Home > Research . Search . Country . Browse . Small Grants

PSC In The News

RSS Feed icon

Thompson says criminal justice policies led to creation of prison gangs like Aryan Brotherhood

Schmitz finds job loss before retirement age contributes to weight gain, especially in men

Kimball says Fed should get comfortable with "backtracking"

Highlights

Overview of Michigan's advanced research computing resources, Monday, June 27, 9-10:30 am, BSRB - Kahn Auditorium

U-M's Data Science Initiative offers expanded consulting services via CSCAR

Elizabeth Bruch promoted to Associate Professor

Susan Murphy elected to the National Academy of Sciences

Next Brown Bag

PSC Brown Bags
will resume fall 2016

Susan A. Murphy photo

SMART Methodology for Constructing Adaptive Intervention

a PSC Research Project

Investigators:   Susan A. Murphy, Daniel Almirall

The long term-goal of this project is to improve clinical practice and thus public health by facilitating the evidence-based construction of efficacious, individualized, adaptive interventions and treatments in drug abuse. Clinicians naturally adapt the level and type of therapy according to patient outcomes such as severity, response to past therapy, risk, stressors, adherence, preference and burden. This project will develop methods for using data to inform and enhance this adaptive clinical practice. Adaptive interventions are composed of operationalized decision rules that input patient outcomes and output recommended alterations in intensity and/or type of therapy.

The construction of adaptive interventions requires addressing questions such as: How do we best use measures of risk and other outcomes to decide when a patient's therapy needs to be intensified or stepped down? What sequence of therapies is best for achieving maximal improvement or preventing drug dependence? Should this sequence of therapies vary by patient outcomes?

To address these questions, this project develops and uses the following methodological innovations. First, the SMART experimental design methodology will be extended for use with time-varying outcomes; in particular this component will provide guidance to researchers on how to match their use of the time-varying outcome in the data analysis of SMART studies to their prevention/clinical goals. Second, this component will generalize a data analytic method from engineering and computer science for use with SMART study data so as to develop adaptive behavioral or combined behavioral-pharmacological interventions. Third, this component will provide methods for using data to develop more flexible adaptive interventions by constructing measures of confidence that can be used to ascertain when there is no evidence to discriminate between two or more successful treatments.

This work will include collaborative research with health scientists interested in constructing adaptive interventions. The goal is to accelerate the improvement of both prevention programs and treatments.

Funding Period: 07/01/2010 to 08/31/2015

Search . Browse