Hypertension, affects approximately 50% of Americans and is a significant contributor to mortality and morbidity from cardiovascular diseases. Despite the availability of numerous medications, administration of right drug for each patient might be challenging since every one has benefits and drawbacks. Machine learning (ML) algorithms possess the capacity to transform hypertension treatment by providing real-time recommendations based on patient-specific characteristics, including demographics, vital signs, health history, and clinical examination documentation.¹

A data-driven model, jointly developed by data scientists and physicians at Boston University, was discussed in a recent study published in BMC Medical Informatics and Decision Making. The model is designed to provide real-time hypertension treatment recommendations tailored to individual patient traits. Developed using deidentified data from 42,752 hypertensive patients at Boston Medical Center, the model demonstrated a 70.3% greater reduction in systolic blood pressure compared to standard of care, showcasing its effectiveness. The algorithm was clinically validated, with the researchers manually reviewing a random sample of 350 cases. The model also showed the benefits of deprescribing—reducing or stopping prescriptions for some patients taking multiple medications—which could be particularly valuable in cases where the medical community is divided on the effectiveness of one drug versus another.

The BU-developed model generates a custom hypertension prescription using an individual patient’s profile, giving physicians a list of suggested medications with an associated probability of success. The researchers’ aim was to highlight the treatment that best controls systolic blood pressure for each patient based on its “Our goal is to facilitate a personalization approach for hypertension treatment based on machine learning algorithms,” says “seeking to maximize the effectiveness of hypertensive medications at the individual level.”³

Another study published in the Journal of Hypertension used ML algorithms to identify new hypertension genes for the early diagnosis of hypertension and the prevention of complications.² The researchers used ML algorithms to analyze data from genome-wide association studies have revealed a number of novel genes linked to hypertension. These findings could lead to the development of new diagnostic tests and targeted therapies for hypertension.

ML algorithms is additionally applicable to develop risk models of incident hypertension using clinical, behavioral, and socioeconomic features. A recent study published in Nature Medicine used ML models such as eXtreme Gradient Boosting (XGBoost) to assess 18 clinical, behavioral, and socioeconomic features and found that ‘age’ and ‘systolic/diastolic BP’ at baseline were particularly important. The study also showed that ML models could be more flexible and capable than traditional models in risk modelling, especially when using more complex data sources like as Electronic Health Records (EHRs) or genetic profiles.⁴

Despite the potential benefits of ML algorithms in hypertension treatment, several challenges remain. These include the need for prospective validation of AI-driven interventions, in order to further refine and evaluate AI-driven solutions for the management of hypertension, comprehensive systems that combine several AI techniques must be developed. Additionally, interdisciplinary cooperation between AI experts, physicians, and healthcare providers are necessary..⁵

In conclusion, recent advancements in machine learning modeling methods have the potential to transform hypertension treatment by providing real-time recommendations based on patient-specific characteristics. ML algorithms can be used to develop risk models of incident hypertension, identify new hypertension genes, and generate custom hypertension prescriptions using individual patient profiles. However, several challenges remain, and multidisciplinary partnerships among AI experts, clinicians, and healthcare providers are crucial to further optimizing and validating AI-driven solutions for hypertension management.