It’s 2020. Consumers are accustomed to seamlessly personalized services that enhance and simplify their lives. Businesses improve efficiency and boost profits using advanced analytical tools to make data-driven decisions, but the healthcare industry is lagging.
The dream of using big data analytics, genomics and other diagnostics technologies to personalize medicine to make informed decisions about optimized and individualized patient care has been around for decades, but isn’t yet broadly realized. Hospitals remain staggeringly inefficient, unable to understand which treatments work best for their patients at the lowest cost. Doctors face a daily deluge of data on patients and increased time demands for data entry, but are not yet equipped with effective tools to access and analyze that data to help improve their patients’ health. Pharmaceutical companies invest hundreds of millions of dollars in promising drugs that ultimately fail in late-stage trials, and then have difficulty identifying appropriate patients once a drug is on the market. Insurance companies find it difficult to accurately predict risks of individual patients and populations, hindering the adoption of accountable care models.
All these factors contribute to unsustainable costs and poor health outcomes in the U.S., and that all traces back to the inability to access and properly understand data. In the end, patients suffer, often unable to find affordable and effective treatments.
How do we solve this? The answers lie in data; we know this to be true. Huge challenges remain, but unlocking this data will generate insights and knowledge that improve the efficiency and effectiveness of the healthcare system.
Healthcare is a very complicated industry with many barriers to progress. Data is siloed across institutions, greatly inhibiting our ability to analyze and understand it. Recent efforts by the U.S. government’s Office of the National Coordinator for Health IT (ONC) and the Centers for Medicare & Medicaid Services (CMS) to enact data-sharing policies that would allow patients more control over their health data has created significant controversy around the balance between privacy risks and accessibility. There are real potential risks, to be sure — patient health data in the hands of bad actors could cause significant harm — but these risks must be weighed against the great benefits that would come from a greater understanding of the data.
In response to this controversy, Stephanie Reel, CIO of Johns Hopkins Health System and a member of ONC’s Health IT Policy Committee’s Information Exchange Workgroup said, “The next big discovery will come from the … use of technology and information. I don’t want us to [be] too careful and too controlling because I think there is some risk that we will not make that next big discovery or cure that dreadful form of cancer.”
Unlocking data and making it accessible will be a major leap forward, but we also need tools to easily analyze it. Data in electronic medical records was structured with billing in mind, not to readily understand patients’ diseases.
However, with the current revolution in machine learning, image analysis and natural language processing, we now have the right big data tools available to organize and make sense of the data. So, now what can be done? Here we will discuss a few examples of how unlocking this data and generating insights can have a direct impact on patients who are suffering, from global pandemics to rare diseases.
For major diseases affecting huge portions of the population, even small improvements in treatment paradigms have an enormous impact on reducing healthcare costs and improving patient outcomes. For example, diseases like heart failure can manifest in varied ways, and there is no one-size-fits-all approach to treatment. Finding the right treatment for a given patient can get them out of the hospital faster and even save their life, yet, currently, why a given patient does well or poorly on a certain treatment regimen often remains a mystery.
However, when we use advanced analytical tools to look across millions of patients, patterns emerge that identify subpopulations that respond to treatments differentially, providing a pathway toward optimization of personalized treatment protocols. In the near future, when a new patient comes to a hospital, their symptoms and diagnostic labs will be compared across millions of similar patients and matched to a treatment that the data proves should have the highest likelihood of success. For many diseases, there may be hidden subpopulations that do not respond to any available drug or standard of care. Uncovering such populations will enable targeted drug development and efficient clinical trials that make novel treatments and cures available faster.
On the opposite end of the spectrum, patients suffering from one of the thousands of rare diseases and disorders frequently suffer for years without diagnosis and, even if they are diagnosed, treatment may not be available. An individual doctor or hospital will see so few of these patients they may never identify the patterns. Analysis of big data pooled across many institutions and rare disease patient registries will help to identify which patients should be referred for genetic screening, so patients can be diagnosed earlier and put on appropriate therapies when available.
Many rare diseases lack approved drugs, but pharmaceutical companies currently face major challenges estimating rare disease population sizes and locations, knowing which diseases to target, and recruiting for clinical trials. Knowing who and where rare disease patients are — finding the needles in the haystack — will greatly accelerate the delivery of new rare disease drugs to patients who currently lack options.
With the barriers surrounding siloed electronic medical records beginning to fall, with genetic and other “omics” data becoming cheaper and more widespread, and with powerful big data analytics tools in hand, we’re now on the cusp of realizing the dream of truly personalized medicine. Delivering this dream will require extensive collaboration among stakeholders across the healthcare industry, with firm focus on and commitment to the patients they serve. Now is the time. The patients have been waiting long enough.