https://orcid.org/0000-0003-2423-3784
Figures
Abstract
Objective
The majority of suicide decedents have had contact with health services close to their death. Some of these contacts include admissions to hospitals for physical and mental health conditions, injury and intentional self-harm. This study aims to establish and quantify the risks of suicide following hospital admission for a range of mental and physical illnesses.
Methods
A retrospective analysis was carried out on existing morbidity and mortality data in Victoria. Data was extracted from the Victorian Admitted Episodes Dataset and the Victorian Suicide Register. Unplanned hospital admissions among adult patients (> = 15 years of age), discharged between 01 January 2011 and 31 December 2016 (2,430,154 admissions), were selected. Standardised Mortality Ratios were calculated for conditions with at least five linked suicides within one year of discharge from hospital.
Results
Forty-three conditions defined at the three-digit level of the International Statistical Classification of Diseases and Related Health Problems 10th Revision, were associated with at least five subsequent suicides (within one year of hospital discharge); 14 physical illnesses, 5 symptoms, signs and abnormal clinical and laboratory findings, 12 mental health conditions, and 12 types of injury and poisonings. The highest Standardised Mortality Ratios were for poisonings (range; 27.8 to 140.0) and intentional self-harm (78.8), followed by mental health conditions (range; 15.5 to 72.9), symptoms, signs and abnormal clinical and laboratory findings (range; 1.4 to 43.2) and physical illnesses (range; 0.7 to 4.9).
Conclusions
Hospital admissions related to mental health conditions and injury and poisonings including self-harm were associated with a greater risk of suicide than physical conditions. Mental health conditions such as depressive episodes, personality disorders and psychotic episodes, injuries caused by intentional-self-harm and poisonings by certain types of drugs, carbon monoxide and hormones such as insulin can be prioritised for targeting suicide prevention initiatives for persons discharged from hospitals.
Citation: Fernando DT, Clapperton A, Berecki-Gisolf J (2022) Suicide following hospital admission for mental health conditions, physical illness, injury and intentional self-harm in Victoria, Australia. PLoS ONE 17(7): e0271341. https://doi.org/10.1371/journal.pone.0271341
Editor: Weihua YUE, Peking University, Institute of Mental Health, CHINA
Received: October 3, 2021; Accepted: June 28, 2022; Published: July 11, 2022
Copyright: © 2022 Fernando et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data cannot be shared publicly because of confidentiality reasons. The data used in this study are de-identified, but however are re-identifiable as it does contain, sex, ethnicity, location etc. among many other specific variables. Furthermore, the databases are linked, which also gives access to more identifiable information. They include personal health data as well as sensitive data (information on suicide) which is not meant for public sharing. The data is not owned or collected by the research team, it was provided by the relevant government department data custodians. The data underlying the results presented in the study are available from the Centre for Victorian Data Linkage (https://www2.health.vic.gov.au/about/reporting-planning-data/the-centre-for-victorian-data-linkage). The authors confirm they did not have any special access privileges that other researchers would not have.
Funding: D Fernando and J Berecki-Gisolf were supported by the Department of Health, State Government of Victoria The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. A Clapperton was supported by an Australian Rotary Mental Health Research Grant. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Suicide is a major public health concern globally, with an estimated 800,000 suicides occurring worldwide every year [1]. In 2018, there were 3,046 suicides registered in Australia which relates to 8 deaths per day [2]. Recent statistics published by the Coroner’s Court of Victoria showed that there were 2,787 suicides in the state of Victoria between 2016 and 2019, which averages to 697 suicides per year [3].
Studies in Western, developed countries have consistently shown that a high proportion of people who die by suicide have a diagnosed, or diagnosable mental health condition at the time of death [4–7]. In addition to mental health conditions, physical illness has consistently been associated with suicide. A recent study found that physical illness was indicated as the main reason for suicide in almost 20% of cases [8], and another study has shown a link between hospital admissions for constipation, gastritis, diabetes, alcoholic liver disease, pneumonia and urinary tract infections and increased risk of suicide [9]. Studies have also shown that certain injuries such as traumatic brain injury are likely to increase the risk of suicide [10, 11], and non-fatal suicide attempts are also strongly correlated with subsequent suicide [12–14].
The majority of people who die by suicide have had contact with health services (i.e., general practitioners, outpatient or inpatient services) proximal to their death [15–20]. These contacts represent potential suicide prevention opportunities [16], through identification of persons at risk of suicide, provision of treatment and/or delivery of interventions.
The main aim of this study is to establish and quantify risks of suicide following hospital admissions for a range of mental health conditions, physical illnesses, injury and intentional self-harm in Victoria.
Methods
Study design
A retrospective analysis of existing morbidity and mortality data in the state of Victoria, Australia was carried out for this study.
Data sources and linkage
Morbidity data was extracted from the Victorian Admitted Episodes Dataset (VAED) and mortality data was extracted from the Victorian Suicide Register (VSR) and the Victorian Death Index (VDI).
The VAED is a unit record data file maintained by the Department of Health and Human Services on all public and private hospital admissions in Victoria, which consists of patient demographics and morbidity information. Morbidity information includes 40 fields related to disease, injury and external causes, coded to the International Statistical Classification of Diseases and Related Health Problems Tenth Revision, Australian Modifications (ICD-10-AM) [21].
The VSR is an ongoing register established by the Coroners Court of Victoria and contains information on all suspected and coroner-determined suicides reported to the Victorian Coroners Court [22].
The VDI contains all death-records for Victoria, and is obtained from the Registrar of Births, Deaths and Marriages in Victoria.
Data provision and linkage was undertaken by the Centre for Victorian Data Linkage (CVDL). Data linkage performed by CVDL used patient-specific identifiers and deterministic data linkage. The Coroners Court of Victoria (CCOV) supplied VSR data to CVDL via secure data exchange portals. CVDL linked the VSR data with the VAED and VDI, and all data were de-identified. The de-identified datasets, containing a linkage identification number, were transferred to researchers at Monash University using the secure data exchange portal.
Study population
Inclusion criteria.
The study population consisted of adults (> = 15 years of age) with unplanned hospital admissions in Victoria, Australia, discharged between 01 January 2011 and 31 December 2016 (3,640,529 episodes) as recorded in the VAED. They were followed up until 31 December 2017 in the VSR and VDI. Consecutive records of inward transfers from other hospitals or statistical separations (changes in care type) within the same hospital were considered to be part of one period of care (POC). All initial POCs, that is the POC that included the index (initial) admission for each patient were retained for final analysis. In cases where the patient was newly admitted subsequent to the initial POC, these records were also retained and included as unique POCs, provided the subsequent admission took place at least one-year post discharge of the previous admission.
Exclusion criteria.
Planned admissions were excluded as some of these were for diagnostic investigations (which are not always indicative of the presence of a particular disease), and the rest were ‘elective’ medical or surgical treatment. These also include kidney patients on routine dialysis treatments (with multiple admissions) and cancer patients on chemo therapy etc. (with multiple admissions), which does not allow for ideal incidence estimation. The following were also excluded: POCs ending with a death in hospital (because the outcome-of-interest was death following discharge), hospital admissions of intersex people (to preserve confidentiality of small cell counts: recorded intersex status was rare), admissions with missing principal diagnosis codes (205 admissions) and admissions with unrealistic time to death (i.e., admission dates occurring after the date of death: 951 records (0.03%), which could be due to data linkage inaccuracy). The final number of episodes retained for the analysis was 2,430,154.
Outcome measure
The study outcome measure was suicide within one year of hospital discharge. The rationale for starting the 1-year follow-up at discharge (and not at admission) was to capture only post-discharge exposure: suicide during the hospital admission was therefore excluded.
Selection of conditions
The conditions considered for this study were physical illnesses (acute and/or chronic); injury; symptoms, signs and abnormal clinical and laboratory findings; and mental health conditions. All conditions were analysed at the three-digit level in the ICD-10-AM ranging from A00-T79 appearing in the principal diagnosis field at admission. Conditions were included if they were associated with at least five suicides within one year of discharge from hospital (43 conditions fulfilled these criteria).
Due to known associations between intentional-self harm and subsequent suicide, admissions for intentional self-harm were also examined. Intentional self-harm admissions were selected if the external cause codes "X60" to "X84" (intentional self-harm) appeared in any of the 40 diagnosis fields in the VAED—–this was necessary as intentional self-harm cannot be coded as a principal diagnosis.
Independent variables
Patients were classified into three age groups (15–24, 25–64, > = 65 years), two gender groups (male and female), three geographic region groups (metropolitan, regional/rural and unknown/interstate/overseas) and four marital status groups (never married, married/de facto, divorced/separated/widowed and not stated). This information was sourced from the hospital admissions data. Socio economic status was classified as per the Socio-Economic Indexes for Areas (SEIFA) [23], based on the patient’s area of residence. The specific SEIFA used in this study was the Index of Relative Socio-Economic Advantage and Disadvantage; with state deciles based on local government areas; decile one indicates relatively the greatest disadvantage and least advantage.
Statistical methods
Descriptive analysis of admissions linked to a suicide within one year was carried out using means, medians and proportions. The main measure used to assess the ‘risk’ of mortality within one year of hospital discharge in the study population was the Standardised Mortality Ratio (SMR). The SMRs compare the observed suicide mortality in the study population in each diagnosis group with the expected suicide mortality in the general population. Suicide mortality rates for the general population was computed using suicide data from the VSR and population data from the Australian Bureau of Statistics [24]. SMRs were calculated using the indirect method adjusting for age (5-year groups) and gender [25].
Survival analysis was carried out on conditions with high SMRs (SMR>5) to assess for competing risks due to non-suicidal deaths. Survival analysis was limited to index POCs. Those alive at one year were censored. Analysis was carried out with and without accounting for competing risks (i.e., death due to causes other than suicide) using the cumulative incidence function method as defined by Fine and Gray [26] and the Cox proportional hazards model, respectively. The independent variables age group, gender, marital status, geographic region and SEIFA were assessed for association with the outcome of suicide using a stepwise forward regression; all except SEIFA were significant to the outcome (assessed using the Akaike Information Criterion [27]). All tests were carried out at the 5% level of significance. SAS software, Version 9.4 [28] and Stata 16.0 (StataCorp) [29] was used to analyse the data.
The study was approved by the Monash University Human Research Ethics Committee (Project no: 14647). Historical administrative data was used. Name, date of birth and other identifiers were removed from the dataset by the data custodians prior to release of the data to the researchers. Due to the magnitude of the dataset, it was impractical to obtain consent.
Results
Overview of study population
A total of 1,737,743 patients with 2,430,154 hospital admissions during the study period 2011 to 2016 were selected for analysis; more than half of patients were aged 25–64 years (56.6%), and 60% were female. Seventy-one conditions related to injury, mental and physical illness were identified at the three-digit level of the ICD-10-AM from these hospital admission records.
Conditions associated with at least five suicides
Forty-three of the 71 conditions were linked to at least five suicides in the 1-year follow-up period: these consisted of 14 physical illnesses, 5 symptoms, signs and abnormal clinical and laboratory findings, 12 mental health conditions, and 12 types of injury and poisonings. Selecting only hospital admissions with one of these 43 conditions appearing in the principal diagnosis field resulted in a sub-set of 756,007 admissions (by 538,910 patients) (Table 1). The most common mental health conditions were disorders due to the use of alcohol, and schizophrenia. The most common physical conditions were pneumonia, acute myocardial infarction and cellulitis. Among symptoms, signs and abnormal clinical and laboratory findings, pain in throat and chest, and abdominal and pelvic pain were the most common conditions. Among injury and poisonings, injuries to the head, and wrist and hand were most common.
Suicides
Over the five-year study period, 756,007 hospital admissions related to the selected 43 conditions (conditions with at least five suicides) were associated with a total of 631 suicides within one year of hospital discharge (Table 2). The mean and median ages of those who died by suicide following hospital admission for the selected 43 conditions were approximately 44 and 43 years respectively, and males accounted for 67.2%.
Schizophrenia, depressive episodes, reaction to severe stress and adjustment disorders, throat and chest pains, nonopioid poisonings, poisoning by antiepileptic, sedative-hypnotic and antiparkinsonism drugs and poisoning by psychotropic drugs were some of the conditions associated with the highest numbers of suicides (>30) during the follow-up period.
SMRs were calculated for each of the 43 conditions: poisonings had the highest SMRs ranging from 27.8 (CI 14.1–49.5) for poisoning by narcotics and psychodysleptics to 140 (CI 61.2–276.9) for carbon monoxide poisoning; 81% of these were intentional-self-harm incidents. SMRs related to mental and behavioural disorders followed, ranging from 15.5 (CI 10.4–22.3) for mental and behavioural disorders due to use of alcohol to 72.9 (CI 58.7–89.6) for depressive episodes (Fig 1). Admissions with a diagnosis code indicating depressive episodes, that also had external cause codes indicating intentional self-harm, had double the SMR (SMR = 150.0 (95% CI 38.2–408.2)) compared to admissions with depressive episodes overall (SMR = 72.9 (95% CI 58.7–89.6)) (not shown in tables). Of the physical illnesses, gastritis had the highest SMR of 4.9 (CI 2.2–9.8). Among Symptoms, signs and abnormal clinical and laboratory findings, admissions related to Symptoms and signs involving emotional state had a high SMR of 43.2 (CI 25.6–68.7) with 50% of these coded to suicidal ideation.
Intentional self-harm
Out of the 43 selected conditions (conditions with > = 5 suicides), external cause codes were particularly relevant to poisoning-related admissions, which were mostly coded as caused by intentional self-harm, and intracranial injuries, which were mostly coded as caused by assaults (not shown in tables). Since intentional self-harm is known to be associated with suicide, self-harm related admissions were selected and further analysed. One-hundred and sixty-seven suicides were associated with intentional self-harm related admissions in the year prior to suicide (Table 3). It is important to note that there is an overlap with the numbers presented in Tables 2 and 3; i.e., of the 167 suicides presented in Table 3, the majority (n = 127) are also included in Table 2. The mean and median age of those who died by suicide following self-harm related hospital admission was approximately 41 years, and 41.3% of those who died were female.
Death by other causes
Death due to other causes pose a competing risk for suicide. To ascertain the impact of competing risks, survival analysis was carried out for 23 conditions (those with an SMR > = 5). The baseline variables considered to have an association with increased hazard of suicide death were age-group, gender, marital status and geographic region. Due to the small number of suicides associated with most conditions, statistically reliable conclusions could not be drawn for most of them. Proportions of death by other causes, and details of the survival analysis is provided in the Table A and B in S1 File. Overall, adjusting for competing risks had no significant impact on the estimated likelihood of suicide for any of the conditions.
Statistical model outputs can be found in the S2 File.
Discussion
This study aimed to assess the risk of suicide following hospital admission for mental health conditions, physical illnesses, injury and poisoning and specifically for intentional self-harm. Standardised Mortality Ratios were used to assess the risk of suicide for each health condition following hospital-discharge relative to the age- and sex-matched suicide risk in the general population. Poisonings (including those resulting from intentional self-harm) had some of the highest SMRs, followed by mental health conditions. These were followed by symptoms and signs and abnormal clinical and laboratory findings involving cognitive functions and emotional states, while the lowest SMRs were for physical illnesses. Age, gender, marital status and geographic region of residence were confounders that were associated with selected conditions as well as the outcome of suicide.
The risk of suicide varied with the type of condition. Separation from hospital for treatment of a mental health condition was associated with a greater risk of suicide (high SMR) within the 1-year follow-up period than for a physical illness. This was expected given previous research has found many mental health conditions to be associated with suicide [30–32]. Similarly, injury and poisoning related hospital admissions also showed a greater risk of suicide in the 1-year follow-up period after discharge when compared to physical illnesses. Again, this is expected given that many of the injury and poisoning admissions were a result of intentional self-harm and there is a known link between suicide attempts and subsequent suicide [12–14]. This study found little, and in some instances no, increased risk of suicide for most of the selected physical illnesses; a similar finding was reported by Roberts, John, Kandalama, Williams, Lyons and Lloyd in a study of similar design [9] in the United Kingdom. Therefore, hospital admission for certain mental health conditions and injuries (in particular poisonings) are still priority groups for interventions; these could be focussed on high-risk groups identified at hospital discharge.
Mental health conditions are a known risk factor for suicide. Certain mental health conditions found to have high SMRs in this study were similar to those found to increase suicide risk in a recent meta-review [30]; these were substance use and schizophrenia, depressive episodes, personality disorders, and bipolar disorders. Further two other studies also found high relative risks for conditions such as depressive disorders, psychotic disorders, mood disorders, personality disorders, substance use disorders, and anxiety disorders [31, 32]. In the current study, suicidal ideation was also found as a secondary diagnosis among many of the admissions with a principal diagnosis of a mental health condition. Suicidal ideation and tendencies (ICD-10-AM code R45.81) are generally coded under Symptoms and signs involving emotional state in the ICD-10-AM (but excludes Signs and symptoms constituting part of a mental disorder) as a principal diagnosis. Mental health conditions could sometimes be masked and appear as Symptoms and signs involving emotional state, which could explain why symptoms and signs involving emotional state showed high SMRs in this study.
Certain types of injuries, including poisonings, are known to be risk factors for suicide. Certain poisonings observed in this study had high SMRs, indicating a high risk of subsequent suicide associated with discharge from hospital following a poisoning-related admission. This is expected as 81% of these poisoning cases resulted from intentional self-harm rather than unintentional incidents. Admissions for poisonings from benzodiazepines, analgesics and antipyretics such as 4-aminopheno derivatives (which commonly include paracetamol), antidepressants and antipsychotics were common among those who died by suicide within 1 year. This study confirms that hospital-admitted poisonings can be linked to subsequent suicide, and provides details regarding the types of poisonings that are associated with suicide, using Victorian hospital admissions data. The current study found carbon monoxide poisoning to have a high SMR. The Victorian government’s assertive outreach trials (HOPE program) target people treated in hospital for suicide attempts. People who have engaged in a suicide attempt using a method that was shown to have particularly high suicide risk in this study, for example carbon monoxide poisoning should be strongly encouraged to participate in these HOPE trials; people who present after carbon monoxide poisoning should be a priority for prevention. This has also been shown in a previous study by Hampson, Rudd and Hauff [33].
Among injuries other than poisonings, those with high SMRs were intracranial injuries (mostly concussive injury), other head injuries and open wounds of forearm. Open wounds of the forearm are likely indicative of intentional self- harm from cutting, which is a known risk factor for subsequent suicide [34]. This once again provides further support for known links between self-harm and suicide.
Certain physical illnesses are known to be associated with suicide. The physical illnesses with an SMR > 3 in this study were gastritis (some of which were alcohol related), constipation, and type 2 diabetes mellitus (SMR just under 3); the same three illnesses were found to have the highest SMRs in a study of similar design conducted by Roberts, John, Kandalama, Williams, Lyons and Lloyd [9] in the United Kingdom. Roberts, John, Kandalama, Williams, Lyons and Lloyd discussed that gastrointestinal diseases were often alcohol related (also seen in the current study) and constipation could be a side-effect of certain medications (e.g., opioids; constipation is also a known adverse effect for some pharmaceuticals to treat epilepsy, anaemia, pain, depression and other mental disorders). They noted that suicide risks were also linked to therapeutic medications such as those used for epilepsy and Parkinson’s disease (also seen in the current study). They noted that endocrine disorders that relate to poor blood glucose control (i.e. potentially indicative of poor chronic disease management) were also linked with suicide risk (also seen in the current study).
Strengths
This study was drawn from state-wide population-level databases, providing reliable estimates of suicide risk following discharge from hospital for a range of physical and mental conditions. Data linkage provided appropriate follow-up. It also provided estimates intrinsic to Victoria, where the physical conditions identified as potential predictors of subsequent suicides were slightly different to what was seen elsewhere in the world (e.g., England and Wales [9]). Locally relevant data is key to informing locally applicable interventions. The competing risk modelling carried out in this study has demonstrated that death due to other causes is not a significant competing risk for suicide.
Limitations
The SMRs computed in this study are limited to incidents serious enough to warrant hospital admission and do not include incidents that did not result in hospital admission.
Sample size was also a study limitation: larger samples lead to more comprehensive lists of conditions (48 conditions), as shown by Roberts, John, Kandalama, Williams, Lyons and Lloyd [9]. Comparatively, the current study consisted of a much smaller sample of admissions; however, 43 conditions were captured.
Selection of a condition is based on the fact that it was associated with at least five suicides. This could occur due to two reasons. The first is that the condition may be common (e.g., throat and chest pain); common conditions are therefore selected not because of a disproportionate risk of suicide, but because the overall case numbers are high, and eventually five suicides are likely to be observed. The second is that the condition can be associated with increased suicide risk. These underlying reasons for inclusion of conditions should be considered when drawing conclusions from this study.
The current study limited the selection of cases to those with an unplanned admission. This may result in loss of information on conditions that require ongoing hospital monitoring (e.g., kidney patients on dialysis). Furthermore, suicide risk among those with unplanned admissions may be different to suicide risk among those with planned admissions. Suicide risk among planned admissions is beyond the scope of this study. Future work in this area can expand the study scope and also include specific illnesses and treatments that fall within this category and retaining planned admissions.
Modelling techniques require a certain number of cases with the outcome (i.e., suicide, in this study) in order to produce reliable estimates of association between exposure variables and outcomes. Suicide is a rare event. In this study, for many of the physical or mental health conditions, there were insufficient numbers of cases with a suicide outcome to provide reliable estimates for the survival analysis that was performed. In interpreting the results, it should be noted that the absence of a statistically significant result does not necessarily demonstrate the absence of a correlation: for some conditions, case numbers were insufficient for statistical testing. Replicating this study in a larger population (e.g. a national study) would help to address this study limitation.
Conclusions
Specified mental health conditions and types of injury and poisonings can be prioritised for targeting suicide prevention initiatives for persons discharged from hospitals: hospitals can therefore serve as a service contact point as well as a suicide risk identification platform. Among mental health conditions, priority can be given to depressive episodes, followed by personality disorders and psychotic episodes. Among poisonings, priority can be given to poisoning by antiepileptic, sedative-hypnotic and antiparkinsonism drugs, poisoning by nonopioid analgesics, antipyretics and antirheumatics, poisoning by unclassified psychotropic drugs, carbon monoxide poisoning and poisoning by hormone pharmaceuticals such as insulin. Intentional self-harm is an injury type that can also be prioritised. Staff working in clinical mental health services and emergency departments, could be advised of the findings relating to high suicide risk associated with particular methods or conditions, to assist with targeted suicide prevention.
Supporting information
S1 File. Deaths by other causes and survival analysis.
https://doi.org/10.1371/journal.pone.0271341.s001
(DOCX)
Acknowledgments
The authors would like to acknowledge the Victorian Department of Health as the source of VAED data and the Coroner’s Court of Victoria as the source of VSR data, and the Centre for Victorian Data Linkage for providing linkage services. A special thank you to Penny Tolhurst at the Department of Health (Victoria). The authors would also like to acknowledge Dr. Stephen Roberts at Swansea University, United Kingdom for sharing their experience from a similar study.
References
- 1.
World Health Organisation. Mental health: Suicide data 2020 [cited 2020 14 October 2020]. Available from: https://www.who.int/mental_health/prevention/suicide/suicideprevent/en/.
- 2.
Australian Institute of Health and Welfare. Suicide and intentional self-harm Canberra: AIHW; 2020 Available from: https://www.aihw.gov.au/reports/australias-health/suicide-and-intentional-self-harm.
- 3.
Coroner’s Court of Victoria. Monthly Suicide Data Report; Report 1. CCOV, Australia: CCOV; 2020 27 August 2020.
- 4. Isometsä ET. Psychological autopsy studies—a review. Eur Psychiatry. 2001;16(7):379–85. pmid:11728849
- 5. Bertolote JM, Fleischmann A. Suicide and psychiatric diagnosis: a worldwide perspective. World Psychiatry. 2002;1(3):181–85. pmid:16946849
- 6. Bertolote JM. Suicide and mental disorders: do we know enough? Br J Psychiatry. 2003;183(5):382–3.
- 7. Bertolote JM, Fleischmann A, De Leo D, Wasserman D. Psychiatric diagnoses and suicide: revisiting the evidence. Crisis. 2004;25(4):147–55. pmid:15580849
- 8. Fegg M, Kraus S, Graw M, Bausewein C. Physical compared to mental diseases as reasons for committing suicide: a retrospective study. BMC Palliat Care. 2016;15(1):14. pmid:26860949
- 9. Roberts SE, John A, Kandalama U, Williams JG, Lyons RA, Lloyd K. Suicide following acute admissions for physical illnesses across England and Wales. Psychol Med. 2018;48(4):578–91. pmid:28714426
- 10. Bryant RA, O’Donnell ML, Forbes D, McFarlane AC Silove D, Creamer M. The course of suicide risk following traumatic injury. J Clin Psychiatry. 2016;77(5):648–53. pmid:27249074
- 11. Teasdale TW, Engberg AW. Suicide after traumatic brain injury: a population study. J Neurol Neurosurg Psychiatry. 2001;71(4):436–40. pmid:11561024
- 12. Van Orden KA, Witte TK, Cukrowicz KC, Braithwaite SR, Selby EA, Joiner TE. The interpersonal theory of suicide. Psychol Rev. 2010;117(2):575–600. pmid:20438238
- 13. Moscicki EK. Epidemiology of Suicidal Behavior. Suicide Life Threat Behav. 1995;25(1):22–35. pmid:7631373
- 14. Mościcki EK. Identification of suicide risk factors using epidemiologic studies. Psychiatr Clin North Am. 1997;20(3):499–517. pmid:9323310
- 15. Pirkis J, Burgess P. Suicide and recency of health care contacts: A systematic review. Br J Psychiatry. 1998;173:462–74. pmid:9926074
- 16.
Bugeja L, Milner A, Pirkis J. Identification and analysis of health service use and pathways to health service contact amongst persons who died from suicide in Victoria, 2009–2010. Melbourne: Coroners Court of Victoria; 2015.
- 17. Luoma JB, Martin CE, Pearson JL. Contact With Mental Health and Primary Care Providers Before Suicide: A Review of the Evidence. Am J Psychiatry. 2002;159:909–16. pmid:12042175
- 18. Schaffer A, Sinyor M, Kurdyak P, Vigod S, Sareen J, Reis C, et al. Population-based analysis of health care contacts among suicide decedents: identifying opportunities for more targeted suicide prevention strategies. World Psychiatry. 2016;15:135–45. pmid:27265704
- 19. Sveticic J, Milner A, De Leo D. Contacts with mental health services before suicide: a comparison of Indigenous with non-Indigenous Australians. Gen Hosp Psychiatry. 2012;34(2):185–91. pmid:22154656
- 20. Barraclough B, Bunch J, Nelson B, Sainsbury P. A hundred cases of suicide: clinical aspects. Br J Psychiatry. 1974;125:355–73. pmid:4425774
- 21.
Australian Consortium for Classification Development. The international statistical classification of diseases and related health problems, tenth revision, australian modification (ICD-10-AM/ACHI/ACS). Darlinghurst, NSW: Independent Hospital Pricing Authority; Numerous years.
- 22. Sutherland G, Milner A, Dwyer J, Bugeja L, Woodward A, Robinson J, et al. Implementation and evaluation of the Victorian Suicide Register. Aust N Z J Public Health. 2018;42(3):296–302.
- 23.
Australian Bureau of Statistics. Census of Population and Housing: Socio-Economic Indexes for Areas (SEIFA), Australia Canberra: Australian Bureau of Statistics; 2011 & 2016 [cited 2021 May 11]. Available from: https://www.abs.gov.au/websitedbs/censushome.nsf/home/seifa.
- 24.
Australian Bureau of Statistics. Quarterly Population Estimates (ERP), by State/Territory, Sex and Age: Australian Bureau of Statistics 2020 [cited 2021 May 11]. Available from: http://stat.data.abs.gov.au/Index.aspx?DataSetCode=ERP_QUARTERLY#.
- 25. Naing NN. Easy way to learn standardization: direct and indirect methods. Malays J Med Sci. 2000;7(1):10–5. pmid:22844209
- 26. Fine JP, Gray RJ. A Proportional Hazards Model for the Subdistribution of a Competing Risk. J Am Stat Assoc. 1999;94(446):496–509.
- 27. Akaike H. A new look at the statistical model identification. IEEE Trans Automat Contr. 1974(19):716–23.
- 28.
SAS Institute. SAS software, Version 9.4 Cary, NC, USA: SAS Institute Inc.; 2002–2012.
- 29.
StataCorp. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP; 2015.
- 30. Chesney E, Goodwin GM, Fazel S. Risks of all-cause and suicide mortality in mental disorders: a meta-review. World Psychiatry. 2014;13:153–60. pmid:24890068
- 31. Moitra M, Santomauro D, Degenhardt L, Collins PY, Whiteford H, Vos T, et al. Estimating the risk of suicide associated with mental disorders: A systematic review and meta-regression analysis. J Psychiatr Res. 2021;137:242–9. pmid:33714076
- 32. Too LS, Spittal MJ, Bugeja L, Reifels L, Butterworth P, Pirkis J. The association between mental disorders and suicide: A systematic review and meta-analysis of record linkage studies. J Affect Disord. 2019;259:302–13. pmid:31450139
- 33. Hampson NB, Rudd RA, Hauff NM. Increased long-term mortality among survivors of acute carbon monoxide poisoning *. Crit Care Med. 2009;37(6).
- 34. Cooper J, Kapur N, Webb R, Lawlor M, Guthrie E, Mackway-Jones K, et al. Suicide after deliberate self-harm: a 4-year cohort study. Am J Psychiatry. 2005;162(2):297–303. pmid:15677594