Precision Health, Digital Health and AI – Paving way for the future of Healthcare
Research and Evidence
The practice of conventional medicine is focused on illness. By design, the system reacts to symptoms, pairing few and sparse measurements with diagnoses informed by standard statistical distributions of large populations. Often these “normal” populations are correlated to age, however, it is well documented that disease and disorders of aging are detectable and preventable, and that individual response to pharmacological intervention can vary widely.[1] An alternative approach is the nascent field of Precision Medicine.[2]
Precision Medicine, by contrast, focuses on health. The field supports the theory that an individual’s unique metabolic, phenotypic, and genotypic makeup, can be used to predict a disease trajectory, and can inform tailored interventions to prevent or effectively reverse flagrant illness.[3] This approach is proactive, and reflects the unique nature of individual physiology via the longitudinal analysis of both hard-wired and dynamic biomarkers.
Longitudinal data is the cornerstone of Precision Medicine
A person is not a paper. Instead of comparing an individual’s parameters to an abstract (and often non-informative) average of a population distribution, Precision Medicine depends on a N-of-1 methodology. This methodology uses the individual’s baseline data to detect possible anomalies due to subclinical disease, disorders of aging such as prediabetes, hyperlipidemia, osteopenia, and even the effectiveness of therapeutic interventions.
As the field of personalized Precision Medicine emerges, (evident by the National Institutes of Health research program in 2019), the value of capturing longitudinal data in individuals that reflects their health trajectory accurately has become increasingly significant.[4]
Longitudinal health data is fundamental to the detection and treatment of all possible diseases and conditions, particularly in Type 2 Diabetes – a progressive disease, where metabolic changes, including but not limited to prediabetes, can occur a decade or more before symptoms, and predict clinical diagnosis. Often these metabolic deviations can be effectively reversed, even optimized, with tailored interventions leading to positive lifestyle changes, such as quality sleep, diet modification, exercise intensification, and enhanced pharmacotherapy, should that be indicated.
Measuring success for the Precision Health approach
In order to assess the success of an intervention, the impact must be accurately evaluated over time. To date, the gold standard for diabetes management in clinical medicine are the universal biomarkers HbA1c and blood glucose.
As an individual’s HbA1c and glucose trends over time are vital to detect and predict a diabetic trajectory and to monitor the success of personalized interventions, lab values need to be consistent and reliable across test centers and time periods. To empirically test this hypothesis, our Founder and Chief Executive Officer, Florence Comite MD and her team set up a prospective study at the Center for Precision Medicine & Health to assess and compare different laboratories’ biomarkers at baseline and specific subsequent intervals based on her background and expertise in clinical research at Yale and the National Institutes of Health. The analysis of this work shows statistically and clinically significant differences in individuals’ carbohydrate metabolism values from the same blood draw when analyzed by different laboratories. For this reason, we consistently use the exact same lab for all individuals over time, to ensure consistency and reliability of data across time and between individuals.[5]
In routine practice, these discrepancies present a roadblock to the widespread adoption of precise, personalized, and proactive health measures in the clinic.
Precision Medicine can unlock unique predictors of future disease
Genetics play a role in susceptibility to metabolic disorders of aging, including Type 2 diabetes. Dr. Comite and her team investigated the relationship between polygenic variants associated with insulin resistance and specific phenotypic biomarkers linked to metabolic syndrome in individuals. They found that Adiponectin is inversely correlated with the number of DNA risk variants.[6] Models utilizing polygenic variants and biomarkers lead to early detection and interventions to reverse and prevent disorders of aging in at-risk individuals.
Digital Health – Bringing Precision Medicine to All
The advent of digital technologies has disrupted several industries, making them more consumer friendly and convenient to use. Air travel, banking and retail have transformed completely in keeping with these technological advances and improved the user experience across the entire value chain. Similarly, digital technologies have the potential to disrupt how healthcare is delivered and received.
Our Co-Founder and Chief Product Officer, Kamal Jethwani, MD MPH, has researched extensively how digital tools may transform the patient experience.
Digital tools are found useful, usable and acceptable to a majority of patients, be it in the case of patients with Atrial Fibrillation, Overactive Bladder, Diabetes or Heart Failure.[7] The ubiquity of these tools in individuals’ lives, coupled with their convenient, user-friendly form factor, makes them an ideal companion in helping people care for themselves.
However, if this use does not translate to actual clinical and experiential value, engagement with these tools drops significantly over time. Several studies show that consistent use of digital tools results in drastic improvements in clinical outcomes as well as cost containment.[8]
Personalization – delivering the promise of Precision Medicine and Health
The most important tenet of Precision Medicine is our ability to personalize our interventions to everyone, where no two individuals would receive the same treatment or intervention. Digital tools can finally deliver on that promise through mass customization in ways that are efficient and scalable. Research has shown that techniques that can personalize messages for consumers can improve both engagement and clinical outcomes.[9]
The exponential progress we are making in the field of artificial intelligence is enabling behavioral changes in healthcare consumers and the use of complex healthcare data in meaningful ways.[10] In the field of personalizing interventions for chronic disease, we have tried various techniques, like reinforcement learning, temporal case based reasoning and product platform approach, to name a few.[11] These techniques will be further magnified by the wearables and genomics industry, each lend volumes of data to inform a consumers’ behavioral genotype and phenotype.[12]
By leveraging clinically proven Comite Center for Precision Medicine & Health algorithms based on the proof-of-concept that integrate N-of-1 health data — family history, personal journey, lifestyle, hormones, metabolism, genomics, wearables — together with the predictive power of AI, and the scalability of digital health tools, Precision Health, Digital Health, and AI — together — are paving the future of health care.
[1] León-Cachón RB, Ascacio-Martínez JA, Barrera-Saldaña HA. (2012). Individual response to drug therapy: bases and study approaches. Rev Invest Clin. Jul-Aug; 64(4):364-76. Review. PubMed PMID: 23227587.
[2] Collins, F. S., & Varmus, H. (2015). A new initiative on precision medicine. The New England Journal of Medicine, 372(9), 793–795.
[3] Klonoff, D. C. (2015). Precision Medicine for Managing Diabetes. Journal of Diabetes Science and Technology, 9(1), 3–7.
[4]Sankar, P. L., & Parker, L. S. (2017). The Precision Medicine Initiative’s All of Us Research Program: An agenda for research on its ethical, legal, and social issues. Genetics in Medicine, 19(7), 743–750; National Institutes of Health (NIH)—All of Us. (n.d.). Retrieved October 15, 2019, from https://allofus.nih.gov/
[5] Comite, F., Santostasi, G., Sansaricq, G., & Fisch, M. (2019). SAT-LB024 Imprecision in Biomarker Analysis: Major Hurdle to Detection and Reversal of Disorders of Aging? Journal of the Endocrine Society, 3 (Supplement_1); Florence Comite, MD, Lorena Martin PhD, Jennifer Braun MPH PA-C, Francisco Carreño-Galvez PhD, Adaobi Onunkwo BA, Joel T. Dudley PhD. Lab Variability: A Road-Block in Precise, Proactive Diabetes Diagnosis and Intervention.
[6] Florence Comite, MD, Kirby Mateja, BS, Lauren Klein, MPAS, PA-C, Gabriella Sansaricq, BS, Giovanni Santostasi, PhD, Christopher Arboleda, MS, Sara Riordan MS, LCGC. (May 8-9, 2019). Select DNA Variants and Biomarkers Predict Metabolic Disorders of Aging. LabRoots, Yorba Linda, CA 92886
[7] Hirschey, J., Bane, S., Mansour, M., Sperber, J., Agboola, S., Kvedar, J., & Jethwani, K. (2018). Evaluating the Usability and Usefulness of a Mobile App for Atrial Fibrillation Using Qualitative Methods: Exploratory Pilot Study. JMIR Human Factors, 5(1), e13; Devoe, C., Bane, S., Hirschey, J., Palacholla, R., Centi, A., Odametey, S., Kvedar, J. (2018).
[8] Agboola S, Jethwani K, Lopez L, Searl M, O’Keefe S, Kvedar J. Text to Move: A Randomized Controlled Trial of a Text-Messaging Program to Improve Physical Activity Behaviors in Patients With Type 2 Diabetes Mellitus. Journal of Medical Internet Research. 2016;18(11):e307; Khateeb, Kholoud; Agboola, Stephen; Jethwani, Kamal; O’Reilly, Peter; Kamarthi, Sagar. The Study of Effectiveness of a Remote Patient Monitoring Program for Heart Failure Patients. (2015).
[9] Mohammadi, R., Centi, A. J., Atif, M., Agboola, S., Jethwani, K., Kvedar, J. C., & Kamarthi, S. (2019). A Neural Network Based Algorithm for Dynamically Adjusting Activity Targets to Sustain Exercise Engagement Among People Using Activity Trackers. BioRxiv, 775908; Agboola, S., Jethwani, K., Lopez, L., Searl, M., O’Keefe, S., & Kvedar, J. (2016). Text to Move: A Randomized Controlled Trial of a Text-Messaging Program to Improve Physical Activity Behaviors in Patients With Type 2 Diabetes Mellitus. Journal of Medical Internet Research, 18(11), e307.
[10] Agboola S.O., Jethwani K. (2017). Technology: Revolutionizing the Delivery of Health Behavior Change Interventions with Integrated Care. In: Maheu M., Drude K., Wright S. (eds) Career Paths in Telemental Health. Springer, Cham.
[11] Javad, M. O. M., Agboola, S. O., Jethwani, K., Zeid, A., & Kamarthi, S. (2019). A Reinforcement Learning–Based Method for Management of Type 1 Diabetes: Exploratory Study. JMIR Diabetes, 4(3), e12905; Jalali, N., Agboola, S., Jethwani, K., Zeid, I., & Kamarthi, S. (2017, February 8). Temporal Case-Based Reasoning for Personalized Hypertensive Treatment. Presented at the ASME 2016 International Mechanical Engineering Congress and Exposition; Boyaci, Ilke; Chin, Jessica; Zeid, Abe; Kamarthi, Sagar V; Agboola, Stephen; et al. Product Platform Approach to Personalized Type 2 Diabetes Mellitus Management. (2014).
[12]Jethwani, K., Kvedar, J., & Kvedar, J. (2010). Behavioral phenotyping: A tool for personalized medicine. Future Medicine, 7(6).
