Adverse social factors and all-cause mortality among male and female patients receiving care in the Veterans Health Administration

All-cause mortality among United States military personnel: Findings from the Millennium Cohort Study, 2001-2021

PURPOSE

The goal of this study was to estimate all-cause mortality among Operations Enduring Freedom, Iraqi Freedom, and New Dawn era service members and veterans and to identify protective and risk factors for mortality.

METHODS

Using 20 years of longitudinal data from the Millennium Cohort Study (2001-2021), sequential Cox proportional hazard models were conducted to examine demographic, military, and health-related characteristics associated with all-cause mortality among service members and veterans.

RESULTS

Among 201,619 participants, 3806 (1.9 %) were deceased by the end of the observation period, with an age- and sex-adjusted incidence of 37.6 deaths per 100,000 person-years. Deployed service members had lower all-cause mortality risk than those who did not deploy. Personnel who experienced combat had higher mortality risk compared with those who did not in unadjusted models; this association was nonsignificant after accounting for health-related factors. Enlisted and Army personnel both had a higher mortality risk, while women and Hispanic individuals had a lower risk. Stressful life events, lower physical health related quality of life, problem drinking, and smoking were also associated with greater mortality risk.

CONCLUSION

These profiles may be useful for developing preventive education and intervention efforts in military and veteran populations to reduce premature mortality.

Differences by Sexual Orientation in Patient-Centered Care Outcomes for Veterans Utilizing Primary Care Services at the Veterans Health Administration

Purpose: This study examined the differences by sexual orientation in patient-centered care outcomes (including health care experiences and health-related screening) of veterans utilizing Veterans Health Administration (VHA) primary care. Methods: VHA's adapted version of the Consumer Assessment of Healthcare Providers and Systems was used to compare the health care experience of primary care services among sexual minority (SM) and heterosexual veterans. Health care experience measures were dichotomized to "always" versus "less" and stratified by SM status. Health-related screening measures were dichotomous. Survey data were weighted using provided sample weights. Descriptive statistics were performed on sociodemographic characteristics. Logistic regression coefficients were represented as adjusted odds ratios (aORs). A total of 66,348 veterans were included in the analytic sample, of which 2.9% (n = 1,935) identified as SM. Sexual orientation was ascertained by self-report measures by veterans. Results: SM veterans were significantly younger (56.95 years vs. 63.43 years, p < 0.001), were less likely to report that their provider showed respect for what they had to say (aOR: 0.76; 95% confidence interval [CI]: 0.61-0.95), that they were asked about difficulties taking care of their health (aOR: 0.81; 95% CI: 0.67-0.96), and their provider listened carefully to them (aOR: 0.71; 95% CI: 0.57-0.87) compared to heterosexual veterans. Conclusion: Health care experiences differed between SM and heterosexual veterans who sought VHA primary care, suggesting the need to increase provider trainings, which may improve cultural competency and promote a more welcoming and inclusive environment.

Using social risks to predict unplanned hospital readmission and emergency care among hospitalized Veterans

OBJECTIVES

(1) To estimate the association of social risk factors with unplanned readmission and emergency care after a hospital stay. (2) To create a social risk scoring index.

DATA SOURCES AND SETTING

We analyzed administrative data from the Department of Veterans Affairs (VA) Corporate Data Warehouse. Settings were VA medical centers that participated in a national social work staffing program.

STUDY DESIGN

We grouped socially relevant diagnoses, screenings, assessments, and procedure codes into nine social risk domains. We used logistic regression to examine the extent to which domains predicted unplanned hospital readmission and emergency department (ED) use in 30 days after hospital discharge. Covariates were age, sex, and medical readmission risk score. We used model estimates to create a percentile score signaling Veterans' health-related social risk.

DATA EXTRACTION

We included 156,690 Veterans' admissions to a VA hospital with discharged to home from 1 October, 2016 to 30 September, 2022.

PRINCIPAL FINDINGS

The 30-day rate of unplanned readmission was 0.074 and of ED use was 0.240. After adjustment, the social risks with greatest probability of readmission were food insecurity (adjusted probability = 0.091 [95% confidence interval: 0.082, 0.101]), legal need (0.090 [0.079, 0.102]), and neighborhood deprivation (0.081 [0.081, 0.108]); versus no social risk (0.052). The greatest adjusted probabilities of ED use were among those who had experienced food insecurity (adjusted probability 0.28 [0.26, 0.30]), legal problems (0.28 [0.26, 0.30]), and violence (0.27 [0.25, 0.29]), versus no social risk (0.21). Veterans with social risk scores in the 95th percentile had greater rates of unplanned care than those with 95th percentile Care Assessment Needs score, a clinical prediction tool used in the VA.

CONCLUSIONS

Veterans with social risks may need specialized interventions and targeted resources after a hospital stay. We propose a scoring method to rate social risk for use in clinical practice and future research.

Scoping review of military veterans involved in the criminal legal system and their health and healthcare: 5-year update and map to the Veterans-Sequential Intercept Model

BACKGROUND

A previous scoping review of legal-involved veterans' health and healthcare (1947-2017) identified studies and their limitations. Given the influx of literature published recently, this study aimed to update the previous review and map articles to the Veterans-Sequential Intercept Model (V-SIM) - a conceptual model used by key partners, including Veterans Health Administration, veteran advocates, criminal justice practitioners, and local governments to identify intercept points in the criminal legal system where resources and programming can be provided. Developing an updated resource of literature is essential to inform current research, discover gaps, and highlight areas for future research.

METHODS

A systematic search of 5 databases identified articles related to legal-involved veterans' health and healthcare published between December 2017 through December 2022. The first and senior authors conducted abstract reviews, full-text reviews, and data extraction of study characteristics. Finally, each article was sorted by the various intercept points from the V-SIM.

RESULTS

Of 903 potentially relevant articles, 107 peer-reviewed publications were included in this review, most related to mental health (66/107, 62%) and used an observational quantitative study design (95/107, 89%). Although most articles did not explicitly use the V-SIM to guide data collection, analyses, or interpretation, all could be mapped to this conceptual model. Half of the articles (54/107, 50%) collected data from intercept 5 (Community Corrections and Support Intercept) of the V-SIM. No articles gathered data from intercepts 0 (Community and Emergency Services Intercept), 1 (Law Enforcement Intercept), or 2 (Initial Detention and Court Hearings Intercept).

CONCLUSIONS

There were 107 articles published in the last five years compared to 190 articles published in 70 years covered in the last review, illustrating the growing interest in legal-involved veterans. The V-SIM is widely used by front-line providers and clinical leadership, but not by researchers to guide their work. By clearly tying their research to the V-SIM, researchers could generate results to help guide policy and practice at specific intercept points. Despite the large number of publications, research on prevention and early intervention for legal-involved veterans is lacking, indicating areas of great need for future studies.

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.

Associations Between Natural Language Processing-Enriched Social Determinants of Health and Suicide Death Among US Veterans

Importance

Social determinants of health (SDOHs) are known to be associated with increased risk of suicidal behaviors, but few studies use SDOHs from unstructured electronic health record notes.

Objective

To investigate associations between veterans' death by suicide and recent SDOHs, identified using structured and unstructured data.

Design, Setting, and Participants

This nested case-control study included veterans who received care under the US Veterans Health Administration from October 1, 2010, to September 30, 2015. A natural language processing (NLP) system was developed to extract SDOHs from unstructured clinical notes. Structured data yielded 6 SDOHs (ie, social or familial problems, employment or financial problems, housing instability, legal problems, violence, and nonspecific psychosocial needs), NLP on unstructured data yielded 8 SDOHs (social isolation, job or financial insecurity, housing instability, legal problems, barriers to care, violence, transition of care, and food insecurity), and combining them yielded 9 SDOHs. Data were analyzed in May 2022.

Exposures

Occurrence of SDOHs over a maximum span of 2 years compared with no occurrence of SDOH.

Main Outcomes and Measures

Cases of suicide death were matched with 4 controls on birth year, cohort entry date, sex, and duration of follow-up. Suicide was ascertained by National Death Index, and patients were followed up for up to 2 years after cohort entry with a study end date of September 30, 2015. Adjusted odds ratios (aORs) and 95% CIs were estimated using conditional logistic regression.

Results

Of 6 122 785 veterans, 8821 committed suicide during 23 725 382 person-years of follow-up (incidence rate 37.18 per 100 000 person-years). These 8821 veterans were matched with 35 284 control participants. The cohort was mostly male (42 540 [96.45%]) and White (34 930 [79.20%]), with 6227 (14.12%) Black veterans. The mean (SD) age was 58.64 (17.41) years. Across the 5 common SDOHs, NLP-extracted SDOH, on average, retained 49.92% of structured SDOHs and covered 80.03% of all SDOH occurrences. SDOHs, obtained by structured data and/or NLP, were significantly associated with increased risk of suicide. The 3 SDOHs with the largest effect sizes were legal problems (aOR, 2.66; 95% CI, 2.46-2.89), violence (aOR, 2.12; 95% CI, 1.98-2.27), and nonspecific psychosocial needs (aOR, 2.07; 95% CI, 1.92-2.23), when obtained by combining structured data and NLP.

Conclusions and Relevance

In this study, NLP-extracted SDOHs, with and without structured SDOHs, were associated with increased risk of suicide among veterans, suggesting the potential utility of NLP in public health studies.

Adaptation of a social risk screening and referral initiative across clinical populations, settings, and contexts in the Department of Veterans Affairs Health System

Identifying and addressing social risks and social needs in healthcare settings is an important step towards achieving health equity. Assessing Circumstances and Offering Resources for Needs (ACORN) is a Department of Veterans Affairs (VA) social risk screening and referral model that aims to systematically identify and address social needs. Since initial piloting in 2018, our team has collaborated with clinical and operations partners to implement ACORN across multiple VA clinical settings while adapting and tailoring the initiative to meet the needs of different populations, specialties, and individuals administering screening. Given ACORN's complexity as a growing initiative with multiple partners and frequent real-time modifications within a large national healthcare system, we recognized a need to systematically document the rationale and process of adaptations over time. We looked to three implementation frameworks-RE-AIM, the Adaptome, and FRAME-to describe the rationale for adaptations, the nature of and context within which adaptations were made, and the details of each adaptation. In this manuscript, we uniquely interweave these three frameworks to document adaptations to ACORN across diverse VA clinical settings, with a focus on how adaptations support the promotion of heath equity in the Veteran population.

Sufficiency of Health Information During Pregnancy: What's Missing and for Whom? A Cross-Sectional Analysis Among Veterans

Background: Women Veterans often experience trauma and physical and mental health conditions that increase risk of adverse pregnancy outcomes. Information provision during pregnancy may facilitate improved outcomes. However, little evidence exists about information women Veterans receive during pregnancy, and their perceptions of it. Materials and Methods: We recruited pregnant Veterans from 15 Veterans Affairs medical centers. Through telephone surveys, women (N = 851) provided information about sociodemographic characteristics, military service, health, and pregnancy experiences. We asked postpartum women whether, during pregnancy, they received sufficient information about nine health topics. We calculated a composite score (range: 0-9) that reflected sufficiency of information received. Multivariable logistic regression models identified determinants of perceived sufficiency of information. Results: Mean age was 32.1 years. Most reported being White (56.3%), non-Hispanic (80.3%), married/living with a partner (85.1%), and employed (54.4%). Most (54.6%) had been diagnosed with depression (54.6%); one-quarter reported current depressive symptoms. Mean sufficiency of information score was 6.9. Topics that women most reported they did not receive sufficient information on included, what to expect during delivery (32.3%) and how their spouse/partner might support them during labor (40.3%). History of depression (β = -0.35, p = 0.03), current depressive symptoms (β = -0.66, p = 0.001), military sexual trauma (β = 0.37, p = 0.03), and experience of violence (β = 0.66, p = 0.03) were associated with lower sufficiency of information scores. Conclusion: Results indicate need for enhanced and tailored provision of information for Veterans during pregnancy, particularly among those with experience of trauma, past depression diagnoses, and current depressive symptoms. This may include optimizing care coordination and increasing access to childbirth education classes and doula support.

Women Veterans' Healthcare Needs, Utilization, and Preferences in Veterans Affairs Primary Care Settings

BACKGROUND

The Veterans Health Administration (VA) is the largest integrated health system in the US and provides access to comprehensive primary care. Women Veterans are the fastest growing segment of new VA users, yet little is known about the characteristics of those who routinely access VA primary care in general or by age group.

OBJECTIVE

Describe healthcare needs, utilization, and preferences of women Veterans who routinely use VA primary care.

PARTICIPANTS

1,391 women Veterans with 3+ primary care visits within the previous year in 12 VA medical centers (including General Primary Care Clinics, General Primary Care Clinics with designated space for women, and Comprehensive Women's Health Centers) in nine states.

METHODS

Cross-sectional survey (45% response rate) of sociodemographic characteristics, health status (including chronic disease, mental health, pain, and trauma exposure), utilization, care preferences, and satisfaction. Select utilization data were extracted from administrative data. Analyses were weighted to the population of routine users and adjusted for non-response in total and by age group.

KEY RESULTS

While 43% had health coverage only through VA, 62% received all primary care in VA. In the prior year, 56% used VA mental healthcare and 78% used VA specialty care. Common physical health issues included hypertension (42%), elevated cholesterol (39%), pain (35%), and diabetes (16%). Many screened positive for PTSD (41%), anxiety (32%), and depression (27%). Chronic physical and mental health burdens varied by age. Two-thirds (62%) had experienced military sexual trauma. Respondents reported satisfaction with VA women's healthcare and preference for female providers.

CONCLUSIONS

Women Veterans who routinely utilize VA primary care have significant multimorbid physical and mental health conditions and trauma histories. Meeting women Veterans' needs across the lifespan will require continued investment in woman-centered primary care, including integrated mental healthcare and emphasis on trauma-informed, age-specific care, guided by women's provider preferences.

Association of lifetime homelessness and justice involvement with psychiatric symptoms, suicidal ideation, and suicide attempt among post-9/11 veterans

Both homelessness and criminal justice involvement can impact mental health symptoms and increase risk for suicide. Despite this, few studies have examined their cumulative impact. Moreover, no studies to date have examined the impact of these social determinants of health on post-9/11 veterans, a population with high rates of housing insecurity and justice involvement. The current study sought to better understand the adverse impacts of homelessness and justice involvement on mental health symptoms and suicide risk among post-9/11 veterans. We carried this out by conducting a secondary analysis of cross-sectional data from a 2018 national survey of men and women post-9/11 veteran users and non-users of Veterans Health Administration (VHA) services (N = 15,067). Gender-stratified Poisson and multivariate regressions examined mental health symptoms and suicide risk based on history of homelessness and justice involvement. Models adjusted for sociodemographics, military-related variables, and trauma exposure. Homelessness and justice involvement were both independently associated with more severe posttraumatic, depressive, and substance use symptoms as well as increased rates of suicidal ideation and attempt relative to those with no history of homelessness or justice involvement. Veterans with a history of both homelessness and justice involvement reported the most severe mental health symptoms and suicide risk. This study found consistent positive associations with mental health symptoms for homelessness and justice-involved veterans. Enhancing and increasing access to services that address complex mental health presentation among those with histories of justice involvement and housing instability remain necessary.

Risk Factors Associated With Nonfatal Opioid Overdose Leading to Intensive Care Unit Admission: A Cross-sectional Study

BACKGROUND

Opioid overdose (OD) and related deaths have significantly increased in the United States over the last 2 decades. Existing studies have mostly focused on demographic and clinical risk factors in noncritical care settings. Social and behavioral determinants of health (SBDH) are infrequently coded in the electronic health record (EHR) and usually buried in unstructured EHR notes, reflecting possible gaps in clinical care and observational research. Therefore, SBDH often receive less attention despite being important risk factors for OD. Natural language processing (NLP) can alleviate this problem.

OBJECTIVE

The objectives of this study were two-fold: First, we examined the usefulness of NLP for SBDH extraction from unstructured EHR text, and second, for intensive care unit (ICU) admissions, we investigated risk factors including SBDH for nonfatal OD.

METHODS

We performed a cross-sectional analysis of admission data from the EHR of patients in the ICU of Beth Israel Deaconess Medical Center between 2001 and 2012. We used patient admission data and International Classification of Diseases, Ninth Revision (ICD-9) diagnoses to extract demographics, nonfatal OD, SBDH, and other clinical variables. In addition to obtaining SBDH information from the ICD codes, an NLP model was developed to extract 6 SBDH variables from EHR notes, namely, housing insecurity, unemployment, social isolation, alcohol use, smoking, and illicit drug use. We adopted a sequential forward selection process to select relevant clinical variables. Multivariable logistic regression analysis was used to evaluate the associations with nonfatal OD, and relative risks were quantified as covariate-adjusted odds ratios (aOR).

RESULTS

The strongest association with nonfatal OD was found to be drug use disorder (aOR 8.17, 95% CI 5.44-12.27), followed by bipolar disorder (aOR 2.69, 95% CI 1.68-4.29). Among others, major depressive disorder (aOR 2.57, 95% CI 1.12-5.88), being on a Medicaid health insurance program (aOR 2.26, 95% CI 1.43-3.58), history of illicit drug use (aOR 2.09, 95% CI 1.15-3.79), and current use of illicit drugs (aOR 2.06, 95% CI 1.20-3.55) were strongly associated with increased risk of nonfatal OD. Conversely, Blacks (aOR 0.51, 95% CI 0.28-0.94), older age groups (40-64 years: aOR 0.65, 95% CI 0.44-0.96; >64 years: aOR 0.16, 95% CI 0.08-0.34) and those with tobacco use disorder (aOR 0.53, 95% CI 0.32-0.89) or alcohol use disorder (aOR 0.64, 95% CI 0.42-1.00) had decreased risk of nonfatal OD. Moreover, 99.82% of all SBDH information was identified by the NLP model, in contrast to only 0.18% identified by the ICD codes.

CONCLUSIONS

This is the first study to analyze the risk factors for nonfatal OD in an ICU setting using NLP-extracted SBDH from EHR notes. We found several risk factors associated with nonfatal OD including SBDH. SBDH are richly described in EHR notes, supporting the importance of integrating NLP-derived SBDH into OD risk assessment. More studies in ICU settings can help health care systems better understand and respond to the opioid epidemic.