Potential for England's Statutory School Entry Assessment to Identify Special Educational Needs and Reveal Structural Inequalities: A Population-Based Study

on November 19, 2024 by guest. Protected by copyright. http://adc.bmj.com/ Arch Dis Child: first published as 10.1136/archdischild-2023-325590 on 12 October 2023. Downloaded from

Potential for England’s statutory school entry assessment to identify special educational needs and reveal structural inequalities: a population-­ based study

Megan L Wood  ‍ ‍ ,1,2 Lydia Gunning,2 Sam Relins,2 Kuldeep Sohal,2 John Wright,2 Mark Mon-­Williams,1,2 Amy L Atkinson3

Abstract Objective  To investigate at a population level whether England’s universal assessment of ’school readiness’ is associated with later identification of special educational needs (SEN). Also, whether ethnic differences exist in SEN identification (white British versus ethnic minority) and whether this varies as a function of school readiness. Method  Analysis included 53 229 individuals aged 5–12 years from the Connected Bradford Database (2012/2013–2019/2020). Logistic regression analyses examined: (1) whether reaching a ’good level of development’ on England’s ’school readiness’ assessment was associated with later SEN identification; and (2) whether interactions exist between school readiness and ethnicity. Results  32 515 of 53 229 children reached a good level of development, of which 3036 (9.3%) were identified as having SEN. In contrast, 10 171 of 20 714 (49.1%) of children who did not reach a good level of development were later identified as having SEN. Children not reaching a good level of development had increased odds of being later identified as having SEN after controlling for covariates (Or: 8.50, 95% Ci: 8.10 to 8.91). In children who did not reach a good level of development, white British children had higher odds of being identified as having SEN compared with ethnic minority peers (Or: 1.22, 95% Ci: 1.11 to 1.34). No ethnic differences of having SEN were observed in children reaching a good level of development (Or: 1.04, 95% Ci: 0.93 to 1.16). Conclusions  School readiness assessments are associated with later SEN identification. Structural inequalities may exist in SEN identification in children not entering formal education ’school ready’. Such assessments could facilitate earlier identification of SEN and reduce structural inequalities in its identification.

►Additional supplemental material is published online only. To view, please visit the journal online (http://​dx.​doi.​ org/​10.​1136/​archdischild-​ 2023-​325590).

What Is Already Known On This Topic

⇒England’s school entry assessments can identify children at risk of special educational needs (SEN). ⇒These assessments can also identify children at 1 School of Psychology, University of Leeds, Leeds, UK 2 Bradford Institute for Health Research, Bradford, UK 3 Department of Psychology, Lancaster University, Lancaster, UK

risk of poor academic attainment.

What This Study Adds

⇒The potential of using statutory school entry assessments to identify children at increased risk of being identified as having SEN at a population-­level. ⇒Structural inequalities may exist in the Correspondence to Dr Megan L Wood, University of Leeds, Leeds, LS2 9JT, UK; ​m.​l.​wood@​leeds.​ac.​uk

identification of SEN in children who do not enter formal education ‘school ready’

How This Study Might Affect Research, Practice Or Policy

Received 29 March 2023 Accepted 13 September 2023 Published Online First 12 October 2023

⇒Using school entry assessments as a screening tool may facilitate earlier identification of SEN. ⇒Address structural inequalities by identifying

children who could benefit for further monitoring or assessment who may otherwise ‘fall through the net’.

formal education,2 despite presenting with indi- cators as early as infancy.3 These delays can result in a range of negative outcomes, including disen- gagement,4 5 school exclusion6 and poor academic achievement5 with consequent long-­term physical and mental health problems.7 Thus, there is a need to identify children with SEN early.8 9

School readiness evaluations are frequently conducted at school entry in various countries (eg, England, Australia, Canada8 10 11). As part of these assessments, children are typically evaluated on a range of academic (eg, numeracy) and non-­ academic abilities (eg, social skills12 13). As children with SEN often experience difficulties in these areas8 (see O’Connor et al5 for a conceptual model), school readiness evaluations may identify children at increased risk of needing support.8 Indeed, there is accumulating evidence that school readiness evaluations can identify children at increased risk of various developmental difficulties, including autism14 and dyslexia.15 More broadly, previous research has also revealed associations between school readiness and SEN at school entry.13 16

Introduction Identification of special educational needs (SEN) is a vital first step to ensuring children get the neces- sary support to enable them to thrive in school. In the UK, SEN is defined as ‘a learning difficulty or disability which calls for special educational provi- sion to be made’1 which includes, but not limited to, ‘autism; social, emotional and mental health needs; or moderate learning difficulties’. Many children with SEN are not identified for some years into their

© Author(s) (or their employer(s)) 2024. Re-­use permitted under Cc By. Published by BMJ.

To cite: Wood ML, Gunning L, Relins S, et al. Arch Dis Child 2024;109:52–57.

52 Wood ML, et al. Arch Dis Child 2024;109:52–57. doi:10.1136/archdischild-2023-325590

on November 19, 2024 by guest. Protected by copyright. http://adc.bmj.com/ Arch Dis Child: first published as 10.1136/archdischild-2023-325590 on 12 October 2023. Downloaded from

Atkinson et al8 used data from the Born in Bradford longitu- dinal cohort to examine whether the school readiness assessment conducted universally in England (the Early Years Foundation Stage Profile; EYFSP) could identify children at risk of later needing support. A strong relationship was observed, suggesting that school entry assessments could be used as a tool to identify children who are at increased risk of needing SEN support but currently ‘under the radar’ (see also Hughes et al9).

Table 1  Sample demographics

Whole sample (n=53 229) SEN (n=13 207)

Not SEN (n=40 022)

SEN Yes 13 207 (24.8%) – – No 40 022 (75.2%) – – Good level of development A range of demographic factors have also been found to be associated with identification of SEN, which may reflect both risk factors in requiring additional need and structural inequali- ties in the needs being identified. For example, previous research suggests that children from more disadvantaged backgrounds are at increased risk of requiring additional support.1 17 18 This may be influenced by features more common in communities with lower socioeconomic position such as increased parental ill-­health and drug/alcohol dependency.19 Conversely, research has shown that children from ethnic minority backgrounds are less likely to have SEN, but this may reflect structural inequali- ties in identification.20–22 These structural inequalities may arise due to variation in views relating to child development between cultures23 or knowledge of and access to services and support within ethnic minority families.20 To date, no existing research has investigated whether such ethnic differences vary as a func- tion of school readiness. This would indicate whether struc- tural inequalities relating to ethnicity are present universally, or whether their effects are limited to particular groups (eg, chil- dren who enter formal education not being ‘school ready’).

Reached 32 515 (61.1%) 3036 (23.0%) 29 479 (73.7%)

Not reached 20 714 (38.9%) 10 171 (77.0%) 10 543 (26.3%)

Ethnicity White British 27 458 (51.6%) 6822 (51.7%) 20 636 (51.6%) Minority 25 771 (48.4%) 6385 (48.3%) 19 386 (48.4%) Free school meal eligibility* Yes 13 658 (25.7%) 4983 (37.7%) 8675 (21.7%) No 39 571 (74.3%) 8224 (62.3%) 31 347 (78.3%) Sex Male 26 583 (49.9%) 8382 (63.5%) 18 201 (45.5%) Female 26 646 (50.1%) 4825 (36.5%) 21 821 (54.5%) English as an additional language Yes 18 914 (35.5%) 4876 (36.9%) 14 038 (35.1%) No 34 315 (64.5%) 8331 (63.1%) 25 984 (64.9%) Percentages may not add up to 100% due to rounding. *Proxy for socioeconomic position. SEN, special educational needs.

(‘reception’; 4–5 years of age). The assessment allowed us to differentiate between children who were judged by teachers as being ‘school ready’, having reached a ‘good level of develop- ment’ (an outcome measure within the EYFSP) and those who were not (having not reached a ‘good level of development’).8 We focused our analyses on the post-­2013 England school read- iness assessments as a different version was used before this (see online supplemental materials 3 for further detail).

The current study had two aims. First, it aimed to use population-­level data to test Atkinson et al’s8 suggestion that school readiness evaluations could be used to identify SEN, after controlling for covariates (eg, sex, ethnicity).24 Second, it aimed to investigate whether ethnic differences exist in the identifica- tion of SEN, and whether this differs as a function of school readiness. We used data from the Connected Bradford dataset24 to investigate this.

For ethnicity, children were categorised as either ‘white British’ or ‘ethnic minority’ based on family-­reported census records (see online supplemental materials 4 for a further break- down). Sex (female, male) was also taken from family-­reported census records. Eligibility for free school meals (no/yes) was used as a proxy for socioeconomic position as children from lower socioeconomic position households are eligible for free school meals in the UK. This was obtained from Department for Educa- tion records. English as an additional language status (no/yes) was teacher-­reported and obtained from the Department for Education records.

Method Study design, setting and participants The study design was longitudinal. Data were collated from the Connected Bradford population-­level linked database for over 800 000 citizens across the Bradford district, UK.24 Education records provided by the Department for Education were linked across multiple time points, spanning academic years 2012– 2013 to 2019–2020. Data were extracted in April 2021. Chil- dren included in the sample were aged 5 (year 1) to 12 (year 7) years old. See online supplemental materials 1 for a description of how the sample was derived. Sample demographics of the final sample in relation to the wider Connected Bradford data- base are included in online supplemental materials 2.

Study size To be included in analyses, individuals were required to have complete data for all variables. Further, as our focus was on chil- dren who fall ‘under the radar’, we removed children already

Variables Outcome variable SEN status (yes/no) was obtained from the Department for Education records and reflected whether children had ever been identified as having SEN. Children were identified as having SEN if they had ever been recorded as having a School Action plan, a School Action Plus plan, an SEN statement, an Educa- tion, Health and Care plan, or in receipt of SEN support.

Table 2  The number of children identified as having special educational needs (SEN) after reception (yes/no) as a function of the good level of development (GLD) measure (reached/not reached)

Identified as SEN*

GLD

Total Yes No

Reached 3036 (9.3%) 29 479 (90.7%) 32 515 (100%)

Not reached 10 171 (49.1%) 10 543 (50.9%) 20 714 (100%)

Predictor variables We used data from the school readiness assessment conducted universally in England (the EYFSP) in the first year of school

Total 13 207 40 022 53 229

*After reception year.

53 Wood ML, et al. Arch Dis Child 2024;109:52–57. doi:10.1136/archdischild-2023-325590

on November 19, 2024 by guest. Protected by copyright. http://adc.bmj.com/ Arch Dis Child: first published as 10.1136/archdischild-2023-325590 on 12 October 2023. Downloaded from

identified as having SEN during reception (ie, before or at the same time as the school readiness assessment; determined by Department for Education records).

Table 3  Classification rates for SEN as the outcome variable

Classification rate Definition Value 95% CIs

Sensitivity % of children identified as having SEN without good development

0.77 0.76 to 0.78

Statistical methods We conducted logistic regression analysis to investigate whether the good level of development (reached, not reached) was asso- ciated with later SEN identification (no, yes). An unadjusted logistic regression model was first conducted, followed by a model adjusting for ethnicity (ethnic minority, white British), eligibility for free school meals (no, yes), sex (female, male) and English as an additional language status (no, yes). Sensitivity, specificity, positive predictive value, negative predictive value and classification rates were also calculated to assess the poten- tial of school entry assessments for flagging unidentified SEN.

Specificity % of children not identified as having SEN with good development

0.74 0.73 to 0.74

Positive predictive value

Probability a child who did not reach good development being identified as having SEN

0.49 0.48 to 0.50

Negative predictive value

Probability a child who reached good development was not identified as having SEN

0.91 0.90 to 0.91

Correct classification rates

% of children correctly classified as at risk/not at risk of being identified as having SEN based on school readiness

0.74 0.74 to 0.75

We then investigated whether interactions exist between the good level of development and ethnicity. A logistic regression model was conducted with good level of development, ethnicity and their interaction, as predictors. This analysis controlled for eligibility for free school meals, sex and English as an additional language status. The reference category for good level of devel- opment was ‘reached’, while the reference category for ethnicity was ‘ethnic minority’. Significant interactions were followed up with planned comparisons (corrected using Bonferroni) to inves- tigate whether an effect of ethnicity was present in children who did, and who did not, reach a good level of development.

SEN, special educational needs.

of development had higher odds of being identified as having SEN compared with children who reached a good level of development. This was observed in both the unadjusted (Or: 9.34, 95% Ci: 8.92 to 9.78, p<0.001) and adjusted (Or: 8.50, 95% Ci: 8.10 to 8.91, p<0.001) models. Several covariates were also associated with SEN status. Children in receipt of free school meals had higher odds of being identified as having SEN compared with children not in receipt of free school meals (Or: 1.73, 95% Ci: 1.64 to 1.81, p<0.001), and males had increased odds of being identified as having SEN compared with females (Or: 1.68, 95% Ci: 1.60 to 1.76, p<0.001). Furthermore, white British children had increased odds of being identified as having SEN relative to children from ethnic minorities (Or: 1.18, 95% Ci: 1.10 to 1.25, p<0.001).

Analyses were conducted in R (V.4.0.2; see online supple- mental materials 5 for R code).25 Participants with missing data for any variable were excluded. In line with previous research (eg, 8 26), we report ORs.

Results The final sample contained data for 53 229 children (see online supplemental materials 1 and 2 for derivation). Table 1 displays the sample characteristics.

Classification rates Classification rates are displayed in table 3. Positive predictive value was relatively low, with approximately one in two children who did not reach a good level of development later identified as having SEN. However, negative predictive value was high, indi- cating few children who reached a good level of development were identified as having SEN.

Is school readiness associated with SEN status? The frequencies and percentages of children identified as having SEN, depending on whether they reached a good level of devel- opment, are displayed in table 2.

The outcomes from the logistic regression models examining whether reaching a good level of development is associated with SEN are presented in figure 1. Children who did not reach a good level

Figure 1  Or and 95% CIs for the association between GLD and SEN status (unadjusted model) and GLD+ethnicity+free school meal eligibility+sex+English as an additional language status (adjusted model). GLD, good level of development; SEN, special educational needs.

54 Wood ML, et al. Arch Dis Child 2024;109:52–57. doi:10.1136/archdischild-2023-325590

on November 19, 2024 by guest. Protected by copyright. http://adc.bmj.com/ Arch Dis Child: first published as 10.1136/archdischild-2023-325590 on 12 October 2023. Downloaded from

Figure 2  Percentage of children identified as having special educational needs by good level of development and ethnicity. Error bars show 95% CIs.

targeted interventions can be implemented within schools to provide children with the support they need to thrive. Second, the use of an early ‘screener’ allows at-­risk children to be iden- tified before they exhibit symptoms of a diagnosable learning difficulty. Thus, school readiness assessments can allow support to be put into place before more serious difficulties occur and prevent a child falling behind their peers.27 It is, however, worth noting that the assessment does appear to overidentify the number of children at risk, with positive predictive value for the assessment being relatively low (0.49; see also 8 9). We there- fore suggest that outcomes from school entry assessments should be considered alongside other information, such as the current teacher’s judgement.8

Do school readiness and ethnicity interact regarding SEN identification? The percentages of children identified as having SEN are displayed in figure 2, as a function of good level of development and ethnicity.

The logistic regression model revealed a significant interac- tion between reaching a good level of development and ethnicity after controlling for covariates (Or: 1.18, 95% Ci: 1.07 to 1.29, p<0.001; see online supplemental materials 6 for full reporting). Post-­hoc tests revealed no ethnic differences in children who reached a good level of development (Or: 1.04, 95% Ci: 0.93 to 1.16, p=1.00). In children who did not reach a good level of development, white British children had increased odds of being identified as SEN compared with children from ethnic minority backgrounds (Or: 1.22, 95% Ci: 1.11 to 1.34, p<0.001).

White British children who were not ‘school ready’ had higher odds of being identified as having SEN relative to ethnic minority peers who were not ‘school ready’. This is in line with previous research which has found ethnic biases in smaller samples.20–22 More broadly, this adds to a growing body of literature indi- cating that individuals from ethnic minority backgrounds are identified later than their ethnic majority counterparts in condi- tions as diverse as autism,28 tuberculosis29 and cancer.30 Impor- tantly, however, there were no ethnic differences in children who entered formal education ‘school ready’, thus indicating that the association between ethnicity and SEN is not universal.

Discussion In a population sample, we found that children assessed as not being ‘school ready’ had higher odds of being identified as having SEN compared with children who were ‘school ready’. We then examined whether the extent to which ethnicity was associated with SEN differed as a function of school readiness. This analysis revealed no ethnic differences in children who reached a good level of development. In contrast, for children who did not reach a good level of development, white British children had higher odds of being identified as having SEN compared with children from ethnic minority backgrounds.

Why might ethnic differences have only emerged in children who were not ‘school ready’? One possibility is that families of children from ethnic minority backgrounds who do not enter formal education ‘school ready’ may be more likely to lack understanding of how to navigate complex systems (eg, schools, the SEN support system) relative to other groups.31 Another possibility is that difficulties at school entry in children from ethnic minority backgrounds could be attributed to factors other than SEN (eg, cultural factors). Use of school readiness

Evidence that reaching a good level of development was asso- ciated with later SEN status illustrates that England’s school readiness assessments can identify children who are at increased risk of being identified as having SEN.8 14 Such findings have important implications for policymakers, schools, children and their families. First, identifying children ‘at risk’ earlier means

55 Wood ML, et al. Arch Dis Child 2024;109:52–57. doi:10.1136/archdischild-2023-325590

on November 19, 2024 by guest. Protected by copyright. http://adc.bmj.com/ Arch Dis Child: first published as 10.1136/archdischild-2023-325590 on 12 October 2023. Downloaded from

assessments to flag these children as ‘at risk’ would facilitate further monitoring or assessments. This may reduce structural inequalities in the identification of SEN. It would be beneficial for further research to investigate this possibility.

regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Open access  This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (Cc By 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/​ licenses/by/4.0/.

There are several limitations with this study. First, the findings are specific to a single district in England (Bradford), which has a large ethnic diversity compared with the rest of the country.32 Future work should therefore examine this at a national level. Furthermore, the current analyses determined the odds of a child later being identified as having SEN. It does not consider the children who were never formally identified. More research is required to determine whether there are structural inequalities in the children who remain ‘under the radar’. In addition, it is possible that other confounding variables exist which were not controlled for. Finally, while this research and previous studies (eg,8 14) have shown the utility of England’s school readiness assessment, it is possible that it may not be as effective at iden- tifying difficulties in all groups of children. For example, it may be difficult to judge domains such as social skills and literacy in children with lower English proficiency33 (although English as an additional language status was controlled for in the present analyses). Teacher-­reported assessments may also be subject to subtle ethnic stereotypes.11 34 Future research should examine these possibilities in relation to school readiness evaluations. Additional work may also consider the longer-­term associations between school readiness and other aspects of learning and development, such as career prospects and Not in Education, Employment or Training status.

ORCID iD Megan L Wood http://orcid.org/0000-0003-1882-2355

References

1. Children’s and families act; 2014.
2. Department for Education. Special educational needs in England; 2022.
3. Crane L, Chester JW, Goddard L, et al. Experiences of autism diagnosis: a survey of

over 1000 parents in the United Kingdom. Autism 2016;20:153–62. 4 Lőrinc M, Ryan L, D’Angelo A, et al. De-­individualising the ’NEET problem’: an

ecological systems analysis. Eur Educ Res J 2020;19:412–27. 5 O’Connor M, Howell-­Meurs S, Kvalsvig A, et al. Understanding the impact of special

health care needs on early school functioning: a conceptual model. Child Care Health Dev 2015;41:15–22. 6 Carrick A, Hamilton CJ. Heated behaviour in the classroom for children with FASD:

the relationship between characteristics associated with Adhd, Odd and ASD, hot executive function and classroom based reward systems. Children (Basel) 2023;10:685. 7 Parker C, Marlow R, Kastner M, et al. Supporting kids, avoiding problems (SKIP)

study: relationships between school exclusion, psychopathology, development and attainment - a case control study. J Child Serv 2016;11:91–110. 8 Atkinson AL, Hill LJB, Pettinger KJ, et al. Can holistic school readiness evaluations

predict academic achievement and special educational needs status? Evidence from the early years foundation stage profile. Learn Instr 2022;77:101537. 9 Hughes C, Foley S, White N, et al. School readiness in children with special educational

In conclusion, using population-­level data, this study indicates that a potentially powerful tool already exists that can address the pressing international need to improve early identification of SEN. It also suggests ethnic differences exist in the identification of SEN in children who do not enter formal education ‘school ready’. Using school readiness assessments may help to address this, by identifying subgroups of children at school entry who may particularly benefit from additional monitoring.

needs and disabilities: psychometric findings from a new screening tool, the brief early skills and support index. Br J Educ Psychol 2018;88:606–27. 10 Brinkman S, Gregory T, Harris J, et al. Associations between the early development

instrument at age 5, reading and numeracy skills at ages 8, 10 and 12: a prospective linked data study. Child Ind Res 2013;6:695–708. 11 Goldfeld S, O’Connor M, Mithen J, et al. Early development of emerging and English-­

proficient bilingual children at school entry in an Australian population cohort. Int J Behav Dev 2014;38:42–51. 12 Department for Education. Special educational needs and disability: an analysis and

summary of data sources; 2022. 13 Goldfeld S, O’Connor M, Sayers M, et al. Prevalence and correlates of special health

Twitter Megan L Wood @DrMeganLWood Acknowledgements  The data are provided by the citizens of Bradford and district and collected by the NHS and other organisations as part of their care and support.

care needs in a population cohort of Australian children at school entry. J Dev Behav Pediatr 2012;33:319–27. 14 Wright B, Mon-­Williams M, Kelly B, et al. Investigating the association between

Contributors  Mlw, Ala and MM-­W designed the study. MLW performed the analysis. Mlw, Lg and ALA drafted the paper. SR and ALA reviewed the analysis. All authors reviewed the final paper. MLW accepts responsibility for the work and acts as the guarantor.

early years foundation stage profile scores and subsequent diagnosis of an autism spectrum disorder: a retrospective study of linked healthcare and education data. BMJ Paediatr Open 2019;3:e000483. 15 Snowling MJ. Early identification and interventions for dyslexia: a contemporary view.

Funding  This work was supported by the National Institute for Health Research (grant number NIHR133648), UK Prevention Research Partnership (grant number Mr/S037527/1), and the Medical Research Council [grant number Mr/W014416/1].

J Res Spec Educ Needs 2013;13:7–14. 16 Janus M. Impact of impairment on children with special needs at school

entry:comparison of school readiness outcomes in Canada, Australia, and Mexico. EEI 2011;21. 17 Croll P. Social deprivation, school-­level achievement and special educational needs.

Disclaimer  The interpretation and conclusions contained in this study are those of the authors alone.

Educ Res 2002;44:43–53. 18 Hollomon HA, Dobbins Dr, Scott KG. The effects of biological and social risk factors

Competing interests  None declared.

on special education placement: birth weight and maternal education as an example. Res Dev Disabil 1998;19:281–94. 19 Department for work and pensions; 2014. 20 Emerson E. Deprivation, ethnicity and the prevalence of intellectual and

Patient consent for publication  Not required.

Ethics approval  This study involves human participants and was approved by HRA CAG and received ethical approval from the Leeds Bradford Research Ethics Committee for Connected Bradford (IRAS ref: 239924, CAG ref: 18/CAG/0091 and REC ref: 18/YH/0200). Participants gave informed consent to participate in the study before taking part.

developmental disabilities. J Epidemiol Community Health 2012;66:218–24. 21 Golson ME, Haverkamp CR, McClain MB, et al. Influences of student race/ethnicity

and gender on autism special education classification considerations. Autism 2022;26:1423–35. 22 Schmaus M. Disproportionate identification of special needs for ethnic and

Provenance and peer review  Not commissioned; externally peer reviewed.

Data availability statement  Data are available upon reasonable request.

language minority students in England - patterns and explanationa. Br J Sociol Educ 2022;43:40–62. 23 Welterlin A, LaRue RH. Serving the needs of immigrant families of children with

Supplemental material  This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-­reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local

autism. Disabil Soc 2007;22:747–60. 24 Sohal K, Mason D, Birkinshaw J, et al. Connected Bradford: a whole system data

linkage. Wellcome Open Res 2022;7:26. 25 R. C. Team. R: A language and environment for statistical computing. Vienna R. C.

Team; 2020.

56 Wood ML, et al. Arch Dis Child 2024;109:52–57. doi:10.1136/archdischild-2023-325590

on November 19, 2024 by guest. Protected by copyright. http://adc.bmj.com/ Arch Dis Child: first published as 10.1136/archdischild-2023-325590 on 12 October 2023. Downloaded from

26 Norbury CF, Gooch D, Baird G, et al. Younger children experience lower levels of 30 Ashing-­Giwa KT, Gonzalez P, Lim JW, et al. Diagnostic and therapeutic delays language competence and academic progress in the first year of school: evidence from a population study. J Child Psychol Psychiatry 2016;57:65–73. 27 Huijsmans MDE, Kleemans T, Kroesbergen EH. The cognitive profiles for different samples of among a multiethnic sample of breast and cervical cancer survivors. Cancer 2010;116:3195–204. 31 Tromans S, Chester V, Gemegah E, et al. Autism identification across ethnic groups: a mathematical learning difficulties and their similarity to typical development: evidence from a longitudinal study. J Exp Child Psychol 2022;214:105288. 28 Mandell DS, Listerud J, Levy SE, et al. Race differences in the age at diagnosis narrative review. AIA 2021;7:241–55. 32 City of Bradford Metropolitan District Council. Census ethnic group, religion and language spoken; 2022. 33 O’Connor M, O’Connor E, Tarasuik J, et al. Academic outcomes of multilingual children among medicaid-­eligible children with autism. J Am Acad Child Adolesc Psychiatry 2002;41:1447–53. 29 Gilmour B, Xu Z, Bai L, et al. The impact of ethnic minority status on in Australia. Int J Speech Lang Pathol 2018;20:393–405. 34 Tenenbaum HR, Ruck MD. Are teachers’ expectations differenet for racial tuberculosis diagnosis and treatment delays in Hunan province, China. BMC Infect Dis 2022;22:90.

minority than for European American students? A meta-­analysis. J Educ Psychol 2007;99:253–73.

57 Wood ML, et al. Arch Dis Child 2024;109:52–57. doi:10.1136/archdischild-2023-325590