Identifying people at increased risk of self-harm : evidence from population-based electronic health records in Hong Kong

Abstract

Globally, self-harm has taken a heavy toll on public health. Risks of self-harm varies by populations. By targeting high-risk populations, limited resources can be more efficiently allocated. Although a plethora of publications has extensively explored the self-harm risk in different vulnerable populations, current evidence is limited by substantial heterogeneities in study populations, research designs, timeframes of investigation, and measures of predictors and outcomes. By using population-based electronic health records from the Clinical Data Analysis and Reporting System in Hong Kong, the research reported in this thesis comprehensively examined self-harm risk among various high-risk populations in Hong Kong. Study one examined the association between seven psychiatric disorders (i.e., depression, alcohol use disorders, personality disorders, bipolar disorders, anxiety disorders, schizophrenia, and drug use disorders) and self-harm and found that people with these psychiatric disorders had a substantially increased risk of subsequent self-harm. The highest risk was observed in the drug use disorder group. Findings highlighted the need to develop more efficient and targeted preventive measures in psychiatric care management. Study two investigated repeated self-harm risk in Hong Kong and found that the highest reoccurrence risk occurred in the first year of the indexed self-harm episode. Four or more previous self-harm events was the strongest risk factor for self-harm repetition. This study suggested that Hong Kong self-harm patients with non-fatal self-harm events should be supported through effective, timely, and ongoing aftercare plans based on their risk profiles, to reduce the risk of self-harm reoccurrence. Study three observed the self-harm risk among people with different drug use disorders (comprising opioid, ketamine, methamphetamine, sedative, hypnotic, or anxiolytic, amphetamine or related stimulant, cocaine, cannabis, hallucinogen, unspecified or other drug, and multidrug) and found that people with these specific drug use disorders carried an elevated risk of self-harm compared to their drug-free counterparts. Risk levels varied by drug class, with the highest risk among ketamine and opioids. This information highlighted the importance of considering drug use disorders when assessing self-harm risk and developing suicide prevention strategies specific to drug use disorders. Study four examined the association between self-harm and antidepressant use (categorized into five classes including tricyclic and related antidepressant drugs, selective serotonin reuptake inhibitors, noradrenergic and specific serotonergic antidepressants, serotonin-norepinephrine reuptake inhibitors, and others) and compared the self-harm risk among different antidepressant classes in adults with depression aged 40 and above. This study found that self-harm risk peaked before the initiated antidepressant treatment and decreased progressively during and after drug exposure, irrespective of age and drug type. There were significant variations in self-harm risk in certain pairs of antidepressant classes. These findings may inform clinicians’ benefit-risk assessments when prescribing antidepressants. This thesis describes the epidemiology of self-harm risk in various vulnerable populations in Hong Kong and identifies several risk factors. These large-scale study findings enable population-level understanding and evaluation of self-harm and suicide risk to promote effective clinical practice and tailored intervention strategies to reduce the healthcare burden.