Understanding the prevalence of behavioural disorders amongst children and adolescents

Study type
Protocol
Date of Approval
Study reference ID
22_001907
Lay Summary

Behavioural disorders are not uncommon in children, affecting about 1 in 20 children at any given
time. They include norm- and rule-breaking behaviour on a scale from defiance, deception and a
rejection of authority figures through to theft, violence and other criminal behaviours. They are
associated with poor lifelong outcomes including low educational attainment, low income,
increased criminality and earlier mortality.
Nevertheless, they are relatively poorly understood with some inconclusive and confusing data on
how this rate varies across different places in England and how it has changed over the last 20
years.
Most families with children affected by behavioural disorders seek help of some kind, with a large
minority approaching their GP. This study will give us the opportunity to add to the evidence base
on how common behavioural disorders are, and where and when the problem is more and less
common, utilising GP data for the first time.

Technical Summary

Objectives:
The present study will add to the evidence base on how many people present to their GP with behavioural disorders,
allowing healthcare funders to plan better interventions.

This study will: describe the prevalence and incidence of behavioural disorders over the last 5 years and use statistical testing and regression modelling to establish whether and how occurrence varies over time and place. This will be repeated for individual types of behavioural disorder.

Exposure: the study is based on a random sample of patients in the 1999 – 2021 birth cohorts, who represent the population at risk of behavioural disorders in the study period of 2017 – 2021 (all people aged under 18)
Outcome: a diagnosis of a behavioural disorder
Methods and data analysis: we will count the number of cases, dividing by time at risk to derive incidence and population at risk to derive prevalence, constructing confidence intervals. Moreover, we will build a panel regression model to compare both incidence and prevalence over time and region. The models will estimate each individual’s risk of disorder onset / continuing disease state, accounting for region, practice and year. Significance of these terms will allow us to assess whether there is significant variation over time and place, accounting for patient demographics.

Health Outcomes to be Measured

Our primary outcome is a diagnosis of a behavioural disorder

Collaborators

Thomas Wagstaff - Chief Investigator - The Health Foundation
Andrew Campbell - Corresponding Applicant - The Health Foundation
Anne Alarilla - Collaborator - The Health Foundation
Arlene Gallagher - Collaborator - The Health Foundation
Jay Hughes - Collaborator - The Health Foundation
Mai Stafford - Collaborator - The Health Foundation

Linkages

Patient Level Index of Multiple Deprivation;Practice Level Rural-Urban Classification