Summary
Amid the rapid acceleration of artificial intelligence (AI) across healthcare, a critical question remains: is this new technology truly aligned with the principles of patient-centred care (PCC)?
While AI offers immense promise, its development often outpaces our understanding of its alignment with the patient outcomes that matter most. This gap necessitates a clear evaluation of AI’s real-world application in PCC.
Our Objective & Method
To address this, our research group has submitted a new systematic review protocol, “Evaluation of Artificial Intelligence methods for Patient-centred care with PROBAST+AI,” for consideration in PLOS One. This protocol outlines a systematic review that aims to critically evaluate how AI methods have actually been applied in advancing PCC.
The methodological contribution is a key part of this work. To our knowledge, this will be the first systematic review to explicitly use the newly enhanced PROBAST+AI tool for this purpose. This tool is essential for rigorously assessing the risk of bias and methodological quality of predictive modelling studies, as it includes new, specific guidance for AI-specific challenges (like algorithm bias and transparency) that older tools miss.
Our protocol, which is registered with PROSPERO (CRD420250631752) and follows PRISMA-P guidelines, will guide us in synthesizing the current evidence.
What We Aim to Discover
By systematically identifying and evaluating these studies, this review will pinpoint the current strengths and, more importantly, the weaknesses in how AI is being used for patient-centred care.
Why This Matters
This work will provide valuable, concrete insights for healthcare providers, policymakers, and AI researchers. By understanding the current gaps, we can help guide the future development of AI, ensuring that these powerful tools are built from the ground up to be transparent, fair, and truly focused on the patient.
