For most of us, the health maintenance, disease prevention, diagnostic and treatment recommendations received throughout our lives are based largely on a combination of our health attributes, our care providers and the population-based clinical evidence deemed most likely to apply in each situation.
This is not to imply that our healthcare providers didn’t recognize each of us as an individual. Or that they didn’t strive to provide the best and most individualized care possible, based on the most current and comprehensive information available to them.
While we can share our individual and family health histories and current concerns – repeatedly – providers are limited by the amount of information the human brain can store and process at any given point in time, and by their available biomedical knowledge.
Fortunately, as EHR technology has become more widely adopted, the volume of patient data captured and processed has vastly increased. Not only are there more data, but these data are more accessible for providers, patients and the automated clinical decision support (CDS) software that offers providers helpful reminders and suggestions.
Still, the clinical, evidence-based knowledge available to the provider is often based on investigations involving large, homogeneous (mostly Caucasian, middle-class) populations. Individuals with rare diseases and conditions, minority and underserved populations, and diverse racial and social profiles rarely are represented in these studies.
This one-size-fits-all approach actually fits very few of us. Wouldn’t it be nice to have healthcare more finely tuned to our own genetics, environments and lifestyles?
Precision medicine is seeking to meet this objective – to make health maintenance, disease prevention, diagnoses and treatment solutions more precisely selectable for each individual.
Healthcare has long recognized the important role that individual variability plays in health.
Before an individual receives a blood transfusion, compatibility testing matches donor and recipient blood types (A, B, O) and rhesus factors. Blood and tissue matching procedures identify the compatibility of organ and stem cell transplants. Genetic testing to determine whether an individual’s DNA includes specific variants associated with treatable conditions has also become commonplace.
For example, women with a family history of breast cancer are tested for BRCA1 and BRCA2 gene variants. And before prescribing warfarin for patients with blood clotting conditions, testing for VKORC1 gene variants that affect warfarin metabolism are performed.
However, developing specific genetic tests with clinical utility, based on validated biomarkers, present many well-documented challenges.
Consider that a single person’s genome (i.e., DNA) comprises approximately 3 billion pairs of nucleotides, organized across 23 chromosomes. Each of these pairs might deviate from the human reference genome in one of several ways; as of March 2018, the National Center for Biotechnology Information identified and cataloged more than 1.8 billion variants. Each genetic test is looking specifically for a match with one or more of these variants – like looking for a needle in a haystack.
Precision medicine looks at the huge population of individuals residing on planet earth as a haystack full of needles – a rich source of valuable biomedical knowledge to be mined and translated into more precise diagnostics and treatments, and healthier humans.
In the U.S., the federal government launched the Precision Medicine Initiative for “health tailored to you.” The initiative then defined this approach to medicine as “an emerging approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person.”
Notice the nuanced difference between healthcare “tailored” for each individual (i.e., personalized medicine) and care that “takes into account” individual variability. This includes far more than the health histories and chief complaints our current EHRs capture, to include individual DNA, living environment and daily health habits. In other words, the initiative sought to include every individual in the clinical research cohorts from which new, more comprehensive, evidence-based biomedical knowledge would emerge.
This more inclusive approach will enable more accurate matching of prevention strategies and treatments for specific conditions with the sub-groups of people for which they are most likely to be beneficial.
This more personalized approach to care will enable pharmaceutical developers to more precisely formulate drugs and to characterize the individuals for whom these drugs are effective. For example, some drugs that fail in clinical trials over large, populations may be shown highly effective for some subset within that population. Today, these drugs would not be sufficiently applicable to be profitably brought to market. Precision medicine would enable these drugs to be specifically developed for and marketed to the types of individuals for whom the drug did show benefit. Researchers, care providers and consumers are hoping that this will become the predominant approach across all of health prevention, maintenance and treatment.
The key to the success is big data – amassing large quantities of genetic, clinical, social, lifestyle and preference data across broad, heterogenous populations. Artificial intelligence algorithms then mine knowledge from the accumulated data.
We need look no further than popular social media to understand how new knowledge can be derived from large accumulations of individual data. To this vast store of personal data, add clinical data captured by EHR technology and DNA data captured through whole-genome sequencing, and precision care becomes realizable.
Whole genome sequencing (WGS) is a process in which a laboratory, using a small DNA sample provided by an individual (e.g., saliva, blood, hair), generates a profile of the 3 billion nucleotides that comprise that individual’s complete genome. Within the U.S., many individuals have provided their saliva samples to personal genomics companies in exchange for insights into their ancestry and inheritance. However, WGS is not performed in routine laboratory testing (e.g., newborn screening), nor is it widely used within U.S. healthcare. However, several other countries, including Iceland, The Netherlands, Japan, Sweden and the United Kingdom, have undertaken broad sequencing programs to more effectively characterize their populations, and they are sharing these data for biomedical research.
This amassing of very large quantities of highly personal information raises new challenges for both security and privacy protection. We need only examine the Department of Health and Human Services’ list of breached protected health information to see the security challenges in protecting clinical information.
Protecting the privacy of genomic data is a whole new challenge. Each individual’s genome is unique (except for identical twins), so a WGS can never be truly unidentifiable. Yet, a discarded coffee cup is likely to contain the material necessary to generate a WGS – which may also reveal private information about the individual’s family members.
Clearly, attaining precision medicine will require skilled management of information security protection, and transparency of, and individual controls over, the collection, sharing and use of personal information. Plus, management of expectations with respect to both associated risks and potential benefits, and legal protections addressing broad, personal data collection and use are needed.
The potential benefits are enormous, and the associated security and privacy risks are daunting. Perhaps we are fortunate that we have the social networking community as a bellwether of the risks associated with big data analytics.
The views and opinions expressed in this blog or by commenters are those of the author and do not necessarily reflect the official policy or position of HIMSS or its affiliates.
December 1–2, 2020 | Digital Conference
While healthcare is beyond the hype and already seeing the influence of machine learning and AI technologies within their workflows, the successful implementation that drives results depends on achieving analytics maturity and ensuring data quality and governance. Join HIMSS and take a holistic, workshop approach with a focus on implementation.