The impact of early special educational needs provision on later hospital admissions, school absence and education attainment: A target trial emulation study of children with isolated cleft lip and/or palate

Abstract

Background: Special educational needs (SEN) provision is designed to help pupils with additional educational, behavioural or health needs. Our aim was to assess the impact of early SEN provision on health and educational outcomes for a well-defined population, pupils with cleft lip and/or cleft palate (CLP) without additional anomalies.

Methods: We used the ECHILD database, which links educational and health records across England. Our target population consisted of children with a recorded diagnosis of CLP without other major congenital anomalies in hospital admission records in ECHILD. We applied a trial emulation framework to define eligibility into our study and investigate the causal impact of SEN provision in the first year of compulsory school (Year 1 - age five/six years) on various health and educational outcomes accumulated by the end of primary education (Year 6 - age ten/eleven years). SEN provision was categorised as: None, SEN Support, and Education and Health Care Plan (EHCP). The outcomes were: unplanned hospital utilisation, medical and unauthorised school absences, persistent absences, and standardised key stage 1 (KS1) and key stage 2 (KS2) mathematics attainment scores. To account for confounding factors affecting the observed associations and estimate the causal effects of early SEN provision on these outcomes, we used three estimating approaches: propensity score-based methods (inverse probability weighting, [IPW]), g-computation, and augmented IPW (AIPW). Causal effects were measured in terms of average treatment effects (ATE) and average treatment effects on the treated (ATT), expressed as rate ratios (RaR) for hospitalisations and absences, risk ratios (RiR) for persistent absences, and mean differences (Δ) for academic scores. Missing values of the confounders were handled via the missing covariate indicator method. We triangulated these results with those obtained by univariable and multivariable regression.

Results: Our study included 6,601 children with CLP and without additional major congenital anomalies. Evaluations involving EHCP were limited by the low numbers of comparative children. Thus, only comparisons of SEN Support (N = 2,009, 31.6%) versus None (N = 4,350, 68.4%) are reported. Observed rates of unplanned hospitalisation (RaRcrude = 1.31, 95% confidence interval (CI): 1.12, 1.52), persistent absence (RiRcrude = 2.21 (1.87, 2.62)) and medical absence (RaRcrude = 1.34 (1.28, 1.40)) were higher amongst children with recorded SEN support, whilst KS1 and KS2 maths scores were lower (Δ crude = -0.85 (-0.90, -0.79) and Δ crude = -0.82 (-0.89, -0.75), respectively). Contrary to the observed relative rates and risks, we found small or no evidence of a causal effect of SEN Support on unplanned hospitalisation (ATE: RaRIPW = 1.16 (1.00, 1.34), RaRg = 0.99 (0.87, 1.12); RaRIAPW = 1.02 (0.87, 1.17) or persistent absences (ATE: RiRIPW = 1.13 (0.92, 1.34); RiRg = 1.08 (0.86, 1.31); RiRAIPW = 1.20 (0.96, 1.45)). We found that SEN support increased rates of medical absences (ATE: RaRIPW = 1.10 (1.04, 1.18); RaRg = 1.09 (1.03, 1.15); RaRAIPW = 1.04 (0.95, 1.13)), decreased those of unauthorised absences (RaRIPW = 0.86 (0.76, 0.97); RaRg = 0.98 (0.86, 1.09); RaRAIPW = 0.80 (0.66, 0.95)) and decreased - but not as extensively as the crude differences suggested- KS1 (ATE: Δ IPW = -0.18 (-0.25, -0.10); Δ g = -0.21 (-0.26, -0.16); Δ AIPW = -0.25 (-0.32, -0.17)) and KS2 maths scores (ATE: Δ IPW = -0.24 (-0.33, -0.15); Δ g = -0.27 (-0.33, -0.21); Δ AIPW = -0.24 (-0.32, -0.17)). Results for the ATT for each of these outcomes were similar to those for the ATE, indicating no observable evidence of heterogeneity of effects by treatment received. Sensitivity analyses confirmed the robustness of these results.

Discussion: In the population of children with CLP without further major congenital anomalies, assignment to receive or not receiving early SEN Support appears to have no harmful impact on the rates of unplanned hospitalisation or persistent absences, but to increase rates of medical absences, whilst reducing rates of unauthorised absences. For the sub-populations of children with key stage results, such hypothetical intervention does not appear to completely reduce the observed disadvantage in KS1 and KS2 mathematics scores. These results relate to the impact of the intention to intervene not the actual delivery of actual SEN Support provision as this information is not available in school administrative records. Furthermore, we cannot discount the impact of unaccounted confounding factors, such as parental education and early home learning environments, particularly for the education attainment results.

Fig 1. Diagram depicting cohort derivation. This figure depicts the derivation of the cohort starting from combining births in hospital episode statistics with evidence of cleft lip and/or palate between 2003 and 2013 who are linkable to the national pupil database between 2008/2009 and 2018/2019. Exclusion criteria were: pupils with multiple links between hospital episode statistics and the national pupil database, had no record in reception class, has no birth record in hospital episode statistics, has a recorded major congenital anomaly in hospital episode statistics, a first phenotypical recording of cleft lip and/or palate after reception, an unknown cleft lip and/or palate type, conflicting date information (e.g., linked mortality data states a death before school start), and a recorded attendance at a special school.

Fig 2. Predicted propensity score distributions by observed special educational needs category for Special Educational Needs Support versus None.

This diagram depicts the density distribution of the probability of receiving SEN Support in reference to None. Predictors used to estimate treatment probability included: gender, gestational age, birthweight category, maternal age, ethnic group, language group, income deprivation affecting children index quintile, free school meal eligibility, academic year, type of cleft lip and/or palate, chronic conditions, early years foundation profile z-score and relative age.

Fig 3. Predicted propensity score distribution by observed special educational needs category for Education and Healthcare Plan versus None.

This diagram depicts the density distribution of the probability of receiving an Education and Healthcare Plan in reference to None. Predictors used to estimate treatment probability included: gender, gestational age, birthweight category, maternal age, ethnic group, language group, income deprivation affecting children index quintile, free school meal eligibility, academic year, type of cleft lip and/or palate, chronic conditions, early years foundation profile z-score.

Fig 4. Predicted propensity score distribution by observed special educational needs category for Education and Healthcare Plan versus Special Education Needs Support.

This diagram depicts the density distribution of the probability of receiving an Education and Healthcare Plan in reference to Special Education Needs Support. Predictors used to estimate treatment probability included: gender, gestational age, birthweight category, maternal age, ethnic group, language group, income deprivation affecting children index quintile, free school meal eligibility, academic year, type of cleft lip and/or palate, chronic conditions, early years foundation profile z-score.