The onset and persistence of life events (LE) such as housing instability, job instability, and reduced social connection have been shown to increase risk of suicide. Predictive models for suicide risk have low sensitivity to many of these factors due to under-reporting in structured electronic health records (EHR) data. In this study, we show how natural language processing (NLP) can help identify LE in clinical notes at higher rates than reported medical codes. We compare domain-specific lexicons formulated from Unified Medical Language System (UMLS) selection, content analysis by subject matter experts (SME) and the Gravity Project, to datadriven expansion through contextual word embedding using Word2Vec. Our analysis covers EHR from the VA Corporate Data Warehouse (CDW) and measures the prevalence of LE across time for patients with known underlying cause of death in the National Death Index (NDI). We found that NLP methods had higher sensitivity of detecting LE relative to structured EHR variables. We observed that, on average, suicide cases had higher rates of LE over time when compared to patients who died of non-suicide related causes with no previous history of diagnosed mental illness. When used to discriminate these outcomes, the inclusion of NLP derived variables increased the concentration of LE along the top 0.1%, 0.5% and 1% of predicted risk. LE were less informative when discriminating suicide death from non suicide related death for patients with diagnosed mental illness.
The objectives of the present research were to examine the prevalence of deliberate self-harm (DSH) among 214 U.S. male Iraq/Afghanistan-era veterans seeking treatment for posttraumatic stress disorder (PTSD) and to evaluate the relationship between DSH and suicidal ideation within this population. Approximately 56.5% (n = 121) reported engaging in DSH during their lifetime; 45.3% (n = 97) reported engaging in DSH during the previous 2 weeks. As hypothesized, DSH was a significant correlate of suicidal ideation among male Iraq/Afghanistan-era veterans, OR = 3.88, p < .001, along with PTSD symptom severity, OR = 1.03, p < .001, and combat exposure, OR = 0.96, p = .040. A follow-up analysis identified burning oneself, OR = 17.14, p = .017, and hitting oneself, OR = 7.93, p < .001, as the specific DSH behaviors most strongly associated with suicidal ideation. Taken together, these findings suggest that DSH is quite prevalent among male Iraq/Afghanistan-era veterans seeking treatment for PTSD and is associated with increased risk for suicidal ideation within this population. Routine assessment of DSH is recommended when working with male Iraq/Afghanistan veterans seeking treatment for PTSD.
Posttraumatic stress disorder (PTSD) is associated with increased smoking initiation, maintenance, and relapse. Dehydroepiandrosterone (DHEA) and DHEA sulfate (DHEAS) are neurosteroids that have been associated with mood measures as well as smoking status, and nicotine is associated with increased DHEA and DHEAS levels. Given the difficulties with mood experienced by smokers with PTSD, the purpose of the current study was to evaluate the association between negative affect and anxiety sensitivity with DHEA and DHEAS levels. Ninety-six smokers with and without PTSD provided blood samples for neurosteroid analyses and completed self-report measures of anxiety sensitivity and electronic diary ratings of negative affect. As expected, PTSD smokers reported higher levels of anxiety sensitivity (F(1,94) = 20.67, partial η2 = 0.18, P < 0.0001) and negative affect (F(1,91) = 7.98, partial η2 = 0.08, P = 0.006). After accounting for age and sex, DHEAS was significantly inversely associated with both anxiety sensitivity (F(3,92) = 6.97, partial η2 = 0.07, P = 0.01) and negative affect (F(3,87) = 10.52, partial η2 = 0.11, P = 0.002) across groups. Effect sizes indicated that these effects are moderate to high. No significant interactions of diagnosis and DHEA(S) levels with mood measures were detected. Given that nicotine is known to elevate DHEA(S) levels, these results suggest that DHEAS may serve as a biomarker of the association between mood and nicotine among smokers. Implications for the results include (1) the use of DHEAS measurement across time and across quit attempts and (2) the potential for careful use of DHEA supplementation to facilitate abstinence during smoking cessation.
