Developing a nationwide registry of UK Veterans seeking help from sector charities-a machine learning approach to stratification

Abstract: The assistance to veterans in the UK is provided by the National Health Service and over 1800 military charities. These charities count services using different definitions and reporting systems, so to date a national registry of service usage does not exist. The aim of the Map Of Need Aggregation ResearCH study is to build a standardized registry of service usage data for the military charity sector. Data are completely anonymized by adopting a Secure Hashing Algorithm. A unique anonymous identifier is generated allowing both privacy protection and avoiding double counts. Data are standardized and linked with an automated process to create an aggregated dataset. The dataset describes the population, using both a priori and machine learning approaches. To date a total of 42 509 veterans with 128 423 needs are included. The mean age was 60.1 years, and 90% were male. 65% were receiving other benefits, 5% were homeless and 1% were in prison. 65% of the needs recorded concerned social wellbeing. 40% of veterans received assistance in at least two different years. The k-means clustering approach returned 4 subgroups of use that were identical to those created using a priori knowledge. The dataset is the most comprehensive source of veteran charity usage data in the UK to date. Service usage is generally homogenous among subgroups, but some differences were highlighted indicating that younger, non-officer veterans may be more at risk of presenting with more complex needs. These first useful insights can help allocate resources to build an effective preventive strategy for more complex cases.

Read the full article
Report a problem with this article

Related articles

  • More for Policy & Practice

    The ask, care, escort suite of trainings: Initial evaluation of the Army’s primary suicide prevention strategy

    Abstract: The U.S. Army’s Ask, Care, Escort (ACE) suicide gatekeeper training has been the annual requirement for all personnel since 2009; however, this training has never been formally evaluated. The present study evaluated three updated versions of ACE: a training for Army leaders (ACE-Suicide Intervention), a training for basic combat trainees (ACE for Basic Combat Training and One Station Unit Training), and a standard training for all personnel (ACE for the Force). Self-report surveys measured pre- to posttraining changes in objective and subjective knowledge and stigma, as well as preparedness, self-efficacy, and likelihood to engage in gatekeeper behaviors. Implementation outcomes, such as training acceptability, suitability, and usability were also assessed. Across these evaluations, participants reported that knowledge and gatekeeper behaviors significantly improved from pre- to posttraining. Implementation metrics revealed a high degree of acceptability and relevance for all three ACE trainings. Overall, the findings of these evaluations suggest important changes in key suicide prevention outcomes following the ACE suite of trainings. Further longitudinal assessment is needed to establish the full effectiveness of gatekeeper interventions in the Army.