Trends in the complexity of self-harm hospitalisations over 11 years at a major regional hospital in Victoria, Australia: a retrospective study

George Mnatzaganian^1*Rebecca Giallo²Reagan Martin¹Angela Crombie³Ziad Nehme⁴Belinda Delardes⁴David Burns¹Susan Furness³Tim Lenten³Renee Sharples¹Lisa Hanson¹Rachel Huxley⁵

• ¹School of Rural Health, College of Science, Health and Engineering, La Trobe University, Melbourne, Victoria, Australia
• ²School of Psychology, Faculty of Health, Deakin University, Melbourne, Victoria, Australia
• ³Bendigo Health, Bendigo, Victoria, Australia
• ⁴Ambulance Victoria, Melbourne, Australia
• ⁵School of Medicine, Faculty of Health, Deakin University, Waurn Ponds, Australia

Background: This population-based study explored the Australian Refined Diagnosis Related Groups (AR-DRG) complexity of self-harm admissions over time at a major regional hospital in Victoria, Australia. It also assessed the prehospital paramedic management of such admitted patients. Methods: Self-harm admissions at the hospital from January 1, 2010, to December 31, 2020, were included, excluding accidental injuries and children under 10. Hospital records were linked with Ambulance Victoria electronic patient care records. Trends in the age-and social status-adjusted AR-DRG complexity of self-harm hospitalisations by sex were analysed using Dickey-Fuller and MacKinnon tests. Odds of presenting with a major complexity were modelled using the Generalised Estimating Equations approach.Findings: Overall, 2,000 individuals (58.6% female, mean age at last admission 36.9 ± 18.6 years), contributing to 2,808 admissions, were included in the study. The proportion of selfharm admissions with a major AR-DRG complexity significantly increased over time, from 9.3% in 2010 to 43.5% in 2020 (p<0.001). Both increased complexity and intensive care unit admissions were observed in both males and females. The use of multiple self-harm methods also rose over time. Of the 2,000 patients, 1,416 (70.8%) sought emergency services assistance within 30 days of hospital admission, with 139 (9.8%) not transported to the emergency department. These non-transported patients had higher odds of presenting to the hospital within one month of the prehospital paramedic assessment with more complex conditions (odds ratio 2.04, 95% confidence interval 1.39-2.99, p<0.001) and longer hospital stays compared to those who were transported.Our findings indicate a trend toward more severe cases of self-harm over time, observed in both males and females, with an increase in the use of multiple self-harm methods. Additionally, our results suggest that paramedic non-transports to the emergency department should be re-evaluated, as these patients experience worse outcomes.