You are here

Article Text

Article menu
Original article
Five-minute Apgar score and educational outcomes: retrospective cohort study of 751 369 children
  1. Emily J Tweed1,
  2. Daniel F Mackay1,
  3. Scott M Nelson2,
  4. Sally-Ann Cooper1,
  5. Jill P Pell1
  1. 1Institute for Health and Wellbeing, University of Glasgow, Glasgow, UK
  2. 2School of Medicine, University of Glasgow, Glasgow, UK
  1. Correspondence to Professor Jill Pell, Institute of Health and Wellbeing, University of Glasgow, 1 Lilybank Gardens, Glasgow G12 8RZ, UK; jill.pell{at}glasgow.ac.uk

Abstract

Background The Apgar score is used worldwide for assessing the clinical condition and short-term prognosis of newborn infants. Evidence for a relationship with long-term educational outcomes is conflicting. We investigated whether Apgar score at 5 min after birth was associated with additional support needs (ASN) and educational attainment.

Methods Data on pregnancy, delivery and later educational outcomes for children attending Scottish schools between 2006 and 2011 were collated by linking individual-level data from national educational and maternity databases. The relationship between Apgar score and overall ASN, type-specific ASN and educational attainment was assessed using binary, multinomial and generalised ordinal logistic regression models, respectively. Missing covariate data were imputed.

Results Of the 751 369 children eligible, 9741 (1.3%) had a low or intermediate Apgar score and 49 962 (6.6%) had ASN. Low Apgar score was independently associated with overall ASN status (adjusted OR for Apgar ≤3, OR 1.52 95% CI 1.35 to 1.70), as well as ASN due to cognitive (OR 1.26, 95% CI 1.09 to 1.47), sensory (OR 2.49 95% CI 1.66 to 3.73) and motor (OR 3.57, 95% CI 2.86 to 4.47) impairments. There was a dose-response relationship between Apgar score and overall ASN status: of those scoring 0–3, 10.1% had ASN, compared with 9.1% of those scoring 4–7 and 6.6% of those scoring 7–10. A low Apgar score was associated with lower educational attainment, but this was not robust to adjustment for confounders.

Conclusions Apgar scores are associated with long-term as well as short-term prognoses, and with educational as well as clinical outcomes at the population level.

  • Neurodevelopment
  • Neurodisability
  • Apgar score
  • Special education
  • Educational status

https://doi.org/10.1136/archdischild-2015-308483

Statistics from Altmetric.com

    Request Permissions

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

    What is already known

    • Neonatal Apgar score is known to predict short-term and long-term survival and neurological outcomes.

    • Whether a low Apgar score is associated with an overall requirement for additional support needs (ASN), different ASN subtypes or long-term educational attainment is unknown.

    What this study adds

    • Low Apgar score was strongly associated in a dose-dependent manner with need for additional educational support in later life, after adjusting for confounding factors.

    • The association was strongest for additional support needs resulting from physical and motor impairments, followed by visual, hearing and cognitive impairments.

    • Low Apgar score was also associated with lower educational attainment, but this was not robust to adjustment for confounders.

    Introduction

    Educational experience and attainment in childhood is a key determinant of health across the life course.1 In Scotland, the education sector has a statutory duty to identify and provide for children with additional support needs (ASN), defined as difficulties in learning requiring different or extra educational support compared with peers of their own age.2 ,3

    To date a number of preschool factors, including prenatal and perinatal characteristics, have been identified as being associated with subsequent ASN.

    The neonatal Apgar score, created by Dr Virginia Apgar in 1952, remains in routine clinical use for determining the need for, and effectiveness of, resuscitation, and as a marker of short-term prognosis.4–6 Low Apgar scores at birth are also consistently associated with an increased risk of subsequent neurological conditions such as cerebral palsy,7–11 epilepsy11–14 and cognitive impairment.15 ,16 However, uncertainty remains regarding the long-term relationship between Apgar score, different types of ASN and educational outcomes, particularly in children without other risk factors such as preterm delivery or low birth weight.

    We carried out a Scotland-wide retrospective cohort study linking birth and educational records to investigate whether Apgar score at 5 min of age was associated with overall, as well as type-specific, ASN and educational attainment.

    Methods

    Data sources

    We linked individual-level data on pregnancy, delivery and later educational outcomes from three national databases: the Scottish Morbidity Record 2 (SMR02), and the Scottish school census and school attainment record.

    The SMR02 collects data on antenatal factors and delivery outcomes for all women discharged from maternity hospitals in Scotland.17 Apgar score at 5 min of age was categorised into three ordinal groups; low (0–3), intermediate (4–6) and normal (7–10).18 Each child's postcode of residence was used to assign a Scottish Index of Multiple Deprivation score as an area-level proxy for socioeconomic status.19

    The school census covers all children attending local authority maintained and grant-aided primary and secondary schools in Scotland and includes mainstream and special schools. Pupil-level data, including ASN status, are submitted annually by each school to the Scottish Government Education Analytical Services division (known as ScotXEd). Identification of children with ASN is a statutory duty of schools in Scotland.2 ,3

    Educational attainment data were obtained from the Scottish Qualifications Authority (SQA), which maintains a database of all children who have been entered for a qualification and the result attained. Using the SQA's established Unified Points Scale, which summates tariff points assigned to each examination result based on the level of qualification and the grade achieved, pupils at different educational stages are stratified into the categories of low, basic, high and very high attainment. Since only a proportion of children included in the school census data had reached the stages at which summative qualifications are undertaken, analysis of educational attainment was carried out on this subgroup only.

    National Health Service (NHS) Information Services Division (ISD) Scotland used probabilistic matching to link records from SMR02 and the education sector. Pupils included in the school census (identified by a Scottish Candidate Number, a unique identifier from the education sector) were matched using date of birth, gender and postcode of residence with the Community Health Index (CHI) database to identify each pupil's CHI number, a unique identifier allocated to all patients registered with a family doctor in Scotland. This CHI number was then linked to the maternal record in the SMR02 database via statutory birth registration records containing child and maternal CHI number. Scottish candidate numbers were used to link pupils from the school census to educational attainment data from the SQA. This method has been previously described:20 a detailed description of the matching algorithm is provided in online supplementary appendix 1.

    The work was enabled by data sharing agreements between the University of Glasgow and ScotXEd, SQA and ISD. Approvals were also obtained from ISD's Privacy Advisory Committee and from the CHI Advisory Group. NHS West of Scotland Research Ethics Service confirmed that NHS ethical approval was not required for this study.

    Inclusion criteria and definitions

    School census data were collected on all children attending Scottish schools between the academic years 2006/2007 and 2011/2012 inclusive. Our analyses were limited to singleton births with birth weight 400–6500 g and estimated gestational age at delivery of 37–43 weeks to women of any parity over 10 years of age and 100–200 cm in height. Infants for whom 5-minute Apgar score was not recorded were excluded. Multiple pregnancies were excluded because it was impossible to determine birth order in the linked data set and hence attribute variables such as Apgar score to the correct sibling.

    Five-minute Apgar score was categorised into three ordinal groups; low (0–3), intermediate (4–6) and normal (7–10).18 The definition of ASN used in this analysis did not include social, cultural, linguistic or emotional factors (eg, bereavement or English as a second language) unlikely to be related to perinatal events and Apgar scores at birth. The category of ‘mental health problems’ contained too few individuals for meaningful subgroup analysis so was also excluded from the definition of ASN. Thus, the five categories of ASN included in the analysis were: cognitive (intellectual disability, other learning disability or dyslexia), visual and/or hearing impairment (visual impairment, hearing impairment or deaf/blindness), physical and/or motor disability (physical or motor impairment, or physical health problem), language or speech disorder, and autism spectrum disorder. Pupils were identified as having ASN if their school census record for any year contained a flag for ASN. Categories were mutually exclusive: children with more than one type of ASN were classified according to their main impairment.

    Statistical analyses

    All analyses were undertaken using Stata V.12.1. Groups were compared using χ2, Kruskal-Wallis and Spearman's rank correlation tests for categorical, continuous and ordinal data, respectively. Statistical significance was assumed at p<0.05. Binary, multinomial and generalised ordinal logistic regression models were used for the outcomes of overall ASN status, type-specific ASN and educational attainment, respectively. Multivariable models were used to adjust for potential confounders: sex; maternal age, height and smoking; parity; marital status; socioeconomic deprivation; presentation; mode of delivery; gestational age and sex-specific, gestation-specific birthweight centile. Population attributable fraction was calculated using the aflogit function in Stata, after adjusting for the covariates listed above. For educational attainment, we tested first for a univariate association with ASN and then constructed multivariable models containing an interaction between Apgar score and ASN or ASN as a stand-alone covariate.

    In the multivariable models, missing values for maternal height and smoking status were imputed using the ICE function in Stata, with the creation of five imputed data sets. Complete data are shown in online supplementary tables S5–S7.

    Results

    Of the 1 011 585 children included in the school censuses conducted between 2006 and 2011 inclusive, 811 860 (80.3%) could be linked to delivery data. Of these, 60 491 (7.5%) were excluded, for the following reasons: multiple pregnancy (n=8585), gestational age (n=45 957), birth weight (n=348), maternal age or height (n=164) and missing Apgar score (n=5437). The study population comprised the remaining 751 369 children. Of these, 331 394 (44.1%) had undertaken summative examinations with the SQA and could be included in analyses of educational attainment. Year of birth ranged from 1989 to 2006, with 464 182 children (61.8%) born before 2000 and 287 186 (38.2%) born during or after 2000. The mean age of children included was 12.6 years (SD=3.8). Figure 1 shows a flow chart of the study cohort.

    Figure 1
    Figure 1

    Flow chart for study cohort. SMR02, Scottish Morbidity Record 2; SQA, Scottish Qualifications Authority.

    Overall, 741 628 (98.7%) children had an Apgar score of 7–10, 6393 (0.9%) had a score of 4–6 and 3348 (0.4%) had a score of 0–3; 49 962 (6.6%) children had a record of ASN (table 1). Of these 49 962 children: 33 031 (66.1%) had cognitive impairment; 5330 (10.7%) had physical or motor impairments; 4986 (10.0%) had autistic spectrum disorder; 4401 (8.8%) had language or speech problems; and 2214 (4.4%) had visual or hearing impairments.

    View this table:
    Table 1

    Characteristics and educational outcomes of children attending Scottish schools between 2006 and 2011, by 5-min Apgar score

    Children with Apgar scores <7 were more likely to be male, breech presentation, of lower socioeconomic quintile, gestational age and sex-specific and gestation-specific birth weight, and to be born by emergency caesarean section or to multiparous mothers. They were also more likely to have ASN: 921 (9.5%) children with an Apgar <7 required ASN during their subsequent schooling in comparison to 49 041 (6.6%) of those with an Apgar of ≥7 (p<0.001). The association between low Apgar score and subsequent ASN was robust to adjustment for confounding factors and there was a dose-response relationship across the three Apgar score groups (table 2). The population attributable fraction for Apgar score <7 in relation to ASN after adjustment for confounding factors was 0.49% (95% CI 0.37% to 0.62%).

    View this table:
    Table 2

    Univariate and multivariable binary logistic regression analysis of the association between 5-min Apgar score and overall additional support needs status

    With regard to specific causes of ASN, Apgar scores less than 7 were independently associated in a dose-response manner with physical/motor, visual/hearing and cognitive disorders, though not with autistic spectrum disorders (table 3). Multivariable analyses also identified an association between risk of language/speech disorders and Apgar score of 4–6, but this was not statistically significant at lower Apgar scores.

    View this table:
    Table 3

    Univariate and multivariable multinomial logistic regression analysis of the association between 5-min Apgar score and type-specific additional support needs

    Among those children who had sat external examinations, low Apgar score was associated with lower educational attainment in univariate analysis, but the association became non-significant when adjusted for potential confounding factors (table 4). ASN was strongly associated with reduced likelihood of high educational attainment (univariate OR 0.22, 95% CI 0.22 to 0.23). However, adjustment for ASN did not change the association between low Apgar and educational attainment, nor was there a significant interaction between Apgar score and ASN.

    View this table:
    Table 4

    Univariate and multivariable generalised ordinal logistic regression analysis of the association between 5-min Apgar score and highest educational attainment*

    Discussion

    A 5-min Apgar score of less than 7 was a significant risk factor for ASN at school even after adjusting for potential confounders. The association was strongest for ASN resulting from physical and motor impairments, followed by visual, hearing and cognitive impairments. There was no association with autistic spectrum disorders.

    The association between suboptimal Apgar score and lower educational attainment was not robust to adjustment for potential confounders and was unaffected by ASN status, despite any ASN requirement greatly reducing the odds of reaching the highest levels of educational attainment. This may reflect the low proportion of children with low Apgar scores among the subgroup of children with ASN, reflected in the low population attributable fraction.

    Previous studies on the association between Apgar score and ASN have reported conflicting results. Some have found no association but had methodological limitations, such as small sample sizes,21 ,22 parental-reported outcomes22 and selection bias.23 In contrast, a number of larger and more robust studies have demonstrated increased risk of learning disability24 or special educational needs schooling14 ,25–27 among children with low Apgar scores at birth. These include a Norwegian birth cohort where children with a low Apgar score were more likely to have ASN and below-average educational performance, as reported by parents;14 three large linkage studies from the USA in which low Apgar score predicted later school-age ASN or learning disability;24–26 and a Swedish linkage study in which lack of graduation grades was used as a proxy for attendance at a special school.27 Another linkage study of Swedish boys reported that low Apgar scores, in the absence of overt encephalopathy, were associated with poorer results on cognitive testing but not school grades,15 which fits with our finding of no relationship between Apgar score and subsequent educational achievement.

    Our finding that low Apgar scores at 5 min after birth were significantly associated only with certain types of ASN is plausible given existing evidence. Previous studies have demonstrated poorer cognitive outcomes15 ,16 among children with low Apgar scores at birth. The association between low Apgar scores and physical/motor causes of ASN may partly reflect the established association with disorders such as cerebral palsy.7–11 Low Apgar score is also a well recognised risk factor for sensorineural hearing loss.28–31 Evidence has previously been lacking on whether low Apgar score is associated with visual impairment in children born at term without encephalopathy,32 while previous studies have produced conflicting results as to whether low Apgar score is associated with speech and language difficulties33 ,34 or autistic spectrum disorders.35

    The study described here has a number of strengths compared with the existing literature on this topic. It is based on a large nationwide cohort of more than 700 000 children, for whom data on ASN and educational performance were collected as a statutory requirement. All eligible children attending mainstream or special schools in Scotland during the study period were included, to enhance generalisability. Unlike previous studies, our analysis looks at the association of Apgar scores with overall ASN status and at a comprehensive range of specific ASN types. We were able to obtain individual-level data on the relationship between Apgar scores and ASN over a long period of follow-up, and to adjust for a number of important potential confounders. Use of quality-assured routine data sources and a high level of completeness of Apgar score data also engender confidence in our results. Finally, the recency of these data is an important strength, given advances in neonatal resuscitation since many previous studies in this field were carried out.

    The majority of pupils could be linked to delivery data: those that could not are likely to have been born outside of Scotland. The study population will not include those children born in Scotland who emigrated or died prior to entering the school system, an unavoidable limitation of the routine data sources.

    Another potential limitation lies in the accuracy of assessment of Apgar scores: previous studies have suggested significant interobserver and intraobserver variability in scoring, particularly among preterm infants and those receiving resuscitation.36 ,37 While we were unable to assess the accuracy of scoring or the extent of resuscitation provided, this study aimed to minimise the impact of these factors by excluding preterm infants and using the 5-min Apgar score to incorporate the effect of any immediate resuscitative measures. Furthermore, any measurement error in Apgar score might be expected to bias towards the null, potentially underestimating the association between Apgar score and ASN. Though modifications to the Apgar score have been proposed to address these shortcomings,38 these are not yet in widespread clinical use.

    Data were not available on adverse antenatal or intrapartum events, neonatal encephalopathy or childhood health problems, which might offer a causal explanation for the association observed. We also excluded multiple pregnancies, preterm deliveries and very low birthweight infants, so results may not be generalisable to children with these characteristics.

    The relationship seen here between Apgar score and ASN does not necessarily imply causality: Apgar score is therefore best characterised as a risk marker for adverse later life outcomes, rather than a causal factor. Furthermore, the population attributable fraction estimate for low Apgar score suggests that this exposure is unlikely to be a major contributor to the population burden of ASN. Nonetheless, our findings are important because they suggest that it may be possible in the future to identify specific modifiable factors in the perinatal period that could improve children’s outcome in the long term.

    Conclusion

    This study has demonstrated a strong and dose-dependent association between low Apgar score at 5 min after birth and later ASN, independent of key confounding variables, and has further investigated this relationship by the type of ASN. Apgar scores are therefore associated with long-term as well as short-term prognoses and with educational as well as clinical outcomes at the population level.

    References

    1. ↵
      1. Ross CE,
      2. Wu CL
      . The links between education and health. Am Sociol Rev 1995;60:719–45. doi:10.2307/2096319
      OpenUrlCrossRefWeb of Science
    2. ↵
      Education (Additional Support for Learning) Act (Scotland). 2004.
    3. ↵
      Education (Additional Support for Learning) Act (Scotland). 2009.
    4. ↵
      1. Casey BM,
      2. McIntire DD,
      3. Leveno KJ
      . The continuing value of the Apgar score for the assessment of newborn infants. N Engl J Med 2001;344:467–71. doi:10.1056/NEJM200102153440701
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Iliodromiti S,
      2. Mackay DF,
      3. Smith GC, et al
      . Apgar score and the risk of cause-specific infant mortality: a population-based cohort study. Lancet 2014;384:1749–55. doi:10.1016/S0140-6736(14)61135-1
      OpenUrlCrossRefPubMed
    6. ↵
      1. Apgar V
      . A proposal for a new method of evaluation of the newborn infant. Curr Res Anesth Analg 1953;32:260–7.
      OpenUrlPubMed
    7. ↵
      1. Nelson KB,
      2. Ellenberg JH
      . Apgar scores as predictors of chronic neurologic disability. Pediatrics 1981;68:36–44.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Moster D,
      2. Lie RT,
      3. Irgens LM, et al
      . The association of Apgar score with subsequent death and cerebral palsy: a population-based study in term infants. J Pediatr 2001;138:798–803. doi:10.1067/mpd.2001.114694
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Thorngren-Jerneck K,
      2. Herbst A
      . Perinatal factors associated with cerebral palsy in children born in Sweden. Obstet Gynecol 2006;108:1499–505. doi:10.1097/01.AOG.0000247174.27979.6b
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Lie KK,
      2. Grøholt E-K,
      3. Eskild A
      . Association of cerebral palsy with Apgar score in low and normal birthweight infants: population based cohort study. BMJ 2010;341:c4990. doi:10.1136/bmj.c4990
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Thorngren-Jerneck K,
      2. Herbst A
      . Low 5-minute Apgar score: a population-based register study of 1 million term births. Obstet Gynecol 2001;98:65–70. doi:10.1016/S0029-7844(01)01370-9
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Ehrenstein V,
      2. Sørensen HT,
      3. Pedersen L, et al
      . Apgar score and hospitalization for epilepsy in childhood: a registry-based cohort study. BMC Public Health 2006;6:23–23. doi:10.1186/1471-2458-6-23
      OpenUrlCrossRefPubMed
    13. ↵
      1. Sun Y,
      2. Vestergaard M,
      3. Pedersen CB, et al
      . Apgar scores and long-term risk of epilepsy. Epidemiology 2006;17:296–301. doi:10.1097/01.ede.0000208478.47401.b6
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Moster D,
      2. Lie RT,
      3. Markestad T
      . Joint association of Apgar scores and early neonatal symptoms with minor disabilities at school age. Arch Dis Child 2002;86:F16–21. doi:10.1136/fn.86.1.F16
      OpenUrlCrossRef
    15. ↵
      1. Odd DE,
      2. Rasmussen F,
      3. Gunnell D, et al
      . A cohort study of low Apgar scores and cognitive outcomes. Arch Dis Child 2008;93:F115–20. doi:10.1136/adc.2007.123745
      OpenUrlCrossRef
    16. ↵
      1. Ehrenstein V,
      2. Pedersen L,
      3. Grijota M, et al
      . Association of Apgar score at five minutes with long-term neurologic disability and cognitive function in a prevalence study of Danish conscripts. BMC Pregnancy Childbirth 2009;9:14. doi:10.1186/1471-2393-9-14
      OpenUrlCrossRefPubMed
    17. ↵
      NHS Information Services Division Scotland. Births in Scottish Hospitals. 2014. http://www.isdscotland.org/Health-Topics/Maternity-and-Births/Births/Background.asp (accessed 20 Dec 2014).
    18. ↵
      American Academy of Pediatrics Committee on Fetus and Newborn & American College of Obstetricians and Gynecologists and Committee on Obstetric Practice. The Apgar score. Pediatrics 2006;117:1444–7. doi:10.1542/peds.2006-0325
      OpenUrlAbstract/FREE Full Text
    19. ↵
      Scottish Government. Scottish Index of Multiple Deprivation. 2014. http://www.scotland.gov.uk/Topics/Statistics/SIMD (accessed 20 Dec 2014).
    20. ↵
      1. Mackay DF,
      2. Smith GCS,
      3. Dobbie R, et al
      . Obstetric factors and different causes of special educational need: retrospective cohort study of 407,503 schoolchildren. Br J Obstet Gynaecol 2013;120:297–307; discussion 07–8. doi:10.1111/1471-0528.12071
      OpenUrlCrossRef
    21. ↵
      1. Blackman JA
      . The value of Apgar scores in predicting developmental outcome at age five. J Perinatol 1988;8:206–10.
      OpenUrlPubMed
    22. ↵
      1. Krebs L,
      2. Langhoff-Roos J,
      3. Thorngren-Jerneck K
      . Long-term outcome in term breech infants with low Apgar score—a population-based follow-up. Eur J Obstet Gynecol Reprod Biol 2001;100:5–8. doi:10.1016/S0301-2115(01)00456-0
      OpenUrlCrossRefPubMed
    23. ↵
      1. Seidman DS,
      2. Paz I,
      3. Laor A, et al
      . Apgar scores and cognitive performance at 17 years of age. Obstet Gynecol 1991;77:875–8.
      OpenUrlPubMedWeb of Science
    24. ↵
      1. Stanton-Chapman TL,
      2. Chapman DA,
      3. Scott KG
      . Identification of early risk factors for learning disabilities. J Early Interv 2001;24:193–206. doi:10.1177/10538151010240030501
      OpenUrlCrossRef
    25. ↵
      1. Andrews H,
      2. Goldberg D,
      3. Wellen N, et al
      . Prediction of special education placement from birth certificate data. Am J Prev Med 1995;11(3 Suppl):55–61.
      OpenUrlPubMedWeb of Science
    26. ↵
      1. Redden SC,
      2. Scandlin DK,
      3. Roth MS, et al
      . The use of public health databases to estimate the risk of special education placement. State Centre for Health Statistics. Raleigh: North Carolina Public Health, 2002:6.
    27. ↵
      1. Stuart A,
      2. Otterblad Olausson P,
      3. Källen K
      . Apgar scores at 5 minutes after birth in relation to school performance at 16 years of age. Obstet Gynecol 2011;118(2 Pt 1):201–8. doi:10.1097/AOG.0b013e31822200eb
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Chu K,
      2. Elimian A,
      3. Barbera J, et al
      . Antecedents of newborn hearing loss. Obstet Gynecol 2003;101:584–8. doi:10.1016/S0029-7844(02)03118-6
      OpenUrlCrossRefPubMed
    29. ↵
      1. Biswas AK,
      2. Goswami SC,
      3. Baruah DK, et al
      . The potential risk factors and the identification of hearing loss in infants. Indian J Otolaryngol Head Neck Surg 2012;64:214–17. doi:10.1007/s12070-011-0307-6
      OpenUrlCrossRefPubMed
    30. ↵
      1. Jiang ZD,
      2. Zang Z,
      3. Wilkinson AR
      . Cochlear function in 1-year-old term infants born with hypoxia-ischaemia or low Apgar scores. J Paediatr Child Health 2012;48:160–5. doi:10.1111/j.1440-1754.2011.02066.x
      OpenUrlPubMed
    31. ↵
      1. Lieu JEC,
      2. Ratnaraj F,
      3. Ead B
      . Evaluating a prediction model for infant hearing loss. Laryngoscope 2013;123:2873–9. doi:10.1002/lary.24033
      OpenUrlCrossRefPubMed
    32. ↵
      1. Mercuri E,
      2. Atkinson J,
      3. Braddick O, et al
      . Visual function in full-term infants with hypoxic-ischaemic encephalopathy. Neuropediatrics 1997;28:155–61. doi:10.1055/s-2007-973693
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Delgado CEF,
      2. Vagi SJ,
      3. Scott KG
      . Identification of early risk factors for developmental delay. Exceptionality 2007;15:119–36. doi:10.1080/09362830701294185
      OpenUrlCrossRef
    34. ↵
      1. Stanton-Chapman TL,
      2. Chapman DA,
      3. Bainbridge NL, et al
      . Identification of early risk factors for language impairment. Res Dev Disabil 2002;23:390–405. doi:10.1016/S0891-4222(02)00141-5
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      1. Guinchat V,
      2. Thorsen P,
      3. Laurent C, et al
      . Pre-, peri- and neonatal risk factors for autism. Acta Obstet Gynecol Scand 2012;91:287–300. doi:10.1111/j.1600-0412.2011.01325.x
      OpenUrlCrossRefPubMed
    36. ↵
      1. O'Donnell CP,
      2. Kamlin CO,
      3. Davis PG, et al
      . Interobserver variability of the 5-minute Apgar score. J Pediatr 2006;149:486–9. doi:10.1016/j.jpeds.2006.05.040
      OpenUrlCrossRefPubMedWeb of Science
    37. ↵
      1. Bashambu MT,
      2. Whitehead H,
      3. Hibbs AM, et al
      . Evaluation of interobserver agreement of apgar scoring in preterm infants. Pediatrics 2012;130:e982–7. doi:10.1542/peds.2012-0368
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. Rudiger M,
      2. Braun N,
      3. Aranda J, et al
      . Neonatal assessment in the delivery room—Trial to Evaluate a Specified Type of Apgar (TEST-Apgar). BMC Pediatr 2015;15:18. doi:10.1186/s12887-015-0334-7
      OpenUrlCrossRefPubMed

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors EJT: conceptualised and designed the study, carried out data analyses, contributed to data interpretation, drafted the initial manuscript, revised the manuscript, and approved the final manuscript as submitted. DFM: conceptualised and designed the study, obtained the data, carried out data analyses, contributed to data interpretation, revised the manuscript and approved the final manuscript as submitted. SMN. S-AC: contributed to data interpretation, revised the manuscript and approved the final manuscript as submitted. JPP: conceptualised and designed the study, obtained the data, contributed to data interpretation, revised the manuscript and approved the final manuscript as submitted.

    • Competing interests None declared.

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

    Linked Articles