Overview: Sequential Multiple Assignment Randomized Trials (SMARTs) randomize participants into different sequences of intervention options based on a set of decision rules about when to adjust a participant’s treatment. This design allows researchers to examine the effects of different sets of interventions and development of adaptive approaches to better meet participants’ needs. SMARTs have four components:
- A sequence of decisions about the points at which to offer different services. For example, researchers need to decide what type of service should be provided next if a participant does not respond positively to the initial service.
- A set of intervention options for each decision point. For example, if a participant does not show the desired outcome at a decision point, two pre-defined options might be either providing more of the same intervention (that is, a higher dosage) or else switching to an alternative intervention. Participants who do not show the desired outcome are then randomized to one of these options.
- The factors that indicate a change in service is needed (that is, the tailoring variables). Examples include signs the participant is not responding to the current intervention or is experiencing unwanted side effects from it.
- A sequence of decision rules that links the first three components. There is one rule for every decision.
Hallmarks: Researchers often use SMARTs to—
- Test ways to tailor when and which service to offer to a specific participant
- Determine how different interventions, or combinations of interventions, lead to different outcomes rather than just looking at outcomes from one service approach
- Decrease attrition rates because participants who might exit the study because they aren’t responding to an initial service get re-randomized into another service that may work better
Benefits for Precision Home Visiting: SMART research can help home visitors identify the most effective way to tailor and sequence interventions to achieve intended outcomes efficiently. The results from SMARTs can also inform how to design interventions, identify target participant groups, and determine follow-up activities.
- Lei, H., Nahum-Shani, I., Lynch, K., Oslin, D., & Murphy, S. A. (2012). A “SMART” design for building individualized treatment sequences. Annual Review of Clinical Psychology, 8, 14.1-14.28. doi: 10.1146/annurev-clinpsy-032511-143152 PMC Journal- In process
- Murphy, S.A. (2005). An experimental design for the development of adaptive treatment strategies. Statistics in Medicine, 24, 1455-1481.
- Laber, E. (2018, April). SMARTs. Presentation from the HARC 2018 Methods Meeting, Chicago, IL. Retrieved from https://www.youtube.com/watch?v=RaVexEpr9vY