MOVING THE SCIENCE FORWARD- PSYCHOMETRIC CONSIDERATIONS AND STUDY DESIGNS FOR UNDERSTANDING MEANINGFUL CHANGE AND CONDUCTING MIXED METHODS RESEARCH
Author(s)
Cheryl Coon, PhD, Outcometrix, Tucson, USA; Stacie Hudgens, MA, Clinical Outcomes Solutions, Tucson, USA; Lori McLeod, PhD, RTI Health Solutions, Research Triangle Park, USA; Tara Symonds, PhD, Clinical Outcomes Solutions Ltd., Folkestone, Kent, UK
Presentation Documents
PURPOSE: The purpose of this workshop is to address two methodological issues facing the regulatory and scientific community in the field of clinical outcome assessment (COA) – derivation of meaningful change thresholds and mixed methods research – and to examine recent deliberations and present recommendations aimed at building scientific consensus on these issues.
DESCRIPTION: A panel of measurement specialists was convened to consider these two issues, which were identified by academics, industry, and regulators as scientifically challenging in the context of developing, evaluating, and interpreting data from COA tools. The issues (and specific questions) considered by this panel included: 1. Proposed methods for determining clinically meaningful change and evaluation of potential anchors. What defines a good anchor? How should potential anchors be evaluated? What parameters should be included in the anchor evaluation? How can distribution-based methods be used to support anchor-based analyses to determine meaningful change at the within patient level? Are qualitative studies a more direct way of deriving meaningful change based on patients’ perspectives? What is the appropriate role for cumulative distribution functions? 2. Definition of mixed methods and mixed methods study designs for COA tool development. What is the definition for mixed methods in this field of research? What is the value of mixed methods research in COA tool development? Which research designs should be considered to provide an efficient way of developing content for new measures? When is a qualitative sample sufficient for exploratory quantitative analysis? The discussion leaders will outline the issues and describe the panel discussions, noting areas where agreement was achieved, where agreement is close but not yet achieved, and where additional research is needed. Discussion leaders will engage attendees to share their perspectives, as well as to provide insight gained from their own research, to move toward scientific consensus.
Conference/Value in Health Info
2016-10, ISPOR Europe 2016, Vienna, Austria
Code
W2
Topic
Methodological & Statistical Research, Patient-Centered Research