AI chatbots like OpenAI's ChatGPT and Google's Gemini are increasingly used for mental health discussions, raising concerns about their safety and efficacy. Recent studies show these chatbots can inadvertently provide harmful advice, such as methods of self-harm, despite initial safeguards. States like Illinois and Nevada have enacted laws to restrict AI from offering direct mental health services, but enforcement is challenging due to the chatbots' disclaimers. Experts emphasize the need for better regulation and ethical design as AI's role in health care expands.

The eHealth Exchange is central to a Utah pilot project aimed at developing a statewide FHIR-based health care data exchange ecosystem. This initiative, driven by new CMS interoperability rules, seeks to improve payer-to-provider and payer-to-payer data exchanges using modern APIs. The project, which began two years ago, is serving as a model for similar efforts in other states, with Washington next in line. Key lessons include the importance of a hub-and-spoke model for efficient data exchange and the potential for streamlined patient identification processes.

The Assistant Secretary for Technology Policy has clarified that TEFCA participants who restrict Qualified Health Information Network choice may violate federal information blocking rules. This guidance affects vendors and health care organizations, with potential penalties including Medicare payment disincentives. Health systems have the freedom to choose their QHIN, and any coercion by QHINs connected to EMR developers is prohibited. The rule also applies to sub-organizations, emphasizing the need for documented free choice.

An AI-enabled smartphone app, the Twin Precision Treatment system, significantly improved glycemic control in adults with type 2 diabetes, according to a study published in NEJM Catalyst Innovations in Care Delivery. The trial, led by Kevin M. Pantalone from the Cleveland Clinic, showed that 71% of participants using the app achieved an HbA1C of less than 6.5% without glucose-lowering drugs, compared to 2.4% in the control group. The app uses real-time data and virtual coaching to provide personalized lifestyle recommendations, resulting in reduced medication dependency and improved quality of life.

A review that synthesized findings from 48 studies published between 2009 and 2024 highlights that patient trust in AI and telemedicine for dermatologic care hinges on clinician oversight and endorsement. While AI offers diagnostic efficiency and telemedicine enhances accessibility, patients prefer these technologies as complementary tools rather than stand-alone solutions. Concerns include loss of human interaction and privacy issues, with patients emphasizing the need for transparent privacy policies and clinician involvement for acceptance.

Eric Brooks and his team at NC State College of Veterinary Medicine have developed a detailed map of RNA expression patterns in early mouse embryos, revealing insights into brain development. By tracking RNA in cells, they can predict cell movements and roles in forming the neural tube, which becomes the brain and spinal cord. This research, involving collaboration with computational biologists from Memorial Sloan Kettering Cancer Center, highlights the role of the Sonic Hedgehog gene in neural tube formation and its implications for developmental defects.

A study in the journal Circulation found that the FonoCheck smartphone app can screen effectively for atrial fibrillation in community settings, showing a positive predictive value of 94%. Researchers from Southeast University in China enrolled 5,312 participants, using the app to record heart sounds for 30 seconds. The app identified 39 suspected A-Fib cases, with 34 confirmed by ECG. Its performance is comparable to wearable devices, but researchers note the prevalence of A-Fib might be underestimated.