Article Text
PDF
Abstract
Introduction Emergency department patients presenting with non-fatal suicidal behaviour face elevated risk of suicide and all-cause mortality, but the extent to which this has changed over time is unknown. This study tracked trends in mortality risks faced by emergency department patients presenting with deliberate self-harm and suicidal ideation in California.
Methods Using statewide linked emergency department and death data, we estimated 2010–2016 trends in suicide and all-cause mortality among emergency department patients with either deliberate self-harm (n=111 658) or suicidal ideation (n=162 959). We also calculated average annual percent changes in age-adjusted mortality rates and compared these to the general California population.
Results Deliberate self-harm and suicidal ideation patients’ age-adjusted suicide rates decreased by approximately 5% per year during the study period; however, their all-cause mortality trends were flat. In the general California population, suicide rate trends were flat while all-cause mortality slightly declined.
Conclusions Suicide mortality unexpectedly declined among self-harming and suicidal patients presenting to California emergency departments. Additional research is needed to understand the reasons behind this decline and inform quality improvement efforts for suicide prevention in hospital settings.
- Suicide/Self?Harm
- Hospital Care
- Mortality
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Emergency department patients presenting with nonfatal suicidal behaviour face extremely elevated risk of suicide and all-cause mortality. Despite calls by the Surgeon General to track trends in mortality among these patients as part of the national suicide prevention strategy, no research has done so.
WHAT THIS STUDY ADDS
Among suicidal emergency department patients presenting 2010–2016 in California, suicide mortality rates decreased by 5% per year; by contrast, suicidal patients’ all-cause mortality trends were flat. Suicide and all-cause mortality in the general California population were largely unchanged.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Our findings suggest that survival in this high-risk patient group can be improved, but more research is needed to understand the source of the improvement.
Introduction
Emergency department (ED) patients, especially those presenting with deliberate self-harm (DSH) and suicidal ideation (SI), face extremely elevated risk of suicide and all-cause mortality.1 Public health stakeholders routinely track and report mortality risk among other patients (eg, cancer patients), but ongoing mortality surveillance among patients presenting with suicidality or other psychiatric conditions is almost non-existent.2 This is despite calls from the Surgeon General’s office and other parties to track and report trends of these patients’ suicide and other mortality risks, in order to to inform timely clinical and public health quality improvement responses.3 4 We aimed to estimate recent trends in mortality rates from suicide and all-cause deaths among all ED patients presenting with DSH and SI in California, which has linked statewide administrative health and death records for many years. For comparison, we also examined mortality trends among the general California population.
Materials and methods
This study was approved by the University of California, Merced Institutional Review Board. The California Department of Public Health linked statewide ED discharge data and death records (including out-of-state deaths) for each calendar year from 2010 to 2016. Probabilistic linkages were based on patient’s social security number, sex, birthdate, race/ethnicity and zip code of residence; linkage rates are estimated to be 98%.5 6 Patients without a matching death record during the follow-up period were assumed alive. All data were analysed using Stata V.17.0 (StataCorp) and the Joinpoint Regression Program V.5.0.2 (National Cancer Institute). Among ED patients aged ≥10 years presenting in January through June of each year, two hierarchical groups were constructed: patients presenting with DSH (International Classification of Diseases, Clinical Modification codes from7) and patients presenting with SI (SI; ICD-9/10 CM codes V.62.84/R45.851). Patients’ first qualifying visit in each annual period was retained for analysis; we excluded patients whose disposition was ‘deceased’.
For suicide8 and all-cause deaths among DSH and SI patients, we calculated mortality rates per 100 000 person-years based on 6-month crude mortality during the first 6 months following ED discharge in each year. A 6-month follow-up period was used9 10 as it allowed us to make the best use of the calendar-year mortality data linkage. To facilitate comparisons with the general population, we retrieved suicide and all-cause deaths and population counts for Californians aged ≥10 years, in calendar-years 2010 through 2016, from the Centers for Disease Control and Prevention’s WISQARS injury fatality and WONDER underlying cause-of-death online databases.11 12 We estimated 6-month mortality rates per 100 000 in these data by halving annual rates.
We first calculated age-adjusted suicide and all-cause mortality rates in each calendar-year separately for DSH patients, SI patients and the California population, using 2010 California population numbers as the reference standard. We then calculated sex-specific crude suicide mortality rates by year for each group. (Calculation of age-adjusted sex-specific rates was not possible given the rare outcome.) Finally, we calculated average annual percent changes (AAPC) in age-adjusted suicide and all-cause mortality rates for all groups using Joinpoint regression, with constant variance and linear trends assumed given the short study period.
Patient and public involvement
This study consisted of secondary analysis of administrative patient records; no patients were involved in the research.
Results
Across all study years, the combined ED sample comprised 111 658 patients with DSH (of whom 544 died by suicide and 1690 of any cause within 6 months) and 162 959 patients with SI (403 suicide deaths and 2838 all-cause deaths). Table 1 shows descriptive characteristics of the patient groups. DSH patients had higher proportions of women (58.6%) and patients aged 10–24 years (34.7%) compared with SI patients (47.2% and 23.6%).
Demographic characteristics of deliberate self-harm and suicidal ideation ED patient groups, 2010–2016, California
DSH patients’ crude 6-month suicide rates per 100 000 person-years were 530.1 in 2010 and 357.8 in 2016. Joinpoint regression analyses using DSH patients’ age-adjusted suicide rates indicated that these rates significantly decreased during the study period, with an AAPC of −5.1% (95% CI −9.7 to –0.1) (figure 1A). On a per 100 000 basis, DSH patients’ crude all-cause mortality rates were 1719.2 in 2010 and 1794.3 in 2016. Although these patients’ age-adjusted all-cause mortality rates trended downward over the study period (figure 1A), this decline did not reach statistical significance (AAPC of −2.7, 95% CI −6.6 to 1.3).
Trends in age-adjusted 6-month suicide and all-cause mortality rates among three groups in5 California, 2010 to 2016: deliberate self-harm emergency department (ED) patients (DSH), suicidal ideation ED patients (SI) and the general California population (CA).
SI patients’ 6-month suicide rates per 100 000 person-years were lower in every year than those of DSH patients (339.0 in 2010 and 206.9 in 2016). Although the magnitude of SI patients’ change in age-adjusted suicide rates was similar to that of DSH patients (AAPC: −5.3%), this decline did not reach statistical significance (95% CI −16.4 to 7.7). As with DSH patients, SI patients’ age-adjusted all-cause mortality rates were substantially higher than those of the general California population, but their all-cause mortality trendline was flat (AAPC of −0.4, 95% CI −2.6 to 1.9).
In the general California population, trends for age-adjusted suicide and all-cause mortality rates from 2010 to 2016 were flat or slightly declining, respectively (suicide: AAPC of 0.5, 95% CI −0.6 to 1.6; all-cause mortality: AAPC of −1.0, 95% CI −1.6 to –0.5).
Joinpoint regression analyses of sex-specific crude suicide rates (online supplemental figure 1A) revealed significant downward trends among female (AAPC: −11.0, 95% CI −16.2 to –5.5) but not male (AAPC: −2.0, 95% CI −6.9 to 3.2) DSH patients. Among SI patients, suicide rate declines were marginally significant for males (AAPC: −7.4, 95% CI −14.6 to 0.4) but flat among women (AAPC: −3.5, 95% CI 16.5 to 11.4). Sex-specific suicide rates for the general California population were flat. Sex-specific all-cause mortality rates (online supplemental figure 1B) were flat for both ED patient groups, but increased slightly for California men (AAPC: 1.3, 95% CI 0.7 to 1.8). We also plotted and tested age-specific (<35 vs ≥35 years) crude suicide rates for all groups; results are available as part of online supplemental figure 2.
Supplemental material
Discussion
In this population-based study in California, age-adjusted suicide mortality rates among ED patients presenting with DSH declined substantially from 2010 to 2016, especially among women. Suicide mortality also trended downwards among male patients presenting with suicide ideation. Suicide rates in both DSH and SI patient groups, however, remained very high in comparison to the general population. DSH and SI patients’ all-cause mortality rates did not appreciably change, indicating that their mortality from other causes of death may have risen. As suicide mortality rates in California’s overall population were flat during this period, our results were somewhat unexpected. The reasons for the observed decline are unknown but are unlikely to be due to an overall reduction in illness severity among the ED population served during this period.13
Our analysis underscores the importance of ongoing surveillance for understanding changes in population mortality burdens among high-risk self-harm and suicidal ED patients.14 Patient survival is a key clinical and public health goal. Improvements in suicide survival rates, such as those observed here, should be tracked and investigated so as to inform quality improvement and intervention efforts15; declines in survival should likewise be tracked and explained.
This study has limitations. Due to the state’s death record linkage procedures, we could only track patient mortality within a calendar year, which led to cohort definitions that under-represent patients seeking care from July to December, and to imprecision in mortality rate calculations given the low base rate of suicide death. We assume that our conclusions would be similar had linkage across calendar years been possible; nevertheless, the imprecision in mortality rates mean that our results should be applied with caution to other populations. Other limitations include potential misclassification related to diagnostic errors in EDs, imprecision in manner of death determinations and death linkages that (at the time of our data request) ended in 2016 and may be differentially accurate according to patient sociodemographic characteristics.
Our finding that suicide mortality significantly declined among ED patients presenting with DSH, and marginally declined among those presenting with SI, demonstrates the feasibility and utility of such public health mortality surveillance and suggests the encouraging possibility that survival in these high-risk individuals has improved. Additional research is needed to identify the sources of these declines, with the goal of informing quality improvement and service delivery-based suicide prevention efforts in hospital settings.
Ethics statements
Patient consent for publication
Ethics approval
The Institutional Review Board of University of California, Merced approved this study with a waiver of informed consent (protocol UCM2020-91).
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors SG-M and MO conceived of the study. SG-M conducted the literature search, acquired the study data, conducted the data analysis and led the interpretation and manuscript writing. MO contributed to the literature search, study design, interpretation of results and manuscript writing. MS contributed to the study design, interpretation of results and manuscript writing.
Funding This project was funded through National Institutes of Health grant R15 MH113108-01 to SGM. The sponsor had no role in the study design; collection, analysis, or interpretation of data; writing of the report, or decision to submit the article for publication. The views expressed here are those of the authors, and not necessarily those of the National Institute of Mental Health, Department of Health and Human Services, or the federal government. No financial disclosures were reported by the authors of this paper.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer-reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.