Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

VA Health Systems Research

Go to the VA ORD website
Go to the QUERI website
HSRD Conference Logo



2019 HSR&D/QUERI National Conference Abstract

Printable View

4051 — Applying a matrix-based multiple-case study approach to examine local influences on VA's multi-site Collaborative Chronic Care Model implementation

Lead/Presenter: Bo Kim,  COIN - Bedford/Boston
All Authors: Kim B (VA Center for Healthcare Organization and Implementation Research, Bedford/Boston, MA; Harvard Medical School, Boston, MA), Connolly SL (VA Center for Healthcare Organization and Implementation Research, Bedford/Boston, MA; Harvard Medical School, Boston, MA), Elwy AR (VA Center for Healthcare Organization and Implementation Research, Bedford/Boston, MA; Brown University Alpert Medical School, Providence, RI) Miller CJ (VA Center for Healthcare Organization and Implementation Research, Bedford/Boston, MA; Harvard Medical School, Boston, MA) Riendeau R (VA Center for Healthcare Organization and Implementation Research, Bedford/Boston, MA; University of Iowa, Iowa City, IA) Stolzmann K (VA Center for Healthcare Organization and Implementation Research, Bedford/Boston, MA) Sullivan JL (VA Center for Healthcare Organization and Implementation Research, Bedford/Boston, MA; Boston University School of Public Health, Boston, MA) Bauer MS (VA Center for Healthcare Organization and Implementation Research, Bedford/Boston, MA; Harvard Medical School, Boston, MA)

Objectives:
From late 2015 through early 2018, we partnered with VA's Office of Mental Health and Suicide Prevention to develop, conduct, and evaluate the implementation of the evidence-based Collaborative Chronic Care Model (CCM) at nine VA facilities. We employed a systematic matrix-based multiple-case study approach to describe the multifaceted heterogeneity and commonalities across the sites and their influences on implementation.

Methods:
We were guided by the i-PARIHS (Integrated Promoting Action on Research Implementation in Health Services) framework to focus on each site's facilitation-, innovation-, recipients-, and context-related characteristics in formulating the comparative multiple-case study. We drew on Yin (2009)'s Case Study Research Methodology to incorporate explicit study tactics toward establishing validity and reliability of our findings. We devised a matrix-based organization of data to enable integrated assessments of multiple data sources, identified influencing factors, and trends across sites.

Results:
The matrix-based multiple-case study approach led to (i) identifying appropriate measures (for construct validity) by reviewing multiple sources of data, (ii) uncovering non-spurious relationships (for internal validity) by conducting pattern matching across sites and addressing rival explanations, (iii) defining the domain of generalizability (for external validity) by using replication logic involving the multiple cases, and (iv) ensuring the replicability of the study (for reliability) by documenting a detailed case study protocol and a thorough case study database. The approach established i-PARIHS-aligned comparative descriptions of each site's characteristics.

Implications:
The matrix-based multiple-case study approach allowed systematic organization, analysis, and presentation of various potential sources of heterogeneity and commonalities across the nine sites. Our approach (i) capitalizes on case study research's strengths in explaining implementation impact within real-world settings, which often include many potential variables of interest whose interrelationships are challenging to rigorously evaluate, and (ii) demonstrates a methodical curation of data/evidence, which is a key component of VA's continued progress towards fostering a learning health care community.

Impacts:
Future work should examine the matrix-based multiple-case study approach's application to studying implementation beyond mental health and VA. The approach can also be considered more generally for formative evaluation of implementation efforts needing simultaneous consideration of wide-ranging data types and sources.