Table 2.

Overview of key findings for school-age outcomes of infants with perinatal brain injury compared with those without brain injury

	NDI	Cognitive	Motor	Speech and language	Behavioural	Hearing†	Vision†	Other
IVH grades 3–4*	6 studies15 17–21 2 comparable studies in meta-analysis17 20 Meta-analysis (2 studies): Increased risk of moderate -severe neurodevelopmental impairment OR 3.15 (95% CI 1.67 to 5.92) I2=0% Van de Bor et al:22 increased prevalence of disability 31% vs 16%	9 studies (15, 20, 21, 24–26, 30, 70) Not comparable Consistently highlighted lower cognitive scores Brouwer et al:25 significantly lower performance IQ but preserved verbal IQ. Lower IQ for those with IVH grade 4 requiring neurosurgery (91±10 vs 98±15) but little difference for those with grade 3 IVH requiring neurosurgery (96±15 vs 98±15). Hollebrandse et al:26 increased risk of cognitive impairment OR 2.68 (95% CI 1.21 to 5.94). Increased risk of academic impairment across all academic domains: reading OR 3.62 (95% CI 1.59 to 8.24); spelling OR 4.48 (95% CI 1.8 to 11.2); arithmetic OR 2.79) 95% CI 1.2 to 6.48) Sherlock et al:21 significantly lower IQ scores after IVH grade 4 vs IVH 1–3 and no brain injury, also seen for several domains: freedom from distractibility, processing speed, reading, spelling and arithmetic. No difference in executive function. Van de Bor:22 increased special education needs at 5, 9 and 14 years aOR 3.99 (95% CI 1.36 to 11.69).	6 studies20 23–26 33 Not comparable All reported increased risk of motor impairment Cerebral palsy 3 comparable studies OR 8.67 (95% CI 5.27 to 14.28) I2=0%.	3 studies20 21 25 Not comparable Van de Bor 2004:22 no significant difference in language scores Sherlock et al:21 downward trend in language scores from no brain injury to each grade of IVH but not statistically significant p=0.12 Hollebrandse et al:26 Increased risk of impaired reading OR 3.62 (95% CI 1.59, 8.24) and spelling OR 4.48 (95% CI 1.8 to 11.2)	3 studies15 24 35 Not comparable Brouwer et al:25 no association with any behavioural domains assessed (internalising, externalising and sleep problems) Adant et al:16 no increased risk of attention deficits, conduct issues or ASD aOR 1.24 (95% CI 0.32, 4.8). Davidovich et al:36 no increased risk of ASD (n=10, 3.9% vs n=103, 2.2% p=0.085)	3 studies21 26 38 Not comparable Outcome too rare for inferential analysis Kaur et al:39 increased risk of hospitalisation for otologic reasons HR 7.87 (95% CI 5.31 to 11.67)	5 studies15 21 26 33 38 Not comparable Outcome too rare for inferential analysis in most studies. Adant et al:16 no increased risk of visual impairment (needing glasses) aOR 0.47 (95% CI 0.13 to 1.69) Klebermass-Schrehof et al:27 increased prevalence of visual impairment (needing glasses or blindness) after IVH grade 3 (45.4%) and IVH grade 4 (90.9%) vs comparators (7.5%). Kaur et al:39 increased risk of hospitalisation for ophthalmic reasons HR 7.87 (95% CI 5.31 to 11.67). Klebermass-Schrehof et al:27 significantly lower VMI scores (67.5 ± 14 vs 76± 26.8; p=0.04)
WMI*	3 studies16 17 22 Not comparable Campbell et al:17 living with no impairment was less common with WMI (n=12, 40%) vs controls (n=487, 76%) Cheong 2018:18 increased risk of survival with major disability after cPVL aOR 9.17 (95% CI 3.57 to 23.53) Vollmer et al:23 Disabling impairments were more common after cPVL at <28 weeks’ gestation (n=3, 75% <28 weeks) vs controls (n=3, 8%) and at over 28 weeks’ gestation (n=6, 50% vs n=14, 6%)	4 studies (16, 29, 32, 70) Not comparable Van den Hout et al:33 50% with PVL had IQ scores <85 vs 11.8% without injury and a lower performance age 4.3 years vs 6.2 years Campbell et al:17 increased risk of moderate-to-severe cognitive impairment aOR 5.07 (95% CI 2.13 to 12.02) Jansen et al:30 WMI predictive of poorer performance on standardised mathematics tests (B 1.856 p=0.003), but not performance on spelling (B 1.076 p=0.075) or reading tests (B 0.241 p=0.483)	Cerebral palsy 1 study16 Campbell 2020: increased risk of cerebral palsy aOR 18.63 (95% CI 7.37 to 47.06)	1 study29 Jansen et al:30 No association between WMI and spelling (B 1.076 p=0.075) or reading performance (B 0.241 p=0.483)	4 studies (16, 35, 36, 71) Not comparable Conflicting results Campbell et al:17 No increased risk of: ADHD (n=3, 10% vs n=97, 15%); anxiety (n=3, 10% vs n=98, 15%); depression (n=7, 23% vs n=100, 16%); or ASD aOR 0.74 (95% CI 0.09 to 5.88) Davidovitch et al:36 No increased risk of ASD after PVL (n=5, 2.5% vs n=88, 2.3% p=0.86) Whitaker et al:37 increased risk of ADHD aOR 6.83 (95% CI 1.26 to 36.91); major depression aOR 2.59 (95% CI 1.02 to 6.58); tic disorders aOR 9.77 (95% CI 1.69 to 56.47); obsessive compulsive disorders aOR 15.32 (95% CI 1.82 to 128.74)	0 studies	1 study32
Stroke	0 studies	6 studies39 41 42 44–46 5 comparable studies in meta-analysis39 41 44–46 Meta-analysis (5 studies): significant mean difference in full scale IQ −24.2 (95% CI −30.73 to –17.67) I2=80% Trauner47 and Gold:42 no significant difference in full scale IQ scores in left vs right-sided strokes Ballantyne et al:40 significantly lower performance IQ (p=0.002) and verbal IQ (p<0.0001). Lower mean scores for reading (p<0.0001), spelling (p=0.001) and arithmetic (p<0.0001) at 7–8 years persisting to 10–12 years Tillema et al:46 reduced verbal IQ scores (mean 84 SD 13.4) vs (mean 108 SD 14.2 p=0.002) Kolk et al:43 poorer attention (across 4 of the 7 assessment sub-domains), visuo-spacial function (across 4 of the 5 subdomains) and memory and learning (across 4 of the 6 subdomains), but normal executive function scores. Those with left-sided strokes had poorer neuropsychological scores. Northam et al:45 most children are in mainstream education (n=28, 93%) but many require additional support (n=12, 40%)	5 studies39 41–44 Combined hemiparesis incidence: 61% (95% CI 39.2% to 82.9%) I2=88%) Kolk et al:43 moderate-to-severe neuromotor impairment in 62% n=13) and significantly lower scores on 5/6 sensorimotor domains of the NEPSY	5 studies39 40 42 44 45 3 comparable studies in meta-analysis Meta-analysis (3 studies): lower receptive language scores-20.88 (95% CI –36.66 to –5.11) I2=88% and lower expressive language scores −20.25 (95% CI –34.36 to –6.13) I2=87% Ballantyne et al 200741 and Ballantyne et al 200840 deficits in receptive language scores at 7–8 years persist at 10–12 years but expressive language scores improved (p=0.012) particularly for children with right-sided strokes (p=0.034) Kolk et al:43 significantly lower scores for 8/9 NEPSY domains including phonologic processing, comprehension of instructions, correct speeded naming, repetition of nonsense words, verbal fluency (semantic and phonetic), oromotor sequences and sentence comprehension	1 study46	1 study43 Martin:44 left-sided strokes predispose children to contralateral auditory neglect and right-sided strokes predispose children to bilateral auditory neglect	1 study39 Ballantyne et al:40 visual field defects are common (n=7, 26%) after perinatal stroke	Seizures 8 studies39 42 43 45 46 5 comparable studies39 42 43 45 46 Combined incidence of seizures: 40.1% (95% CI 26.8% to 53.3%) I2=56%
Meningitis	3 studies47–49 Not comparable All reported increased risk of neurodevelopmental impairment Bedford 2011: increased prevalence of neuromotor disability (n=45, 16% vs n=2, 0.1%) Stevens et al:50 Risk of severe disability seen in Bedford 2011 at 5 years of age persisted until 9–10 years (n=12, 10.8% vs n=0, 0%) Horvath-Puho et al:49 increased risk of any neurodevelopmental impairment after GBS meningitis in the Netherlands RR 5.30 (95% CI 2.57 to 10.89) and Denmark RR 7.80 (95% CI 4.42 to 13.77) at 5 years of age persisting to 11 years in the Netherlands RR 2.99 (95% CI 1.83 to 4.88) and 15 years in Denmark RR 3.15 (95% CI 1.82 to 5.46)	1 study49 Stevens 2003:50 significantly lower mean cognitive scores (mean 88.8 (95% CI 85 to 92) vs mean 99.4 (95% CI 97 to 102))	1 study49 Stevens et al:50 significantly higher motor impairment scores (mean 7.1 (95% CI 5.9 to 8.5) vs mean 5 (95% CI 4.3 to 5.8))	0 studies	0 studies	2 studies (49, 72) Martinez-Cruz 2008: increased odds of neonatal meningitis among preterm infants with sensorineural hearing loss OR 4.37 (95% CI 1.7 to 10.9 50Stevens 2003: 3.6% (n=4) had hearing loss compared with none in the control group.	1 study49 Stevens et al:50 Bilateral visual impairment was common after neonatal meningitis (n=18, 17%)
HIE	0 studies	3 studies30 50 51 (two of the same population) Not comparable Koc et al:31 preterm infants with HIE significantly more likely to have below average IQ scores (n=8, 89% vs n=24, 30% p=0.001) Lee-Kelland et al51 and Tonks et al:52 report lower full scale IQ scores after moderate to severe HIE (mean difference −13.62 (95% CI −20.53 to –6.71)) and poorer perceptual reasoning, working memory and processing speed. Children with previous HIE more likely to receive additional classroom support OR 10 (95% CI 1.16 to 86)	2 studies50 51 (of the same population) Lee-Kelland et al51 and Tonks et al:52 significantly lower motor scores (mean difference −2.12 (95% CI −3.93 to –0.30)) after moderate-severe HIE (for children without cerebral palsy)	2 studies50 51 (of the same population) Lee-Kelland et al51 and Tonks et al:52 significantly lower verbal comprehension scores (mean difference −8.8 (95% CI –14.25 to –3.34)) after moderate to severe HIE.	2 studies50 51 (of the same population) Lee-Kelland et al51 and Tonks et al:52 higher behavioural difficulty scores (median score 12 IQR (6.5, 13.5 vs median score 6 IQR (2.25, 10) p=0.005)	0 studies	0 studies
Kernicterus	0 studies

*Does not include studies where infants with IVH grades 3–4 cannot be separated from those with WMI or those with IVH 1–2.

†Does not include studies using hearing or visual outcomes only as part of their composite outcome.

ADHD, attention-deficit/hyperactivity disorder; aOR, adjusted OR; ASD, autism spectrum disorder; cPVL, cystic PVL; GBS, group B Streptococcus; HIE, hypoxic-ischaemic encephalopathy; IVH, intraventricular haemorrhage; NDI, Neurodevelopmental impairment; PVL, periventricular leukomalacia; RR, Risk ratio; VMI, visual motor integration; WMI, white matter injury.