Measurement and Modeling of Within-Person Variability in Cannabis Protective Behavioral Strategies: A Novel Approach Using Scale Development, Daily Data, and Machine Learning Methods

大麻保护行为策略中人内变异性的测量和建模：一种使用量表开发、每日数据和机器学习方法的新方法

• 批准号: 10604567
• 负责人: Kirstyn Nicole Smith-LeCavalier
• 金额: $ 4.77万
• 依托单位: UNIVERSITY OF WASHINGTON
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-03-16 至 2026-03-15
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10604567
• 关键词: Address; Age; Alcohol consumption; Alcohols; Algorithms; Behavior; Behavioral; Brain; Cannabis; Complement; Cross-Sectional Studies; DSM-IV; Data; Data Collection; Data Pooling; Data Science; Development; Education; Epidemiology; Etiology; Feedback; Focus Groups; Foundations; Frequencies; Future; Gambling; Goals; Grant; Harm Reduction; Impairment; Individual; Intervention; Intervention Studies; Judgment; Knowledge; Learning; Legal; Literature; Machine Learning; Manuscripts; Marijuana; Marijuana Dependence; Measurement; Measures; Mediating; Memory; Memory impairment; Methodology; Methods; Modeling; Motor; National Institute of Drug Abuse; Outcome; Persons; Phase; Prevalence; Prevention; Psychometrics; Public Health; Recommendation; Reporting; Research; Research Design; Research Personnel; Research Training; Resources; Review Literature; Risk Factors; Sampling; Sex Behavior; Social Change; Surveys; Techniques; Technology; Training; Validation; Work; Writing; addiction; condoms; design; improved; insight; interest; machine learning method; machine learning model; marijuana use; multilevel analysis; novel; novel strategies; personalized approach; phase 2 study; predictive modeling; psychosocial; random forest; research data dissemination; response; substance use; symposium; tool; undergraduate student; young adult

PROJECT SUMMARY/ABSTRACT Cannabis is the most commonly used controlled substance in the US.1,2 Young adults (YA; ages 18-25) report highest rates of use, and recent epidemiological surveys show an increase in both proportion of YA using cannabis and frequency of use among those that use.3,4 As frequent and heavy cannabis use is associated with a variety of short- and long-term unwanted physical and psychosocial outcomes (e.g., altered brain development, impaired judgement and memory, poor educational outcomes),5–7 there is a need for approaches to help individuals reduce use and/or use-related harms. One approach to mitigating substance- related harm is using protective behavioral strategies (PBS). PBS for cannabis consist of strategies an individual can use before, during, after, or instead of using to reduce use or consequence.8,9 In retrospective assessments, frequency of PBS use is associated with lower past 30-day cannabis use and consequences and mediates the relation between various risk factors and cannabis outcomes,8,10–15 highlighting PBS as a promising means of reducing cannabis use and harms. However, research on PBS-focused interventions is mixed9,16,17; this mixed support may be due to gaps in the PBS literature. Specifically, the majority of PBS research has consisted of cross-sectional, between-person retrospective designs, thus we lack understanding about which strategies work for whom under what circumstances. Emerging PBS research suggests both between- and within-person (i.e., daily) variability in whether an individual uses any PBS, and if so, which strategies they use. This suggests a need for a daily measure of cannabis PBS to increase understanding of how and when individuals utilize PBS and under what circumstances PBS are or are not effective. To address these gaps, the proposed F31 will take a novel, multimethod approach incorporating scale development work, a daily data study design, and machine learning methods. Specific Aims include (1) developing and validating a daily measure of cannabis PBS; and (2) utilizing a daily data design and machine learning techniques to develop models predicting PBS efficacy (reductions in use/consequences) for each strategy for a given individual. To complement these aims, the applicant will receive training in (1) etiology, prevention, intervention, and harm reduction methods for substance use, with a focus on cannabis PBS; (2) psychometric development and quantitative methods including multilevel modeling and machine learning; (3) daily data study design and methodology; and (4) research dissemination, including manuscript/grant writing and conference presentations. Study findings will have important implications for future PBS intervention research. Specifically, results can be used to better understand cannabis PBS on a daily level and improve future technology-based PBS interventions to reduce cannabis-related harms.

项目摘要/摘要 大麻是美国最常用的受控物质。1,2年轻人（Ya; 18-25岁） 报告最高的使用率和最近的流行病学调查显示，YA的两种比例都在增加 使用大麻和使用频率。 与各种短期和长期不必要的身体和社会心理结局相关（例如， 大脑发育，判断力和记忆力受损，教育成果不佳），5-7需要 帮助个人减少使用和/或使用相关危害的方法。一种减轻物质的方法 - 相关危害使用受保护的行为策略（PBS）。用于大麻的PBS包括策略 回顾性可以在，期间，之后，之后或不使用减少使用或后果。8,9 评估，PBS使用的频率与过去30天的大麻使用和后果有关 并调解了各种危险因素与大麻结局之间的关系，8,10-15强调PBS作为一个 减少大麻使用和危害的有希望的方法。但是，针对PBS的干预措施的研究是 混合9,16,17;这种混合支持可能是由于PBS文献中的差距。具体而言，大多数PBS 研究对横截面，人之间的回顾性设计是一致的，因此我们缺乏了解 关于哪些策略在什么情况下为谁起作用。新兴的PBS研究表明 个人是否使用任何PBS的人之间（即每天）之间的可变性，如果是，则 他们使用的策略。这表明需要每天衡量大麻PBS以增加对 个人如何以及何时利用PBS以及在什么情况下PBS是有效的。解决 这些差距，拟议的F31将采用一种新颖的多方法方法，以增加规模开发工作， 每日数据研究设计和机器学习方法。具体目的包括（1）开发和验证 每天测量大麻PBS； （2）使用日常数据设计和机器学习技术 开发预测给定策略的PBS效率（减少使用/后果）的模型 个人。为了补充这些目标，申请人将接受（1）病因，预防， 干预和减少药物使用方法，重点是大麻PBS； （2）心理测量学 开发和定量方法，包括多级建模和机器学习； （3）每日数据研究 设计和方法论； （4）研究传播，包括手稿/赠款写作和会议 演讲。研究发现将对未来的PBS干预研究具有重要意义。具体来说， 结果可以用来更好地理解每天的大麻PB，并改善未来的基于技术的PBS PBS干预措施以减少与大麻相关的危害。