Health service use of young people after mental health admission far from home: a retrospective cohort study in England.

Abstract

Background: Severe mental illness requiring inpatient admission is associated with poorer mental health and physical and social outcomes. Out-of-area admissions persist in child and adolescent mental health services (CAMHS) but are associated with longer length of stay and increased family burden. However, their association with future health service use is unknown. Using linked mental health and national health-care data, we examined patterns of health service use up to 60 months following out-of-area psychiatric admission versus within-area psychiatric admission in young people younger than 18 years.

Methods: The Maudsley Clinical Record Interactive Search (CRIS) database was used to identify all young people (aged <18 years) who had at least one mental health-related admission between Jan 1, 2009, and Dec 31, 2023, to South London and Maudsley NHS Foundation Trust in England. Admission was categorised as out of area (>50 miles [80 km or 60 min away from home address) or within area. Linkage to the Hospital Episode Statistics enabled analysis of health service use (emergency department presentations, and general hospital admissions for non-mental health-related or mental health-related reasons) in the 1 month, 12 months, and 60 months after discharge. Cox regression was used to investigate associations between admission out of area and subsequent health service use, adjusting for demographic, clinical, and admission characteristics.

Findings: The outcomes of 2287 young people were analysed (median age at index admission 16·3 years [IQR 15·3-17·3]; 1435 [63%] were female, 827 [36%] were male, and 21 [1%] identified as other). Of those with relevant data, 852 were admitted out of area and 1423 were admitted within area. Out-of-area admission was associated with subsequent emergency department presentation (adjusted hazard ratio [HR] 1·13 [95% CI 1·001-1·26]) and non-mental health-related general hospital admissions at 60 months (1·12 [1·0001-1·26]) despite adjusting for demographic, clinical, and admission characteristics. However, there was no association with mental health-related general hospital admissions at 60 months (adjusted HR 1·02 [95% CI 0·91-1·15]). Other factors such as ethnicity, diagnosis, previous hospital presentations, and being in the care of the state were also associated with presentations to emergency departments and general hospital admissions.

Interpretation: Admission out of area to a CAMHS ward is associated with increased likelihood of non-mental-health general hospital and emergency department use in the 60 months after discharge, raising important service planning needs in both physical and mental health care.

Funding: Royal College of Psychiatrists.