JMIR Publications

ABSTRACT

Background:

The following co-design research method details the iterative process developed to identify health professional recommendations for the graphical user interface (GUI) of an artificial intelligence (AI) enabled risk prediction tool. Driving the decision to include a co-design process is the belief that choices regarding the aesthetic and functionality of an intervention are best made by its’ intended users, and that engaging these users in its design will also ease their concerns regarding adoption and use.

Objective:

The aim of this research is to identify health professional design and uptake recommendations for the GUI of an AI enabled predictive risk tool.

Methods:

Three research phases each consisting of two workshops will be held with health professionals between mid-2023 and mid-2024. Six health professionals will be sought per workshop, resulting in a total enrolment of 36 at the conclusion of the research. Seven workshop activities have been scheduled across the three workshops, these include: Context of use, Notifiers, Format, AI survey – Likert, Prototype, AI survey – Written, and Testing. The first six of these activities will be repeated in each workshop to enable the iterative development and refinement of GUI aesthetic and function recommendations. Where the last activity (Testing) will be performed in the final workshop to examine health professional thoughts on the final iteration of the GUI. Each activity is comprised of one or more research tasks to produce qualitative and quantitative results. Textual data will be examined via inductive thematic analysis or via deductive thematic analysis in accordance with the ‘Nonadoption, Abandonment and Challenges to the Scale-up, Spread and Sustainability’ (NASSS), visual data will be examined in accordance with ‘framework of interactivity’, and, quantitative data examined using descriptive statistics.

Results:

Project registration with the Australia and New Zealand Clinical Trial Registry has been requested (#384098). Finalised design recommendations are expected in early- to mid-2024, with a results manuscript to be submitted in mid-2024. This research has human research ethics approval from the South Australian Department of Health and Wellbeing (#2022/HRE00131) and the University of South Australia (application ID#204143).

Conclusions:

Understanding whether an intervention is needed within a particular situation is just the start, designing an intervention so that it is used within that situation is even more important. The co-design process involves end-users in process of creating an intervention to include functions and to align with existing work practices. Indeed, interventions that fail to do this at the least may be disliked, and at the worst may be dangerous. Clinical Trial: Australia and New Zealand Clinical Trial Registry; (Request number: 384098)