{"id":7032993,"date":"2026-08-08T03:07:12","date_gmt":"2026-08-08T03:07:12","guid":{"rendered":"https:\/\/peraltafinancing.com\/uncategorized\/reimagining-the-fdas-role-in-digital-medicine\/"},"modified":"2026-08-08T03:07:12","modified_gmt":"2026-08-08T03:07:12","slug":"reimagining-the-fdas-role-in-digital-medicine","status":"publish","type":"post","link":"https:\/\/fivemor.com\/?p=7032993","title":{"rendered":"Reimagining the FDA\u2019s Role in Digital Medicine"},"content":{"rendered":"


\n<\/p>\n

\n

In addition to evaluating the
\nsafety of software as a medical device (SaMD), the agency needs to devote more
\nresources to evaluating its efficacy and quality.<\/h4>\n
<\/a><\/div>\n

John Halamka, M.D., president, Mayo
\nClinic Platform, and Paul Cerrato, senior research analyst and communications
\nspecialist, Mayo Clinic Platform, wrote this article.<\/i><\/p>\n

The FDA\u2019s approach to software as a
\nmedical device (SaMD) has been evolving. Consider a few examples.<\/p>\n

In 2018, IDx-DR, a software system used
\nto improve screening for retinopathy, a common complication of diabetes that
\naffects the eye, became the first AI-based medical device to receive US
\nFood and Drug Administration clearance<\/a> to \u201cdetect greater than a mild level
\nof \u2026 diabetic retinopathy in adults who have diabetes.\u201d To arrive at that
\ndecision, the agency not only reviewed data to establish its safety, it also
\ntook into account prospective studies, an essential form of evidence that
\nclinicians look for when trying to decide if a device or product is worth
\nusing. The software was the first medical device approved by the FDA that does
\nnot require the services of a specialist to interpret the results, making it a
\nuseful tool for health care providers who may not normally be involved in eye
\ncare. The FDA clearance emphasized the fact that IDx-DR is a screening tool not
\na diagnostic tool, stating that patients with positive results should be
\nreferred to an eye care professional. The algorithm built into the IDx-DR
\nsystem is intended to be used with the Topcon NW400 retinal camera and a cloud
\nserver that contains the software.<\/p>\n

Similarly, FDA looked at a
\nrandomized prospective trial before approval of a machine learning-based
\nalgorithm that can help endoscopists improve their ability to detected smaller,
\neasily missed colonic polyps. Its recent clearance of GI Genius by Medtronic
\nwas based on a clinical trial published in Gastroenterology<\/i><\/a>, <\/i>in which investigators in Italy evaluated data from 685 patients,
\ncomparing a group that underwent the procedure with the help of the computer-aided detection (CADe) system to a group who acted as controls. Repici et al
\nfound that the adenoma detection rate was significantly higher in the CADe
\ngroup, as was the detection rate for polyps 5 mm or smaller, which led to the
\nconclusion: \u201cIncluding CADe in colonoscopy examinations increases detection of
\nadenomas without affecting safety.\u201d<\/p>\n

Their findings raise several
\nquestions: is it reasonable to assume that a study of 600+ Italians would apply
\nto a U.S. population, which has different demographic characteristics? More
\nimportantly, were the 685 patients representative of the general public, including
\nadequate numbers of persons of color and those in lower socioeconomic groups?
\nWhile the Gastroenterology <\/i>study did report enough female patients,
\nthere is no mention of these other marginalized groups. \u00a0<\/p>\n

An independent 2021 analysis<\/a>
\nof FDA approvals has likewise raised several concerns about the effectiveness
\nand equity of several recently approved AI algorithms. Eric Wu from Stanford
\nUniversity and his colleagues examined the FDA\u2019s clearance of 130 devices and
\nfound the vast majority were approved based on retrospective studies (126 of
\n130). And when they separated all 130 devices into low- and high-risk subgroups
\nusing FDA guidelines, they found none of the 54 high-risk devices had been evaluated
\nby prospective trials. Other shortcomings documented in Wu\u2019s analysis included
\nthe following:<\/p>\n