Today's edition · Saturday, November 7, 2026
Every story verifiedNo politics. Ever.No pop-ups
TodayLocalWorldHeroesAnimalsScienceCommunityHealth & TechMapQuizKindnessAnnouncementsKidsTopicsExplainersOn this dayCalendarGood NeighborHometown Heroes

TodayHealth & TechTennessee

Health & TechNashville, Tennessee

In a randomized trial, an AI alert more than halved the time to an autoimmune diagnosis

Vanderbilt and Johns Hopkins researchers built the open-source AURA tool into health records. In a two-year randomized trial of about 1,960 patients, those with the alert were diagnosed in a median 43 days versus 138.

Verified · Source: Johns Hopkins Engineering for Professionals (NEJM AI study)Nov 6, 2026 · 1 min read
How to help
A physician in a white coat reading papers beside an older desktop computer
Illustrative photo of a clinician with a computer. This is not the AURA trial, and this physician is not at Vanderbilt or Johns Hopkins. Ernie Branson (Photographer) / Wikimedia Commons · Public domain

Read the original

The reporting lives at Johns Hopkins Engineering for Professionals (NEJM AI study). Our summary is checked against it, and we link it on every story.

Researchers at Vanderbilt, in Nashville, and at Johns Hopkins built an open-source tool called AURA into electronic health records, Johns Hopkins Engineering for Professionals reported. The study is published in NEJM AI. In a two-year randomized trial of about 1,960 patients, people whose clinicians received the alert were diagnosed in a median of 43 days, compared with 138 days without it.

That is a shorter wait for a diagnosis, more than cut in half on the median. It is not a treatment and not a cure. The alert flags patients who may be at high risk for autoimmune disease so testing can start sooner. The 43 and 138 figures are medians in this trial, not a promise for every patient.

More good news

Other stories in the same neighborhood of kindness.