Implementation and Dissemination Simulation Laboratory (SimLab)
Principal Investigator:
Peter Hovmand, PhD, PI
Funder:
Timeframe: 09/2008-06/2009
Affiliation:
Center for Mental Health Services Research (CMHSR)
Project Staff: Erin Leaver, BS, Project Coordinator
Eric Stiens, MSW Candidate, Research Assistant
Project Contact: Erin Leaver, BS
(314) 935-9796
Project Update as of 09/23/2009
Stage:
Pilot Stage
Description:
Mission Statement
The mission of the Dissemination & Implementation Simulation Laboratory (SimLab) is to support research on the dissemination and implementation of innovations in the health and human services sector using modeling and computer simulation methods. These methods can help organizations and policy makers understand the impact of organizational and system changes on outcomes. The innovations can range from changes in specific treatments to changes at the policy level; and, can involve efforts to disseminate and implement innovations at one or more levels from the individual provider to an entire retion. The work of the SimLab will occur in stages. The timeframe above and aims below reflect the first stage of work.
Aims for Year 1
To pursue this mission, the SimLab will develop some basic infrastructure to support modeling and computer simulation activities. The goal of year 1 will be to develop these methods by working on researcher led projects that demonstrate the potential applications of modeling and computer simulation to dissemination and implementation research with a primary focus in year 1 on system dynamics and agent based modeling approaches. Specific aims for year 1 are:
Aim 1: Develop the System Dynamics Analysis (SDA) package for R. The SDA package for R will be a set of open source routines that enable automated testing and analysis of simulation models. This provides a means for ensuring that computer simulations of social systems conducted as part of the analysis can be readily checked and replicated by other researchers. The SDA package will also provide tools for testing and developing implementation strategies, and provide a foundation for developing future tools (e.g., statistical power analysis, testing research designs).
Aim 2: Develop the State Transition Analysis (STA) package for R. STA generates state transition maps from administrative services data, data for model parameters and distributions, and time series for model testing that can be used for both agent based and system dynamics modeling. This will shorten the time to construct datasets for simulation modeling and testing.
Aim 3: Develop the Group Model Building Scripts Collection. Group model building is a way to develop computer simulation models with participants including other experts, practitioners, policy makers, providers, and consumers. The use of “scripts” is becoming a best practice in group model building. Scripts are step-by-step protocols for leading the group model building process. Each script defines a 5-45 minute group exercise. Scripts can be combined in different ways to fit the specific project requirements, and thus provide researchers with a means of adapting and tailoring group model building exercises to diverse problems and participants. The Group Model Building Scripts Collection will provide researchers using both system dynamics and agent based modeling approaches with a menu of participatory methods for developing models of dissemination and implementation problems.
Aim 4: Develop the Dissemination and Implementation Research Model Library. This model library will contain generic structures that can be combined to build models of dissemination and implementation problems specific to each research problem. Each structure in the library will be rigorously tested, robust, and reflect the best computer modeling practices. The library will allow novice to intermediate modelers to (e.g., graduate students, investigators new to simulation research) to develop their own preliminary computer simulation models.
Aim 5: Convene meetings of partnering agencies and researchers. A critical step to developing both the capacity of the SimLab and developing computer simulation models of dissemination and implementation problems that are relevant involves bringing potential agency partners together with researchers interested in developing simulation models. To support this, we will convene meetings with key stakeholders that are topic focused (e.g., system transformation, schools, diabetes) and host the Social Systems Modeling Group. This will help both disseminate the results of the SimLab efforts and help facilitate collaborative modeling and simulation projects with maximum relevance to health and human services.
Associated Publications
Hovmand, P.S. & Ford, D.N. (accepted). Sequence and timing of three community interventions to domestic violence. American Journal of Community Psychology.
Hovmand, P.S. & Ford, D.N. (in press). Victims arrested for domestic violence: unintended consequences of arrest policies. System Dynamics Review.
Hovmand, P.S. & Gillespie, D.F. (available online). Implementation of evidence based practice and organizational performance. Journal of Behavioral Health Services & Research.
Hovmand, P.S., Jonson-Reid, M,, & Drake, B. (2007). Mapping service networks. Journal of Technology and Human Services, 25 (4), 1-22.