LGBTQ+ young people are disproportionately affected by self-harm and suicide and are active social media users, yet little is known about their suicide-related online experiences. We analysed data from a quota-sampled survey of Australian social media users aged 15–25 (N = 885; LGBTQ+ n = 249). Compared with their non-LGBTQ+ peers, LGBTQ+ youth had over fourfold higher odds of exposure to self-harm and suicide-related content, and elevated odds of creating such content and seeking help on social media. More LGBTQ+ youth reported worsened mood and self-harm following exposure. These findings suggest that engagement-driven algorithms may create a feedback loop in which help-seeking amplifies harmful exposure. Enforceable platform accountability frameworks that reduce harmful exposure without limiting support pathways are needed.

Background Formulas for the extraction of continuous and binary effect sizes that are entered into a meta-analysis are readily available. Only some formulas for the extraction of count outcomes have been presented previously. The purpose of this methodological article is to present formulas for extracting effect sizes and their standard errors for studies of count outcomes with person-time denominators.Methods Formulas for the calculation of the number of events in a study and the corresponding person time in which these events occurred are presented. These formulas are then used to estimate the relevant effect sizes and standard errors of interest. These effect sizes are rates, rate ratios and rate differences for a two-group comparison and rate ratios and rate differences for a difference-in-difference design.Results Two studies from the field of suicide prevention are used to demonstrate the extraction of the information required to estimate effect sizes and standard errors. In the first example, the rate ratio for a two-group comparison was 0.957 (standard error of the log rate ratio, 0.035), and the rate difference was -0.56 per 100,000 person years (standard error 0.44). In the second example, the rate ratio for a difference-in-difference analysis was 0.975 (standard error of the log rate ratio 0.036) and the rate difference was -0.30 per 100,000 person years (standard error 0.42).Conclusions The application of these formulas enables the calculation of effect sizes that may not have been presented in the original study. This reduces the need to exclude otherwise eligible studies from a meta-analysis, potentially reducing one source of bias.

BACKGROUND AND OBJECTIVES:Medical regulators play an important part in protecting the public from harm. This includes receiving notifications about general practitioners (GPs) who are impaired in their ability to provide safe care because of a health condition. We sought to understand the prevalence and characteristics of health-related notifications to inform efforts to support GP wellbeing while protecting patient safety. METHOD:We linked 10 years of notifications to the Australian Health Practitioner Regulation Agency (Ahpra) with workforce demographics. We described characteristics of health-related notifications and calculated the odds of receiving a notification. RESULTS:GPs received 40% of notifications about doctors from 2012 to 2022. Health impairments represented 2.8% of notifications. The most frequent health- related concerns were drug or alcohol use for male and rural GPs, mental health for female GPs and cognitive issues for older GPs. Over a third (38.8%) of health-related notifications were raised by another health practitioner, and 17.3% led to restrictions on practice. DISCUSSION:Health-related concerns are a small proportion of notifications about GPs but can have serious consequences. Targeted interventions may support GP wellbeing.

This study investigates the epidemiology of adolescent suicide in India, addressing the limited research on the subject. Data on adolescent suicide (14-17 years) by sex and state were obtained from the National Crimes Records Bureau for 2014-2019, which included acquiring unpublished data from 2016 to 2019. Crude suicide rates for the period 2014-2019 were calculated by sex and state. Rate ratios (RRs) by sex and state were also calculated to assess changes over time, comparing suicide rates from 2017-2019 to 2014-2016. Female adolescent suicide rates, which ranged between 9.04 and 8.10 per 100,000 population, were consistently higher than male adolescent suicide rates, which ranged between 8.47 and 6.24 per 100,000 population. Compared to the first half of the study period (2014-2016), adolescent suicide rates significantly increased between 2017 and 2019 among less developed states (RRs = 1.06, 95% uncertainty interval [UI] = 1.03-1.09) and among females in these states (RRs = 1.09, 95% UI = 1.05-1.14). Male suicide rates aligned with global averages, while female rates were two to six times higher than in high-income and Southeast Asian countries. Findings highlight the urgent need for comprehensive surveillance and targeted suicide prevention strategies to address this critical public health issue.

Background Restricting access to means is a highly effective suicide prevention strategy for some methods. We evaluated the effectiveness of nets installed at the Golden Gate Bridge to prevent suicides by jumping at this site. Methods We used Poisson regression analyses to model suicide before, during and after the installation of safety nets at the Golden Gate Bridge between January 2000 and December 2024. We also modelled the number of times a third party intervened with someone showing signs of imminent suicide risk on the bridge. Results There were 681 suicides at the site. There were 2.48 suicides per month before installation of the safety nets, 1.83 during installation and 0.67 after installation. During the installation of the nets, suicides declined by 26% (rate ratio (RR)=0.74, 95% CI 0.60 to 0.90) and after installation by 73% (RR=0.27, 95% CI 0.13 to 0.54). There were 2901 instances where a third party intervened, 8.22 per month before installation, 14.42 during installation and 11.00 after installation. The number of interventions by a third party increased during installation by 75% (RR=1.75, 95% CI 1.62 to 1.90) and after installation by 34% (RR=1.34, 95% CI 1.12 to 1.60). Conclusions The early evidence indicates the installation of safety nets on the Golden Gate Bridge is associated with an immediate and substantial reduction in suicides at the site. This finding highlights the value of installing nets on this bridge and the importance of barriers as a strategy to prevent suicides by jumping.

In Australia and internationally, it is men who predominately die by suicide. Men are less likely than women to seek help for their mental health difficulties and this may contribute to their higher suicide rates. We created a 4-minute music video encouraging Australian men to seek help for mental health difficulties (Boys Do Cry). We aimed to assess in a randomised controlled trial (RCT) whether the Boys Do Cry video increased men’s intentions to seek help for mental health difficulties from baseline (T1) to post-intervention (1 week = T2). We conducted an online single-blind, two-arm RCT comparing the effects of Boys Do Cry against a control video. Analyses were undertaken on an intent-to-treat basis using linear mixed effects models with variables for trial arm, occasion of measurement and their interaction. Intervention effectiveness was assessed by comparing the mean difference between arms in change of the total score on the General Help-Seeking Questionnaire (GHSQ) from T1 to T2. 476 participants were randomised (intervention = 243; control = 233). At T1, GHSQ means were similar (intervention = 45.28; control = 45.70). After viewing the videos, GHSQ means increased in both arms (intervention = 47.33; control = 46.59), with no evidence of a difference in scores at T2 (modelled mean difference = 0.62, 95

OBJECTIVE:There is concern that suicides in Australia have been increasing, especially among young women. Our aim was to describe the long-term trends in suicide rates, identifying any change points among all suicides and among age-sex population groups. METHOD:We extracted annual Australian suicide data from 2000 to 2022 from the General Record of Incidence of Mortality. We used join point regression to model change points in age-standardised suicide rates over time for the total population, males, females and for eight age-sex strata. RESULTS:Age-standardised population suicide rates declined by 4.4% per year (95% confidence interval: -6.1 to -2.6) between 2000 and 2005 and then rose by 2.0% per year until 2018 (95% confidence interval: 1.4 to 2.5). No change was observed after this. Similar patterns were observed for males and females. When disaggregated by age and sex, a more complex picture emerged. Suicides declined in the early part of the century for some groups but not others. Most age-sex groups experienced an increase in suicides from the mid-to-late 2000s. Suicides declined between 2020 and 2022 for 15- to 24-year-old men by 16.0% per year (95% confidence interval: -24.0 to -7.4). In all other age-sex groups, suicides either plateaued or continued to rise. CONCLUSION:The broad trends in suicides observed in the total population and in males and females mask more complex patterns occurring in some age and sex groups. Understanding these long-term patterns is critical to informing interventions to reduce suicide.

BACKGROUND:There is some evidence that restricting access does not lead to displacement effects at other sites. However, most of these studies do not consider the similarities and the distances between intervention and comparison sites. We aimed to examine the effectiveness of a barrier in preventing suicidal behaviours at one bridge and its impact on nearby bridges with similar environments. METHODS:We used eight-year annual suicide deaths and emergency service call-outs for suicidal events at one bridge in South Korea where a safety barrier was installed in December 2016 and four nearby bridges. We estimated rate ratios to evaluate the effectiveness at the intervention bridge only, and the impact of this intervention at the four comparison bridges, The pre-intervention period was January 2013 to December 2016, and the post-intervention period was January 2017 to December 2020. FINDINGS:A total of 287 suicides and 2245 emergency call-outs for suicidal events were utilised for analyses. Restricting access to means at the intervention bridge was effective in reducing suicides at the intervention bridge (RR: 0.37, 95 % CI: 0.26-0.54) and was not associated with an increase at the four nearby comparison bridges (pooled RR (pRR): 1.37, 95 % CI: 0.99-1.90). In the post intervention period, there was an increase in call-outs at the four comparison bridges (pRR: 1.47, 95 % CI: 1.14-1.90), whereas there was a decrease in call-outs at the intervention bridge (RR: 0.89, 95 % CI: 0.80-0.99). INTERPRETATION:Our findings underscore the need for initiatives to streamline a broader approach in site-based intervention, by not only considering intervention sites but also similar sites in close proximity.

Rates of self-harm and suicide appear to be increasing in young people and many attribute this to social media use. However, high quality studies examining young people’s experiences of self-harm and suicide-related content on social media, and the impact on wellbeing, are lacking. An online national cross-sectional survey was conducted between January and March 2024. Quota sampling was used. Participants from across Australia were recruited from the Roy Morgan Single Source Panel, a panel managed by Pureprofile and via snowball sampling. Descriptive statistics were used to examine respondents’ experiences; logistic regressions examined differences between young people and adults. Three thousand five hundred forty-nine individuals (895 young people; 2,654 adults) completed the survey. Just over half had been exposed to self-harm or suicide-related content on social media. Young people were more likely to be exposed than adults (Adjusted OR 3.81; 95

Suicidal ideation is a critical risk factor and intervention target in suicide prevention research, yet its descriptive reporting remains inconsistent. Most studies report prevalence rates or mean scores, often omitting the conditional mean, the average score specifically among those experiencing suicidal ideation. This limits our ability to assess the true severity of ideation in affected populations and hampers cross-study comparability. We propose a standardised approach to descriptive reporting incorporating three key metrics: overall mean, conditional mean, and prevalence rate. These metrics provide a more comprehensive picture of the extent, severity, and prevalence of suicidal ideation. Where only categorical responses are available, we recommend reporting both prevalence and the full breakdown of response categories to preserve interpretability. The adoption of these standards would improve the precision of meta-analyses, enable more meaningful comparisons across studies, and support the development of targeted, evidence-based interventions. We call on researchers, journal editors, and reviewers to implement this reporting standard to strengthen the empirical foundation of suicide prevention efforts.

OBJECTIVE:To examine factors associated with the choice of public location over home to die by suicide. METHODS:This study used a case-control design. Data on suicides that occurred between 2001 and 2017 in Australia were extracted from the National Coronial Information System. Cases were suicides that occurred in public places and controls were suicides that occurred at home. Logistic regression models were used to estimate the associations between suicide location and several sociodemographic variables, depending on whether hotel rooms were included in or excluded from public places. RESULTS:In total, 25.2% of 42,656 suicides occurred in public places including hotel rooms, 69.3% at home, and 5.4% in nonpublic places other than at home (e.g., inpatient ward or correctional facilities). Excluding suicides in hotel rooms from public places, 1.6% points of suicides in public places moved to nonpublic places other than at home. In multivariable regression models regardless of scenarios, males (compared with females) had higher odds of dying by suicide in public places, while those who were divorced/separated/widowed (compared with married people), those who were older (aged 30-54, and aged 55 and above, compared with under 30 years), and those who were unemployed or not in the labor force (compared with employed people) had lower odds of suicide in public places. CONCLUSION:The findings should be used to inform the design of strategies to prevent suicides in public places.

BACKGROUND:Children and adolescents exposed to the youth justice system have poor health profiles, but little is known about their subsequent mortality. We aimed to examine mortality outcomes in a large, state-wide cohort of young people in Australia who had contact with the youth justice system. METHODS:We linked youth justice records in the state of Queensland, Australia from July 1, 1993, to June 30, 2014, with adult correctional records and the National Death Index, for records up to Jan 31, 2017. We calculated all-cause and cause-specific crude mortality rates per 100 000 person-years, and age-standardised and sex-standardised mortality ratios with 95% CIs. Calculations were performed for the whole cohort and in subgroups defined by sex, Indigenous status, and youth justice history. We used survival analysis to identify demographic and criminal justice factors associated with all-cause mortality. FINDINGS:Of 49 011 individuals in the study sample, 321 were excluded due to data linkage or data quality issues and 20 were excluded as they did not have an age or date of birth recorded, which resulted in 48 670 (99·3%) participants. 11 897 (24·4%) participants were female, 36 773 (75·6%) were male, and 13 250 (27·2%) were Indigenous. During a median of 13·5 years (IQR 8·4-18·4) of follow-up, we observed 1431 (2·9%) deaths among the 48 670 participants. Median age at end of follow-up was 28·6 years (IQR 23·6-33·6). The most common causes of death were suicide (495 [34·6%]), transport accidents (244 [17·1%]), and accidental drug poisoning (209 [14·6%]). The all-cause crude mortality rate was 218·9 deaths (95% CI 207·9-230·6) and the all-cause standard mortality ratio was 4·2 (3·9-4·4). In multivariable analyses, mortality rates were higher for males (adjusted hazard ratio [aHR] 1·5 [95% CI 1·3-1·7]); those who had been subject to community supervision (aHR 1·3 [1·1-1·5]), or detention (aHR 2·1 [1·8-2·4]) versus charge only; and those under adult correctional supervision in the community (aHR 1·9 [1·5-2·4]) versus unsupervised. More than half of the observed deaths occurred before 25 years of age, and very few (1·6%) occurred in custody. INTERPRETATION:Justice-involved young people are at markedly increased risk of premature death from largely preventable causes. Reducing the burden of preventable death among these young people will require coordinated, multi-sectoral responses that extend beyond the criminal justice system. FUNDING:National Health and Medical Research Council, Australia.

BACKGROUND:This update outlines amendments to the MAPSS trial protocol in response to both the COVID-19 pandemic and broader feasibility considerations. While many changes were necessary to navigate pandemic-related disruptions-such as school closures and remote learning-others were made to improve feasibility in the school-based setting, independent of COVID-19. The protocol was updated to align with public health guidelines and general school practices, ensuring feasibility for continued school participation and increased support and flexibility for school communities. METHODS:Key adjustments included changes to participant-facing documentation, including implementing digital consent processes, extending timelines for participant recruitment and participation, and enhancing remote engagement strategies. Modifying outcome measures and risk management protocols, including adverse event reporting, ensured both participant safety and data continuity. Lastly, new study measures were added, such as a custom-designed questionnaire to assess study acceptability. Despite the updates, the core trial design, eligibility criteria, and primary outcomes remained unchanged. CONCLUSION:These protocol amendments reflect the pragmatic challenges of conducting school-based mental health research during a global pandemic and as part of ongoing efforts to enhance feasibility. The adjustments enabled the trial to proceed safely, prioritizing participant engagement and maintaining alignment with public health measures. TRIAL REGISTRATIONS:Australian New Zealand Clinical Trial Registry, ACTRN12621000279820, originally registered on 12 March 2021, and currently awaiting update approval. Australian New Zealand Clinical Trial Registry, ACTRN12621000770864, originally registered on 21st June 2021, and updated on 16th February 2025.

BACKGROUND:There has been rapid expansion in the development of machine learning algorithms to predict suicidal behaviours. To test the accuracy of these algorithms for predicting suicide and hospital-treated self-harm, we undertook a systematic review and meta-analysis. The study was registered (PROSPERO CRD42024523074). METHODS AND FINDINGS:We searched PubMed, PsycINFO, Scopus, EMBASE, IEEE, Medline, CINALH and Web of Science from database inception until 30 April 2025 to identify studies using machine learning algorithms to predict suicide, self-harm and a combined suicide/self-harm outcome. Studies were included if they examined suicide or hospital-treated self-harm outcomes using a case-control, case-cohort or cohort study design. Studies were excluded if they used self-reported outcomes or examined outcomes using other study designs. Accuracy was assessed using statistical methods appropriate for diagnostic accuracy studies. Fifty-three studies met the inclusion criteria. The area under the receiver operating characteristic curves ranged from 0.69 to 0.93. Sensitivity was 45%-82% and specificity was 91%-95%. Positive likelihood ratios were 6.5-9.9 and negative likelihood values were 0.2-0.6. Using in-sample prevalence values, the positive predictive values ranged from 6% to 17%. Using out-of-sample prevalence values at an LR+ value of 10, the positive predictive value was 0.1% in low prevalence populations, 17% in medium prevalence populations and 66% in high prevalence populations. The main study limitations were the exclusion of relevant studies where we could not extract sufficient information to calculate accuracy statistics and between-study differences in the follow-up time over which the outcomes were observed. CONCLUSIONS:The accuracy of machine learning algorithms for predicting suicidal behaviour is too low to be useful for screening (case finding) or for prioritising high-risk individuals for interventions (treatment allocation). For hospital-treated self-harm populations, management should instead include three components for all patients: a needs-based assessment and response, identification of modifiable risk factors with treatment intended to reduce those exposures, and implementation of demonstrated effective aftercare interventions.

OBJECTIVE:The objective was to examine response rates, types of callers and their probability of being answered, prevalence of at-risk callers, and to calculate national call rates. METHODS:Data on all calls to the Danish, national telephone helpline for suicide prevention during July 2019 to December 2022 were analysed. A measure of unique calls was developed to account for repeat calls not being answered. We examined the probability of calls being answered by caller types using logistic regression and calculated national call rates for individuals aged ⩾15 years. RESULTS:Overall, 526,533 calls were made by 31,317 individuals, and 131,621 unique calls were identified, of which 48.9% were answered. First-time callers (95.1%) accounted for 5.7% of calls. We found that 0.1% of callers accounted for 61.8% of all calls. This group of daily callers (>1000 calls each year) consisted of 8-12 unique callers and was more likely to be answered (odds ratio = 24, 95% confidence interval = [23, 25] vs first-time callers), often hung up (49.1% vs first-time callers: 4.4%), and received 33.0% of the total counselling time. The yearly national call and caller rates were 893 calls and 212 unique callers per 100,000 inhabitants, respectively. CONCLUSIONS:Correcting for repeated unanswered calls provided an informative estimate of the response rate. The call distribution was highly skewed; a small group of daily callers accounted for most calls and were more likely to be answered. These callers frequently hung up before a conversation was initiated. National call rates facilitate cross-country comparisons.

BACKGROUND:Several countries have implemented National Suicide Prevention Strategies (NSPSs), some of which include specific targets relating to reductions in suicide and/or suicidal behaviour. This study aimed to: (1) identify the proportion of NSPSs that include targets; and (2) assess whether these targets were achieved. METHODS:A comprehensive review of NSPSs was conducted using relevant databases to identify those with defined targets. Official agencies were consulted to verify target achievement and to gather data on suicidal behaviour. RESULTS:We identified 54 national strategies for analysis, of which 27 (50 %) included specific, quantitative targets. Eleven of these strategies had passed their target date and had available data for assessment. Of these eleven, only two countries, Chile and Israel, achieved their targets. For the remaining nine, outcomes varied: non-statistically significant reductions were observed in Bulgaria and Panama. Ireland also saw a reduction that, while numerically surpassing its target, was not statistically significant. Rates remained stable in England, and increased in Northern Ireland, the United States, Scotland, and Uruguay. LIMITATIONS:Although the data were sourced from official records, potential biases may have affected the quality and accuracy of the analysis. CONCLUSIONS:A significant proportion of NSPSs lack specific targets and even among those that have specific targets, achieving these remains a challenge for many. However, the presence of specific targets could at least facilitate the identification of barriers hindering the attainment of these targets.

The Suicide Stigma Assessment Scale is a 44-item measure of three components of public suicide stigma: stereotypes, prejudice, and discrimination. We examined the validity and reliability of the Suicide Stigma Assessment Scale for use with young adults in the Indian context. The study comprised translation and back-translation, cognitive interviews, and assessments of content validity. Scale reliability and confirmatory factor analyses were undertaken using a sample of 375 participants aged 18-29 years drawn from the general population in Chennai, India. Cronbach alpha scores indicated that the three scales (stereotypes [.76], prejudice [.71], and discrimination [.88]) each had acceptable internal consistency. All three scales had acceptable scale-level content validity ratings, and the majority of individual items (n = 33, 75%) were rated as "quite relevant" or "very relevant" by at least four of the five panel members. However, 11 items had poor item-level content validity ratings. Confirmatory factor analyses yielded mediocre model fit parameters for each of the three scales, indicating a poor fit. Further examination of internal consistency and confirmatory factor analyses at the subscale level indicated that the most promising subscales were the Weak, Distressed, Anger, and Avoidance subscales. The Suicide Stigma Assessment Scale is a novel scale developed in the United States for measuring public suicide stigma. While some of the subscales were promising, our results suggest the current full scale in its current form requires further investigation and adaptation to be a good cultural fit for use in India.