Improving explainability of post-separation suicide attempt prediction models for transitioning service members: insights from the Army Study to Assess Risk and Resilience in Servicemembers — Longitudinal Study
Abstract Risk of U.S. Army soldier suicide-related behaviors increases substantially after separation from service. As universal prevention programs have been unable to resolve this problem, a previously reported machine learning model was developed using pre-separation predictors to target high-ris...
Saved in:
| Main Authors: | , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Publishing Group
2025-01-01
|
| Series: | Translational Psychiatry |
| Online Access: | https://doi.org/10.1038/s41398-025-03248-z |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571324210348032 |
|---|---|
| author | Emily R. Edwards Joseph C. Geraci Sarah M. Gildea Claire Houtsma Jacob A. Holdcraft Chris J. Kennedy Andrew J. King Alex Luedtke Brian P. Marx James A. Naifeh Nancy A. Sampson Murray B. Stein Robert J. Ursano Ronald C. Kessler |
| author_facet | Emily R. Edwards Joseph C. Geraci Sarah M. Gildea Claire Houtsma Jacob A. Holdcraft Chris J. Kennedy Andrew J. King Alex Luedtke Brian P. Marx James A. Naifeh Nancy A. Sampson Murray B. Stein Robert J. Ursano Ronald C. Kessler |
| author_sort | Emily R. Edwards |
| collection | DOAJ |
| description | Abstract Risk of U.S. Army soldier suicide-related behaviors increases substantially after separation from service. As universal prevention programs have been unable to resolve this problem, a previously reported machine learning model was developed using pre-separation predictors to target high-risk transitioning service members (TSMs) for more intensive interventions. This model is currently being used in a demonstration project. The model is limited, though, in two ways. First, the model was developed and trained in a relatively small cross-validation sample (n = 4044) and would likely be improved if a larger sample was available. Second, the model provides no guidance on subtyping high-risk TSMs. This report presents results of an attempt to refine the model to address these limitations by re-estimating the model in a larger sample (n = 5909) and attempting to develop embedded models for differential risk of post-separation stressful life events (SLEs) known to mediate the association of model predictions with post-separation nonfatal suicide attempts (SAs; n = 4957). Analysis used data from the Army STARRS Longitudinal Surveys. The revised model improved prediction of post-separation SAs in the first year (AUC = 0.85) and second-third years (AUC = 0.77) after separation, but embedded models could not predict post-separation SLEs with enough accuracy to support intervention targeting. |
| format | Article |
| id | doaj-art-b1778a04203547baa8ec55d2c5fe509a |
| institution | Kabale University |
| issn | 2158-3188 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Publishing Group |
| record_format | Article |
| series | Translational Psychiatry |
| spelling | doaj-art-b1778a04203547baa8ec55d2c5fe509a2025-02-02T12:43:31ZengNature Publishing GroupTranslational Psychiatry2158-31882025-01-0115111010.1038/s41398-025-03248-zImproving explainability of post-separation suicide attempt prediction models for transitioning service members: insights from the Army Study to Assess Risk and Resilience in Servicemembers — Longitudinal StudyEmily R. Edwards0Joseph C. Geraci1Sarah M. Gildea2Claire Houtsma3Jacob A. Holdcraft4Chris J. Kennedy5Andrew J. King6Alex Luedtke7Brian P. Marx8James A. Naifeh9Nancy A. Sampson10Murray B. Stein11Robert J. Ursano12Ronald C. Kessler13VISN 2 MIRECC, Department of Veterans AffairsVISN 2 MIRECC, Department of Veterans AffairsDepartment of Health Care Policy, Harvard Medical SchoolSoutheast Louisiana Veterans Health Care SystemDepartment of Health Care Policy, Harvard Medical SchoolDepartment of Psychiatry, Harvard Medical SchoolDepartment of Health Care Policy, Harvard Medical SchoolDepartment of Statistics, University of WashingtonNational Center for PTSD, VA Boston Healthcare SystemDepartment of Psychiatry, Center for the Study of Traumatic Stress, Uniformed Services University of the Health SciencesDepartment of Health Care Policy, Harvard Medical SchoolDepartment of Psychiatry, University of California San DiegoDepartment of Psychiatry, Center for the Study of Traumatic Stress, Uniformed Services University of the Health SciencesDepartment of Health Care Policy, Harvard Medical SchoolAbstract Risk of U.S. Army soldier suicide-related behaviors increases substantially after separation from service. As universal prevention programs have been unable to resolve this problem, a previously reported machine learning model was developed using pre-separation predictors to target high-risk transitioning service members (TSMs) for more intensive interventions. This model is currently being used in a demonstration project. The model is limited, though, in two ways. First, the model was developed and trained in a relatively small cross-validation sample (n = 4044) and would likely be improved if a larger sample was available. Second, the model provides no guidance on subtyping high-risk TSMs. This report presents results of an attempt to refine the model to address these limitations by re-estimating the model in a larger sample (n = 5909) and attempting to develop embedded models for differential risk of post-separation stressful life events (SLEs) known to mediate the association of model predictions with post-separation nonfatal suicide attempts (SAs; n = 4957). Analysis used data from the Army STARRS Longitudinal Surveys. The revised model improved prediction of post-separation SAs in the first year (AUC = 0.85) and second-third years (AUC = 0.77) after separation, but embedded models could not predict post-separation SLEs with enough accuracy to support intervention targeting.https://doi.org/10.1038/s41398-025-03248-z |
| spellingShingle | Emily R. Edwards Joseph C. Geraci Sarah M. Gildea Claire Houtsma Jacob A. Holdcraft Chris J. Kennedy Andrew J. King Alex Luedtke Brian P. Marx James A. Naifeh Nancy A. Sampson Murray B. Stein Robert J. Ursano Ronald C. Kessler Improving explainability of post-separation suicide attempt prediction models for transitioning service members: insights from the Army Study to Assess Risk and Resilience in Servicemembers — Longitudinal Study Translational Psychiatry |
| title | Improving explainability of post-separation suicide attempt prediction models for transitioning service members: insights from the Army Study to Assess Risk and Resilience in Servicemembers — Longitudinal Study |
| title_full | Improving explainability of post-separation suicide attempt prediction models for transitioning service members: insights from the Army Study to Assess Risk and Resilience in Servicemembers — Longitudinal Study |
| title_fullStr | Improving explainability of post-separation suicide attempt prediction models for transitioning service members: insights from the Army Study to Assess Risk and Resilience in Servicemembers — Longitudinal Study |
| title_full_unstemmed | Improving explainability of post-separation suicide attempt prediction models for transitioning service members: insights from the Army Study to Assess Risk and Resilience in Servicemembers — Longitudinal Study |
| title_short | Improving explainability of post-separation suicide attempt prediction models for transitioning service members: insights from the Army Study to Assess Risk and Resilience in Servicemembers — Longitudinal Study |
| title_sort | improving explainability of post separation suicide attempt prediction models for transitioning service members insights from the army study to assess risk and resilience in servicemembers longitudinal study |
| url | https://doi.org/10.1038/s41398-025-03248-z |
| work_keys_str_mv | AT emilyredwards improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT josephcgeraci improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT sarahmgildea improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT clairehoutsma improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT jacobaholdcraft improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT chrisjkennedy improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT andrewjking improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT alexluedtke improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT brianpmarx improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT jamesanaifeh improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT nancyasampson improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT murraybstein improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT robertjursano improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy AT ronaldckessler improvingexplainabilityofpostseparationsuicideattemptpredictionmodelsfortransitioningservicemembersinsightsfromthearmystudytoassessriskandresilienceinservicememberslongitudinalstudy |