Abstract / Summary
Artificial intelligence/machine learning (AI/ML)-based in vitro diagnostic software is increasingly authorized for clinical use in digital pathology and hematology morphology. However, it remains unclear how often performance evidence is publicly available. We performed a systematic landscape analysis identifying 77 CE-marked or FDA-authorized AI/ML-based IVD software devices from 29 manufacturers, including 68 digital pathology and 9 hematology morphology devices, identified through a multi-source search of FDA databases, EUDAMED, PubMed, Google Scholar, manufacturer portfolios, and clinical conference proceedings. Publicly available device-specific performance evidence was identified for 30 devices (39%), while 47 devices (61%) had no identified publicly available performance evidence. Evidence availability was higher among devices with dual EU and USA authorization than EU-only devices (8/8, 100% vs 22/69, 32%; p < 0.001), and among hematology morphology devices than digital pathology devices (7/9, 78% vs 23/68, 34%; p = 0.02). In total, 15 distinct performance metrics were reported, covering discrimination, agreement, correlation, and prognostic categories. However, their use varied across publications, limiting direct cross-device comparison. While performance evaluation results do not require publication, our findings indicate a gap between evidence availability and regulatory authorization, alongside heterogeneous use of metrics. These factors may undermine confidence in innovative medical devices and hinder informed clinical adoption.