Using the English national health service dataset for research into mental health service use among children and young people

BackgroundRoutinely collected administrative data, such as the Mental Health Services Data Set (MHSDS) in England, provide opportunities to investigate determinant and patterns of mental health service use across the population. Persistent challenges with data completeness and consistency, however, limit their value for robust population-level research. ObjectiveTo assess the methodological robustness and research utility of the MHSDS for investigating patterns and determinants of mental health service use among children and young people (CYP) in England between 2016 and 2023. MethodsWe evaluated the completeness and consistency of key sociodemographic variables (gender, ethnicity, location, and socioeconomic status) over time and across local authorities. To assess the impact of data quality on analytical validity, we modelled the likelihood of care contact attendance using sociodemographic covariates, comparing estimates from complete-case analyses with those from models treating missing or conflicting data as distinct categories. Spatial and temporal patterns of CYP in contact with (referred to) mental health services and attendance rates were examined annually and by local authority. FindingsFrom 2016 to 2023, over 4.7 million CYP were in contact with mental health services, of whom 62.4% had at least one attended contact. Missing data were substantial, persistent, often co- occurred in subpopulations with distinct attendance patterns. The proportion of CYP referred to services increased from 1.9% (371,655) in 2016 to 9.7% (1,699,899) in 2023, while attendance rates remained stable at 59.8%. Spatial analyses highlighted regional variations, with higher attendance rates in Northern England ConclusionsData completeness influences the validity of research using the MHSDS. Addressing this issue is essential for producing equitable and reproducible evidence in CYP mental health research. Clinical ImplicationMethodological improvements in handling missing data and temporal inconsistencies will strengthen the interpretability, reproducibility, and equity relevance of MHSDS-based mental health research.