What is the causal effect of placement in Mainstream versus Special secondary school on the health and education outcomes of adolescents with Down syndrome?

1. Julia Shumway (Great Ormond Street Institute of Child Health, UCL)
2. Jill Ellis (East London NHS Foundation Trust)
3. Bianca De Stavola (Great Ormond Street Institute of Child Health, UCL)
4. Ruth Gilbert (Great Ormond Street Institute of Child Health, UCL)
5. Vincent Nguyen (Great Ormond Street Institute of Child Health, UCL)
6. Vaso Totsika (Division of Psychiatry, University College London)
7. Ania Zylbersztejn (Great Ormond Street Institute of Child Health, UCL)

Correspondence: julia.shumway.22@ucl.ac.uk

Background

Pupils with Down syndrome (DS) have unique health and learning needs which require additional support in education settings. Parents of pupils with Down syndrome (DS) choose between enrolling their children in Mainstream schools or schools with specialise in providing additional support (Special schools). Despite a lack of evidence on which type of secondary school better serves pupils with DS, there are policy proposals to cut English education costs by supporting pupils with special education needs in Mainstream rather than Special schools.

Research questions

To understand the influence of Mainstream vs Special secondary schools on health and education outcomes for pupils with DS, we asked two questions:

1. What are the health and education trajectories of pupils with DS prior to entering secondary school?
2. What is the causal impact of Mainstream vs Special secondary school on the health and education of pupils with DS?

Methods

We used linked health and education data for England, available as part of the Education and Child Health Insights from Linked Data (ECHILD) project, to follow children with DS from birth through adolescence. First, we calculated the cumulative incidence of hospital-recorded chronic health conditions (CHCs) and multimorbidity. Next, we examined health, education, and socioeconomic patterns in enrolment and transfer from Mainstream to Special primary school. Finally, we planned a target trial emulation (TTE) approach to compare the causal impact of Mainstream vs Special school; we assessed the viability of this approach by using propensity scores (PS) to evaluate the comparability of pupils with DS by Year 7 school placement.

Result

By age 11, children with DS had an 88% cumulative risk for having at least one hospital-recorded CHC, and a quarter of children had multiple CHCs affecting four or more body systems. While 67% of children with DS enrolled in Mainstream school at age 5, only 49% remained by age 11. Socioeconomic deprivation was associated with enrolling in Special school at age 5 and with transfer from Mainstream to Special school by age 11, even after controlling for differences in sex, health history, and skills assessments.

Among pupils attending Mainstream primary school in Year 6, 43% enrolled in Mainstream secondary school in Year 7. There was strong overlap between the PS distribution of characteristics among pupils who enrolled in Mainstream (mean PS: 0.62, standard deviation [SD]: 0.14) and Special (mean PS: 0.52, SD: 0.15) secondary schools.

Conclusion

Among pupils with DS entering secondary school, most have at least one CHC, and many have multimorbidity. Those who are more socioeconomically deprived are less likely to have commenced or sustained enrolment in Mainstream primary school. Having verified the viability of a TTE approach, we are using PS-based methods to estimate the causal effect of Mainstream vs Special secondary school on health and education outcomes for pupils with DS.