What Is an Open-Source Model? Forking a Path to Definition
Speaker(s)
Moderator: Stephanie Harvard, PhD, The University of British Columbia, Vancouver, BC, Canada
Speakers: Koen Degeling, PhD, Lumen Value & Access, Hengelo, OV, Netherlands; Eric Jutkowitz, PhD, Brown University, Providenice, RI, USA; Blythe Adamson, PhD, MPH, Flatiron Health, New York, NY, USA
The HEOR community is increasingly adopting open-source practices, especially in developing and sharing open-source models. New initiatives enable analysts to make code and other research outputs publicly available. However, uncertainty remains about what ‘open-source’ means in the HEOR context. In particular, what constitutes – or should constitute – an ‘open-source model’ has not been specified. In this forum session, led by ISPOR’s Open Source Models (OSM) Special Interest Group (SIG), the panellists will explore challenges in identifying practical definitions for ‘open-source’ – and other terms – in HEOR. The purpose of this forum is to provide the first dedicated opportunity for stakeholders to focus on this issue of definition. The OSM SIG surveyed its members to inform this forum, soliciting views on the matter of definitions in the OSM context. The survey explored the meaning and scope of open-source in HEOR, the practical value of agreed definitions, and the role of the SIG in promoting their adoption. Building on these findings, the panellists will discuss software choice, code, data, updating, certification, licensing, and sharing platforms. They will consider how definitions align with the purpose and objectives of open science in HEOR, such as transparency, verification, validation, and reuse. The forum will also explore how our community can learn from other contexts, such as software development, and review the different perspectives that ought to be considered. The moderator will provide an overview of the topic and frame the challenge of definition for open-source in HEOR. Each panellist will then speak for 10 minutes, providing their perspective and sharing related findings from the SIG member survey. At least 20 minutes will be reserved for audience discussion, including polling. The OSM SIG will develop a manuscript on this topic, informed by audience contributions at this forum.
Code
138
Topic
Methodological & Statistical Research