Mood instability and mental health service use in autism and attention-deficit/hyperactivity disorder: a natural language processing analysis of CRIS electronic healthcare records from 21 906 children and adolescents.

Abstract

Background: Children and young people (CYP) with neurodevelopmental diagnoses such as autism spectrum disorder (ASD) and attention-deficit/hyperactivity disorder (ADHD) have high child and adolescent mental health service (CAMHS) needs. Mood instability is a common and impairing emotion dysregulation-related symptom linked to increased adult psychiatric service use; however, its role in CAMHS trajectories remains unclear. We aimed to examine whether baseline mood instability was significantly associated with time to discharge and annual CAMHS use in CYP with ASD and/or ADHD.

Methods: We applied natural language processing (NLP) to extract mentions of mood instability within 3 months of ASD or ADHD index diagnosis from electronic health records of 21 906 CYP referred to CAMHS between 2008 and 2022. We used accelerated failure time models and negative binomial regression to assess associations between baseline mood instability and time to discharge and annual CAMHS use, adjusting for clinical and sociodemographic confounders.

Findings: Mood instability was associated with increased annual CAMHS use across ASD (adjusted incidence rate ratio (aIRR) 1.24, 95% CI 1.08 to 1.42), ADHD (aIRR 1.47, 95% CI 1.30 to 1.67) and ASD+ADHD (aIRR 1.27, 95% CI 1.12 to 1.44) groups. While mood instability had no significant effect on discharge timelines in autistic children with or without ADHD, it was linked to reduced time to discharge in the ADHD group (aTR 0.76, 95% CI 0.69 to 0.84). Associations were most pronounced in those not receiving ADHD medication in the ADHD group (aIRR 1.67, 95% CI 1.47 to 1.89; aTR 0.70, 95% CI 0.61 to 0.79).

Conclusions: Mood instability was significantly associated with elevated CAMHS use in CYP with neurodevelopmental conditions, with differential effect across diagnostic groups. This may reflect both variations in clinical expression of mood instability and configuration of neurodevelopmental CAMHS provision.

Clinicalimplications: These findings suggest the importance of assessing emotion dysregulation in care planning and pathway allocation in neurodevelopmental CAMHS. NLP offers a time- and cost-efficient approach to surface and structure clinical data from electronic CAMHS records for scalable clinical research on complex constructs such as mood instability.