10.1371/journal.pmed.0020297

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Artificial intelligence (AI) tools are entering clinical practice at unprecedented speed. 1,357 AI/ML-enabled medical devices have received U.S. FDA clearance or approval, yet their impact on patient outcomes remains largely untested. We conducted a systematic analysis of all FDA-cleared AI/ML-enabled medical devices through December 5, 2025 using the FDA device database and the ACR Data Science Institute catalogue, with linked searches of ClinicalTrials.gov and PubMed to identify registered trials and publications. Of 1,357 cleared AI devices, only 34 (2.5%) were linked to registered prospective trials, 12 (0.9%) posted results, 12 (0.9%) had peer-reviewed publications, and only 3 (0.2%) evaluated patient-centered outcomes such as mortality, morbidity, or readmissions. Most studies (62%) employed observational designs with small, homogenous cohorts, limited subgroup analyses, and frequent exclusion of vulnerable populations. Structural barriers (including misaligned financial incentives, reliance on predicate-based regulatory pathways, and logistical challenges of multi-center trials) discourage rigorous evaluation. Internationally, FDA clearance often functions as a gateway for global deployment, raising ethical concerns when under-validated tools are introduced into low- and middle-income countries without contextual validation or safeguards. Regulatory approval has outpaced clinical validation, creating an ecosystem where innovation advances without accountability. The finding that only 0.2% of cleared devices have undergone evaluation for patient-centered outcomes reveals a profound validation gap and points to the need for evidence standards capable of keeping pace with the speed of regulatory clearance. Readiness should no longer be defined by FDA clearance alone, but by demonstrated, durable, and equitable benefit to patients.

Abstract: PubMed · Datensatz