"1.2 billion people worldwide rely on their smartwatches to keep track of their health, unaware that the technology is inherently flawed. It doesn’t accurately read the heart rate of darker skinned people" (IPG Health, Ref 1). This quote is from a recap of the 2024 Cannes Lions International Festival of Creativity, New York, June 26 2024: Healthcare's tragic flaw: Big bias in big data.

Are you aware of the potentials and challenges of big data in healthcare?

In 2013, Joel Selanikio M.D., a data scientist, discussed a big data revolution in healthcare in a TED talk (Ref 2). He briefly pointed out that utilizing the large amount of data collected by healthcare workers would lead to advances in healthcare. He may not have been the pioneer of this idea, but he highlighted the importance of the standardization of data. More recently, in 2020, a paper in Nature medicine (Ref 3) reinforced the concepts of data standardization and the healthcare revolution.

Medical data is being produced at a large scale. In the paper by Shilo, et. al, they indicated that along with technological advances in data generation and data analysis methodology, the healthcare revolution is coming. The authors characterized multiple properties of medical data in order to address some potentials and challenges.

• Properties of the medical data (Figure 1. from Ref 3: The different axes of health data. The complexity of large health datasets can be represented by distinct axes, each encompassing a quantifiable property of the data.)

• Potentials

  • Disease diagnosis, prevention and prognosis

  • Modeling disease progression

  • Genetic and environmental influence on phenotypes

  • Target identification

  • Improvement of health processes

  • Disease phenotyping

  • Precision medicine

• Challenges

  • Small sample sizes and low population heterogeneity

  • Extracting characteristics of patients from EHRs not straightforward

  • Continuously updated, accurate, well-calibrated and delivered prediction model

  • Incompleteness and irregularity of data

  • Heterogeneity of patient comorbidities and medication usage for modeling

The discussion of the potential of big data resources is ongoing. We must strive to advance the understanding of health, disease, and analytical methodologies.

References

1. https://ipghealth.com/news/healthcare-s-tragic-flaw-big-bias-in-big-data

2. https://www.ted.com/talks/joel_selanikio_the_big_data_revolution_in_health_care?subtitle=en

3. Shilo, S., Rossman, H. & Segal, E. Axes of a revolution: challenges and promises of big data in healthcare. Nat Med 26, 29–38 (2020). https://doi.org/10.1038/s41591-019-0727-5