Recurrent Mental Health Presentations to Public Hospital Services: A Focus on Borderline Personality Disorder

David Smith, Paul Cammell, Malcolm Battersby, Dianna Bartsch, Jeremy Stevenson, Tarun Bastiampillai

Research output: Contribution to journalArticlepeer-review


Objective: To investigate associations between patients with borderline personality disorder (BPD)–related symptoms and their hospital presentations as well as the effect of inpatient length of stay (LOS) on time to hospital re-presentation.

Methods: A retrospective cohort design was used to investigate mental health emergency department (ED) visits and inpatient admissions. The cohort comprised 13,320 men and 12,290 women with a follow-up period between January 1, 2014, and December 31, 2019.

Results: Across all presentations in the study period, approximately 4% of mental health patients were discharged from ED or inpatient admission with primary diagnosis of BPD. Both male and female patients with BPD were at higher risk of hospital re-presentation when compared to patients with any other type of mental disorder (P < .01). Patients with BPD who had LOS > 14 days in their first inpatient admission were, on average, more likely to experience a repeat ED or inpatient presentation 58 days sooner than patients who had LOS < 2 days (P = .036).

Conclusions: Findings suggest the need for (a) more accurate recording of BPD and related presentations, (b) more in-depth investigations of BPD care pathways, and (c) identifications of subpopulations who may benefit from a specific inpatient length of stay.
Original languageEnglish
Article number23m03559
Number of pages18
JournalPrimary Care Companion for CNS Disorders
Issue number1
Publication statusPublished - 9 Jan 2024


  • borderline personality disorder (BPD)
  • Hospital presentation
  • public hospital
  • Length of stay (LOS)


Dive into the research topics of 'Recurrent Mental Health Presentations to Public Hospital Services: A Focus on Borderline Personality Disorder'. Together they form a unique fingerprint.

Cite this