Discovering Tailoring Variables in Childhood Mental Health Treatment Research

描述（由申请人提供）：儿童精神疾病通常需要临床医生做出个性化的治疗决定。在儿童和青少年心理健康（CAMH）研究和实践的一个令人兴奋的趋势中，人们对制定个性化决策规则（IDRs）的兴趣和认识日益增加。IDRs使用社会人口统计信息、病史、合并症、家庭/人际测量或生物标志物作为输入，然后推荐后续治疗的类型、方式、强度或交付作为输出。IDR为个性化医疗提供了一个框架：个性化水平与IDR中用作输入的有用信息（裁剪变量）的数量成比例地增加。开发用于治疗精神疾病儿童的idr是干预研究的重要“下一步”。大规模临床试验已经为许多常见的CAMH疾病建立了有效的干预措施。然而，尽管有强有力的证据表明治疗方法平均有效，但对于哪些儿童更有可能或更少可能从一种治疗或另一种治疗中受益，人们知之甚少。在idr发挥其潜力之前，关于如何利用现有数据协助其发展，还有许多方法学方面的工作要做。制定有效idr的关键步骤是确定适合的变量

DESCRIPTION (provided by applicant): Childhood psychiatric conditions often require that clinicians make individualized treatment decisions. In an exciting trend in child and adolescent mental health (CAMH) research and practice, there is growing interest and awareness in developing individualized decision rules (IDRs). IDRs use sociodemographic information, illness history, comorbidities, family/interpersonal measures, or biomarkers as inputs and then recommend the type, modality, and intensity or delivery of subsequent treatment(s) as outputs. IDRs provide a framework for personalized medicine: the level of personalization increases in proportion to the amount of useful information (tailoring variables) used as inputs in an IDR. The development of IDRs for the treatment of children with mental illness is an important "next step" in interventions research. Large scale clinical trials have established efficacious interventions fr many of the common CAMH disorders. However, despite a strong evidence base for treatments that are effective on average, little is known about which children are more or less likely to benefit from one treatment or another. Before IDRs can fulfill their potential, much methodological work remains to be done regarding how to use currently available data to assist in their development. A key step in developing effective IDRs is identifying tailoring variables to use as inputs. Tailoring variables are special types of pre-treatment measures that pinpoint which treatment is best for whom. Existing methods do not explicitly (a) identify which measures are most useful (tailoring variable selection), or (b) combine them in an optimal fashion (tailorin variable feature construction). The specific aims of this study are: Aim 1 (Methods). To develop and evaluate a framework for (a) tailoring variable selection and (b) tailoring variable feature construction for individualizing CAMH treatment. The new method will be used to address scientific questions such as: (a) What are the most important measures (e.g., demographic, severity and duration of illness, comorbidity, parental psychopathology, psychosocial environment including family and peer relationships, biological) to use for tailoring treatment? (b) How do we combine them in order to guide treatment decisions (e.g., individual psychotherapy treatment, medication, or their combination)? Aim 2 (Application): To illustrate and evaluate the method by applying it to the Child/Adolescent Anxiety Multimodal Study randomized trial data to construct an IDR for choosing between SSRI, CBT, or SSRI+CBT for children with anxiety disorders. We will apply the new methodology to identify variables and features pinpointing children who are more or less likely to benefit from SSRI, CBT, or combined SSRI+CBT. Results will be used to construct a proposed IDR for pediatric anxiety disorders. We will compare how average symptom and severity outcomes differ on the IDR vs SSRI for all vs CBT for all vs SSRI+CBT for all. This study will provide researchers in CAMH with a novel tool for building IDRs. This study has significant public health impact because it aims to develop tools to personalize treatment for children and adolescents with mental health disorders. This is a high-priority area of research for the National Institute of Mental Health.