Claude, ChatGPT confidently missed skin cancers in darker skin

SAN DIEGO — Two popular AI chatbots demonstrated limited accuracy in identifying skin cancers in darker skin tones, yet expressed relatively high confidence in their diagnoses, researchers reported.
For the study, presented at the American Society for Dermatologic Surgery Annual Meeting, researchers asked Claude and ChatGPT AI models to rate their diagnostic confidence on a scale of 1% to 100%. For lesions in Fitzpatrick skin types V to VI — the darkest skin colors — the models reported confidence levels of 65% to 70% when their diagnoses were correct and 60% to 65% when they were incorrect.

How patients, employees perceive value-based kidney care services

DENVER — Patient and employee-reported perspectives about a value-based kidney care model highlight strengths and limitations in value-based services, according to study data presented at ASN Kidney Week.
In contrast to traditional fee-for-service models, which provide reimbursement based on service complexity and volume, value-based kidney care models aim to provide interdisciplinary care to manage chronic kidney disease earlier in the treatment paradigm, according to Umayr Ahmed, student at the Perelman School of Medicine at the University of Pennsylvania, and colleagues.
To better understand

Large language models vary on thyroid nodule biopsy decisions

Large language models varied on their decision of whether to biopsy text descriptions of thyroid nodules, according to study data published in The Journal of Clinical Endocrinology & Metabolism.
Using hypothetical clinical vignettes of thyroid nodules that varied in size and risk, researchers surveyed thyroid cancer specialists and queried multiple large language models to determine whether they would choose to biopsy. Clinicians only unanimously agreed on their decision in half of the clinical vignettes, but similar variability in response was seen among the large language models.
“When

ETH47 reduces symptoms in asthma challenged with rhinovirus

After a rhinovirus challenge, numerical reductions in respiratory symptoms were found with a nasally delivered messenger RNA therapeutic vs. placebo in adults with asthma on stable maintenance therapy, according to topline data.
ETH47 (Ethris) is an interferon lambda-encoding messenger RNA (mRNA) therapy, according to a company press release.
“This is a proof-of-concept study showing very encouraging and consistent signals in terms of virus positivity, upper and lower respiratory symptoms and development of colds and exacerbations,” Sebastian Johnston, MBBS, PhD, professor of respiratory

Medicaid expansion linked to lower colorectal cancer mortality

States with Medicaid expansion had significantly lower colorectal cancer mortality than those without Medicaid expansion.
An evaluation of more than 650,000 patients showed individuals of all races and stages of disease benefited from Medicaid expansion, although Hispanic patients and those with stage IV colorectal cancer had the greatest improvements in survival.
“Medicaid expansion in the United States has become a partisan issue,” Oluwasegun Akinyemi, MD, PhD, MSc, health services researcher at Clive O. Callender, MD, Outcomes Research Center at Howard University College of Medicine, told

Data support 5-year surveillance interval in high-risk adenoma

Incidence of colorectal cancer appeared similar regardless of whether patients underwent surveillance colonoscopy at 3 or 5 years after high-risk adenoma removal, according to interim results of an ongoing, 10-year noninferiority study.
Although current guidelines advise colonoscopy 3 years after polyp removal in high-risk patients, limited data support this recommendation. Extending surveillance intervals could ease both patient and healthcare system burden while also reducing costs, researchers wrote.
“Endoscopy units are overwhelmed with surveillance colonoscopies after positive results of

Litifilimab improves multiple lupus skin outcomes at 52 weeks

ORLANDO — Patients with cutaneous lupus erythematosus treated with litifilimab demonstrated continuous improvement in both clinician and patient skin assessment scores through week 52, according to data presented at ACR Convergence.
“The AMETHYST trial represents the first cutaneous lupus study designed in this way with 52-week results,” Joseph F. Merola, MD, MMSc, FAAD, professor of internal medicine in the division of rheumatic diseases at the University of Texas Southwestern Medical Center, told attendees. “I hope you are as excited as I am.”
Litifilimab