Early screening for autism spectrum disorder (ASD) can enhance educational and health outcomes for affected children. This narrative systematic review explores school-based screening tools used around the world to identify children with ASD and explore the differences across socio-demographic groups.

Systematic review of electronic databases (EMBASE, MEDLINE, PsycINFO, Cochrane and Scopus) in October 2024 of papers published between 2011 and 2024.

Mainstream school-based settings globally.

Children aged 4–16 years old attending mainstream school.

School-based screening tools for ASD, including all types of informant and format of tools reported in eligible studies.

Primary outcomes included prevalence of screen positives, sensitivity and specificity of the screening tools. Secondary outcomes included participants’ sex, socioeconomic status and ethnicity, and the relation of this to the primary outcomes.

Of 7765 eligible articles, 14 studies were included in this review. We identified eight different school-based ASD screening tools. Study populations ranged from 103 to 16 556 children, with sensitivity and specificity varying by screening tool used, age group, setting and ASD prevalence. The percentage of children screening positive for ASD ranged from 0.7% to 8.5%. Studies were conducted in Europe (n=6), Western Pacific (n=4), the Americas (n=3) and Eastern Mediterranean (n=1) regions. No studies explicitly explored accuracy or validity outcomes based on ethnicity or socioeconomic status. Half of the 14 studies (n=7) reported the sensitivity and specificity of the screening tools; sensitivity ranged from 58% to 94% and specificity from 61% to 100%. There was insufficient evidence to recommend any single ASD screening tool.

ASD screening tools vary widely across the globe, with limited standardisation. Evidence is lacking on how ethnicity and socioeconomic status affect their effectiveness in schools. Given the dearth of scientific evidence in this field, collaboration among educators, researchers and policymakers is needed to establish the evidence base for universal screening, identify optimal tools, coordinate their use and ensure their validation for specific populations.

To estimate the sociodemographic and geographical variation in prescribing selective serotonin reuptake inhibitors (SSRIs) and medications for attention-deficit/hyperactivity disorder (ADHD) to children and young people (CYP) in North West London, UK.

Cross-sectional population-based study.

General practices in North West London, UK, with data for the period 2020–2022 obtained from the Discover Now platform, which covers approximately 95% of the local population.

762 390 CYP aged 5–24 years in the year 2022.

Primary outcome: Prescription rates of SSRIs and ADHD medications. Secondary outcomes: Associations between prescription rates and sociodemographic factors, including age, gender, geographical area (local authority), ethnicity and socioeconomic deprivation (measured using the Index of Multiple Deprivation).

The total sample comprised 762 390 CYP. 2.20% of the sample were prescribed an SSRI (95% CI 2.17% to 2.24%) and 0.50% an ADHD medication (95% CI 0.49% to 0.52%) in years 2020–2022. High deprivation was associated with the highest rates of an SSRI prescription (2.5%). In contrast, low deprivation was associated with the highest rates of an ADHD medication prescription (0.70%). This divergent pattern was evident in some London boroughs and not in others. The relationship between level of area deprivation and prescription rates also differed by borough. Overall, the sociodemographic factors could not explain most of the variation in prescription rates (Pseudo R² 0.18 for SSRI and 0.06 for an ADHD medication).

Prescriptions for common mental disorders and ADHD for CYP from North West London varied by sociodemographic characteristics and London borough of residence, potentially exacerbating mental health inequalities. To monitor and address these inequalities, more extensive use of linked electronic health records should be undertaken; for example, data on mental health diagnosis and service utilisation are needed to investigate the relationship between diagnosis and treatment over time.