Precision Public Health: From Individual Data to Population Intelligence
Author: Peehu Sachdeva, Market Strategy Lead, Axia Medicine
Published by Axia Medicine | November 2025
This is the third article of a series. Previous entries:
Healthcare produces more data than ever before. Genomes are sequenced in days, wearables capture signals by the second, and electronic records document millions of encounters. Yet, as introduced in The Genomic Paradox, technological progress alone does not ensure clinical impact. And as The Multimodal Shift demonstrated, institutional structures prevent data being combined in effective ways for care. Because data does not move with patients, preventable harm persists.
An analysis of 248 EHR-related malpractice claims found that over 80% resulted in medium to severe harm, with nearly a third contributing to deaths. Cases include medication errors when discharge instructions never reached the next provider and neurological deterioration missed because paper and digital systems could not integrate.
These are not isolated incidents. These failures share a common denominator: they occur wherever responsibility for data cannot extend seamlessly beyond the edge of an institution. This design choice is embedded in every workflow and database, and is what makes health data fragmentation inevitable.
COVID’s Integration Failure
The pandemic revealed this problem at population scale. A National Academies assessment concluded that U.S. genomic epidemiology suffered from insufficient funding, poor coordination, and limited integration capacity, blocking correlation of viral sequences with clinical and epidemiological data.
Each separate discipline worked within its own institutional boundary, replicating at national scale the same ownership logic that fragments clinical care.
Genomicists sequenced variants. Epidemiologists mapped the spread. Social scientists showed how housing and occupation shaped risk. Yet these insights rarely connected. Genetic variants linked to severe disease were identified, but existing systems could not map them onto the communities most exposed. Vaccine acceptance data remained siloed from hospitalisation patterns. Environmental exposures were seldom integrated.
To echo the The Genomic Paradox, the failure was not data scarcity. It was infrastructure built for institutional control rather than integration.
Clinical Research at a Standstill
The same fragmentation undermines research. Over half of clinical trials are discontinued due to recruitment failure. More than 80% miss enrollment timelines, and in rare diseases, nearly a third of Phase III trials collapse for the same reason.
The FOR-DMD trial for Duchenne Muscular Dystrophy shows the cost. Planned to open 40 sites by 2011, it began recruitment only in 2013, an 18-month delay caused by disconnected registries and databases. Each day of delay translates to lost opportunity and millions of dollars in costs; meanwhile, 30 million Americans live with rare diseases, yet most remain unaware of trials that could benefit them.
The problem is not scientific capability but visibility, when patients remain invisible to the very system built to study them.
What Population Scale Demands
As explored in our previous article, precision medicine requires data beyond just health records: genomics, proteomics, imaging, behavioural, environmental and socioeconomic data, all contribute to delivering personalized Care in the most efficient way.
Projects like UK Biobank demonstrate the impact of population-scale, multimodal datasets: when data is integrated across genetics, imaging, and clinical records, discoveries in cardiovascular disease and dementia follow at unprecedented pace.
However, such national initiatives each took more than a decade and billions of investment. If every therapeutic area and every country required comparable efforts, the industry’s progress would be measured in decades.
The cost of inaction is vast. U.S. racial and ethnic health inequities cost $451 billion in 2018, with education-related disparities adding $978 billion. Without change, combined costs could surpass $1 trillion annually by 2040. Precision public health therefore reframes equity as not only a moral imperative but an economic necessity.
Patient-Driven Infrastructure
The common thread is fragmentation caused by institution-centric design. Adding more applications cannot fix this. Patient-driven infrastructure offers a structural solution.
When patients own their complete multimodal data, information travels with them across care. Medication errors from missing discharge notes become preventable. Clinical changes cannot be lost between systems. Recruitment shifts from institutional matching to patient-initiated participation. The FOR-DMD delay would not have occurred if patients could choose to make themselves visible to relevant research.
These failures share a common denominator: they occur wherever responsibility for data stops at the edge of an institution.
At population scale, patient-contributed data with preserved demographic, geographic, and socioeconomic context allows integration by design. Genetic susceptibility overlays with occupational exposure, vaccine acceptance aligns with outcomes, and environmental factors cluster with clinical patterns. Integration happens at the source, not through fragile institutional negotiations.
Across health systems worldwide, what’s missing is not data but a structural foundation that starts with the patient as the persistent layer of the system.
When integration begins at the individual and remains portable across every interaction, systemic failures stop propagating.
From Constraint to Capability
What has really been commoditized is data capture. What remains scarce is integration at scale with context and agency preserved. That distinction determines whether health systems prevent harm, whether trials recruit efficiently, and whether public health learns from crises rather than repeats them.
Precision public health will not be achieved by more apps. As demonstrated in the following article, precision medicine requires infrastructure that makes fragmentation structurally impossible and finally builds health systems that can see patients and populations in full.
The core question is not how to collect more data, but how to build systems where health data disconnection is impossible by design; a Precision Medicine ecosystem where the integration of health information begins with the patient.
References
Crow, R.A., Hart, K.A., McDermott, M.P., Tawil, R., Martens, W.B., Herr, B.E., McColl, E., Wilkinson, J., Kirschner, J., King, W.M., Eagle, M., Brown, M.W., Hirtz, D., Lochmuller, H., Straub, V., Ciafaloni, E., Shieh, P.B., Spinty, S., Childs, A.-M. and Manzur, A.Y. (2018). A checklist for clinical trials in rare disease: obstacles and anticipatory actions-lessons learned from the FOR-DMD trial. Trials, [online] 19(1), p.291. doi:https://doi.org/10.1186/s13063-018-2645-0.
Desai, M. (2020). Recruitment and retention of participants in clinical studies: Critical issues and challenges. Perspectives in Clinical Research, [online] 11(2), p.51. doi:https://doi.org/10.4103/picr.picr_6_20.
Dorantes-Gilardi, R., Ivey, K.L., Costa, L., Matty, R., Cho, K., Gaziano, J.M. and Albert-László Barabási (2025). Quantifying the impact of biobanks and cohort studies. Proceedings of the National Academy of Sciences, [online] 122(16), pp.e2427157122–e2427157122. doi:https://doi.org/10.1073/pnas.2427157122.
Graber, M.L., Siegal, D., Riah, H., Johnston, D. and Kenyon, K. (2019). Electronic Health Record–Related Events in Medical Malpractice Claims. Journal of Patient Safety, 15(2), p.1. doi:https://doi.org/10.1097/pts.0000000000000240.
LaVeist, T.A., Pérez-Stable, E.J., Richard, P., Anderson, A., Isaac, L.A., Santiago, R., Okoh, C., Breen, N., Farhat, T., Assenov, A. and Gaskin, D.J. (2023). The Economic Burden of Racial, Ethnic, and Educational Health Inequities in the US. JAMA, [online] 329(19), pp.1682–1692. doi:https://doi.org/10.1001/jama.2023.5965.
National Institutes of Health (NIH). (2025). Rare Diseases. [online] Available at: https://www.nih.gov/about-nih/nih-turning-discovery-into-health/promise-precision-medicine/rare-diseases.
Niemi, M.E.K., Daly, M.J. and Ganna, A. (2022). The human genetic epidemiology of COVID-19. Nature Reviews Genetics. doi:https://doi.org/10.1038/s41576-022-00478-5.
Olaker, V.R., Fry, S., Terebuh, P., Davis, P.B., Tisch, D.J., Xu, R., Miller, M.G., Dorney, I., Palchuk, M.B. and Kaelber, D.C. (2024). With big data comes big responsibility: Strategies for utilizing aggregated, standardized, de‐identified electronic health record data for research. Clinical and Translational Science, 18(1). doi:https://doi.org/10.1111/cts.70093.
Prosperi, M., Min, J.S., Bian, J. and Modave, F. (2018). Big data hurdles in precision medicine and precision public health. BMC Medical Informatics and Decision Making, 18(1). doi:https://doi.org/10.1186/s12911-018-0719-2.
Ranum, D. (2019). Electronic Health Records Continue to Lead to Medical Malpractice Suits. [online] Thedoctors.com. Available at: https://www.thedoctors.com/articles/electronic-health-records-continue-to-lead-to-medical-malpractice-suits.
