JMIR Briefs Examine AI Screening, GLP-1 Ads and Postpartum Psychosis Detection — Research Library

Key Highlights

JMIR Publications has released five News and Perspectives features on digital health, spanning AI screening for gender-based violence, GLP-1 advertising, patient views on AI-assisted communication, resuscitation training and postpartum psychosis detection.
ClarIA, developed by Vital Strategies, scans clinical notes and medical history for signs of gender-based violence and has so far prompted eight formal reports.
Experts are examining whether wearables can register irritability, insomnia and racing thoughts, described as early warning signs of postpartum psychosis.

Context and Core Dynamics

ClarIA, developed by the global health organization Vital Strategies, searches patient clinical notes and medical history for indicators of gender-based violence and flags possible unreported cases to health care professionals, an effort to move care from reactive to proactive; it has so far produced eight formal reports. A separate feature examines the psychological footprint of ubiquitous GLP-1 advertising, especially online campaigns that route users toward telehealth services. A cited study finds such advertising commonly emphasizes emotional transformation and personal growth, tying weight to self-worth and potentially exploiting body image insecurities. Research remains limited, but clinicians and patients already describe a real impact.

Two features address the terms on which patients meet algorithmic tools. Evidence on AI-assisted clinical communication suggests acceptance rises when AI makes care more responsive and falls when patients cannot tell who is responsible for a message or whether a clinician engaged with it. In resuscitation, a proof-of-concept study found that ChatCPR, a model generating CPR guidance, exceeded human dispatchers in meeting American Heart Association guidelines, while the Elevate SMART Manikin supports practice without an instructor present.

The fifth feature considers postpartum psychosis, which can produce severe symptoms and, in the worst cases, tragedy. Experts describe how wearables might help clinicians identify discrepancies pointing to subtle shifts in mood and behavior that serve as early warning signs, and note that postpartum mobile health initiatives, including therapeutic AI chatbots, could widen access to mental health care.

Caregiving Application and Social Impact

For caregiving organizations, the features signal a design principle rather than a product. Continuous passive sensing of the kind proposed for postpartum monitoring is the same architecture used in eldercare and post-discharge rehabilitation, where deviation from an individual baseline carries more signal than population thresholds. Screening tools such as ClarIA show how documentation generated in routine care can be repurposed to surface risk that overstretched staff may miss. The resuscitation findings matter for home care, where a bystander is often the first responder.

Ethics and Compliance Note

Each feature carries a compliance dimension. Screening clinical notes for violence depends on sensitive inferences drawn from records patients did not create for that purpose, and requires a lawful basis, minimum necessary access, audit trails and a defined human reviewer before any flag becomes an intervention. Wearable monitoring in perinatal and elder care raises questions of consent, data minimization and secondary use. Where AI drafts or transmits clinical communication, accountability must rest with a named professional.