Dengue NS1 antibodies are associated with clearance of viral NS1.

BACKGROUND: Dengue vascular permeability syndrome is the primary cause of death in severe dengue infections. The protective versus potentially pathogenic role of dengue NS1 antibodies are not well understood. The main goal of this analysis was to characterize the relationship between free NS1 concentration and NS1 antibody titers in primary and secondary dengue infection in order to better understand the presence and duration of NS1 antibody complexes in clinical dengue infections. METHODS: Hospitalized participants with acute dengue infection were recruited from Northern Colombia between 2018 to 2020. Symptom assessment including dengue signs and symptoms, chart review and blood collection was performed. Primary versus secondary Dengue was assessed serologically. NS1 titers and anti-NS1 antibodies were measured daily. RESULTS: Patients with secondary infection have higher antibody titers than those in primary infection, and we find a negative correlation between anti-NS1 antibody titer and NS1 protein. We demonstrate that in a subset of secondary infection, there are indeed NS1 antibody-antigen complexes at the admission day during the febrile phase that are not detectable by the recovery phase. Furthermore, dengue infection status is associated with higher circulating sialidases. DISCUSSION: The negative correlation between antibody and protein suggests that antibodies may play a role in clearing this viral protein.

Sexual and Gender Minority Status and Suicide Mortality: An Explainable Artificial Intelligence Analysis.

Objectives: Suicide risk is elevated in lesbian, gay, bisexual, and transgender (LGBT) individuals. Limited data on LGBT status in healthcare systems hinder our understanding of this risk. This study used natural language processing to extract LGBT status and a deep neural network (DNN) to examine suicidal death risk factors among US Veterans. Methods: Data on 8.8 million veterans with visits between 2010 and 2017 was used. A case-control study was performed, and suicide death risk was analyzed by a DNN. Feature impacts and interactions on the outcome were evaluated. Results: The crude suicide mortality rate was higher in LGBT patients. However, after adjusting for over 200 risk and protective factors, known LGBT status was associated with reduced risk compared to LGBT-Unknown status. Among LGBT patients, black, female, married, and older Veterans have a higher risk, while Veterans of various religions have a lower risk. Conclusion: Our results suggest that disclosed LGBT status is not directly associated with an increase suicide death risk, however, other factors (e.g., depression and anxiety caused by stigma) are associated with suicide death risks.

Self-injurious unnatural death among Veterans with HIV: a nested case-control study.

OBJECTIVE: People living with HIV (PWH) are at increased risk of suicide and death from unintentional causes compared with people living without HIV. Broadening the categorization of death from suicide to self-injurious unnatural death (SIUD) may better identify a more complete set of modifiable risk factors that could be targeted for prevention efforts among PWH. DESIGN: We conducted a nested case-control study using data from the Veterans Aging Cohort Study (VACS), a longitudinal, observational cohort of Veterans from 2006-2015. A total of 5,036 Veterans with HIV, of whom 461 died by SIUD, were included in the sample. METHODS: SIUD was defined using the International Classification of Disease 10th revision cause of death codes. Cases (n = 461) included individuals who died by SIUD (intentional, unintentional, and undetermined causes of death). Controls (n = 4,575) were selected using incidence density sampling, matching on date of birth ± one year, race, sex, and HIV status. SIUD and suicide was estimated using conditional logistic regression. RESULTS: A previous suicide attempt, a diagnosis of an affective disorder, recent use of benzodiazepines, psychiatric hospitalization, and living in the western US significantly increased the risk of suicide and SIUD. Risk factors that appear more important for SIUD than for suicide included a drug use disorder, alcohol use disorder, Hepatitis C, VACS Index 2.0, current smoking, and high pain levels (7-10). CONCLUSION: Limiting studies to known suicides obscures the larger public health burden of excess deaths from self-injurious behavior. Our findings demonstrate the benefit of expanding the focus to SIUD for the identification of modifiable risk factors that could be targeted for treatment.

Non-affirmation minority stress, internalized transphobia, and subjective cognitive decline among transgender and gender diverse veterans aged 45 years and older.

OBJECTIVES: To examine the associations of two measures of minority stress, non-affirmation minority stress and internalized transphobia, with subjective cognitive decline (SCD) among transgender and gender diverse (TGD) veterans. METHOD: We administered a cross-sectional survey from September 2022 to July 2023 to TGD veterans. The final analytic sample included 3,152 TGD veterans aged ≥45 years. We used a generalized linear model with quasi-Poisson distribution to calculate prevalence ratios (PR) and 95% confidence intervals (CIs) measuring the relationship between non-affirmation minority stress and internalized transphobia and past-year SCD. RESULTS: The mean age was 61.3 years (SD = 9.7) and the majority (70%) identified as trans women or women. Overall, 27.2% (n = 857) reported SCD. Adjusted models revealed that TGD veterans who reported experiencing non-affirmation minority stress or internalized transphobia had greater risk of past-year SCD compared to those who did not report either stressor (aPR: 1.09, 95% CI: 1.04-1.15; aPR: 1.19, 95% CI: 1.12-1.27). CONCLUSION: Our findings demonstrate that proximal and distal processes of stigma are associated with SCD among TGD veterans and underscore the need for addressing multiple types of discrimination. Above all, these results indicate the lasting sequelae of transphobia and need for systemic changes to prioritize the safety and welfare of TGD people.

Risk factors for suicide among veterans living with and without HIV: a nested case-control study.

The rate of suicide among people with HIV (PWH) remains elevated compared to the general population. The aim of the study was to examine the association between a broad range of risk factors, HIV-specific risk factors, and suicide. We conducted a nested case-control study using data from the Veterans Aging Cohort Study (VACS) between 2006 and 2015. The risk of suicide was estimated using conditional logistic regression and models were stratified by HIV status. Most risk factors associated with suicide were similar between PWH and people without HIV; these included affective disorders, use of benzodiazepines, and mental health treatment. Among PWH, HIV-specific risk factors were not associated with suicide. A multiplicative interaction was observed between a diagnosis of HIV and a previous suicide attempt. Among PWH, a high prevalence of psychiatric, substance use disorders and multimorbidity contribute to the risk of suicide.

RESUMEN: La tasa de suicidio entre las personas con VIH (PWH) sigue siendo elevada en comparación con la población general. El objetivo del estudio fue examinar la asociación entre un amplio rango de factores de riesgo, los riesgos específicos del VIH y el suicidio. Realizamos un estudio anidado de casos y controles usando datos del Veterans Aging Cohort Study (VACS) entre 2006­2015. El riesgo de suicidio fue estimado mediante regresión logística condicional y los modelos se estratificaron por estado serológico. La mayoría de los factores de riesgo asociados con el suicidio fueron similares entre las PWH y las personas sin VIH; estos incluyeron trastornos afectivos, uso de benzodiazepinas y tratamiento de salud mental. Entre las PWH, los factores de riesgo específicos del VIH no se asociaron con el suicidio. Se observó una interacción multiplicativa entre un diagnóstico de VIH y un intento de suicidio previo. Entre las PWH, una alta prevalencia de trastornos psiquiátricos, por consumo de sustancias y multimorbilidad contribuyen al riesgo de suicidio.

Metformin prescription for U.S. veterans with prediabetes, 2010-2019.

Affecting an estimated 88 million Americans, prediabetes increases the risk for developing type 2 diabetes mellitus (T2DM), and independently, cardiovascular disease, retinopathy, nephropathy, and neuropathy. Nevertheless, little is known about the use of metformin for diabetes prevention among patients in the Veterans Health Administration, the largest integrated healthcare system in the U.S. This is a retrospective observational cohort study of the proportion of Veterans with incident prediabetes who were prescribed metformin at the Veterans Health Administration from October 2010 to September 2019. Among 1,059,605 Veterans with incident prediabetes, 12,009 (1.1%) were prescribed metformin during an average 3.4 years of observation after diagnosis. Metformin prescribing was marginally higher (1.6%) among those with body mass index (BMI) ≥35 kg/m2, age <60 years, HbA1c≥6.0%, or those with a history of gestational diabetes, all subgroups at a higher risk for progression to T2DM. In a multivariable model, metformin was more likely to be prescribed for those with BMI ≥35 kg/m2 incidence rate ratio [IRR] 2.6 [95% confidence intervals (CI): 2.1-3.3], female sex IRR, 2.4 [95% CI: 1.8-3.3], HbA1c≥6% IRR, 1.93 [95% CI: 1.5-2.4], age <60 years IRR, 1.7 [95% CI: 1.3-2.3], hypertriglyceridemia IRR, 1.5 [95% CI: 1.2-1.9], hypertension IRR, 1.5 [95% CI: 1.1-2.1], Major Depressive Disorder IRR, 1.5 [95% CI: 1.1-2.0], or schizophrenia IRR, 2.1 [95% CI: 1.2-3.8]. Over 20% of Veterans with prediabetes attended a comprehensive structured lifestyle modification clinic or program. Among Veterans with prediabetes, metformin was prescribed to 1.1% overall, a proportion that marginally increased to 1.6% in the subset of individuals at highest risk for progression to T2DM.

Relationship Between Pain and LGBT Status Among Veterans in Care in a Retrospective Cross-Sectional Cohort.

Background: Pain assessment is performed in many healthcare systems, such as the Veterans Health Administration, but prior studies have not assessed whether pain screening varies in sexual and gender minority populations that include individuals who identify as lesbian, gay, bisexual, and/or transgender (LGBT). Objective: The purpose of this study was to evaluate pain screening and reported pain of LGBT Veterans compared to non-LGBT Veterans. Methods: Using a retrospective cross-sectional cohort, data from the Corporate Data Warehouse, a national repository with clinical/administrative data, were analyzed. Veterans were classified as LGBT using natural language processing. We used a robust Poisson model to examine the association between LGBT status and binary outcomes of pain screening, any pain, and persistent pain within one year of entry in the cohort. All models were adjusted for demographics, mental health, substance use, musculoskeletal disorder(s), and number of clinic visits. Results: There were 1,149,486 Veterans (218,154 (19%) classified as LGBT) in our study. Among LGBT Veterans, 94% were screened for pain compared to 89% among those not classified as LGBT (non-LGBT) Veterans. In adjusted models, LGBT Veterans' probability of being screened for pain compared to non-LGBT Veterans was 2.5% higher (95% CI 2.3%, 2.6%); risk of any pain was 2.1% lower (95% CI 1.6%, 2.6%); and there was no significant difference between LGBT and non-LGBT Veterans in persistent pain (RR = 1.00, 95% CI (0.99, 1.01), p = 0.88). Conclusions: In a nationwide sample, LGBT Veterans were more likely to be screened for pain but had lower self-reported pain scores, though adjusted differences were small. It was notable that transgender and Black Veterans reported the greatest pain. Reasons for these findings require further investigation.

Identifying suicide documentation in clinical notes through zero-shot learning.

Background and Aims: In deep learning, a major difficulty in identifying suicidality and its risk factors in clinical notes is the lack of training samples given the small number of true positive instances among the number of patients screened. This paper describes a novel methodology that identifies suicidality in clinical notes by addressing this data sparsity issue through zero-shot learning. Our general aim was to develop a tool that leveraged zero-shot learning to effectively identify suicidality documentation in all types of clinical notes. Methods: US Veterans Affairs clinical notes served as data. The training data set label was determined using diagnostic codes of suicide attempt and self-harm. We used a base string associated with the target label of suicidality to provide auxiliary information by narrowing the positive training cases to those containing the base string. We trained a deep neural network by mapping the training documents' contents to a semantic space. For comparison, we trained another deep neural network using the identical training data set labels, and bag-of-words features. Results: The zero-shot learning model outperformed the baseline model in terms of area under the curve, sensitivity, specificity, and positive predictive value at multiple probability thresholds. In applying a 0.90 probability threshold, the methodology identified notes documenting suicidality but not associated with a relevant ICD-10-CM code, with 94% accuracy. Conclusion: This method can effectively identify suicidality without manual annotation.

Implementer report: ICD-10 code F44.5 review for functional seizure disorder.

OBJECTIVE: The study aimed to measure the validity of International Classification of Diseases, 10th Edition (ICD-10) code F44.5 for functional seizure disorder (FSD) in the Veterans Affairs Connecticut Healthcare System electronic health record (VA EHR). METHODS: The study used an informatics search tool, a natural language processing algorithm and a chart review to validate FSD coding. RESULTS: The positive predictive value (PPV) for code F44.5 was calculated to be 44%. DISCUSSION: ICD-10 introduced a specific code for FSD to improve coding validity. However, results revealed a meager (44%) PPV for code F44.5. Evaluation of the low diagnostic precision of FSD identified inconsistencies in the ICD-10 and VA EHR systems. CONCLUSION: Information system improvements may increase the precision of diagnostic coding by clinicians. Specifically, the EHR problem list should include commonly used diagnostic codes and an appropriately curated ICD-10 term list for 'seizure disorder,' and a single ICD code for FSD should be classified under neurology and psychiatry.

Intimate partner violence among lesbian, gay, and bisexual veterans.

The present study describes intimate partner violence (IPV) perpetration and victimization alongside theoretically associated variables in a sample of lesbian, gay, and bisexual veterans. We conducted bivariate analyses (chi-square tests and independent t test) to examine whether the frequencies of IPV perpetration and victimization varied by demographic characteristics, military sexual trauma, alcohol use, and mental health symptoms. Out of the 69 lesbian, gay, and bisexual (LGB) veterans who answered the questions on IPV, 16 (23.2%) reported some form of IPV victimization in the past year, and 38 (55.1%) reported past-year perpetration. Among the 43 veterans who reported psychological IPV, roughly half (48.9%) reported bidirectional psychological IPV, 39.5% reported perpetration only, and 11.6% reported victimization only. LGB veterans who reported bidirectional psychological IPV in their relationships were younger and reported greater symptoms of posttraumatic stress disorder symptoms and depression. The results presented here call for universal screening of IPV perpetration and victimization to both accurately assess and ultimately intervene among all veterans. Inclusive interventions are needed for all genders and sexual orientations, specifically interventions that do not adhere to gendered assumptions of perpetrators and victims. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Fall-related Injuries and Opioid Administration Among Veterans With Dementia in US Department of Veterans Affairs Community Living Centers.

OBJECTIVES: Opioid use is associated with fall-related injuries (FRI) among older adults, especially those with dementia. We examined FRI following changes in national opioid safety initiatives over 3 regulatory periods [preinitiatives baseline (period 1): October 2012 to June 2013; post-Veteran Affairs (VA) opioid safety initiative (period 2): January 2014 to November 2015; post-VA and CDC opioid prescribing guidelines (period 3): March 2017 to September 2018] among Department of VA Community Living Center (CLC) long-stay residents with dementia. DATA: VA provided and purchased care records, Medicare claims, CLC Minimum Data Set (MDS) assessments. VA bar-code medication administration data, VA outpatient prescription refill data, and Medicare Part D data were used to capture medication from inpatient, outpatient, and Medicare sources. SETTINGS AND PARTICIPANTS: A total of 12,229 long-stay CLC residents with dementia between October 2012 and September 2018. METHODS: We applied Veteran-regulatory period level (1) generalized linear model to examine the unadjusted and adjusted trends of FRI, and (2) difference-in-difference model with propensity score weighting to examine the relationship between opioid safety initiatives and FRI in 3 regulatory periods. We applied propensity score weighting to enable the cohorts in periods 2 and 3 had similar indications for opioid administration as in period 1. RESULTS: FRI prevalence per month among CLC residents with Alzheimer disease and related dementias decreased from 3.1% in period 1 to 1.6% and 1.2% in periods 2 and 3, and the adjusted probability of FRI was 17% and 40% lower in periods 2 and 3 compared with period 1. The any, incident, and continued opioid administration were significantly associated with higher FRI, whereas the differences in FRI probabilities between opioid and nonopioid users had no significant changes over the 3 regulatory periods. CONCLUSIONS: FRI was reduced among CLC residents with Alzheimer disease and related dementias receiving care in VA CLCs over the 3 regulatory periods, but the FRI reduction was not significantly associated with opioid safety initiatives. Other interventions that potentially targeted falls are likely to have helped reduce these fall events. Future studies could examine whether opioid use reduction ultimately benefitted nursing home residents by focusing on other possible outcomes or whether such reduction only resulted in more untreated pain.

Association between early nonpharmacological management and follow-up for low back pain in the veterans health administration.

Background: Low back pain (LBP) is a common reason individuals seek healthcare. Nonpharmacologic management (NPM) is often recommended as a primary intervention, and earlier use of NPM for LBP shows positive clinical outcomes. Our purpose was to evaluate how timing of engagement in NPM for LBP affects downstream LBP visits during the first year. Methods: This study was a secondary analysis of an observational cohort study of national electronic health record data. Patients entering the Musculoskeletal Diagnosis/Complementary and Integrative Health Cohort with LBP from October 1, 2016 to September 30, 2017 were included. Exclusive patient groups were defined by engagement in NPM within 30 days of entry ("very early NPM"), between 31 and 90 days ("early NPM"), or not within the first 90 days ("no NPM"). The outcome was time, in days, to the final LBP follow-up after 90 days and within the first year. Cox proportional hazards regression was used to model time to final follow up, controlling for additional demographic and clinical covariables. Results: The study population included 44,175 patients, with 16.7% engaging in very early NPM and 13.1% in early NPM. Patients with very early NPM (5.2 visits, SD=4.5) or early NPM (5.7 visits, SD=4.6) had a higher mean number of LBP visits within the first year than those not receiving NPM in the first 90 days (3.2 visits, SD = 2.5). The very early NPM (HR=1.50, 95% CI: 1.46-1.54; median=48 days, IQR=97) and early NPM (HR=1.27, 95% CI: 1.23-1.30; median=88 days, IQR=92) had a significantly shorter time to final follow-up than the no NPM group (median=109 days, IQR=150). Conclusions: Veterans Health Administration patients receiving NPM for LBP within the first 90 days after initially seeking care demonstrate a significantly faster time to final follow-up visit within the first year compared to those who do not.

Opioid use and opioid use disorder in mono and dual-system users of veteran affairs medical centers.

Introduction: Efforts to achieve opioid guideline concordant care may be undermined when patients access multiple opioid prescription sources. Limited data are available on the impact of dual-system sources of care on receipt of opioid medications. Objective: We examined whether dual-system use was associated with increased rates of new opioid prescriptions, continued opioid prescriptions and diagnoses of opioid use disorder (OUD). We hypothesized that dual-system use would be associated with increased odds for each outcome. Methods: This retrospective cohort study was conducted using Veterans Administration (VA) data from two facilities from 2015 to 2019, and included active patients, defined as Veterans who had at least one encounter in a calendar year (2015-2019). Dual-system use was defined as receipt of VA care as well as VA payment for community care (non-VA) services. Mono users were defined as those who only received VA services. There were 77,225 dual-system users, and 442,824 mono users. Outcomes were three binary measures: new opioid prescription, continued opioid prescription (i.e., received an additional opioid prescription), and OUD diagnosis (during the calendar year). We conducted a multivariate logistic regression accounting for the repeated observations on patient and intra-class correlations within patients. Results: Dual-system users were significantly younger than mono users, more likely to be women, and less likely to report white race. In adjusted models, dual-system users were significantly more likely to receive a new opioid prescription during the observation period [Odds ratio (OR) = 1.85, 95% confidence interval (CI) 1.76-1.93], continue prescriptions (OR = 1.24, CI 1.22-1.27), and to receive an OUD diagnosis (OR = 1.20, CI 1.14-1.27). Discussion: The prevalence of opioid prescriptions has been declining in the US healthcare systems including VA, yet the prevalence of OUD has not been declining at the same rate. One potential problem is that detailed notes from non-VA visits are not immediately available to VA clinicians, and information about VA care is not readily available to non-VA sources. One implication of our findings is that better health system coordination is needed. Even though care was paid for by the VA and presumably closely monitored, dual-system users were more likely to have new and continued opioid prescriptions.

Graded chronic pain scale revised: validation in a Veteran sample.

OBJECTIVE: The Graded Chronic Pain Scale (GCPS) is frequently used in pain research and treatment to classify mild, bothersome, and high impact chronic pain. This study's objective was to validate the revised version of the GCPS (GCPS-R) in a US Veterans Affairs (VA) healthcare sample to support its use in this high-risk population. METHODS: Data were collected from Veterans (n = 794) via self-report (GCPS-R and relevant health questionnaires) and electronic health record extraction (demographics and opioid prescriptions). Logistic regression, adjusting for age and gender, was used to test for differences in health indicators by pain grade. Adjusted odds ratio (AOR) with 95% confidence intervals (CIs) were reported with CIs not including an AOR of 1 indicating that the difference exceeded chance. RESULTS: In this population, the prevalence of chronic pain (pain present most or every day, prior 3 months) was 49.3%: 7.1% with mild chronic pain (mild pain intensity and lower interference with activities); 23.3% bothersome chronic pain (moderate to severe pain intensity with lower interference); and 21.1% high impact chronic pain (higher interference). Results of this study mirrored findings in the non-VA validation study; differences between bothersome and high impact were consistent for activity limitations and present but not fully consistent for psychological variables. Those with bothersome chronic pain or high impact chronic pain were more likely to receive long-term opioid therapy compared to those with no/mild chronic pain. CONCLUSIONS: Findings highlight categorical differences captured with the GCPS-R, and convergent validity supports use of the GCPS-R in US Veterans.

Trends in Risk-Adjusted Initiation and Reduction of Opioid Use among Veterans With Dementia in US Department of Veterans Affairs Community Living Centers.

OBJECTIVE: Proper initiation and reduction of opioids is important in providing effective and safe pain relief to Veterans with dementia, including in Community Living Centers (CLCs). We examined the trends in aggregated monthly risk-adjusted opioid administration days and dosage over 3 opioid safety regulatory periods: pre-Opioid Safety Initiative period (October 1, 2012-June 30, 2013; period 1), pre-CDC Clinical Practice Guideline period (January 1, 2014-November 30, 2015, period 2) and post-Veterans Affairs Clinical Practice Guideline period (March 1, 2017-September 30, 2018; period 3). DESIGN: A retrospective study between October 1, 2012, and September 30, 2018. SETTINGS AND PARTICIPANTS: 4995 long-stay CLC residents with dementia who had incident (incident cohort, n = 2609) or continued (continued opioid cohort, n = 2386) opioid administration in CLCs. METHODS: CLC Minimum Data Set (MDS) assessments data and bar-code medication administration data were used. Opioid initiation was examined for incident opioid cohort and reduction was examined using continued opioid cohort. We first computed aggregated monthly risk-adjusted opioid administration days, opioid with benzodiazepine administration days and opioid dosage, and then examined risk-adjusted incident and continued opioid administration trends over the regulatory periods controlling for facility-level characteristics. RESULTS: Among the incident opioid cohort, compared to period 1, there were 1.9 and 2.1 fewer risk-adjusted opioid administration days per month in periods 2 and 3, respectively; 1.5 fewer risk-adjusted days per month with opioid and benzodiazepine administration in both periods 2 and 3; and 2.2 and 3.7 morphine milligrams equivalent per day (MMED) lower risk-adjusted dosage in periods 2 and 3, respectively. Among the continued opioid cohort, compared to period 1, there were 1.6 and 2.9 fewer risk-adjusted days with opioid and benzodiazepine administration days per month in periods 2 and 3, respectively, and 5.3 MMED lower risk-adjusted dosage per month in period 3. CONCLUSIONS AND IMPLICATIONS: CLC providers initiated and reduced opioid administration in fewer days and at lower dosage among Veterans with dementia across the regulatory periods. The result was likely due to systemic efforts from health care professionals, CLC administrators, and policy makers or VA central office, aiming to reduce opioid misuse and improve quality of care in nursing home residents with dementia. What is still unknown is whether pain was well controlled or nonpharmacologic treatments were utilized.

Complementary and Integrative Health Approaches and Pain Care Quality in the Veterans Health Administration Primary Care Setting: A Quasi-Experimental Analysis.

Background: Complementary and integrative health (CIH) approaches have been recommended in national and international clinical guidelines for chronic pain management. We set out to determine whether exposure to CIH approaches is associated with pain care quality (PCQ) in the Veterans Health Administration (VHA) primary care setting. Methods: We followed a cohort of 62,721 Veterans with newly diagnosed musculoskeletal disorders between October 2016 and September 2017 over 1-year. PCQ scores were derived from primary care progress notes using natural language processing. CIH exposure was defined as documentation of acupuncture, chiropractic or massage therapies by providers. Propensity scores (PSs) were used to match one control for each Veteran with CIH exposure. Generalized estimating equations were used to examine associations between CIH exposure and PCQ scores, accounting for potential selection and confounding bias. Results: CIH was documented for 14,114 (22.5%) Veterans over 16,015 primary care clinic visits during the follow-up period. The CIH exposure group and the 1:1 PS-matched control group achieved superior balance on all measured baseline covariates, with standardized differences ranging from 0.000 to 0.045. CIH exposure was associated with an adjusted rate ratio (aRR) of 1.147 (95% confidence interval [CI]: 1.142, 1.151) on PCQ total score (mean: 8.36). Sensitivity analyses using an alternative PCQ scoring algorithm (aRR: 1.155; 95% CI: 1.150-1.160) and redefining CIH exposure by chiropractic alone (aRR: 1.118; 95% CI: 1.110-1.126) derived consistent results. Discussion: Our data suggest that incorporating CIH approaches may reflect higher overall quality of care for patients with musculoskeletal pain seen in primary care settings, supporting VHA initiatives and the Declaration of Astana to build comprehensive, sustainable primary care capacity for pain management. Future investigation is warranted to better understand whether and to what degree the observed association may reflect the therapeutic benefits patients actually received or other factors such as empowering provider-patient education and communication about these approaches.

Enhancing Clinical Data Analysis by Explaining Interaction Effects between Covariates in Deep Neural Network Models.

Deep neural network (DNN) is a powerful technology that is being utilized by a growing number and range of research projects, including disease risk prediction models. One of the key strengths of DNN is its ability to model non-linear relationships, which include covariate interactions. We developed a novel method called interaction scores for measuring the covariate interactions captured by DNN models. As the method is model-agnostic, it can also be applied to other types of machine learning models. It is designed to be a generalization of the coefficient of the interaction term in a logistic regression; hence, its values are easily interpretable. The interaction score can be calculated at both an individual level and population level. The individual-level score provides an individualized explanation for covariate interactions. We applied this method to two simulated datasets and a real-world clinical dataset on Alzheimer's disease and related dementia (ADRD). We also applied two existing interaction measurement methods to those datasets for comparison. The results on the simulated datasets showed that the interaction score method can explain the underlying interaction effects, there are strong correlations between the population-level interaction scores and the ground truth values, and the individual-level interaction scores vary when the interaction was designed to be non-uniform. Another validation of our new method is that the interactions discovered from the ADRD data included both known and novel relationships.

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.