Free cookie consent management tool by TermsFeed Policy Generator

Intersection of Machine Learning and Health Economics

Our team presented multiple workshops on machine learning for health economics and outcomes research
Thirty percent of the world’s data volume is generated by the healthcare industry. By 2025, the global datasphere will explode to 175 zettabytes and the healthcare industry contribution will grow to 36%. (1) Through its ability to make these large volumes of real-world evidence usable, machine learning is an area of growing interest in healthcare because of its potential to identify customized, precise medical solutions and streamline treatment. However, its use in health economics and outcomes research (HEOR) field is just beginning to emerge.
In their workshop, presented at ISPOR Asia Pacific Conference in 2020 and ISPOR Annual Meeting in 2021, Turgay Ayer, Jag Chhatwal, and Selin Merdan of Value Analytics Labs discussed the intersection of machine learning and HEOR. They discussed:
  • commonly used methods for machine learning, including support vector machines and random forests,
  • how machine learning differs from artificial intelligence,
  • current application of machine learning in HEOR, including identification of risk factors, precision medicine, and machine-learning based cost-effectiveness analysis, and
  • future trends in the field such as inclusion of large mobility data and social networks data into HEOR and modeling, and use of data from wearable devices in HEOR.

They also discussed challenges in field of machine learning and HEOR. These include lack of comprehensive datasets and infrastructure to access those datasets, lack of familiarity with machine learning methods, lack of transparency and reproducibility. They emphasized that we should keep our expectations from machine learning field realistic—these methods don’t necessarily generate new insights that otherwise are not known from traditional statistical methods.

Finally, Dr. Ayer added that computer-based technologies have disrupted several industries. Taking an example of disruption in retail industry, he mentioned healthcare industry follows retail industry with about a two-decade lag time, and we anticipate that computer-based technologies will have similar disruptive effects in HEOR space in the years to come.

If you have a large data challenge and would like to use machine learning, Value Analytics Labs would love to collaborate.  Simply contact us through our online form or email

  1. Reinsel D, Gantz J, Rydning J. The Digitization of the World from Edge to Core. Published online 2018:28.

©2023 Value Analytics Labs. All rights reserved | Terms of Service | Privacy Policy | Design by 9piksel