Personalized Medicine Study Finds Machine Learning-Based Approach to Statin Therapy Outperforms Traditional High-Risk Method

Strategic Insights -

Current study results on drugs and therapies - personalized medicine have been published, highlighting the potential of machine learning to personalize treatment and improve population health outcomes. Researchers from the Graduate School of Public Health utilized the Shizuoka Kokuho Database to investigate heterogeneity in statin treatment effects, employing a 1:1 propensity score-matched cohort design to evaluate the effect of statins in preventing a composite outcome of cardiovascular and cerebrovascular events and all-cause mortality. The study found that a machine learning-based high-benefit approach achieved a number needed to treat (NNT) of 15.1, significantly outperforming the traditional high-risk approach.

Key Takeaways:

  • A machine learning-based approach to statin therapy, known as the high-benefit approach, outperformed the traditional high-risk approach in preventing cardiovascular and cerebrovascular events and all-cause mortality.
  • The high-benefit approach achieved a number needed to treat (NNT) of 15.1, compared to 29.5 for the high-risk approach.
  • The study found substantial heterogeneity in treatment effects, highlighting the need for personalized treatment strategies.
  • The researchers utilized a 1:1 propensity score-matched cohort design to evaluate the effect of statins in a sample of 8,792 individuals.
  • The study's findings demonstrate the potential of machine learning to enhance statin therapy by personalizing treatment and minimizing unnecessary medication.

Statistics:

  • 8,792 individuals were included in the propensity score-matched cohort.
  • The mean age of the sample was 67.4 years, with 68.6% women.
  • The high-benefit approach achieved a number needed to treat (NNT) of 15.1 (95% CI: 9.4-23.4).
  • The NNT for the high-risk approach was 29.5 (95% CI: 17.2-235.3).

Sources:

  • Watanabe, R., et al. (2025). Machine learning-based high-benefit approach versus traditional high-risk approach in statin therapy: the Shizuoka Kokuho database study. Scientific Reports, 15(1), 27627.