A Software Product That Empowers Individuals Affected By Substance Use Disorders and Their Care Teams with Health and Social Resources

一款软件产品，可为受药物使用障碍影响的个人及其护理团队提供健康和社会资源

• 批准号: 10360396
• 负责人: Tom Lee
• 金额: $ 42.55万
• 依托单位: MEDICAL INNOVATORS COMPANY, LLC
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-04-15 至 2023-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10360396
• 关键词: Address; Affect; Algorithms; Artificial Intelligence; Businesses; Capital; Caring; Client; Communities; Community Health; Computer software; Cost Savings; Counseling; County; Coupled; Data; Distal; Eligibility Determination; Employment; Enrollment; Gender; Health; Health Personnel; Health Services; Health Services Accessibility; Housing; Individual; Intake; Internet; Interview; Law Enforcement; Lead; Link; Manuals; Mental Health; Modeling; Natural Language Processing; Outcome; Outpatients; Patient Participation; Patient Triage; Patients; Peer Review; Pharmaceutical Preparations; Phase; Privatization; Procedures; Process; Publications; Race; Recovery; Relapse; Reporting; Resources; Screening procedure; Services; Sex Bias; Small Business Innovation Research Grant; Social Impacts; Social Work; Stereotyping; Substance Use Disorder; Technology; Testing; Texas; Text Messaging; Time; Translating; Transportation; United States; Vendor; Work; Youth; behavioral health; child protective service; cohort; community partnership; digital; high risk; improved; mental health center; non-compliance; novel; opioid overdose; patient screening; phase 1 study; phase 2 study; point of care; programs; prototype; racial bias; recruit; scale up; screening; service programs; service providers; service utilization; social; social factors; social health determinants; societal costs; substance use treatment; tool; treatment program; treatment risk; young adult

Project Summary Access and participation in community resource programs such as transportation, housing and medication assistance - also known as social determinants of health (SDOH) is inextricably linked to a successful treatment and recovery in substance use disorders (SUD). Therefore, health and social service providers dedicate a significant amount of time to curate local community resource listings or “referral binders”. These “referral binders” are often highly duplicated and fragmented across organizations and by using expensive non-scalable solutions technology vendors in this domain have not fully addressed this problem. Furthermore, participation in such community programs is also hindered by the absence of highly efficient program eligibility screening tools. Undoubtedly, these shortcomings contribute to lack of direct access to recovery capital for individuals affected by SUD. We report two feasibility outcomes from our SBIR Phase I study. 1) A novel co-creation led business model that leverages local partnerships with subject matter expertise agencies on SUD and SDOH - can provide enhanced access to community resource programs at the point-of-care. 2) A highly adaptive digital assisted SDOH screening tool powered by novel conversational artificial intelligence (AI) and natural language processing (NLP) technologies can enhance patients’ participation in SDOH related community programs. This is achieved by using these technologies to perform program eligibility screenings through both web and text messaging channels coupled with patient triaging within a case work team. The purpose of this Phase II study is to: 1) Optimize and scale the community resource co-creation business model established in Phase I by onboarding eight co-creation partners in Texas within 2 years. 2) Examine the relationship between the conversational AI and NLP powered SDOH screening tool piloted in Phase I with SUD outcomes among youths and young adults. Specifically, we will partner with the Harris Center for Mental Health and IDD - the largest outpatient mental health provider in Texas to examine whether youths screened and enrolled into a treatment program using our technology report improved SUD outcomes. As AI and NLP are core pillars of our technology we will also investigate and subsequently correct for potential societal biases and stereotypes (e.g. race, gender) encoded in such technologies. This effort will avoid such biases to inadvertently determine outcomes in our downstream prediction tasks. The key Phase II milestones include; 1) Demonstrable evidence that the co-creation business model leads to enhanced access to SDOH and SUD community resources at the point-of-care. 2) An understanding on whether screening and eventual participation in a treatment program as supported by our conversational AI technology can lead to reduced encounters with law enforcement including CPS among youths with mental health and SUD. In summary, the proposed Phase II work will lead to enhanced access to SUD recovery capital and facilitate patients’ participation in local health and SDOH community programs.

项目概要 获得和参与社区资源计划，例如交通、住房和药物援助 - 也称为 健康的社会决定因素 (SDOH) 与药物滥用障碍 (SUD) 的成功治疗和康复有着千丝万缕的联系。 因此，健康和社会服务提供者投入大量时间来策划当地社区资源列表或 “推荐粘合剂”。这些“推荐活页夹”通常在各个组织中高度重复和分散，并且使用昂贵的 该领域的不可扩展解决方案技术供应商尚未完全解决这个问题。此外，参与此类 缺乏高效的计划资格筛选工具也阻碍了社区计划的实施。毫无疑问，这些 这些缺陷导致受 SUD 影响的个人无法直接获得恢复资金。 我们报告了 SBIR 第一阶段研究的两个可行性结果。 1) 一种利用本地资源的新颖的共同创造主导的商业模式 与 SUD 和 SDOH 主题专业机构建立伙伴关系 - 可以提供更多对社区资源的访问 护理点计划。 2) 高度自适应的数字辅助 SDOH 筛查工具，由新颖的对话人工技术提供支持 智能（AI）和自然语言处理（NLP）技术可以增强患者对 SDOH 相关社区的参与 程序。这是通过使用这些技术通过网络和短信执行计划资格筛选来实现的 渠道与案例工作团队内的患者分类相结合。 第二阶段研究的目的是： 1）优化和规模化建立的社区资源共创商业模式 第一阶段是在 2 年内在德克萨斯州引入八个共同创造合作伙伴。 2）检查对话式AI之间的关系 NLP 驱动的 SDOH 筛查工具在第一阶段进行了试点，并在青少年中取得了 SUD 结果。具体来说，我们将 与德克萨斯州最大的门诊心理健康提供者哈里斯心理健康中心和 IDD 合作，检查是否 使用我们的技术报告筛选并参加治疗计划的青少年改善了 SUD 的结果。以AI和NLP为核心 我们还将调查并纠正潜在的社会偏见和刻板印象（例如种族、 性别）编码在此类技术中。这项努力将避免这种偏见无意中确定我们下游的结果 预测任务。 第二阶段的关键里程碑包括： 1) 有明显证据表明，共同创造商业模式可以增强获得信息的机会 SDOH 和 SUD 护理点社区资源。 2) 了解筛选和最终参与是否 我们的对话式人工智能技术支持的治疗计划可以减少与执法部门的接触，包括 患有心理健康和 SUD 的青少年的 CPS。总之，拟议的第二阶段工作将提高 SUD 的可及性 回收资金并促进患者参与当地健康和 SDOH 社区计划。