Mental health difficulties affect the well-being of doctors and compromise the delivery of healthcare. However, large-scale data on doctors’ mental health needs are limited.

Describe patterns of self-referrals for mental health support among doctors in England and explore associations with demographic factors, speciality, neurodevelopmental and mental health indicators.

Observational study using data from doctors who self-referred for mental health difficulties to a national service in England over a 4-year period. Logistic regression was used to explore associations between speciality and mental health indicators.

Of the 16 815 doctors who self-referred during the study period, 80% were under the age of 49 and 70.6% were female with the two largest ethnicities being 65.1% White and 22.7% Asian. Women were more likely to report higher scores for depression (odds ratio 0.90, 95% CI = 0.84 to 0.97), anxiety (odds ratio 0.78, 95% CI = 0.72 to 0.84) and psychological distress (odds ratio 0.78, 95% CI = 0.70 to 0.87), but males were more likely to screen positive for attention-deficit hyperactivity disorder (ADHD) symptoms. Doctors in general practice accounted for 46.3% of referrals. Compared with them, doctors in most other specialities had higher odds of elevated mental health scores across all measures, including ADHD.

The findings highlight a significant mental health burden among self-referring doctors, particularly for females and doctors in non-general practice specialities. Tailored and easily accessible support strategies that account for both demographic and professional contexts are essential to address the diverse mental health needs of the medical workforce.

Mental health issues are prevalent among children and young people (CYP) with chronic conditions like epilepsy, yet few access evidence-based psychological therapies. Evidence from the Mental Health Intervention for Children with Epilepsy (MICE) trial supports the effectiveness of a personalised modular psychological intervention, but cost-effectiveness is unknown.

To assess the cost-effectiveness of the MICE intervention compared with assessment-enhanced usual care at 12-months follow-up, taking a health and social care perspective.

We conducted a within-trial economic evaluation. Outcomes were the Strengths and Difficulties Questionnaire (SDQ; primary) and quality-adjusted life years (QALYs; secondary) for CYP, caregivers, and CYP and caregivers combined. Sensitivity analyses examined missing data and intervention-costing assumptions.

Cost-effectiveness results for the SDQ indicated that MICE had a higher probability of being cost-effective compared with control at a willingness to pay ≥£368 per unit improvement. For QALYs, MICE had a lower probability of being cost-effective for CYP compared with control (35 to 42%) across the £20 000–£30 000 per QALY threshold range. However, at the upper threshold this finding was reversed in sensitivity analyses with missing data imputed (45 to 58%) and with MICE costed at 75%, assuming the intervention partly substituted standard services (46 to 55%). Furthermore, MICE had a higher probability of being cost-effective for caregiver QALYs (52 to 63%) and combined CYP and caregiver QALYs (62 to 75%).

MICE appears to be cost-effective compared with assessment-enhanced usual care when considering QALYs for CYP and caregivers combined, though uncertainty exists across willingness-to-pay thresholds.