95%CI 49 to 63.7%) (Johnson 2011).

Special school or special class

Moderate quality evidence from one study (n=169) showed that among children born at <26 weeks of gestation the prevalence of those in special class or special school was 15.1% (95%CI 8.3 to 24.5%) at 11 years age (Farooqi 2007).

Low quality evidence from one study (n=2901) showed that among children born at 24–28 weeks of gestation the prevalence of those in an institution or special school or special class (parent reported) was 9.4% (95%CI 6.5 to 13.0%) at 8 years age (Larroque 2011).

Special education needs provision/support at school

Low quality evidence from one study (n=219) showed that among children born at <26 weeks of gestation the prevalence of SEN provision (teacher reported) was 61.4% (95%CI 54.5 to 68.8%) at 11 years age (Johnson 2011). In the same study, among children who had SEN provision in mainstream school only (teacher reported) the prevalence was 55.4% (95%CI 47.9 to 62.7).

Low quality evidence from one study (n=2901) showed that among children born at 24–28 weeks of gestation the prevalence of support in mainstream school (parent reported) was 22.7% (95%CI 18.3 to 27.5%) at 8 years age (Larroque 2011). In the same study, the prevalence of children who had special care since 5 years age or support at school (parent reported) was 69.7% (95%CI 64.5 to 74.5%) at 8 years age. The prevalence of children who had special care since 5 years for more than one developmental problem (orthoptic, speech therapy, PT, OT or psychology) was 65.4% (95%CI 60.1 to 70.4%) at 8 years age (Larroque 2011).

Special education needs (overall and individual problems)

Low quality evidence from one study (n=152757) showed that among children born at 28–32 weeks of gestation the prevalence of SEN was 12.8% (95%CI 11.7 to 14%) at 5–18 years age (Mackay 2010).

Low quality evidence from one study (n=237894) showed that among children born at 28–32 weeks of gestation the prevalence of sensory SEN was 0.49% (95%CI 0.29 to 0.79%), physical or motor SEN was 2.8% (95%CI 2.3 to 3.5%), language SEN was 0.38% (95%CI 0.2 to 0.6%), intellectual SEN was 4.8% (95%CI 4.1 to 5.6%), specific learning difficulties SEN was 1.4% (95%CI 1.1 to 1.9%), ASD SEN was 1.0% (95%CI 0.7 to 1.4%), and social, emotional behavioural SEN was 0.9% (95%CI 0.6 to 1.3%) at 5–18 years (Mackay 2013).

School difficulties (special schooling or low grades)

Low quality evidence from one study (n=2855) showed that among children born at 29–32 weeks of gestation who were small for gestational age, the prevalence of school difficulties was 28% (95%CI 19.8 to 37.6%) at 8 years age compared to a prevalence of 23.1% (95%CI 15.4 to 32.4%) among children who were MGA (Guellec 2011).

Special school or special class

Low quality evidence from one study (n=2901) showed that among children born at 29–30 weeks of gestation the prevalence of those in an institution or special school or special class (parent reported) was 5.2% (95%CI 3.2 to 7.9%) at 8 years age (Larroque 2011).

Support at school

Low quality evidence from one study (n=2901) showed that among children born at 29–30 weeks of gestation the prevalence of support in mainstream school (parent reported) was 10.3% (95%CI 7.5 to 13.8%) at 8 years age (Larroque 2011). In the same study, the prevalence of children who had special care since 5 years age or support at school (parent reported) was 53.6% (95%CI 48.5 to 58.7%) at 8 years age.

Special school or special class

Low quality evidence from one study (n=2901) showed that among children born at 24–32 weeks of gestation the prevalence of those in an institution, special school or special class (parent reported) was 5.2% (95%CI 4.1 to 6.5%) at 8 years age (Larroque 2011). For those children born at 31–32 weeks of gestation, the prevalence of the same outcome was 3.3% (95%CI 2.1 to 4.8%) at 8 years age.

Support at school

Low quality evidence from one study (n=2901) showed that among children born at 24–32 weeks of gestation the prevalence of those supported at school in mainstream class (parent reported) was 15.4% (95%CI 13.6 to 17.4%) at 8 years age (Larroque 2011). For those children born at 31–32 weeks of gestation, the prevalence of the same outcome was 14.7% (95%CI 12.2 to 17.5%) at 8 years age.

In the same study among children born at 24–32 weeks of gestation, the prevalence among those who had special care since 5 years age or support at school (parent reported) was 58.5% (95%CI 55.9 to 61.1%) at 8 years age. Among children born at 31–32 weeks of gestation, the prevalence of the same outcome was 55.7% (95%CI 52 to 59.4%) at 8 years age (Larroque 2011).

Attainment (Foundation Stage Profile [FSP] or Key Stage 1 [KS1])

Moderate quality evidence from one study (n=8728) showed that among children born at 23–31 weeks of gestation the prevalence of those not attaining a good overall level of achievement (teacher reported FSP) was 66.7% (95%CI 55.5 to 76.6%) at 5 years age (Quigley 2012). In the same study, among children who did not attain in all three scales of personal, social and emotional development (teacher reported FSP) the prevalence was 42.9% (95%CI 32.1 to 54.1%). Among children who did not attain in all 4 scales of communication, language and literacy, the prevalence was 61.9% (95%CI 43.5 to 65.7%). The prevalence was 54.8% (95%CI 43.5 to 65.7%) among children who did not attain in all 3 scales of mathematical development a 5 years age (Quigley 2012).

Moderate quality evidence from one study (n=18818) showed that among children born at <32 weeks of gestation not achieving level 2 or more in reading, writing or maths (teacher reported KS1) the prevalence was 42% (95%CI 30.2 to 54.5%) at 7 years age (Chan 2014). In the same study, the prevalence among children not achieving level 2 or more in speaking and listening was 29% (95%CI 18.7 to 41.2%) and for science, the prevalence was 24.6% (95%CI 15.1 to 36.5%) (Chan 2014).

Special education needs (overall and individual problems)

Low quality evidence from one study (n=152757) showed that among children born at 33–36 weeks of gestation the prevalence of SEN was 7.1% (95%CI 6.7 to 7.5%) at age 5–18 years age (Mackay 2010).

Very low quality evidence from one study (n=741) showed that among children born at 32–36 weeks of gestation the prevalence of SEN was 34.5% (95%CI 29.3 to 40%) at 8 years age (Odd 2012).

Individualised programme

Low quality evidence from one study (n=17565) showed that among children born at 32–36 weeks of gestation the prevalence of those who were enrolled on an individualised education programme (ECLS-K data) was 9.1% (95%CI 7.1 to 11.4%) at kindergarten stage (3 years age), 12% (95%CI 9.7 to 14.6%) at 1^st grade (6–7 years), 13.6% (95%CI 11.1 to 16.5%) at 3^rd grade (8–9 years) and 16.4% (95%CI 12.9 to 20.4%) at 5^th grade (10–11 years) (Chyi 2008).

Low quality evidence from one study (n=17565) showed that among children born at 32–33 weeks of gestation the prevalence of those who were enrolled on an individualised education programme was 13% (95%CI 8 to 19%) at kindergarten stage (3 years age), 17.8% (95%CI 12 to 25%) at 1^st grade (6–7 years), 19.7% (95%CI 13.3 to 27.5%) at 3^rd grade (8–9 years) and 18.1% (95%CI 10.9 to 27.4%) at 5^th grade (10–11 years) (Chyi 2008).

Low quality evidence from one study (n=17565) showed that among children born at 34–36 weeks of gestation the prevalence of those who were enrolled on an individualised education programme was 8% (95%CI 6 to 10.6%) at kindergarten stage (3 years age), 10.5% (95%CI 8.2 to 13.3%) at 1st grade (6–7 years), 12.1% (95%CI 9.5 to 15.2%) at 3rd grade (8–9 years) and 12.2% (95%CI 9.2 to 15.8%) at 5th grade (10–11 years) (Chyi 2008).

Special education enrolment

Low quality evidence from one study (n=17565) showed that among children born at 32–36 weeks of gestation the prevalence of those who were enrolled on a special education programme (ECLS-K data) was 6.9% (95%CI 5.4 to 8.7%) at kindergarten stage (3 years age), 7.4% (95%CI 5.8 to 9.3%) at 1st grade (6–7 years), 10% (95%CI 8 to 12.3%) at 3rd grade (8–9 years) and 11.1% (95%CI 8.8 to 13.8%) at 5th grade (10–11 years) (Chyi 2008).

Low quality evidence from one study (n=17565) showed that among children born at 32–33 weeks of gestation the prevalence of those who were enrolled on a special education programme (ECLS-K data) was 8% (95%CI 4.7 to 12.7%) at kindergarten stage (3 years age), 11.9% (95%CI 7.7 to 17.3%) at 1st grade (6–7 years), 14.4% (95%CI 9.2 to 21%) at 3rd grade (8–9 years) and 14.5% (95%CI 8.8 to 22%) at 5th grade (10–11 years) (Chyi 2008).

Attainment (FSP or KS1)

Moderate quality evidence from one study (n=8728) showed that among children born at 32–36 weeks of gestation the prevalence of those not with a good level of overall achievement (FSP, teacher reported) was 59% (95%CI 54.8 to 63.1%) at 5 years age (Quigley 2012). In the same study, the prevalence among children who did not achieve in all 3 scales of personal, social and emotional development (FSP, teacher reported) was 31.6% (95%CI 27.8 to 35.6%). For those who did not achieve in all 4 scales of communication, language and literacy, the prevalence was 49.1% (95%CI 47.7 to 50.4%), and for mathematical development (not achieving in all 3 scales) the prevalence was 37.5% (95%CI 33.5 to 41.6%) at 5 years age (Quigley 2012).

Moderate quality evidence from one study (n=8728) showed that among children born at 32–33 weeks of gestation the prevalence of those not with a good level of overall achievement (FSP, teacher reported) was 60.9% (95%CI 50.1 to 70.9%) at 5 years age (Quigley 2012). In the same study, the prevalence among children who did not achieve in all 3 scales of personal, social and emotional development (FSP, teacher reported) was 32.6% (95%CI 23.2 to 43.2%). For those who did not achieve in all 4 scales of communication, language and literacy, the prevalence was 57.6% (95%CI 46.9 to 67.9%), and for mathematical development (not achieving in all 3 scales) the prevalence was 40.2% (95%CI 30.1 to 51%) at 5 years age (Quigley 2012).

Moderate quality evidence from one study (n=8728) showed that among children born at 34–36 weeks of gestation the prevalence of those not with a good level of overall achievement (FSP, teacher reported) was 58.6% (95%CI 54 to 63.1%) at 5 years age (Quigley 2012). In the same study, the prevalence among children who did not achieve in all 3 scales of personal, social and emotional development (FSP, teacher reported) was 31.4% (95%CI 27.3 to 35.8%). For those who did not achieve in all 4 scales of communication, language and literacy, the prevalence was 54.1% (95%CI 49.5 to 58.7%), and for mathematical development (not achieving in all 3 scales) the prevalence was 36.9% (95%CI 32.6 to 33.5%) at 5 years age (Quigley 2012).

Moderate quality evidence from one study (n=18818) showed that among children born at 32–33 weeks of gestation not achieving level 2 or more in reading, writing and mathematics (KS1, teacher reported), the prevalence was 26.9% (95%CI 16.8 to 39.1%) at 7 years age (Chan 2014). For those children not achieving level 2 or more in reading, writing, speaking/listening and science, the prevalence was 19.4% (95%CI 10.8 to 30.9%), 23.9% (14.3 to 35.9%), 16.4% (95%CI 8.5 to 27.5%) and 16.4% (95%CI 8.5 to 27.5%), respectively (Chan 2014).

Moderate quality evidence from one study (n=18818) showed that among children born at 34–36 weeks of gestation not achieving overall level 2 or more in reading, writing and mathematics was 23.3% (95%CI 19.1 to 28.1%) at 7 years age (Chan 2014). For those children not achieving level 2 in reading, writing, speaking and listening, mathematics, or science, the prevalence was 18.1% (95%CI 14.2 to 22.4%), 20.6% (95%CI 16.5 to 25.1%), 13.1% (95%CI 9.8 to 17%), 8.6% (95%CI 5.9 to 12%), and 11.7% (95%CI 8.5 to 15.4%), respectively (Chan 2014).

Very low quality evidence from one study (n=13978) showed that among children born at 32–36 weeks of gestation not achieving level 2 or more in reading, writing or maths (teacher reported KS1) the prevalence was 29% (95%CI 25.4 to 33%) at 5–7 years age (Peacock 2012). For those children not achieving level 2 in reading, writing and mathematics (individual items of KS1), the prevalence was 22.2% (95%CI 19 to 25.7%), 22.7% (95%CI 19.4 to 26.2%), and 18.1% (95%CI 15.1 to 21.5%) respectively (Peacock 2012).

Overall special education needs

Low quality evidence from one study (n=722) showed that among children born at <37 weeks of gestation the prevalence of SEN was 35.5% (95%CI 32 to 39.1%) at 8 years age (Odd 2013).

Moderate quality evidence from one study (n=775) showed that among children born at <37 weeks of gestation the prevalence of SEN was 24.3% (95%CI 21.1 to 27.7%) at 14 to 16 years age (Odd 2016).

Low quality evidence from one study (n=722) showed that among children born at <37 weeks of gestation the prevalence of low achievement (KS1) was 31.4% (95%CI 28.1 to 35%) at 8 years age (Odd 2013).

Any cerebral palsy

Moderate to low quality evidence from four studies (sample size ranging from 141 to 373) showed that among children born at 22–27 weeks GA the prevalence of any CP varied from 7% (95% CI 4.6 to 10.10) to 11.3% (95%CI: 6.6 to 17.8) at 2 years (corrected age), 5 years and 8 years (corrected) (Leversen 2010; Leversen 2011; Roberts 2011; Anderson 2011).

Moderate quality evidence from four studies (sample size ranging from 75 to 244) showed that among children born at <27 weeks GA the prevalence of any CP varied from 14.7% (95%CI 7.6 to 24.7% to 24.7% (95%CI 15.6 to 35.8%) at age range 12 months CA to 9 years (Mikkola 2005; Stahlmann 2009; Sutton 1999; De Groote 2007).

Moderate to low quality evidence from four studies (sample size ranging from 275 to 331,154) showed that among children born at <28 weeks GA the prevalence of any CP varied from 6.7% (95%CI 5.1 to 8.6) to 16.6% (95%CI 12.5 to 21.3) (Larroque 2008; Ancel 2006; Glinianaia 2011; Anderson 2003).

Moderate quality evidence from one study (n=1718) showed that among children born at 24–27 weeks GA the prevalence of any CP was 14.7% (95%CI 10.6–19.5%) at 5 years age (Foix-Helias 2008).

Low quality evidence from one study (n=104) showed that among children born at 22–26 weeks GA the prevalence of any CP was 11.5% (95% CI 6.1–19.3%) at 18 months CA (Tommiska 2003)

Low quality evidence from one study (n=283) showed that among children born 22–25 weeks GA the prevalence of any CP was 17.7% (95% CI 13.4–22.6%) at a median age of 30 months (Wood 2000).

Moderate to very low quality evidence from three studies (sample size ranging from 19 to 189) showed that among children born at a mean GA range of 25.4 (±1) to 26.5 (±2) weeks the prevalence of any CP was 7.3% (95% CI 3.8–12.4%) to 37% (95%CI 16–62%) at age 2 years to 8 years (Hutchinson 2013; Doyle 2011; Rieger-Fackeldey 2010).

Low quality evidence from one study (n=219) showed that among children born at 23–27 weeks GA the prevalence of any CP was 11% (95%CI 7.2–15.9%) at 2 years age (Anon 1997).

Moderate quality evidence from one study (n=142) showed that among children born at a mean GA of 27 weeks, the prevalence of CP was 19.0% (95%CI 12.9 to 26.5%) at 4 years age (Salakorpi 2001).

Mild cerebral palsy

Moderate to low quality evidence from two studies (sample size ranging from 77 to 456) showed that among children born at <27 weeks GA the prevalence of mild CP across the two studies (10.4% (95%CI 4.6 to 19.5) and 2.9% (95% CI 1.5 to 4.8)) at 2.5 years CA and 3 years age (De Groote 2007; Serenius 2013).

Moderate cerebral palsy

Moderate quality evidence from one study (n=241) showed that among children born at <26 weeks the prevalence was 7.1% (95%CI 4.2 to 11.1) at 6 years (Marlow 2005). The prevalence was varied in two studies of moderate to low quality in children (sample size ranging from 456 to 576) born at <27 weeks GA (2.6% (95%CI 1.5 to 4.3)) and 2.9% (95%CI 1.5 to 4.8)) (Moore 2012; Serenius 2013), whereas prevalence of CP was 13% (95% CI 6.4 to 22.6) in one study (at <27 weeks GA) (De Groote 2007).

Moderate to severe cerebral palsy

Moderate to low quality evidence from two studies (sample size ranging from 88 to 241) showed that among children born at <26 weeks GA the prevalence of CP (moderate/disabling or both ambulatory/non-ambulatory) was varied, with a prevalence of 6.8% (95%CI 2.5 to 14.3) at 11 years (Farooqi 2011) and 13.3% (95%CI 9.3 to 18.2) at 6 years (Marlow 2005). There was also variation of prevalence of moderate to severe CP in children born at <27 weeks GA at 2.5 years corrected age (4.2% (95%CI 2.5 to 6.4)) and at 3 years (7.8% (95%CI 5.8 to 10.3)) in two studies of moderate and low quality (Serenius 2013; Moore 2012).

Moderate quality evidence from one study (n=3785) showed that among children born at 22–26 weeks GA the prevalence of moderate to severe CP (non-ambulatory or needing aids) was 11% (95%CI 9.8 to 12.4) at 18–22 months corrected age (Vohr 2005).

Severe cerebral palsy

Moderate to low quality evidence from two studies (sample size ranging from 77 to 456) showed that among children born at <27 weeks GA the prevalence of severe CP was 1.3% (95%CI 0.03 to 7%) at age 2.5 years CA to 3 years (Serenius 2013; De Groote 2007).

Moderate quality evidence from two studies (sample size ranging from 75 to 241) showed that among children born at <26 weeks and <27 weeks GA the prevalence of non-ambulatory CP was 6.2% (95%CI 3.5 to 7.4%) at 6 years age (Marlow 2005), and 10.7% (95%CI 4.7 to 19.9%) at 7–9 years age (Stahlmann 2009).

Low quality evidence from one study (n=576) showed that among children born at <27 weeks GA the prevalence for severe CP (GMFCS level 3–5) was 5.2% (95% CI 3.5–7.4%) at 3 years age (Moore 2012). Moderate quality evidence from one study (n=306) showed that among children born at 22–27 weeks GA the prevalence for severe CP (GMFCS level4–5) was 3.3% (95%CI 1.6–5.9%) at 5 years age (Leversen 2011).

Low quality evidence from one study (n=283) showed that among children born at 22–25 weeks GA the prevalence of severe diplegia was 4.2 % (95%CI 2.2 to 7.3), severe hemiplegia was 0.4% (95%CI 0.01 to 2), and severe quadriplegia was 3.9% (95%CI 2 to 6.9) at 30 months corrected age (Wood 2000).

Low quality evidence from one study (n=1718) showed that among children born at 24–27 weeks GA the prevalence for severe CP (unable to walk or only with aids) was 4.9% (95% CI 2.6 to 8.2%) at 5 years age (Foix-Helias 2008).

Low quality evidence from one study (n=1506) showed that among children born at <28 weeks GA the prevalence for severe motor impairment (GMFCS level 5, no self-mobility) was 1.9% (95%CI 1.1–3.1) at 10 years age (Joseph 2016b).

Any cerebral palsy

Moderate to low quality evidence from three studies (sample size ranging from 1812 to 331,154) showed that among children born at 28–31 weeks the prevalence of any CP was varied, ranging from 5.9% (95%CI 4.9 to 7) to 9.5% (95%CI 7.8 to 11.4) across the three studies at 2–8 years (Larroque 2008; Ancel 2006; Glinianaia 2011).

Moderate to low quality evidence from two studies (sample ranging from 1455 to 1781) showed that among children born at 28–32 or 30–31 weeks, there was no difference in prevalence (7.7% (95%CI 5.8 to 9.9) and 7.9% (95%CI 6.6 to 9.3)) at 5 years (Marret 2007; Foix-Helias 2008). However, moderate quality evidence from one study (n=3785) showed that among children born at 27–32 weeks GA the prevalence of CP was higher (11.6% (95%CI 10 to 13.3) at 18–22 months corrected age (Vohr 2005).

Moderate quality evidence from one study (n=3785) showed that among children born at 22–32 weeks GA the prevalence of CP was 16% (95%CI 14.9 to 17.2) at 18–22 months corrected age (Vohr 2005). However, the prevalence of CP was lower (4.3% (95%CI 2.2 to 7.5)) in low quality evidence from one study (n=259) among children born at 23–32 weeks GA (Andrews 2008). There was minimal difference in prevalence in GA groups including 24–32 weeks (prevalence 8.9% (95%CI 7.6 to 10.3)) (Foix-Helias 2008), 25–32 weeks GA (prevalence 13.2 (95%CI 8.4 to 19.3)) (Burguet 1999), or <31(prevalence 16% (95%CI 14.9 to 17.2)), <32 (prevalence 11% (95%CI 6.5 to 17), or <33 weeks GA (prevalence 8.8% (95%CI 7.5 to 10.2)) (Vincer 2014; Toome 2012; Larroque 2008).

Mild cerebral palsy

Low quality evidence from one study (n=801) showed that among children born at <31 weeks GA the prevalence of 6.7% (95%CI 5.1 to 8.7) for mild CP (GMFCS level1) at 12–42 months corrected age (Vincer 2014).

Moderate to severe cerebral palsy

Low quality evidence from one (n=801) showed that among children born at <31 weeks GA the prevalence of moderate to severe CP (GMFCS level 2–5) was 3.4% (95%CI 2.2–4.9%) at 12–42 months corrected age (Vincer 2014).

Low quality evidence from one study (n=155) showed that among children born at <32 weeks GA the prevalence of moderate to severe CP (GMFCS level 2–5) was 8.4% (95%CI 4.5–13.9%) at 2 years CA (Toome 2012).

Low quality evidence from one study (1455) showed that among children born at 30–31 weeks GA the prevalence of 5.7% (95%CI 4.1 to 7.7) for moderate to severe CP (bilateral spastic CP) at 5 years (Marret 2007).

Moderate quality evidence from one study (n=3785) showed that among children born at 27–32 weeks GA the prevalence for moderate to severe CP (non-ambulatory or needing aids) was 7% (95%CI 5.8 to 8.4) at 18–22 months corrected age (Vohr 2005).

Moderate quality evidence from one study (n=3785) showed that among children born at 22–32 weeks GA the prevalence of moderate to severe CP (non-ambulatory or needing aids) was 9.4% (95%CI 8.5–10.4%) at 18–22 months corrected age (Vohr 2005).

Severe cerebral palsy

Moderate quality evidence from one study (n=1781) showed that among children born at 28–32 weeks GA the prevalence of severe CP (unable to walk or only with aids) was 2.4% (95%CI 1.7 to 3.4) at 5 years (Foix-Helias 2008). In the same study, the prevalence at 24–32 weeks was 2.8% (95%CI 2.1 to 3.7).

Any cerebral palsy

Low quality evidence from one study (n=1455) showed that among children born at 32–34 weeks GA the prevalence of any CP type was 3.4% (95%CI 2.3 to 5) at 5 years (Marret 2007).

Moderate to low quality evidence from three studies (sample size ranging from 741 to 331,154) showed that among children born at 32–26 weeks GA the prevalence of any CP was similar (range from 0.8% (95%CI 0.7 to 0.9) to 1% (95%CI 0.8 to 1.1) across the studies at age up to 7 or 8 years (^{Odd 2013}; Hirvonen 2014; Glinianaia 2011).

Moderate to severe cerebral palsy

Low quality evidence from one study (n=1455) showed that among children born at 32–34 weeks GA the prevalence of CP (bilateral spastic CP) was 2.2% (95% CI 1.3 to 3.5) at 5 years (Marret 2007).

Moderate quality from one study (n=53,078) showed that among children born at 32–36 weeks GA found the prevalence of CP (other types) was 0.35% (95%CI 0.3 to 0.4) at up to 7 years (Hirvonen 2014).

Low quality evidence from one study (n=331,154) showed that among children born at <37 weeks GA the prevalence of spastic-bilateral or unilateral CP was 1.3% (95%CI 1.1 to 1.5) and 0.4% (95%CI 0.3 to 0.5) respectively at up to 8 years (Glinianaia 2011).

Low quality evidence from one study (n=104) showed that among children born at 22.3–34.9 weeks GA/bw <1000g the prevalence of CP (ataxia/athetosis) was 1% (95%CI 0.1 to 3.4) at 18 months corrected age (Tommiska 2003).

Children born small for gestational age

Low quality evidence from one study (n=2357) showed that among children born at 24–28 weeks GA and small for gestational age, the prevalence of any CP was 18% (95%CI 5.2–40.3%). In the same study, the prevalence was 3.2% (95%CI 0.9–8%) at 5 years age (Guellec 2011).

Hemiplegia

Low quality evidence from one study (n=283) showed that among children born at 22–25 weeks GA the prevalence of hemiplegia was 1.8% (95%CI 0.6–4.1%) at median 30 months (Wood 2000). In the same study, the prevalence of severe hemiplegia was 0.4% (95%CI 0.01–2%).

Very low quality evidence from one study (n=167) showed that among children born at 25–32 weeks GA the prevalence of hemiplegia was 1.2% (95%CI 0.2–4.3%) at 2 years (corrected age) (Burguet 1999).

Low quality evidence from one study (n=77) showed that among children born at <27 weeks GA the prevalence of hemiplegia was 3.9% (95%CI 0.8–11%) at 3 years age (De Groote 2007).

Moderate quality evidence from one study (n=142) showed that among children born at a mean GA of 27 weeks, the prevalence of hemiplegia was 5.6% (95%CI 2.5 to 10.8%) at 4 years (Salakorpi 2001).

Low quality evidence from one study (n=1455) showed that among children born at gestational age ranging from 30 to 33 weeks the prevalence of hemiplegia ranged from 0.4% to 0.8% (95%CI range 0.01 – 4.1%) at 5 years age (Marret 2007).

Moderate quality evidence from one study (n=53,078) showed that among children born at <32 weeks GA the prevalence of hemiplegia was 1.3 % (95%CI 1–1.6%) at age up to 7 years (Hirvonen 2014). In the same study the prevalence of hemiplegia CP was 0.5% (95%CI 0.4–0.8%) at 32–33 weeks GA, 0.14% (95%CI 0.11–0.19%) at 34–36 weeks GA, and 0.2% (95%CI 0.16–0.25%) at 32–26 weeks GA (Hirvonen 2014).

Diplegia

Low quality evidence from one study (n=104) showed that among children born at 22.3 to 34.9 weeks GA the prevalence of diplegia was 7.2% (95%CI 4.1–11.6%) at 18 months corrected age (Tommiska 2003).

Low quality evidence from one study (n=283) showed that among children born at 22–25 weeks GA the prevalence of diplegia was 9.5% (95%CI 6.4–13.6 %) at median 30 months (Wood 2000). In the same study, the prevalence of severe diplegia was 4.2% (95%CI 2.2–7.3%).

Very low quality evidence from one study (n=167) showed that among children born at 25–32 weeks GA the prevalence of spastic diplegia was 6% (95%CI 2.9–10.7%) at 2 years (corrected age) (Burguet 1999).

Low quality evidence from one study (n=77) showed that among children born at <27 weeks GA the prevalence of diparesis was 11.7% (95%CI 5.5–21%) at 3 years age (De Groote 2007).

Low quality evidence from one study (n=155) showed that among children born at <32 weeks GA the prevalence of spastic diplegia was 4.5% (95%CI1.8–9.1%) at 2 years (corrected age) (Toome 2012).

Moderate quality evidence from one study (n=53,078) showed that among children born at <32 weeks GA the prevalence of diplegia was 3.4 % (95%CI 2.9–3.8%) at age up to 7 years (Hirvonen 2014). In the same study the prevalence of diplegia CP was 0.7% (95%CI 0.5–0.9%) at 32–33 weeks GA, 0.13% (95%CI 0.10–0.17%) at 34–36 weeks GA, and 0.2% (95%CI 0.17–0.26%) at 32–26 weeks GA (Hirvonen 2014).

Triplegia

Low quality evidence from one study (n=77) showed that among children born at <27 weeks GA the prevalence of triparesis was 2.6% (95%CI 0.3–9.1%) at 3 years age (De Groote 2007).

Diplegia or tetraplegia

Moderate quality evidence from one study (n=142) showed that among children born at a mean GA of 27 weeks, the prevalence of bilateral spastic CP (diplegia or tetraplegia) was 10.6% (6.0 to 16.8%) at 4 years (Salakorpi 2001).

Tetraplegia

Low quality evidence from one study (n=104) showed that among children born at 22.3 to 34.9 weeks GA the prevalence of tetraplegia was 1.9% (95%CI 0.5–4.9%) at 18 months corrected age (Tommiska 2003).

Very low quality evidence from one study (n=167) showed that among children born at 25–32 weeks GA the prevalence of tetraplegia was 1.2% (95%CI 0.2–4.3%) at 2 years (corrected age) (Burguet 1999).

Quadriplegia

Low quality evidence from one study (n=283) showed that among children born at 22–25 weeks GA the prevalence of quadriplegia was 4.2% (95%CI 2.2–7.3 %) at median 30 months (Wood 2000). In the same study, the prevalence of severe quadriplegia was 3.9% (95%CI 2.0–6.9%).

Low quality evidence from one study (n=77) showed that among children born at <27 weeks GA the prevalence of quadriplegia was 5.2% (95%CI 1.4–12.8%) at 3 years age (De Groote 2007).

Moderate quality evidence from one study (n=53,078) showed that among children born at <32 weeks GA the prevalence of quadriplegia was 0.6 % (95%CI 0.4–0.8%) at age up to 7 years (Hirvonen 2014). In the same study the prevalence of quadriplegia was 0.2% (95%CI 0.1–0.3%) at 32–33 weeks GA, 0.04% (95%CI 0.02–0.06%) at 34–36 weeks GA, and 0.06% (95%CI 0.04–0.08%) at 32–26 weeks GA (Hirvonen 2014).

Dystonic or athetoid type

Moderate quality evidence from one study (n=142) showed that among children born at a mean GA of 27 weeks, the prevalence of dystonic or athetoid CP was 2.8% (95%CI 0.8 to 7.1%) at 4 years (Salakorpi 2001).

Any cerebral palsy

Low quality evidence from one study (n=244) showed that among children born at 23 weeks GA the prevalence of any CP was 100% (95%CI 25 to 100%) at 12 months corrected age. However, the prevalence was 19.10% (95%CI 12 to 27.9%) for children who were born at 27 weeks GA (Sutton 1999).

Low quality evidence from one study (n=104) showed that among children born at 22–23 weeks GA the prevalence of any CP was 20% (95%CI 0.5 to 71.6%) compared to a prevalence of 10.6% (95%CI 3.6 to 23.10%) in children who were born at 26 weeks GA, assessed at the age of 18 months corrected age (Tommiska 2003).

Low quality evidence from one study (n=1954) showed that among children born at 24–25 weeks GA the prevalence of any CP was 19.4% (95%CI 10.4 to 31.4%) compared to a prevalence of 4.4% (95%CI 2.9 to 6.6%) in children who were born at 32 weeks GA, assessed at the age of 2 years (Ancel 2006).

Moderate quality evidence from one study (n=1812) showed that among children born at 24–25 weeks GA the prevalence of any CP was 18.3% (95%CI 9.5 to 30.4%) compared to a prevalence of 4.1% (95%CI 2.6 to 6.2%) in children who were born at 32 weeks GA, assessed at the age of 5 years (Larroque 2008).

Low quality evidence from one study (n=1455) showed that among children born at 30 weeks GA the prevalence of any CP was 6.3% (95%CI 3.8 to 9.7%) compared to a prevalence of 3.7% (95%CI 1.2 to 8.4%) in children who were born at 34 weeks GA, assessed at the age of 5 years (Marret 2007).

Moderate quality evidence from one study (n=6347) showed that among children born at <32 weeks GA the prevalence of any CP was 8.7% (95%CI 8.0 to 9.4%) compared to a prevalence of 0.56% (95%CI 0.49 to 0.64%) in children born at 34–36 weeks GA, assessed at up to the age of 7 years (Hirvonen 2014).

Moderate cerebral palsy

Low quality evidence from one study (n=576) showed that among children born at 24 weeks the prevalence of moderate CP was 4.1% (95%CI 1.1 to 10.1%) compared to a prevalence of 2% (95%CI 0.7 to 4.6%) in children who were born at 26 weeks GA, assessed at 3 years age (Moore 2012).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks the prevalence of moderate CP was 12.5% (95%CI 2.7 to 32.4%) compared to a prevalence of 5.6% (95%CI 2.4 to 10.7%) in children who were born at 25 weeks GA, assessed at 6 years age (Marlow 2005).

Moderate quality evidence from one study (n=306) showed that among children born at 23–25 weeks GA the prevalence of moderate CP was 4.6% (95%CI 1.3 to 11.4%) compared to a prevalence of 2.0% (95%CI 0.4 to 5.7%) in children born at 26–27 weeks GA, assessed at 5 years age (Leversen 2011).

Moderate to severe cerebral palsy

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks the prevalence of moderate to severe CP (GMFCS 2–5) was 10.5% (95%CI 2.9 to 24.8%) compared to a prevalence of 6.4% (95%CI 3.7 to 10.2%) in children who were born at 26 weeks GA, assessed at 3 years age (Moore 2012).

Low quality evidence from one study (n=1455) showed that among children born at 30 weeks GA the prevalence of moderate to severe CP (bilateral spastic CP) was 4.2% (95%CI 2.2 to 7.2%) compared to a prevalence of 1.5% (95%CI 0.2 to 5.3%) in children who were born at 34 weeks GA, assessed at the age of 5 years (Marret 2007).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks the prevalence of moderate to severe CP (ambulatory or non-ambulatory) was 16.7% (95%CI 4.7 to 37.4%) compared to a prevalence of 9.7% (95%CI 5.4 to 15.8%) in children who were born at 25 weeks GA, assessed at 6 years age (Marlow 2005).

Moderate quality evidence from one study (n=6347) showed that among children born at <32 weeks GA the prevalence of moderate to severe CP (other types) was 3.5% (95%CI 3.0 to 4.0%) compared to a prevalence of 0.25% (95%CI 0.2 to 0.3%) in children born at 34–36 weeks GA, assessed at up to the age of 7 years (Hirvonen 2014).

Severe cerebral palsy

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks the prevalence of severe CP (GMFCS 3–5) was 10.5% (95%CI 2.9 to 24.8%) compared to a prevalence of 4.4% (95%CI 2.2 to 7.7%) in children who were born at 26 weeks GA, assessed at 3 years age (Moore 2012).

Moderate quality evidence from one study (n=306) showed that among children born at 23–25 weeks GA the prevalence of severe CP (GMFCS 4–5) was 9.2% (95%CI 4.1 to 17.3%) compared to a prevalence of 1.3% (95%CI 0.2 to 4.7%) in children born at 26–27 weeks GA, assessed at 5 years age (Leversen 2011).

Moderate quality evidence from one study (n=1455) showed that among children born at ≤23 weeks the prevalence of severe CP (non-ambulatory) was 4.2% (95%CI 0.1 to 21.1%) compared to a prevalence of 4.2% (95%CI 1.5 to 8.9%) in children who were born at 25 weeks GA, assessed at 6 years age. The prevalence among children born at 24 weeks was higher (11% (95%CI 4.9 to 20.5%) (Marlow 2005).

Any cerebral palsy

Low quality evidence from one study (n=2858) showed that among children born at <28 weeks GA the rate of any CP was 112.7 per 1000 survivors (95%CI 50 to 210) (Drummond 2002).

Moderate quality evidence from one study (n=94,466 live births) showed that among children born at <28 weeks GA the rate of any CP was 71.4 per 1000 livebirths (95%CI 42 to 112 per 1000 live births) at 4 to 8 years age (Himmelmann 2014).

Low quality evidence from one study (n=46) showed that among children born at <28 weeks GA the rate of any CP was 72.3 per 1000 live births (95%CI 39 to 120.3 per 1000 live births) at age 4–7 years (Nordmark 2001).

Moderate quality evidence from one study (n=975) showed that among children born at <28 weeks GA the rate of any CP in 1992–1994 was 131 per 1000 live births (95% CI 90–183/1000 live births) at age 2 years (confirmed at 3 years age) (Robertson 2007). In the same study, the rate of any CP decreased with the time points (years). From 1995–1997 and 1998–2000, the rate was 69 per 1000 live births (95%CI 41 to 108 per 1000 live births). From 2001–2003 the rate was 19 per 1000 live births (95%CI 6 to 44 per 1000 live births). Over the whole 11 years of the study, the rate was 70 per 1000 live births (95%CI 55 to 88 per 1000 live births) at 2 years age (Robertson 2007).

Severe cerebral palsy

Moderate quality evidence from one study (n=975) showed that among children born at <28 weeks GA the rate of non-ambulatory CP in 1992–1994 was 59 per 1000 live births (95% CI 32–99 per 1000 live births) at age 2 years (confirmed at 3 years age) (Robertson 2007). In the same study, the rate of any CP decreased with the time points in years. From 1995–1997 the rate was 16 per 1000 livebirths (95%CI 5–41 per 1000 livebirths) and from 1998–2000, the rate was 8 per 1000 live births (95%CI 1 to 29 per 1000 live births). From 2001–2003 the rate was 8 per 1000 live births (95%CI 1 to 27 per 1000 live births). Over the whole 11 years of the study, the rate was 22 per 1000 live births (95%CI 13 to 33 per 1000 live births) at 2 years age (Robertson 2007).

Any cerebral palsy

Low quality evidence from one study (n=2858) showed that among children born at 28–32 weeks GA the rate of any CP was 56.3 per 1000 neonatal survivors (95%CI 33 to 90) (Drummond 2002).

Moderate quality evidence from one study (n=94,466 live births) showed that among children born at 28–32 weeks GA the rate of any CP was 39.6 per 1000 livebirths (95%CI 25 to 59 per 1000 live births) at 4 to 8 years age (Himmelmann 2014).

Low quality evidence from one study (n=46) showed that among children born at 28–31 weeks GA the rate of any CP was 32.2 per 1000 live births (95%CI 18.1 to 52.2 per 1000 live births) at age 4–7 years (Nordmark 2001).

Any cerebral palsy

Low quality evidence from one study (n=189) showed that among children (1991–1996 cohort in Norway) born at 33–36 weeks GA the rate of any CP was 13.8 per 1000 livebirths at earliest age of 4 years (Andersen 2011). In the same study the prevalence of any CP among children (1991–1998 cohort in Italy) was 8.8 per 1000 livebirths whereas in cohorts from Spain and Ireland the rate was 4 per 1000 livebirths (Andersen 2011).

Low quality evidence from one study (n=2858) showed that among children born at 32–36 weeks GA the rate of any CP was 9.6 per 1000 survivors (95%CI 6 to 14) (Drummond 2002).

Moderate quality evidence from one study (n=94,466 live births) showed that among children born at 32–36 weeks GA the rate of any CP was 6.4 per 1000 livebirths (95%CI 4 to 9 per 1000 live births) at 4 to 8 years age (Himmelmann 2014). For children born at <37 weeks GA, the rate of any CP was 13 per 1000 live births (95%CI 10 to 16 per 1000 live births).

Low quality evidence from one study (n=46) showed that among children born at 32–36 weeks GA the rate of any CP was 4.6 per 1000 live births (95%CI 2.7 to 7.3 per 1000 live births) at age 4–7 years (Nordmark 2001).

Diplegia or tetraplegia

Moderate quality evidence from one study (n=94,466 live births) showed that among children born at <37 weeks GA the rate of bilateral spastic CP was 7.5 per 1000 livebirths (95%CI 5 to 10 per 1000 live births) at 4 to 8 years age (Himmelmann 2014).

Moderate intellectual disability

Moderate to low quality from 4 studies (sample size ranging from 165 to 576) showed that among children born at a range of 23 to 27 weeks GA the prevalence of intellectual disability (BSIDIII −2SD to −3SD) ranged from 6.4 (95%CI 4.6 to 8.8) to 24% (95%CI 20 to 29) (Doyle 2011; Moore 2012; Anon 1997; Serenius 2013). One further low quality study (n=77) used the Dutch version of BSIDII, which showed that the prevalence of intellectual disability was 10.4% (95%CI 4.6 to 19.5) (MDI 55–69) (De Groote 2007).

Moderate quality evidence from two studies (sample size ranging from 75 to 1508) showed that among children born at 24–27 weeks GA or <27 weeks GA the prevalence of intellectual disability (K-ABC 55–69) was 14.9% (95%CI 10.5 to 20.2) and 10.7% (95%CI 4.7 to 19.9) at 5 years and 7–9 years respectively (Foix-Helias 2008; Stahlmann 2009).

Moderate quality from one study (n=241) showed that among children born at <26 weeks GA the prevalence of intellectual disability (IQ −2 to −3SD on K-ABC, GMDS or NEPSY) was 19.9% (95%CI 15.1 to 25.5) at 6 years (Marlow 2005).

Moderate quality evidence from one study (n=306) showed that among children born at 22–27 weeks GA the prevalence of intellectual disability (Full scale IQ WPPSI-R 55–70) was 4.9% (95%CI 2.8 to 8) at 5 years (Leversen 2011). Low quality evidence from one study (n=141) showed that the prevalence (WISC-IV −2SD to −3SD) was 8.5% (95%CI 4.4 to 14.1) in children born in the same gestational age range but assessed at 8 years (Roberts 2010).

Moderate to severe intellectual disability

Moderate to low quality evidence from 5 studies (sample size ranging from 19 to 1508) showed that among children born at GA range 24 to 28 weeks GA the prevalence of intellectual disability (MPC <70 or IQ <70 K-ABC) ranged from 17.6% (95%CI 12.8 to 23.2) to 41% (95%CI 18 to 67) at a range of 5–9 years (Beaino 2011; Foix-Helias 2008; Larroque 2008; Rieger-Fackeldey 2010; Stahlmann 2009).

Moderate to low quality evidence from 5 studies (sample size ranging from 77 to 3785) showed that among children born at a GA range 22–27 weeks GA the prevalence of intellectual disability (BSID <−2SD or MDI <70) ranged from 15.2%(95%CI 10.1 to 21.6) to 39% (95%CI 37 to 41) at 18–36 months (Doyle 2011; Moore 2012; Anon 1997; Vohr 2005; De Groote 2007).

Moderate to low quality evidence from two studies (sample size ranging from 203 to 1455) showed that among children born at 22–27 or <27 weeks GA the prevalence of intellectual disability (WPPSI-R IQ <70) was 11.8% (95%CI 6.2 to 19.7) and 5.6% (95%CI 3.3 to 8.8) respectively at 5 years (Mikkola 2005; Leversen 2011).

Low quality from one study (n=141) showed that among children born at 22–27 weeks GA the prevalence of intellectual disability (WISC-IV IQ <−2SD) was 14.6% (95%CI 9.3 to 21.4) at 8 years corrected age (Roberts 2010). Low quality evidence from one other study showed that the prevalence (using WISC-III <70) in 275 children born at <28 weeks GA was 5.1% (95%CI 2.8 to 8.4) (Anderson 2003).

Moderate quality evidence from one study (n=241) showed that among children born at <26 weeks the prevalence of intellectual disability (IQ <−2SD [K-ABC, GMDS or NEPSY]) was 40.7% (95%CI 34.4 to 47.2) at 6 years (Marlow 2005).

Low quality evidence from one study (n=244) showed that among children born at <27 weeks GA the prevalence of intellectual disability (Griffiths <2SD) was 10.4% (95%CI 5.8 to 16.8) at 12 months corrected age (Sutton 1999).

Low quality evidence from one study (n=1506) showed that among children born at <28 weeks GA the prevalence of intellectual disability (verbal, DAS II <=2SD) was 17% (95%CI 14.5 to 19.5) and 15% (95%CI 12.7 to 17.6) for non-verbal reasoning (DAS II <=2SD) at 10 years (Joseph 2016b).

Severe intellectual disability

Moderate quality evidence from two studies (sample size ranging from 75 to 1508) showed that among children born at <27 weeks or 24–27 weeks GA the prevalence of intellectual disability (IQ <55, K-ABC) was 14.7% (95% CI 7.6 to 24.7) and 2.7% (95%CI 1 to 5.8) at 5–9 years (Stahlmann 2009; Foix-Helias 2008).

Moderate to low quality evidence from 5 studies (sample size ranging from 77 to 576) showed that among children born at GA range 23 to 27 weeks the prevalence of intellectual disability (BSIDIII <−3SD or MDI <55) ranged from 3.6% (95%CI 1.4 to 7.8) to 18.2% (95%CI 10.3 to 28.6) across the studies (Moore 2012; Anon 1997; De Groote 2007; Serenius 2013; Doyle 2011).

Low quality evidence from one study (n=141) showed that among children born at 22–27 weeks GA the prevalence of intellectual disability (IQ <−3SD, WISC-IV) was 6.3% (95%CI 2.9 to 11.5) at 8 years corrected age (Roberts 2010).

Moderate quality evidence from one study (n=306) showed that among children born at 22–27 weeks GA the prevalence of intellectual disability (IQ <55, WPPSI-R) was 2.9% (95%CI 1.4 to 5.5) at 5 years (Leversen 2011).

Moderate quality evidence from one study (n=241) showed that among children born at <26 weeks GA the prevalence of intellectual disability (IQ <−3SD, K-ABC, GMDS or NEPSY) was 20.8% (95%CI 15.8 to 26.4) at 6 years (Marlow 2005).

Moderate quality evidence from one study (n=142) showed that among children born at a mean GA of 27 weeks, the prevalence of intellectual disability (IQ <71 WPPSI) was 4.2% (95%CI 1.6 to 9.0%) At 4 years (Salakorpi 2001).

Moderate intellectual disability

Moderate quality evidence from one study (n=1508) showed that among children born at 28–32 weeks GA the prevalence of intellectual disability (MPC 55–69) was 8.7% (95CI 7.2 to 10.4) at 5 years (Foix-Helias 2008). In the same study, the prevalence in children born at 24–32 weeks GA was 9.6% (95%CI 8.2 to 11.2).

Moderate to severe intellectual disability

Moderate to low quality evidence from 4 studies (sample size ranging from 1455 to 1812) showed that among children born at a gestational age range of 28–32 weeks the prevalence of intellectual disability (MPC <70, K-ABC) was similar across the studies (range 8.9% (95%CI 7.3 to 10.7) to 12.1% (95%CI 10 to 14.4)) at 5 years (Beaino 2011; Marret 2007; Foix-Helias 2008; Larroque 2008).

A number of studies reported intellectual disability in children born at <32 weeks of gestation. One study of moderate quality in 3785 children born at 22–32 weeks GA found that the prevalence for intellectual disability (MDI <70, BSIDII) was 33.8% (95%CI 32.3 to 35.4) at 18–22 months corrected age (Vohr 2005).

Low quality evidence from two studies (sample size ranging from 203 to 259) showed that among children at 23–32 weeks or mean GA 27.3 (2.1) the prevalence of intellectual disability (IQ<70, WISC-IV or DAS, or IQ<70, WPPSI-R) was 15.8% (95%CI 11.6 to 20.9) and 9.4% (95%CI 5.7 to 14.2) respectively at 5 years (Andrews 2008; Mikkola 2005).

Moderate quality evidence from two studies (sample size ranging from 1508 to 1812) showed that among children born at 24–32 weeks and <33 weeks GA the prevalence was the same (11.9% (95%CI 10.3 to 13.7)) at 5 years (Foix-Helias 2008; Larroque 2008).

Moderate quality evidence from one study (n=402) showed that among children born at <32 weeks GA/<1500g the prevalence of intellectual disability (IQ<−2SD, revised Amsterdam Child Intelligence Test) was 6.2% (95%CI 4.1 to 9) at 5 years (de Kleine 2003).

Moderate quality evidence from one study (n=3785) showed that among children born at 27–32 weeks GA the prevalence of intellectual disability (MDI <70, BSIDII) was 25.9% (95%CI 23.7 to 28.2) at 18–22 months corrected age (Vohr 2005). Another study reported a prevalence of 17% (95%CI 11 to 24) at <32 weeks GA (Cognitive delay, <2SD BSID) (Toome 2012).

Low quality evidence from one study (n=347) showed that among children born at <33 weeks GA the prevalence of intellectual disability (DQ <70, Brunet-Lezine) was 2.3% (95%CI 1 to 4.5) at 2 years (corrected age) (Charkaluk 2010).

Severe intellectual disability

Moderate quality from one study (n=1508) showed that among children born at 28–32 weeks GA the prevalence of intellectual disability (MPC <55) was 2.3% (95%CI 1.5 to 3.2) at 5 years (Foix-Helias 2008). In the same study, the prevalence in children born at 24–32 weeks GA was 2.3% (95%CI 1.6 to 3.2).

Moderate to severe intellectual disability

Low quality evidence from one study (n=646) showed that among children born at 32–34 weeks GA the prevalence of intellectual disability (MPC<70) was 7.6% (95%CI 5.7 to 9.9) at 5 years (Marret 2007).

Moderate intellectual disability

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of moderate intellectual disability (BSIDII −2 to −3 SD) was 13.2% (95%CI 4.4 to 28.1%) compared to a prevalence of 4.4% (95%CI 2.2 to 7.7%) in children born at 26 weeks GA, assessed at 3 years age (Moore 2012).

Moderate quality evidence from one study (n=306) showed that among children born at 23–25 weeks GA the prevalence of moderate intellectual disability (full scale IQ 55–70, WPPSI R) was 6.9% (95%CI 2.6 to 14.4%) compared to a prevalence of 2.6% (95%CI 0.7 to 6.6%) in children born at 26–27 weeks GA, assessed at 5 years age (Leversen 2011).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks GA the prevalence of intellectual disability (IQ −2 to −3 SD, KABC GMDS or NEPSY) was 33.3% (95%CI 15.6 to 55.3%) compared to a prevalence of 18.8% (95%CI 12.7 to 26.1%) in children born at 25 weeks GA, assessed at 6 years age (Marlow 2005).

Moderate to severe intellectual disability

Low quality evidence from one study (n=244) showed that among children born at 23 weeks GA the prevalence of moderate to severe intellectual disability (major developmental delay, Griffiths <2SD) was 100% (95%CI 25 to 100%) compared to a prevalence of 3.9% (95%CI 0.81 to 11%) in children born at 26 weeks GA, assessed at 12 months corrected age (Sutton 1999).

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of moderate to severe intellectual disability (cognitive impairment BSIDIII ≤−2SD) was 31.6% (17.5 to 48.7%) compared to a prevalence of 12.0% (95%CI 8.2 to 16.6%) in children born at 26 weeks GA, assessed at 3 years (Moore 2012).

Low quality evidence from one study (n=1503) showed that among children born at 24–26 weeks GA the prevalence of moderate to severe intellectual disability (MPC<70, KABC) was 15.7% (95%CI 9.2 to 24.2) compared to a prevalence of 8.9% (95%CI 6.2 to 12.0%) in children born at 31–32 weeks GA, assessed at 5 years (Beaino 2011).

Moderate quality evidence from one study (n=306) showed that among children born at 23–25 weeks GA the prevalence of moderate to severe intellectual disability (full scale IQ <70, WPPSI-R) was 9.2% (95%CI 4.1 to 17.3%) compared to a prevalence of 2.6% (95%CI 0.7 to 6.6%) in children born at 26–27 weeks GA, assessed at 5 years (Leversen 2011).

Moderate quality evidence from one study (n=1534) showed that among children born at 24–25 weeks GA the prevalence of moderate to severe intellectual disability (MPC <70, KABC) was 12.5% (95%CI 4.7 to 25.3%) compared to a prevalence of 10.7% (95%CI 7.5 to 14.6%) in children born at 32 weeks GA. However, the prevalence was higher in children born at 26 weeks GA (prevalence 21.1% (95%CI 11.4 to 33.9%), 27 weeks (prevalence 18.6% (95%CI 12.1 to 26.9%), and 28 weeks GA (prevalence 20.7% (95%CI 14.5 to 28%) (Larroque 2008).

Low quality evidence from one study (n=1455) showed that among children born at 30 weeks GA the prevalence of moderate to severe intellectual disability (MPC <70, KABC) was 9.9% (95%CI 6.5 to 14.3%) compared to a prevalence of 5.3% (95%CI 2.0 to 11.2%) in children born at 34 weeks GA, assessed at 5 years (Marret 2007).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks GA the prevalence of moderate to severe intellectual disability (IQ≤-=2SD, KABC GMDS or NEPSY) was 58.3% (95%CI 36.6 to 77.9%) compared to a prevalence of 35.4% (95%CI 27.6 to 43.8%) in children born at 25 weeks GA, assessed at 5 years (Marlow 2005).

Severe intellectual disability

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of severe intellectual disability (cognitive impairment, BSIDIII <−3SD) was 18.4% (95%CI 7.7 to 34.3%) compared to a prevalence of 7.6% (95%CI 4.6 to 11.6%) in children born at 26 weeks GA, assessed at 3 years age (Moore 2012).

Moderate quality evidence from one study (n=306) showed that among children born at 23–25 weeks GA the prevalence of severe intellectual disability (full scale IQ <55, WPPSI-R) was 4.6% (95%CI 1.3 to 11.4%) in children born at 26–27 weeks GA, assessed at 5 years age (Leversen 2011).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks GA the prevalence of severe intellectual disability (IQ <−3SD, KABC, GMDS or NEPSY) was 25.0% (95%CI 9.8 to 46.7%) compared to a prevalence of 16.7% (95%CI 11 to 23.8%) in children born at 25 weeks GA, assessed at 6 years (Marlow 2005).

Moderate and severe speech and/or language disorder

Moderate quality evidence from one study (n=456) showed that among children born at <27 weeks GA the prevalence of moderate language impairment (−2 to −3SD BSIDIII) was 9.4% (95%CI 6.7 to 12.7) (Serenius 2013).

Low quality evidence from one study (n=576) showed that among children born at <27 weeks GA the prevalence of moderate communication impairment (−2SD to −3SD BSIDIII) was 5.4% (95%CI 3.7 to 7.6) at 3 years age (Moore 2012). In the same study, there was a prevalence of 11.6% (95%CI 9.1 to 14.5) in children with moderate to severe impairment (<=2SD BSIDIII).

Low quality evidence from one study (n=283) showed that among children born at 22–25 weeks GA the prevalence of severe speech/communication impairment ranged from 1.10% to 5.3% depending on whether they could communicate by a systemised method or not at 30 months (median) (Wood 2000).

Low quality evidence from one study (n=241) showed that among children born at <=25+6 weeks GA the prevalence for total severe impairment (PLS <2SD) was 15.6% (95%CI 10.8 to 21.4) at a median age of 6 years (Wolke 2008). However, the prevalence of severe communication impairment and severe language impairment in children (sample size ranging from 456 to 576) born at <27 weeks was lower in two studies of moderate to low quality (6.30% (95%CI 4.4 to 8.6) and 6.60% (95%CI 4.4 to 9.5) respectively) at the age of 2.5 to 3 years age (Serenius 2013; Moore 2012).

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of moderate communication impairment (−2 to −3 SD BSID III) was 10.5% (95%CI 2.3 to 24.8) compared to 4.4% (95%CI 2.2 to 7.7) at 26 weeks GA (at the age of 3 years). A similar trend was observed when severe communication impairment was assessed (<−3SD BSIDIII), with prevalence increasing with decreasing gestational age by week. At 22–23 weeks GA, the prevalence was 15.8% (95%CI 6 to 31.3) (Moore 2012) compared to the prevalence at 26 weeks GA, which was 4% (95%CI 1.9 to 7.2) (Moore 2012).

For moderate to severe impairment, there was a similar trend, prevalence in the 22–23 GA group was 26.5% (95%CI 13.4 to 43.1) compared to 8.4% (95% CI 5.3 to 12.5) in the 26 weeks GA group (Moore 2012).

Moderate speech and language disorder

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of moderate speech/language disability (communication impairment, BSIDII −2 to −3 SD) was 10.5% (95%CI 2.9 to 24.8%) compared to a prevalence of 4.4% (95%CI 2.2 to 7.7%) in children born at 26 weeks GA, assessed at 3 years (Moore 2012).

Moderate to severe speech and language disorder

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of moderate to severe speech/language disability (communication impairment, BSIDII ≤−2 SD) was 26.3% (95%CI 13.4 to 43.1%) compared to a prevalence of 8.4% (95%CI 5.3 to 12.5%) in children born at 26 weeks GA, assessed at 3 years (Moore 2012).

Severe speech and language disorder

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of severe speech/language disability (communication impairment, BSIDII <−3 SD) was 15.8% (95%CI 6.0 to 31.3%) compared to a prevalence of 4.0% (95%CI 1.9 to 7.2%) in children born at 26 weeks GA, assessed at 3 years (Moore 2012).

Moderate vision impairment

Moderate quality evidence from one study (n=241) showed that among children born at <26 weeks GA the prevalence of visual impairment (impaired but not blind) was 4.6% (95%CI 2.3 to 8) at 6 years age (Marlow 2005).

Low quality evidence from one study (n=576) showed that among children born at <27 weeks GA the prevalence of visual impairment (functionally impaired vision) was 5.9% (95%CI 4.1 to 8;2) at 3 years age (Moore 2012).

Moderate quality evidence from one study (n=456) showed that among children born at <27 weeks GA the prevalence of visual impairment (moderate impairment) was 2.9% (95% CI 1.5 to 4.8) at 2.5 years corrected age (Serenius 2013).

Moderate to severe vision impairment

Moderate quality evidence from one study (n=3785) showed that among children born at 22–26 weeks GA the prevalence of unilateral blindness was 2.7% (95%CI 2 to 3.4) at 18–22 months corrected age (Vohr 2005).

Moderate quality evidence from one study (n=242) showed that among children born at <28 weeks GA the prevalence of moderate to severe visual deficiency (<3/10, one or both eyes) was 7% (95%CI 4.1 to 11) at 5 years age (Larroque 2008).

Moderate quality evidence from one study (n=241) showed that among children born at <26 weeks GA the prevalence of visual impairment (impaired or blind) was 7.1% (95%CI 4.2 to 11.1) at 6 years age (Marlow 2005).

Low quality evidence from one study (n=576) showed that among children born at <27 weeks GA the prevalence of impaired vision (blind or functionally impaired) was 6.9% (95%CI 5 to 9.3) at 3 years (Moore 2012).

Low quality evidence from one study (n=77) showed that among children born at <27 weeks GA the prevalence of visual impairment (little useful vision) was 9.1% (95%CI 3.7 to 17.8) at 3 years age (De Groote 2007).

Low quality evidence from one study (n=88) showed that among children born at <28 weeks the prevalence of severe visual impairment (uni- or bilateral blindness or visual acuity <20/200 without glasses in at least one eye) was 12.5% (95%CI 6.4 to 21.3) at 11 years (Farooqi 2011).

Moderate quality evidence from two studies (n=306) showed that among children born at either 22–27 weeks GA or 23–25 weeks the prevalence for severe visual impairment was 0.3% (95%CI 0.01 to 1.8) and 1.2% (95%CI 0.03 to 6.2) respectively at 5 years (Leversen 2011).

Low quality evidence from one study (n=283) showed that among children born at 22–25 weeks GA the prevalence of severe visual impairment (blind or perceives light) was 2.5% (95%CI 1 to 5) at 30 months (median) (Wood 2000).

Moderate quality evidence from one study (n=411) showed that among children born at <27 weeks GA the prevalence of visual impairment (blind or able to only fixate and follow light binocularly) was 3.1% (95%CI 1.6 to 5.3) at 30 months corrected age (Holmstrom 2014).

Low quality evidence from one study (n=77) showed that among children born at <27 weeks GA the prevalence of visual impairment (no useful vision) was 2.6% (95%CI 0.9 to 9.1) at 3 years age (De Groote 2007).

Low quality evidence from two studies (sample size ranging from 189 to 219) showed that among children born at 23–27 weeks GA and 22–27 weeks GA the prevalence for blindness (<6/60 in both eyes) was 2.3% (95%CI 0.8 to 5.3) and 1.6% (95%CI 0.3 to 4.6) at 2 years and 8 years (corrected) respectively (Anon 1997; Anderson 2011).

Moderate to low quality evidence from three separate studies (sample size ranging from 306 to 373) showed that among children born at 22–27 weeks GA and also 23–25 weeks GA the prevalence for blindness was varied, ranging from 5.8% (95%CI 1.9 to 12.9) in the lower GA group (Leversen 2011), and 1.6% (95%CI ranged from 0.5 to 3.8) in the two 22–27 GA groups (Leversen 2010; Leversen 2011).

Moderate to very low quality evidence from 8 studies (sample size ranging from 19 to 3785) showed that among children born at various gestational ages (ranging from <26 weeks to <28 weeks) the prevalence of blindness was varied, ranging from 0.9% (95%CI 0.24 to 2.3) to 11% (95%CI 1.3 to 33) (Vohr 2005; Roberts 2010; Marlow 2005; Moore 2012; Hutchinson 2013; Serenius 2013; Anderson 2003; Rieger-Fackeldey 2010).

Low quality evidence from one study (n=1506) showed that among children born at <28 weeks GA the prevalence of severe visual impairment (functional blindness) was 0.8% (95%CI 0.3 to 1.7) at 10 years (Joseph 2016b).

Moderate to severe vision impairment

Low quality evidence from one study (n=1455) showed that among children born at 30–31 weeks GA the prevalence of visual impairment (visual acuity <3/10 in both eyes) was 1.5% (95%CI 0.7 to 2.8) at 5 years (Marret 2007).

Moderate quality evidence from one study (n=3785) showed that among children born at 27–32 weeks GA found that the prevalence of visual impairment (unilateral blindness) was 1.3% (95%CI 0.8 to 2) at 18–22 months corrected age (Vohr 2005).

Moderate quality evidence from one study (n=971) showed that among children born at 28–31 weeks GA the prevalence of moderate to severe visual deficiency (<3/10 in one or both eyes) was 2.1% (95%CI 1.3 to 3.2) at 5 years age (Larroque 2008).

Severe vision impairment

Moderate quality evidence from on study (n=3785) showed that among children born at 27–32 weeks GA the prevalence of visual impairment (bilateral blindness) was 0.7% (95%CI 0.3 to 1.2) at 18–22 months corrected age (Vohr 2005). In the same study, the prevalence of bilateral blindness in children born at 22–32 weeks GA was 1.2% (95%CI 0.9 to 1.6) (Vohr 2005).

Low quality evidence from one study (n=93) showed that among children born at <32 weeks GA the prevalence of visual impairment (best eye blind or only able to fixate a torch) was 1.1% (95%CI 0.03 to 5.9) at 2.5 years corrected age (Hreinsdottir 2013).

Moderate to severe vision impairment

Low quality evidence from on study (n=1455) showed that among children born at 32–24 weeks GA the prevalence of visual impairment (visual acuity <3/10 in both eyes) was 1.7% (95%CI 0.9 to 3) at 5 years age (Marret 2007).

Moderate vision impairment

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of moderate visual impairment (functionally impaired vision) was 15.8% (95%CI 6.0 to 31.3%) compared to a prevalence of 3.2% (95%CI 1.4 to 6.2%) in children born at 26 weeks GA, assessed at 3 years (Moore 2012).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks GA the prevalence of moderate visual impairment (visually impaired, not blind) was 8.3% (95%CI 1.0 to 27.0%) compared to a prevalence of 2.8% (95%CI 0.8 to 7.0%) in children born at 25 weeks GA, assessed at 6 years age (Marlow 2005).

Moderate quality evidence from one study (n=494) showed that among children born at 22–23 weeks GA the prevalence of visual impairment (any; best estimated visual acuity <20/40) was 23.8% (95%CI 12 to 40) compared to a prevalence of 13.4% (95%CI 6.9 to 22.7) at 24 weeks GA, prevalence of 7% (95%CI 3.4 to 12.6) at 25 weeks GA, and a prevalence of 5.1% (95%CI 2.1-1-.2) at 26 weeks GA (Hellgren 2016).

Moderate to severe vision impairment

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of moderate to severe visual impairment (functionally impaired vision) was 18.4% (95%CI 7.7 to 34.3%) compared to a prevalence of 4.4% (95%CI 2.2 to 7.7%) in children born at 26 weeks GA, assessed at 3 years (Moore 2012).

Low quality evidence from on study (n=1455) showed that among children born at 30 weeks GA the prevalence of moderate to severe visual impairment (visual acuity <3/10 in both eyes) was 0.7% (95%CI 0.1 to 2.6) compared to a prevalence of 0.8% (95%CI 0.02 to 4.1%) in children born at 34 weeks GA. The prevalence was higher at GA 31 weeks (2.2% (95%CI 0.8 to 4.3%), and 33 weeks GA (2.3% (95%CI 0.5 to 6.5%), assessed at 5 years age (Marret 2007).

Moderate quality evidence from one study (n=1817) showed that among children born at 24–25 weeks GA the prevalence of moderate to severe visual impairment (<3/10 one or both eyes) was 9.3% (95%CI 3.1 to 20.3%) compared to a prevalence of 1.9% (95%CI 0.9 to 3.5%) in children born at 32 weeks GA, assessed at 5 years age (Larroque 2008).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks GA the prevalence of moderate to severe visual impairment (visually impaired, or blind) was 16.7% (95%CI 4.7 to 37.4%) compared to a prevalence of 3.5% (95%CI 1.1 to 7.9%) in children born at 25 weeks GA, assessed at 6 years age (Marlow 2005).

Severe vision impairment

Moderate quality evidence from one study (n=411) showed that among children born at 22–23 weeks GA the prevalence of severe visual impairment (blind or able to only fixate and follow light binocularly) was 4.8% (95%CI 0.6 to 16.2%) compared to a prevalence of 1.4% (95%CI 0.2 to 4.8%) in children born at 26 weeks GA, assessed at 30 months corrected age (Holmstrom 2014).

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of visual impairment (blindness) was 2.6% (95%CI 0.1 to 13.8%) compared to a prevalence of 1.2% (95%CI 0.3 to 3.5%) in children born at 26 weeks GA, assessed at 3 years (Moore 2012).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks GA the prevalence of severe visual impairment (blindness) was 8.3% (95%CI 1.0 to 27.0%) compared to a prevalence of 0.7% (95%CI 0.02 to 3.8%) in children born at 25 weeks GA assessed at 6 years age (Marlow 2005).

Moderate hearing impairment

Moderate quality evidence from one study (n=241) showed that among children born at <26 weeks GA, the prevalence of hearing loss (corrected with hearing aids) was 2.9% (95%CI 1.2 to 5.9) when assessed at 6 years age (Marlow 2005).

Low quality evidence from one study (n=576) showed that among children born at <27 weeks GA the prevalence of hearing loss (improved by aids) was 5.2% (95%CI 3.5 to 7.4) when assessed at 3 years age (Moore 2012).

Low quality evidence from one study (n=77) showed that among children born at < 27 weeks GA the prevalence of hearing impairment (but useful hearing) was 3.9% (95%CI 0.8 to 11) (De Groote 2007).

Moderate to severe hearing impairment

Low quality evidence from one study (n=141) showed that among children born at 22–27 weeks GA the prevalence of hearing impairment was 2.1% (95%CI 0.4 to 6) at 8 years corrected age (Roberts 2010).

Moderate quality evidence from one study (n=241) showed that among children born at <26 weeks the prevalence of moderate to severe hearing impairment was 5.8% (95%CI 3.2 to 9.6) at 6 years (Marlow 2005). In another study of low quality with 576 children born at <27 weeks GA the prevalence for severe hearing impairment was 5.4% (95%CI 3.7 to 7.6) at 3 years (Moore 2012).

Low quality evidence from one study (n=19) showed that among children born at mean 25.4 weeks GA the prevalence of hearing impairment (requiring hearing aid) was 11% (95%CI 1.3 to 33) at 5 years age (Rieger-Fackeldey 2010). Ten other studies (sample size ranging from 77 to 3785) of moderate to very low quality assessing hearing impairment or deafness (requiring hearing aids) in children born at a range of 22–28 weeks GA found that the prevalence was lower but varied, ranging from 0.7% (95%CI 0.14 to 2) to 5.7% (95%CI 1.9 to 12.8) (Farooqi 2011; Leversen 2011; Vohr 2005; Doyle 2011; Anderson 2011; De Groote 2007; Hutchinson 2013; Wood 2000; Serenius 2013; Anderson 2003).

Severe hearing impairment

Low quality evidence from one study (n=283) showed that among children born at 22–25 weeks GA the prevalence of severe hearing impairment (uncorrected without hearing aid) was 5.3% (95%CI 3.0 to 8.6) at 30 months (median) (Wood 2000).

Low quality evidence from one study (n=373) showed that among children born at 22–27 weeks GA the prevalence of deafness was 0.8% (95%CI 0.1 to 2.7) at 2 years (corrected age) (Leversen 2010). In another study (n=401) of low quality, the prevalence of deafness was 0.9% (95%CI 0.1 to 3.3) in children assessed at 2 years (Anon 1997). Prevalence of deafness was 0.2% (95%CI 0.01 to 1.2) in children (n=456) born at <27 weeks GA (moderate quality, Serenius 2013). At 5 years age, the prevalence of deafness was 1.0% (95%CI 0.2 to 2.8) in children (n=306) born at 22–27 weeks GA (moderate quality study, Leversen 2011).

Low quality evidence from one study (n=261) showed that among children born at <28 weeks GA the prevalence of severe hearing deficiency (>70 decibels in one or both ears or hearing aid) was 0.8% (95%CI 0.1 to 2.7) at 5 years age (Larroque 2008).

Low quality evidence from one study (n=576) showed that among children born at <27 weeks GA the prevalence of profound sensorineural hearing loss (not improved by aids) was 0.2% (95%CI 0.1 to 1) at 3 years age (Moore 2012). In another moderate quality study (n=241) children born at <26 weeks GA found that the prevalence of profound sensorineural hearing loss was 2.9% (95%CI 1.2 to 5.9) at 6 years age (Marlow 2005).

Moderate to severe hearing impairment

Moderate quality evidence from one study (n=3785) showed that among children born at 27–32 weeks GA the prevalence of permanent hearing loss (amplification in both ears) was 1.4% (95%CI 0.9 to 2.1) at 18–22 months corrected age (Vohr 2005).

Low quality evidence from one study (n=1455) showed that among children born at 30–31 weeks GA the prevalence for hearing loss >70 decibels was 0.30% (95%CI 0.04 to 1.1) at 5 years (Marret 2007).

Severe hearing impairment

Moderate quality evidence from one study (n=1020) showed that among children born at 28–31 weeks GA the prevalence for severe hearing deficiency (>70 decibels in one or both ears or hearing loss) was 0.5% (95%CI 0.2 to 1.1) at 5 years age (Larroque 2008).

Moderate hearing impairment

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of moderate hearing impairment (hearing loss improved by aids) was 5.3% (95%CI 0.6 to 17.8%) compared to a prevalence of 5.2% (95%CI 2.8 to 8.7%) in children born at 26 weeks GA, assessed at 3 years (Moore 2012).

Moderate quality evidence from one study (n=241) showed that among children born at 24 weeks GA the prevalence of moderate hearing impairment was 2.7 (95%CI 0.3 to 9.6%) compared to a prevalence of 3.5% (95%CI 1.1 to 7.9%) in children born at 25 weeks GA, assessed at 6 years (Marlow 2005).

Moderate to severe hearing impairment

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of moderate hearing impairment (hearing loss improved by aids) was 7.9% (95%CI 1.7 to 21%) compared to a prevalence of 5.2% (95%CI 2.8 to 8.7%) in children born at 26 weeks GA, assessed at 3 years (Moore 2012).

Moderate quality evidence from one study (n=306) showed that among children born at 23–25 weeks GA the prevalence of moderate to severe hearing impairment (hearing aid in both ears) was 2.3% (0.3 to 8.1%) compared to a prevalence of 1.3% (95%CI 0.2 to 4.7%) in children born at 26–27 weeks GA, assessed at 5 years (Leversen 2011).

Low quality evidence from one study (n=1455) showed that among children born at 30 weeks GA the prevalence of moderate to severe hearing impairment (hearing loss >70 decibels or aids in one or both ears) was 0.3% (95%CI 0.01 to 1.9%) compared to a prevalence of 1.5% (95%CI 0.2 to 5.3%) in children born at 34 weeks GA, assessed at 5 years (Marret 2007).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks GA the prevalence of moderate to severe hearing impairment was 4.2% (95%CI 0.1 to 21.1%) compared to a prevalence of 4.9% (95%CI 2.0 to 9.8%) in children born at 25 weeks GA, assessed at 6 years (Marlow 2005).

Severe hearing impairment

Low quality evidence from one study (n=576) showed that among children born at 22–23 weeks GA the prevalence of severe hearing impairment (profound sensorineural hearing loss not improved by aids) was 2.6% (95%CI 0.1 to 13.8%), assessed at 3 years (Moore 2012).

Moderate quality evidence from one study (n=1817) showed that among children born at 24–25 weeks GA the prevalence of severe hearing impairment (>70 decibels in one or both ears or hearing aid) was 1.7% (95%CI 0.04 to 9.2%) compared to a prevalence of 0.2% (95%CI 0.01 to 1.1%) in children born at 32 weeks GA, assessed at 5 years (Larroque 2008).

Moderate quality evidence from one study (n=241) showed that among children born at ≤23 weeks GA the prevalence of severe hearing impairment (profound sensorineural hearing loss) was 4.2% (95%CI 0.1 to 21.1%) compared to a prevalence of 1.4% (95%CI 0.1 to 4.9%) in children born at 25 weeks GA, assessed at 6 years (Marlow 2005).