A Population-Level Approach to Suicide Prevention

Overdose, Behavioral Health Services, and Medications for Opioid Use Disorder After a Nonfatal Overdose

Importance

Recognizing and providing services to individuals at highest risk for drug overdose are paramount to addressing the drug overdose crisis.

Objective

To examine receipt of medications for opioid use disorder (MOUD), naloxone, and behavioral health services in the 12 months after an index nonfatal drug overdose and the association between receipt of these interventions and fatal drug overdose.

Design, Setting, and Participants

This cohort study was conducted in the US from January 2020 to December 2021 using claims, demographic, mortality, and other data from the Centers for Medicare & Medicaid Services, the Centers for Disease Control and Prevention, and other sources. The cohort comprised Medicare fee-for-service beneficiaries aged 18 years or older with International Statistical Classification of Diseases, Tenth Revision, Clinical Modification codes for a nonfatal drug overdose. Data analysis was performed from February to November 2023.

Exposures

Demographic and clinical characteristics, substance use disorder, and psychiatric comorbidities.

Main Outcomes and Measures

Receipt of MOUD, naloxone, and behavioral health services as well as subsequent nonfatal and fatal drug overdoses.

Results

The cohort consisted of 136 762 Medicare beneficiaries (80 140 females [58.6%]; mean (SD) age of 68.2 [15.0] years) who experienced an index nonfatal drug overdose in 2020. The majority of individuals had Hispanic (5.8%), non-Hispanic Black (10.9%), and non-Hispanic White (78.8%) race and ethnicity and lived in metropolitan areas (78.9%). In the 12 months after their index nonfatal drug overdose, 23 815 beneficiaries (17.4%) experienced at least 1 subsequent nonfatal drug overdose and 1323 (1.0%) died of a fatal drug overdose. Opioids were involved in 72.2% of fatal drug overdoses. Among the cohort, 5556 (4.1%) received any MOUD and 8530 (6.2%) filled a naloxone prescription in the 12 months after the index nonfatal drug overdose. Filling a naloxone prescription (adjusted odds ratio [AOR], 0.70; 95% CI, 0.56-0.89), each percentage of days receiving methadone (AOR, 0.98; 95% CI, 0.98-0.99) or buprenorphine (AOR, 0.99; 95% CI, 0.98-0.99), and receiving behavioral health assessment or crisis services (AOR, 0.25; 95% CI, 0.22-0.28) were all associated with reduced adjusted odds of fatal drug overdose in the 12 months after the index nonfatal drug overdose.

Conclusions and Relevance

This cohort study found that, despite their known association with reduced risk of a fatal drug overdose, only a small percentage of Medicare beneficiaries received MOUD or filled a naloxone prescription in the 12 months after a nonfatal drug overdose. Efforts to improve access to behavioral health services; MOUD; and overdose-prevention strategies, such as prescribing naloxone and linking individuals to community-based health care settings for ongoing care, are needed.

Prognosis and risk of suicide after cancer diagnosis

INTRODUCTION

Suicide rates are elevated after cancer diagnosis. Existential distress caused by awareness of one's impending death is well-described in patients with cancer. The authors hypothesized that suicide risk is associated with cancer prognosis, and the impact of prognosis on suicide risk is greatest for populations with higher baseline suicide risk.

METHODS

The authors identified patients (≥16 years old) with newly diagnosed cancers from 2000 to 2019 in the Surveillance, Epidemiology, and End Results database, representing 27% of US cancers. Multiple primary-standardized mortality ratios (SMR) were used to estimate the relative risk of suicide within 6 months of diagnosis compared to the general US population, adjusted for age, sex, race, and year of follow-up. Suicide rates by 20 most common cancer sites were compared with respective 2-year overall survival rates (i.e., prognosis) using a weighted linear regression model.

RESULTS

Among 6,754,704 persons diagnosed with cancer, there were 1610 suicide deaths within 6 months of diagnosis, three times higher than the general population (SMR = 3.1; 95% confidence interval, 3.0-3.3). Suicide risk by cancer site was closely associated with overall prognosis (9.5%/percent survival deficit, R2  = 0.88, p < .0001). The association of prognosis with suicide risk became attenuated over time. For men, the risk of suicide increased by 2.8 suicide deaths per 100,000 person-years (p < .0001) versus 0.3 in women (p < .0001). The risk was also higher for persons ≥60 old and for the White (vs. Black) race.

CONCLUSIONS

Poorer prognosis was closely associated with suicide risk early after cancer diagnosis and had a greater effect on populations with higher baseline risks of suicide. This model highlights the need for enhanced psychiatric surveillance and continued research in this patient population.

Toward a generalized developmental model of psychopathological liabilities and psychiatric disorders

BACKGROUND

Most psychiatric disorders are associated with several risk factors, but a few underlying psychopathological dimensions account for the common co-occurrence of disorders. If these underlying psychopathological dimensions mediate associations of the risk factors with psychiatric disorders, it would support a trans-diagnostic orientation to etiological research and treatment development.

METHOD

An analysis was performed of the 2012-2013 National Epidemiologic Survey on Alcohol and Related Conditions III (NESARC-III), a US nationally representative sample of non-institutionalized civilian adults, focusing on respondents who were aged ⩾21 (n = 34 712). Structural equation modeling was used to identify the psychopathological dimensions underlying psychiatric disorders; to examine associations between risk factors, psychopathological dimensions and individual disorders; and to test whether associations of risk factors occurring earlier in life were mediated by risk factors occurring later in life.

RESULTS

A bifactor model of 13 axis I disorders provided a good fit (CFI = 0.987, TLI = 0.982, and RMSEA = 0.011) including an overall psychopathology factor as measured by all 13 disorders and 2 specific factors, one for externalizing disorders and one for fear-related disorders. A substantial proportion of the total effects of the risk factors occurring early in life were indirectly mediated through factors occurring later in life. All risk factors showed a significant total effect on the general psychopathology, externalizing and fear-related factors. Only 23 of 325 direct associations of risk factors with psychiatric disorders achieved statistical significance.

CONCLUSION

Most risk factors for psychiatric disorders are mediated through broad psychopathological dimensions. The central role of these dimensions supports trans-diagnostic etiological and intervention research.

Contextual factors associated with country-level suicide mortality in the Americas, 2000-2019: a cross-sectional ecological study

Background

The suicide mortality rate in the Region of the Americas has been increasing, while decreasing in all other World Health Organization regions; highlighting the urgent need for enhanced prevention efforts. Gaining a better understanding of population-level contextual factors associated with suicide may aid such efforts. We aimed to evaluate the contextual factors associated with country-level, sex-specific suicide mortality rates in the Region of the Americas for 2000-2019.

Methods

Annual sex-specific age-standardized suicide mortality estimates were obtained from the World Health Organization (WHO) Global Health Estimates database. To investigate the sex-specific suicide mortality rate trend over time in the region, we performed joinpoint regression analysis. We then applied a linear mixed model to estimate the effects of specific contextual factors on the suicide mortality rate across countries in the region over time. All potentially relevant contextual factors, obtained from the Global Burden of Disease Study 2019 covariates and The World Bank, were selected in a step-wise manner.

Findings

We found that the mean country-level suicide mortality rate among males in the region decreased as health expenditure per capita and the proportion of the country with a moderate population density increased; and increased as the death rate due to homicide, prevalence of intravenous drug use, risk-weighted prevalence of alcohol use, and unemployment rate increased. The mean country-level suicide mortality rate among females in the region decreased as the number of employed medical doctors per 10,000 population and the proportion of the country with a moderate population density increased; and increased when relative education inequality and unemployment rate increased.

Interpretation

Although there was some overlap, the contextual factors that significantly impacted the suicide mortality rate among males and females were largely different, which mirrors the current literature on individual-level risk factors for suicide. Taken together, our data supports that sex should be considered when adapting and testing suicide risk reduction interventions, and when developing national suicide prevention strategies.

Funding

This work received no funding.

Evaluation of a Model to Target High-risk Psychiatric Inpatients for an Intensive Postdischarge Suicide Prevention Intervention

Importance

The months after psychiatric hospital discharge are a time of high risk for suicide. Intensive postdischarge case management, although potentially effective in suicide prevention, is likely to be cost-effective only if targeted at high-risk patients. A previously developed machine learning (ML) model showed that postdischarge suicides can be predicted from electronic health records and geospatial data, but it is unknown if prediction could be improved by adding additional information.

Objective

To determine whether model prediction could be improved by adding information extracted from clinical notes and public records.

Design, Setting, and Participants

Models were trained to predict suicides in the 12 months after Veterans Health Administration (VHA) short-term (less than 365 days) psychiatric hospitalizations between the beginning of 2010 and September 1, 2012 (299 050 hospitalizations, with 916 hospitalizations followed within 12 months by suicides) and tested in the hospitalizations from September 2, 2012, to December 31, 2013 (149 738 hospitalizations, with 393 hospitalizations followed within 12 months by suicides). Validation focused on net benefit across a range of plausible decision thresholds. Predictor importance was assessed with Shapley additive explanations (SHAP) values. Data were analyzed from January to August 2022.

Main Outcomes and Measures

Suicides were defined by the National Death Index. Base model predictors included VHA electronic health records and patient residential data. The expanded predictors came from natural language processing (NLP) of clinical notes and a social determinants of health (SDOH) public records database.

Results

The model included 448 788 unique hospitalizations. Net benefit over risk horizons between 3 and 12 months was generally highest for the model that included both NLP and SDOH predictors (area under the receiver operating characteristic curve range, 0.747-0.780; area under the precision recall curve relative to the suicide rate range, 3.87-5.75). NLP and SDOH predictors also had the highest predictor class-level SHAP values (proportional SHAP = 64.0% and 49.3%, respectively), although the single highest positive variable-level SHAP value was for a count of medications classified by the US Food and Drug Administration as increasing suicide risk prescribed the year before hospitalization (proportional SHAP = 15.0%).

Conclusions and Relevance

In this study, clinical notes and public records were found to improve ML model prediction of suicide after psychiatric hospitalization. The model had positive net benefit over 3-month to 12-month risk horizons for plausible decision thresholds. Although caution is needed in inferring causality based on predictor importance, several key predictors have potential intervention implications that should be investigated in future studies.

County-level estimates of suicide mortality in the USA: a modelling study

BACKGROUND

Suicide is one of the leading causes of death in the USA and population risk prediction models can inform decisions on the type, location, and timing of public health interventions. We aimed to develop a prediction model to estimate county-level suicide risk in the USA using population characteristics.

METHODS

We obtained data on all deaths by suicide reported to the National Vital Statistics System between Jan 1, 2005, and Dec 31, 2019, and age, sex, race, and county of residence of the decedents were extracted to calculate baseline risk. We also obtained county-level annual measures of socioeconomic predictors of suicide risk (unemployment, weekly wage, poverty prevalence, median household income, and population density) and state-level prevalence of major depressive disorder and firearm ownership from US public sources. We applied conditional autoregressive models, which account for spatiotemporal autocorrelation in response and predictors, to estimate county-level suicide risk.

FINDINGS

Estimates derived from conditional autoregressive models were more accurate than from models not adjusted for spatiotemporal autocorrelation. Inclusion of suicide risk and protective covariates further reduced errors. Suicide risk was estimated to increase with each SD increase in firearm ownership (2·8% [95% credible interval (CrI) 1·8 to 3·9]), prevalence of major depressive episode (1·0% [0·4 to 1·5]), and unemployment rate (2·8% [1·9 to 3·8]). Conversely, risk was estimated to decrease by 4·3% (-5·1 to -3·2) for each SD increase in median household income and by 4·3% (-5·8 to -2·5) for each SD increase in population density. An increase in the heterogeneity in county-specific suicide risk was also observed during the study period.

INTERPRETATION

Area-level characteristics and the conditional autoregressive models can estimate population-level suicide risk. Availability of near real-time situational data are necessary for the translation of these models into a surveillance setting. Monitoring changes in population-level risk of suicide could help public health agencies select and deploy targeted interventions quickly.

FUNDING

US National Institute of Mental Health.