Who Uses Which Suicide Attempt Method? A Decade of Emergency Room Data From Japan Reveals Striking Patterns

When someone survives a suicide attempt and is rushed to an advanced emergency and critical care medical center, that moment represents one of the most important—and most fleeting—opportunities in all of suicide prevention. The patient is alive, medically stabilized or being stabilized, and about to be seen by a psychiatrist. What clinicians do in those hours can shape whether a second attempt ever happens. Yet the psychiatric assessment that follows is often guided more by intuition and broad risk categories than by detailed evidence about who attempts suicide in what way, and why that might matter. A new retrospective study from Nara Medical University Hospital in Japan, published in BMC Psychiatry, offers one of the most granular contemporary pictures to date of how suicide attempt methods distribute across sex, age, and psychiatric diagnosis among patients arriving at a high-acuity emergency setting.

The research team, led by Minobu Ikehara and colleagues in the Department of Psychiatry at Nara Medical University School of Medicine, examined a full decade of clinical data, spanning April 2014 to March 2024. Their cohort consisted of 344 patients who had been transported by emergency services to the hospital’s advanced emergency and critical care center following a suicide attempt and who were subsequently referred for formal psychiatric evaluation. This is a crucial detail: the study does not capture every suicide attempt in the community, but rather the subset severe enough to require critical care transport. That makes the findings especially relevant to general hospitals and emergency departments, which serve as the de facto front line of post-attempt psychiatric care in Japan and in many other countries.

Methodologically, the study took a careful, classification-driven approach. Both psychiatric diagnoses and suicide attempt methods were coded using categories derived from the International Classification of Diseases, 10th Revision, or ICD-10, the diagnostic standard used widely across Japanese clinical settings. The primary analyses asked a deceptively simple question: do suicide attempt methods differ systematically by sex, by age group, and by psychiatric diagnosis? The researchers then went a step further with secondary analyses that compared patients whose attempts were classified as high lethality against those classified as low lethality, testing whether medical severity tracked cleanly with any of the demographic or diagnostic variables. Statistical associations were expressed as odds ratios with confidence intervals, the standard epidemiological machinery for quantifying how much more likely one method is within a given subgroup compared with the rest of the cohort.

The headline finding is that drug overdose was the most common method overall, and that it was significantly more frequent among female patients. This pattern echoes a well-documented international literature in which women who attempt suicide disproportionately use poisoning, particularly pharmaceutical overdose, while men more often use methods with higher intrinsic fatality. But the Japanese data add texture to that familiar story. Substance use disorders, for instance, were more frequent among male patients in the cohort, a diagnostic skew that carries its own clinical implications, since intoxication and impulsivity are known to interact dangerously in the moments surrounding an attempt. The sex differences, in other words, are not just about the physical method chosen; they extend into the underlying psychiatric landscape that clinicians must navigate during the post-attempt evaluation.

Age emerged as another powerful organizing variable. Drug overdose was more frequent among patients younger than 20 years, suggesting that adolescents and young adults in this emergency population gravitated toward the most accessible method—medications found in the home medicine cabinet or obtained by prescription. At the opposite end of the life course, pesticide poisoning was more frequent among patients aged 65 and older. This age-related shift is particularly sobering in the Japanese context, where agricultural chemicals are present in rural households and where late-life suicide remains a persistent public health concern. Pesticide ingestion is associated with high case fatality and with severe systemic toxicity, including organophosphate poisoning syndromes that demand intensive respiratory and neurological management. The fact that older patients in this cohort disproportionately arrived via this route means that emergency physicians and psychiatrists evaluating elderly survivors are, on average, dealing with a medically graver and methodologically distinct population than the one they encounter in younger patients.

Perhaps the most clinically provocative finding concerns psychiatric diagnosis. Jumping from height was more frequent among patients with schizophrenia and other psychotic disorders. This aligns with a broader international evidence base showing that psychotic illnesses are associated with elevated suicide risk and, in some studies, with the use of more violent or higher-lethality methods. The reasons are likely multifactorial: psychotic symptoms such as command hallucinations or delusional conviction can drive behavior with unusual directness, and the chronicity and social dislocation associated with severe mental illness may compound risk. For liaison psychiatry teams, the practical message is that the diagnostic formulation reached during the emergency evaluation should inform how the attempt itself is understood, not merely the treatment plan that follows it.

The secondary analyses on lethality add an important layer of nuance, and arguably constitute the study’s most consequential contribution to clinical practice. Patients in the high-lethality group were older on average than those in the low-lethality group, consistent with the pesticide finding among older adults and with the general observation that method lethality tends to rise with age. But—and this is the critical caveat—high- and low-lethality attempts were observed across all diagnostic categories. In other words, no psychiatric diagnosis guaranteed a low-lethality attempt, and no diagnosis made a high-lethality attempt inevitable. Medical severity, as measured in the critical care setting, did not map neatly onto diagnostic labels. A patient with a diagnosis often considered lower risk could still arrive in the intensive care unit, and a patient with severe mental illness could survive a medically minor attempt.

This is precisely why the authors argue against relying solely on categorized lethality when structuring post-attempt assessment. Their conclusion is that evaluation in general hospital settings should integrate demographic and diagnostic context, the observed suicide attempt method, and the actual medical severity of the presentation, rather than compressing all of that information into a single lethality category. The distinction matters because lethality classifications can be deceptively reassuring. A low-lethality label may reflect rescue, timing, or simple pharmacological luck rather than low suicidal intent, and a high-lethality label may reflect method availability as much as determination. By keeping method, demographics, diagnosis, and medical course in view simultaneously, clinicians can build a more honest and more individualized risk picture during the narrow window in which intervention is possible.

The study’s limitations are worth stating plainly, and the authors are transparent about them. It is a single-center retrospective analysis, drawing on records from one advanced emergency and critical care medical center in Nara Prefecture. That design means the findings describe patients who reached this particular hospital, and referral patterns, regional demographics, and local access to pesticides or medications could all shape the observed distributions. Retrospective designs also depend on the quality of recorded clinical data, and ICD-10-based categorizations, while standardized, necessarily flatten some clinical complexity. The cohort of 344 patients, while substantial for a decade of single-center recruitment, limits the statistical power available for fine-grained subgroup comparisons, which is why the reported associations are best read as directional signals to be tested in larger, multicenter samples rather than as definitive effect sizes.

Even with those caveats, the study lands at a moment when emergency psychiatry is under intense scrutiny worldwide. Suicide rates in Japan have remained a persistent national concern, with documented stresses including economic disruption and the psychological aftermath of the COVID-19 pandemic, and emergency departments everywhere report rising volumes of psychiatric presentations. What this research offers is a template for evidence-based triage: know that a young woman arriving after overdose presents a different clinical configuration than an older man arriving after pesticide ingestion, and that a patient with psychosis who has jumped requires an assessment framework attuned to the specific risks of that diagnosis. None of these observations is a substitute for careful, individual clinical judgment—but together they give that judgment a firmer empirical footing. In suicide prevention, where every post-attempt contact may be the last chance to intervene, that footing is not a luxury. It is the difference between assessment as ritual and assessment as genuine, life-saving medicine.

Subject of Research: Associations between suicide attempt methods and sex, age, and psychiatric diagnosis among patients referred for psychiatric evaluation at a Japanese emergency and critical care medical center

Article Title: Suicide attempt methods among patients referred for psychiatric evaluation at an advanced emergency and critical care medical center in Japan: associations with sex, age, and psychiatric diagnosis: a single-center retrospective study

Article References: Ikehara, M., Takada, R., Okumura, K., Nishi, Y., Noriyama, Y., Hamano, Y., Mizui, R., Fujimoto, Y., Doi, F., Fukui, H., Honda, Y., Araki, S., Sunami, Y., Yamanaka, K., Urushitani, S., Nishikawa, N., Yamamuro, K., & Okada, T. (2026). Suicide attempt methods among patients referred for psychiatric evaluation at an advanced emergency and critical care medical center in Japan: associations with sex, age, and psychiatric diagnosis: a single-center retrospective study. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08681-y

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08681-y

Keywords: suicide attempt, emergency psychiatry, drug overdose, pesticide poisoning, psychiatric diagnosis, schizophrenia, lethality, ICD-10, Japan, retrospective study, substance use disorder, geriatric psychiatry