Epidemiologic and Machine Learning Approaches to Frame Suicide Prevention Strategies Among Juvenile Justice Youth - 2021

项目名称 制定青少年自杀预防策略的流行病学和机器学习方法 正义青年 项目摘要/摘要 自杀是美国10-24岁年轻人的第二大死因 被发现自杀行为发生率较高的是被关押在少年司法系统中的青少年。而当 在减少青少年教养所内青少年自杀方面取得了进展，最低限度 已经考虑到青少年在被监禁后自杀的风险。之前曾被监禁 年轻人面临着重新融入社区的众多挑战，这可能会增加他们患 自杀行为。估计还表明，参与司法系统的60%至80%的青年 严重的心理健康问题，以及在系统中花费的时间会加剧这些情况。未被满足的人 参与司法系统的青年对心理健康服务的需求也是一个严重问题。尽管 在这一弱势群体中，公认的风险，循证自杀预防策略并不是 作为为出狱青年提供的例行重返社会服务的一部分。人们对此知之甚少 在少年司法系统中实施这些战略的成功办法。为了解决这一差距， 建议的研究使用创新的机器学习技术来开发风险预测模型 结合了青年的特点和与监禁(暴力、 受害、隔离/隔离做法、保健服务)以更准确地评估#年的自杀风险 少年犯被监禁。在这些发现和结构化实施科学框架的指导下， 该提案还将与主要利益攸关方进行干预前评估，以验证 通知干预选择的机器学习模型。我们会考虑潜在的 将模型整合到实践中的促进者和障碍，以及何时、如何和在哪里干预 少年司法程序。实现这项拟议研究的目标将提供有针对性的干预 关于在少年司法系统中预防高危青年自杀的建议。这项建议 还将支持鲁赫博士的培训，他致力于减少青年自杀的研究事业 涉足司法系统。鲁赫博士的培训计划包括：(1)获得机器学习技能和 预测建模技术，以更准确地识别自杀风险并为有针对性的预防提供信息 司法系统中的青年(2)加强预防自杀干预措施的知识，包括保健 服务系统，以了解卫生保健做法和政策可能如何促进或阻碍干预 针对参与司法系统的青年和(3)加强执行科学和 先进的定性研究方法，弥合研究与实践之间的差距。这条调查路线 将通过引入创新的自杀技术方法来推动青少年自杀研究领域的发展 在显著高危和服务不足的人群中进行预防，同时也产生了关于 潜在资源有限的少年司法系统中的独特执行挑战。

PROJECT TITLE Epidemiologic and Machine Learning Approaches to Frame Suicide Prevention Strategies Among Juvenile Justice Youth PROJECT SUMMARY / ABSTRACT Suicide is the second leading cause of death among youth aged 10-24 years in the U.S. One population found to have higher rates of suicidal behavior is youth incarcerated in the juvenile justice system. While progress has been made to reduce suicide for youth within juvenile correctional facilities, minimal consideration has been given to the risk for suicide in youth after incarceration. Previously incarcerated youth face numerous challenges reintegrating back into the community which can increase their risk for suicidal behavior. Estimates further suggest that 60% to 80% of youth involved in the justice system have significant mental health issues, and time spent in the system can exacerbate these conditions. The unmet need for mental health services by youth involved with the justice system is also a serious problem. Despite the recognized risk in this vulnerable population, evidence-based suicide prevention strategies are not integrated as part of routine reentry services for youth released from confinement. Even less is known about successful approaches to implement these strategies in juvenile justice systems. To address this gap, the proposed study uses innovative machine learning techniques to develop a risk prediction model incorporating youth characteristics and contextual factors associated with confinement (violence, victimization, segregation /isolation practices, health care services) to more accurately assess suicide risk in youth following incarcerated. Guided by these findings and a structured implementation science framework, this proposal will also conduct a pre-intervention assessment with key stakeholders to validate the utility of the machine learning model to inform intervention selection. Consideration will be given to potential facilitators and barriers to integrating the model into practice, and when, how, and where to intervene in the juvenile justice process. Achieving the aims of this proposed study will provide targeted intervention recommendations for suicide prevention among at-risk youth in the juvenile justice system. This proposal will also support the training of Dr. Ruch, who is devoted to a research career to reduce suicide in youth involved with justice system. Dr. Ruch’s training plan includes: (1) acquiring skills in machine learning and forecast modeling techniques to more accurately identify suicide risk and inform targeted preventions for youth in the justice system (2) enhancing knowledge of suicide prevention interventions, including health service systems to understand how health care practices and policies may facilitate or impede intervention for youth involved with the justice system and (3) strengthening skills in implementation science and advanced qualitative research methods to bridge the gap between research and practice. This line of inquiry will further the field of youth suicide research by introducing innovative technological approaches for suicide prevention in a significantly high-risk and underserved population, while also generating new insights about the distinct implementation challenges in the potentially resource constrained juvenile justice system.