Classification & Assessment in Psychiatry · Paper A
Self-harm: epidemiology
Self-harm is one of the strongest known predictors of future completed suicide, which is why every presentation needs proper assessment, not just physical treatment. Take a young man who presents to A&E after cutting himself for tension relief, e.g. superficial forearm cuts with no expressed wish to die: his risk of eventual suicide is still markedly higher than someone with no self-harm history, even though this particular episode wasn't intended to be lethal. In practice, apparent low lethality never replaces a full psychosocial assessment.
Repetition of self-harm is common, often occurring within a year of an initial episode. Self-cutting is frequently done for tension relief rather than lethal intent, while self-poisoning is the commonest method presenting to hospital. Previous self-harm and mental illness are strong risk factors for repetition; male sex is specifically a stronger risk factor for completed suicide, even though self-harm itself is more commonly reported by women. Say a low-toxicity overdose is followed by regret, e.g. a small paracetamol overdose taken impulsively and immediately regretted: neither feature makes a future episode predictable or safely dismissible. Method and stated intent must inform a broader psychosocial assessment rather than a simple risk score.
Self-harm is deliberate, non-fatal harm to one's own body, and it is one of the strongest known predictors of future completed suicide (death intentionally caused by one's own actions). This single fact underpins why every self-harm presentation to a hospital or clinic warrants a full assessment, not simply physical treatment of the injury and discharge. Understanding the epidemiology, meaning the patterns of who self-harms, by what method, and what happens afterwards, helps clinicians judge risk accurately rather than relying on gut feeling or on how serious an episode looks at first glance.
Self-harm itself is not a formal diagnosis under either ICD-11 or DSM-5 diagnostic criteria; it is a behaviour that can accompany many different underlying diagnoses, including depression, psychosis, a personality disorder, or no diagnosable mental illness at all. This matters because a thorough risk formulation must look beyond diagnostic labels to the individual's own account, circumstances and stated intent, rather than assuming that a particular diagnosis automatically explains or predicts a particular episode.
The caveat that catches people out is judging future risk from how serious the current episode looks. Take a patient who takes a small, low-toxicity paracetamol overdose during a brief argument and immediately regrets it, e.g. ten tablets swallowed impulsively and then vomited up within the hour: this patient may describe the episode as 'not serious', yet low apparent lethality of the method used does not equate to low future risk, and it should never replace a full psychosocial assessment. Repetition of self-harm is common, and a substantial proportion of people who self-harm will repeat within the following twelve months, making the period immediately after an episode a genuinely high-risk window for structured follow-up and safety planning.
Previous self-harm is itself one of the strongest risk factors for further self-harm, alongside comorbid mental illness such as depression, psychosis or an underlying personality disorder. Male sex is a specifically important risk factor for eventual completed suicide, even though self-harm episodes themselves are more commonly reported by women; this apparent contradiction, sometimes called the gender paradox in suicide research, partly reflects differences in method lethality and help-seeking behaviour between the sexes. Imagine risk as something that accumulates rather than resets after each episode: every prior act of self-harm adds to the background risk level the next presentation is assessed against, it does not wash away once the wound has healed.
Self-cutting is frequently carried out for tension relief (using the physical act to reduce overwhelming emotional distress) rather than with lethal intent; many patients describe it as a coping strategy rather than as a genuine suicide attempt. Self-poisoning, most often with over-the-counter or prescribed medication, is the commonest method of self-harm presenting to hospital in the United Kingdom. A young man presents to the emergency department after cutting his forearm for tension relief, with no expressed wish to die and no previous history of self-harm; because any history of self-harm carries a markedly raised future suicide risk compared with the general population, and because male sex and any coexisting mental illness both raise that risk further if present, this presentation still requires a full psychosocial assessment rather than discharge with wound care alone.
Self-harm needs to be distinguished from a genuine suicide attempt with clear lethal intent, even though the two overlap and one can shade into the other over time; the stated and inferred intent, the method chosen, and the degree of planning versus impulsivity all help separate them, though intent can be genuinely mixed or difficult for the patient themselves to articulate. Non-suicidal self-injury is the term some classification systems and clinicians use specifically for self-harm without any suicidal intent, most often for tension relief, as distinct from a self-harm episode carried out with some degree of suicidal intent; the two are managed with overlapping but not identical approaches. Self-harm should also be distinguished from accidental injury or risk-taking behaviour without any element of deliberate self-directed harm, such as substance-related accidents, since the deliberate, self-directed element is what defines self-harm in the first place.
Every presentation of self-harm should trigger a structured psychosocial assessment, covering the circumstances and apparent intent of the episode, ongoing suicidal ideation, protective and risk factors, and any existing mental illness. NICE guidance on self-harm explicitly recommends against relying on brief numerical risk-scoring checklists to decide who needs admission or follow-up, because these scoring tools perform poorly at predicting an individual's true future risk; instead, a full biopsychosocial formulation, built from the person's own account, history and current circumstances, should guide the decision. Screening for depression, substance misuse and psychosis should form part of this assessment, since treating an underlying condition can meaningfully reduce future risk of repetition.
Management is stratified by clinical need rather than by method or apparent severity of the current episode alone. Immediate physical treatment of the injury or overdose comes first, often in the emergency department, followed by a full psychiatric assessment before any decision about discharge. For people who self-harm repeatedly, psychological therapies, including dialectical behaviour therapy (a structured therapy combining individual and group skills training, originally developed for chronic self-harm in borderline personality disorder) and problem-solving-based interventions, are recommended first-line approaches by NICE. Medication is not used to treat self-harm behaviour directly, but treating an underlying depressive or psychotic illness with appropriate first-line agents is an important part of comprehensive management. Safety planning, actively involving the patient in identifying personal triggers, coping strategies and sources of support, is now standard practice recommended across NICE and Royal College of Psychiatrists guidance: this is where the epidemiology actually earns its keep clinically: knowing that repetition risk peaks soon after an episode, and that low apparent lethality does not mean low risk, is exactly what should shape how tightly that safety plan and follow-up are scheduled.
The pearl is the gender paradox itself: women report more self-harm episodes, but men complete suicide more often, largely because of the methods each group tends to use. A question that gives you sex alongside method and asks about future suicide risk, rather than self-harm frequency, is testing whether you know which way that paradox cuts.
Self-harm is a behaviour, not a diagnosis, and its single strongest signal is history repeating itself: every prior episode raises future risk regardless of how minor the current one looks.