C-DIAS RP 1: A community-driven modeling approach for identifying and implementing evidence-based interventions and implementation strategies to reduce overdose deaths.

C-DIAS RESEARCH PROJECT 1: PROJECT SUMMARY/ABSTRACT
While effective treatments exist, and substantial new funding has been distributed to states and territories to
expand access, the US continues to be plagued with opioid and stimulant related deaths. Nonlethal overdoses,
emergency room and hospitalization utilization also continue to escalate. Research Project 1 proposes to
develop an innovative model-driven decision support system to aid counties to prioritize and optimize the use
of available resources to address opioid and stimulant deaths. This new implementation strategy differs from
and extends data-driven decision making, which by itself does not produce information on projected impact. In
particular, this system examines a multitude of different options for implementation strategies that fit a county’s
needs and capacities, and identifies optimal solutions that are robust to changing contexts. This model-driven
decision support (MDDS) implementation strategy integrates three components: 1) an ongoing community-
research partnership, which is essential if decision makers are to feel the predictions are locally relevant and
trustworthy; 2) integrated administrative, interview, and survey data, which are required to account for the
locally unique contextual factors; and 3) a modular agent-based model (ABM) that integrates data, represents
alternative implementation strategies, and simulates their behavior on overdose deaths under varying
conditions. The first aim focuses on extending an existing ABM that takes as input local data and produces
projections across diverse strategies’ impact on overdose deaths and disparities. Response surface
methodology and mediation modeling are used to summarize findings. Front-end tools are developed along
with training so the system can be used by other counties. The second aim examines the generalizability of
this approach in three diverse counties: Pinellas County FL, Vermilion County IL, and Santa Clara County CA.
We have active partnerships with all three communities. If there is no change in mortality in these 3 counties,
more than 3,000 individuals are expected to die from overdose in the coming 5 years. The third aim focuses on
advancing an approach for improving care through monitoring and feedback, both within these communities
and across using a learning collaborative. This aim also formally evaluates the MDDS using the Stages of
Implementation Completion (SIC) as well as measuring acceptability and the degree to which the communities
are using this approach differently from one based on data only. At the PREPARE phase of the implementation
process, Research Project 1 is bi-directionally integrated with C-DIAS, and uses common measures with the
other C-DIAS Research Projects. In pursuit of C-DIAS specific aim 2, in concert with the Research Core, this
project will serve as a vehicle to test and advance ABM and cost modeling decision supports for policymakers
and implementation endeavors nationwide. Project findings will be widely influential because of C-DIAS
dissemination platforms, including open science assess to ABM tools, resulting in greater impact than
otherwise possible if conducted independently.

Grant Number: 5P50DA054072-04
NIH Institute/Center: NIH
Principal Investigator: C. BROWN