Table 1.

Eight strategies in the evidence base for Personalizing Mental Health Interventions

Strategy	Description	Example
Therapies Adapted for Specific Subgroups	Empirically supported therapies (ESTs) adapted to improve outcomes or engagement in subgroups of individuals expected to respond poorly to ESTs.	Parent-Child Interaction Therapy (PCIT), an EST for disruptive behavior, adapted for Mexican American families; outperformed nonadapted PCIT at follow-up (McCabe & Yeh, 2009; McCabe et al., 2005, 2012).
Therapies Targeting Youths’ Environments	ESTs that alter or leverage environments (e.g., family, school, peers) thought to impact youth outcomes; therapists conduct treatment at least partly within these environments using formats tailored to patient needs, based on individualized goals.	Multisystemic Therapy, an EST for both delinquent and substance-abusing adolescents that is widely disseminated (Henggeler, 2011; Henggeler & Schaeffer, 2010; Schoenwald, 2010).
Modular Therapies	ESTs organized into self-contained modules that can be used multiple times or not at all, and combined as needed; decision-making flowcharts guide which modules to use and when to use them for a particular patient.	Modular Approach to Therapy for Children with Anxiety, Depression, Trauma, or Conduct Problems; outperformed standard ESTs at posttreatment and usual care at posttreatment and follow-up (Chorpita & Weisz, 2009; Weisz et al., 2012; Chorpita et al., 2013).
Sequential, Multiple Assignment, Randomized Trials (SMARTs)	A trial design that randomizes individuals to a first-stage treatment or assessment condition, assesses response, then potentially randomizes individuals to next-stage treatment options based on their response; generates evidence for constructing decision rules in sequencing treatments.	A SMART of minimally verbal children with autism found superior outcomes for communication intervention augmented by a speech-generating device (vs. non-augmented intervention), and, for nonresponders after three months, intensified augmented intervention (vs. intensified non-augmented intervention; Kasari et al., 2014).
Measurement Feedback Systems	A system of administrating assessments of treatment outcomes and progress indicators that are psychometrically sound, sensitive to clinical change, brief, and clinically useful; then storing and displaying the data in meaningful formats to provide feedback about how well treatment is working.	The Youth Outcome Questionnaire and Youth Outcome Questionnaire Self-Report have identified youths at-risk of treatment failure; Youth-Clinical Support Tools pinpoint obstacles and suggest solutions (Burlingame et al., 2001; Cannon et al., 2010; Ridge et al., 2009; Warren & Lambert, 2012; Warren et al., 2012).
Meta-analyses Comparing Treatments for Specific Patient Characteristics	Research syntheses of randomized trials comparing alternative treatment strategies or types directly (i.e., within the same trial) among patients with specific characteristics.	A meta-analysis compared psychotherapy, medication, and combination psychotherapy-medication for subgroups of depressed adults and found sufficient evidence to recommend medication for dysthymia and combination treatment for older adults and outpatients (Cuijpers et al., 2012).
Data-mining Decision Trees	Models that guide decision-making based on multiple characteristics of individuals; developed through data mining, an exploratory approach for detecting and interpreting patterns in data.	The Distillation and Matching Model mined data from youth psychotherapy trials to develop a tool to select efficacious treatments, or their elements, based on patient characteristics; produced medium-large pre-post effects as part of a comprehensive service model (Chorpita & Daleiden, 2013; Chorpita et al., 2005; Southam-Gerow et al., 2013).
Individualized Metrics	Indices that quantify the benefit each patient is expected to receive from alternative interventions by accounting for one or more characteristics of the patient.	Probability of treatment benefit (PTB) was computed for a randomized trial of youth anxiety treatments at different levels of pretreatment symptom severity; PTB differed across treatments only for severe anxiety, with highest PTB for combination CBT-SSRI (Beidas et al., 2014; Lindhiem et al., 2012).