Page 16 - hypertension_newsletter

REFLECTIONS
Hypertension
Hypertension Global Newsletter #10 2026

determinants of health. This enhanced data extraction allows AI to detect secondary causes of hypertension (like primary
aldosteronism) significantly earlier than traditional rule-based screening. Furthermore, the review highlights that AI models Hypertension
incorporating continuous, high-frequency data from wearable devices can quantify variability patterns, identify abnormal trajectories,
and detect early deviations from an individual’s baseline that may signal loss of BP control or poor adherence. Despite these
promising applications, several challenges limit the current clinical impact of wearable-based AI systems in hypertension. Clinical
language varies across institutions, specialties, and individual clinicians, necessitating careful model training and validation. Data
privacy concerns, computational requirements, and the need for continuous model updating further complicate deployment. Data
quality is also a concern, as consumer-grade devices vary widely in sensor accuracy and sampling frequency.

The review also indicates that machine learning is being applied to genetic profiles to predict an individual’s pharmacological
response to renin-angiotensin system inhibitors, beta blockers, and CCBs. By simulating these complex gene-drug interactions,
AI aims to help clinicians pinpoint the most effective targeted therapy for a patient, thereby minimizing the standard trial-and-error
prescribing process.

The authors emphasize AI’s capacity to transform hypertension CLINICAL PEARLS FROM THE FACULTY
care into a dynamic and personalized approach by integrating
complex, longitudinal, and multimodal data. AI can identify latent
risks and optimize patient-specific interventions before irreversible
organ damage occurs; however, most applications currently
remain in early translational stages and require rigorous validation
in diverse populations. To turn this technological promise into
clinical reality, predictive tools need to be seamlessly integrated
into clinician workflows and thoughtfully designed to close
existing gaps in health equity. AI should be viewed as an enabling
technology that augments human judgment rather than replacing
clinicians or established guidelines.
WATCH
VIEW COMMENTARY FROM PROF.
PEDREROS GUERRA DISCUSSING
CLICK HERE THE CLINICAL RELEVANCE OF THE
FOR THE LINK TO FULL ARTICLE ARTICLE.

Artificial intelligence in cardiovascular pharmacotherapy: Applications and
perspectives.
Costa F, et al. Eur Heart J. 2025 Oct 1;46(37):3616-3627.

CVD remains a primary driver of global mortality, and optimizing pharmacotherapy for these conditions is highly complex due
to vast amounts of patient data and individual biological variability. Traditional statistical models used in clinical practice rely
on explicit assumptions and often take a “one-size-fits-all” or basic risk-stratified approach to prescribing. AI and its subfield,
machine learning (ML), offer a flexible, powerful alternative capable of analyzing massive, multidimensional data streams,
such as continuous physiological monitors, high-resolution imaging, and EHRs, without rigid prior assumptions. AI has the
potential to revolutionize CV care by identifying which patients will benefit most from specific therapies, optimizing drug dosing,
predicting adverse effects, and dramatically accelerating the discovery of novel CV drugs.

TABLE OF CONTENTS