JMIR Publications

ABSTRACT

Background:

AI health technologies have the potential to transform existing clinical workflows and ultimately improve patient outcomes. AI health technologies have shown potential for a range of clinical tasks such as diagnostics, prognostics, and therapeutic decision making such as drug dosing. There is however an urgent need to ensure that AI health technologies remain safe for all populations. Recent literature demonstrates the need for rigorous performance error analysis to identify issues such as algorithmic encoding of spurious correlations (e.g. protected characteristics), or specific failure modes that may lead to patient harm. Guidelines for reporting of studies evaluating AI health technologies (e.g. CONSORT-AI) require mention of performance error analysis, however there is still a lack of understanding around how performance errors should be analysed in clinical studies, and what harms authors should aim to detect and report.

Objective:

This systematic review will assess the frequency, severity of AI errors and patient harms in randomised controlled trials (RCTs) investigating AI interventions in clinical settings. The review will also explore how performance errors are analysed including whether analysis includes investigation of subgroup level outcomes.

Methods:

This systematic review will identify and select randomised controlled trials assessing AI interventions. Search strategies will be deployed in MEDLINE, EMBASE, Cochrane CENTRAL and clinical trials registries to identify relevant articles. RCTs identified in bibliographic databases will be cross-referenced with clinical trials registries. The primary outcomes of interest are the frequency and severity of AI errors, patient harms and reported adverse events. Quality assessment of RCTs will be based on RoB2. Data analysis will include comparison of error rates and patient harms between study arms and a meta-analysis of the rates of patient harm in control versus intervention arms will be conducted if appropriate.

Results:

The project was registered on PROSPERO in February 2023. Preliminary searches have been completed and the search strategy has been designed in consultation with an information specialist (see appendices 1 and 2). Abstract screening will start in August 2024.

Conclusions:

Evaluations of AI health technology have shown promising results, however reporting of studies has been variable. Detection, analysis and reporting of performance errors and patient harms is vital to robustly assess the safety of AI interventions in RCTs. Scoping searches have illustrated that reporting of harms is variable, often with no mention of adverse events. The findings of this systematic review will identify the frequency and severity of AI performance errors and patient harms, and generate insights into how errors should be analysed to account for both overall and subgroup performance. Systematic review registration PROSPERO CRD42023387747.