The management of many health disorders often entails a sequential, individualized approach whereby treatment is adapted and re-adapted over time in response to the specific needs and evolving status of the individual. Adaptive interventions provide one way to operationalize the strategies (e.g., continue, augment, switch, step-down) leading to individualized sequences of treatment. An adaptive intervention is a sequence of decision rules that specify whether, how or when (timing) to individualize treatment in the course of an individual’s care. Often, a wide variety of critical questions must be answered when developing a high-quality adaptive intervention. Yet, there is often insufficient empirical evidence or theoretical basis to address these questions. The Sequential Multiple Assignment Randomized Trial (SMART)—a type of research design—was developed explicitly for the purpose of building high-quality adaptive interventions. SMART designs represent an attractive alternative to the standard randomized clinical trial when the overarching aim is to construct (as opposed to evaluate) a high-quality adaptive intervention.
The overarching goals of this two-day workshop are to (a) provide an introduction to adaptive interventions; (b) help you to gain the background needed to plan a Sequential Multiple Assignment Randomized Trial (SMART); and (c) help you learn how to implement data analytic methods with SMART study data to construct adaptive interventions.
General Description of Workshop
At this workshop, participants will be provided with a hard copy of all lecture notes, select computer exercises and output, and suggested reading lists for future reference. Three different formats will be used. First, all materials will be presented following the standard didactic format with a slideshow. Second, there will be practice exercises designed to help participants connect the material with their own research area. These practice exercises are focused on SMART study design principles, and they are aimed at helping to prepare participants to write a grant proposal that uses a SMART design to build an adaptive intervention. Third, there will be computer exercises using SAS®. Computer code and simulated data examples will be supplied by the instructors. The computer exercises will help investigators learn how to implement typical primary and secondary analyses using data arising from a SMART and to interpret the results. Throughout the workshop, ample time will be set aside for Q&A and discussion about how the concepts learned in class can be applied in participants’ research.
The prerequisites for this workshop are (1) familiarity with the basic principles of experimental (e.g., randomized trial) design, and (2) graduate-level statistics training for the behavioral, management, social or health sciences up through linear regression (usually two semesters of course work). Basic familiarity with the SAS® programming language is necessary for participation in the computer exercises (see next section).
Participants are strongly encouraged to bring a laptop so that they can participate in the computer exercises. To conduct analyses at the workshop, SAS® Version 9 for Windows must be installed on the laptop prior to arrival. In addition, approximately one week prior to the workshop, participants will be sent an email requesting that they download and install the free SAS® procedure PROC QLEARN. They also will be asked to download and run a few lines of code to verify that SAS® and PROC QLEARN are working on their computer. We regret that we cannot provide IT support at the workshop.
Sample of Topics to be Covered
- When and why adaptive interventions are needed
- How adaptive interventions differ from fixed (one-stage) interventions
- The critical components of adaptive interventions: decision points, tailoring variables, intervention options and decision rules
- The role of theory in developing adaptive interventions
- Examples of adaptive interventions from the literature
- The difference between a moderator variable and a tailoring variable
- SMART study principles, including how to provide a rationale for designing a SMART
- How SMARTs differ from standard randomized clinical trials
- Different types of SMART designs
- How to choose the sample size for a SMART (statistical power considerations)
- Common types of primary and secondary scientific aims in a SMART
- Data analytic strategies used to examine primary and secondary scientific aims in a SMART