Validation of an Electronic Health Record–Based Suicide Risk Prediction Modeling Approach Across Multiple Health Care Systems

Barak-Corren Y, Castro VM, Nock MK, et al. Validation of an electronic health Record–Based suicide risk prediction modeling approach across multiple health care systems. JAMA Netw Open. 2020;3(3):e201262. https://doi.org/10.1001/jamanetworkopen.2020.1262.

Importance  Suicide is a leading cause of mortality, with suicide-related deaths increasing in recent years. Automated methods for individualized risk prediction have great potential to address this growing public health threat. To facilitate their adoption, they must first be validated across diverse health care settings.

Objective  To evaluate the generalizability and cross-site performance of a risk prediction method using readily available structured data from electronic health records in predicting incident suicide attempts across multiple, independent, US health care systems.

Design, Setting, and Participants  For this prognostic study, data were extracted from longitudinal electronic health record data comprising International Classification of Diseases, Ninth Revision diagnoses, laboratory test results, procedures codes, and medications for more than 3.7 million patients from 5 independent health care systems participating in the Accessible Research Commons for Health network. Across sites, 6 to 17 years’ worth of data were available, up to 2018. Outcomes were defined by International Classification of Diseases, Ninth Revision codes reflecting incident suicide attempts (with positive predictive value >0.70 according to expert clinician medical record review). Models were trained using naive Bayes classifiers in each of the 5 systems. Models were cross-validated in independent data sets at each site, and performance metrics were calculated. Data analysis was performed from November 2017 to August 2019.

Main Outcomes and Measures  The primary outcome was suicide attempt as defined by a previously validated case definition using International Classification of Diseases, Ninth Revision codes. The accuracy and timeliness of the prediction were measured at each site.

Results  Across the 5 health care systems, of the 3 714 105 patients (2 130 454 female [57.2%]) included in the analysis, 39 162 cases (1.1%) were identified. Predictive features varied by site but, as expected, the most common predictors reflected mental health conditions (eg, borderline personality disorder, with odds ratios of 8.1-12.9, and bipolar disorder, with odds ratios of 0.9-9.1) and substance use disorders (eg, drug withdrawal syndrome, with odds ratios of 7.0-12.9). Despite variation in geographical location, demographic characteristics, and population health characteristics, model performance was similar across sites, with areas under the curve ranging from 0.71 (95% CI, 0.70-0.72) to 0.76 (95% CI, 0.75-0.77). Across sites, at a specificity of 90%, the models detected a mean of 38% of cases a mean of 2.1 years in advance.

Conclusions and Relevance  Across 5 diverse health care systems, a computationally efficient approach leveraging the full spectrum of structured electronic health record data was able to detect the risk of suicidal behavior in unselected patients. This approach could facilitate the development of clinical decision support tools that inform risk reduction interventions.

Follow us on Facebook to stay updated on all the library news.

Lock to Live—An Interactive Web-Based Lethal Means Safety Decision Aid for Suicidal Adults: Pilot Randomized Controlled Trial

Betz ME, Knoepke CE, Simpson S, et al. Lock to Live—An interactive web-based lethal means safety decision aid for suicidal adults: Pilot randomized controlled trial. J Med Internet Res. 2020;22(1). https://doi.org/10.2196/16253.

Background: Counseling to reduce access to lethal means such as firearms and medications is recommended for suicidal adults but does not routinely occur. We developed the Web-based Lock to Live (L2L) decision aid to help suicidal adults and their families choose options for safer home storage.

Objective: This study aimed to test the feasibility and acceptability of L2L among suicidal adults in emergency departments (EDs).

Methods: At 4 EDs, we enrolled participants (English-speaking, community-dwelling, suicidal adults) in a pilot randomized controlled trial. Participants were randomized in a 13:7 ratio to L2L or control (website with general suicide prevention information) groups and received a 1-week follow-up telephone call.

Results: Baseline characteristics were similar between the intervention (n=33) and control (n=16) groups. At baseline, many participants reported having access to firearms (33/49, 67%), medications (46/49, 94%), or both (29/49, 59%). Participants viewed L2L for a median of 6 min (IQR 4-10 min). L2L also had very high acceptability; almost all participants reported that they would recommend it to someone in the same situation, that the options felt realistic, and that L2L was respectful of values about firearms. In an exploratory analysis of this pilot trial, more participants in the L2L group reported reduced firearm access at follow-up, although the differences were not statistically significant.

Conclusions: The L2L decision aid appears feasible and acceptable for use among adults with suicide risk and may be a useful adjunct to lethal means counseling and other suicide prevention interventions. Future large-scale studies are needed to determine the effect on home access to lethal means.

The eIF2α Kinase GCN2 Modulates Period and Rhythmicity of the Circadian Clock by Translational Control of Atf4

Pathak, S. S., Liu, D., Li, T., de Zavalia, N., Zhu, L., Li, J., . . . Cao, R. (2019). The eIF2α kinase GCN2 modulates period and rhythmicity of the circadian clock by translational control of Atf4. https://doi.org/10.1016/j.neuron.2019.08.007

The integrated stress response (ISR) is activated in response to diverse stress stimuli to maintain homeostasis in neurons. Central to this process is the phosphorylation of eukaryotic translation initiation factor 2 alpha (eIF2α). Here, we report a critical role for ISR in regulating the mammalian circadian clock. The eIF2α kinase GCN2 rhythmically phosphorylates eIF2α in the suprachiasmatic circadian clock. Increased eIF2α phosphorylation shortens the circadian period in both fibroblasts and mice, whereas reduced eIF2α phosphorylation lengthens the circadian period and impairs circadian rhythmicity in animals. Mechanistically, phosphorylation of eIF2α promotes mRNA translation of Atf4. ATF4 binding motifs are identified in multiple clock genes, including Per2, Per3, Cry1, Cry2, and Clock. ATF4 binds to the TTGCAGCA motif in the Per2 promoter and activates its transcription. Together, these results demonstrate a significant role for ISR in circadian physiology and provide a potential link between dysregulated ISR and circadian dysfunction in brain diseases.

Decreased Nociceptin Receptors Are Related to Resilience and Recovery in College Women Who Have Experienced Sexual Violence: Therapeutic Implications for Posttraumatic Stress Disorder

Narendran R, Tollefson S, Fasenmyer K, et al. Decreased nociceptin receptors are related to resilience and recovery in college women who have experienced sexual violence: Therapeutic implications for posttraumatic stress disorder. Biological Psychiatry. 2019. https://doi.org/10.1016/j.biopsych.2019.02.017.

Background

Posttraumatic stress disorder (PTSD) is a stress disorder that develops in only some individuals following a traumatic event. Data suggest that a substantial fraction of women recover after sexual violence. Thus, the investigation of stress and antistress neuropeptides in this sample has the potential to inform the neurochemistry of resilience following trauma. Nociceptin is an antistress neuropeptide in the brain that promotes resilience in animal models of PTSD.

Methods

[¹¹C]NOP-1A positron emission tomography was used to measure the in vivobinding to nociceptin receptors in 18 college women who had experienced sexual violence irrespective of whether they met DSM-5 diagnostic criteria for PTSD. [¹¹C]NOP-1A data from 18 healthy control subjects were also included to provide a contrast with the sexual violence group. [¹¹C]NOP-1A total distribution volume (V_T) in the regions of interest were measured with kinetic analysis using the arterial input function. The relationships between regional V_T and Clinician-Administered PTSD Scale for DSM-5 total symptom and subscale severity were examined using correlational analyses.

Results

No differences in [¹¹C]NOP-1A V_T were noted between the sexual violence and control groups. V_T in the midbrain and cerebellum were positively correlated with PTSD total symptom severity in the past month before positron emission tomography. Intrusion/re-experiencing and avoidance subscale symptoms drove this relationship. Stratification of subjects by a DSM-5 PTSD diagnosis and contrasting their V_T with that in control subjects showed no group differences.

Conclusions

Decreased midbrain and cerebellum nociceptin receptors are associated with less severe PTSD symptoms. Medications that target nociceptin should be explored to prevent and treat PTSD.