Precision Medicine and Clinical Decision Support at the Point of Care

Healthcare Gets Personal

Skyrocketing healthcare costs. An aging population. Widespread obesity. It’s a volatile trifecta that threatens to cripple the American economy unless the practice and delivery of care changes dramatically.

The statistics are staggering. Americans spent over $3.4 Trillion this year on healthcare. Medicare pays $50 billion per year for doctor and hospital bills incurred in the last 60 days of life. Today, one in three Americans is obese and there are over 5 million Americans with Alzheimer’s: it’s the most expensive disease in the nation costing us $259 billion this year alone (burdening taxpayers $20 million per hour). By 2030, one in five Americans (~72 million) will be over the age of 65. By 2050, the number of people with Alzheimer’s will triple and costs are predicted to exceed $1 Trillion to manage this disease alone.

Healthcare as we know it today is teetering dangerously on the edge of collapse. We need a solution. Precision medicine appears to be our best hope.

The Promise of Precision Medicine

For those “Trekkies” out there, the concept of pointing a tricoder at a person to secure a full genomic, phenotypic and clinical work-up and suggest exactly the right treatment for the patient may not seem that far-fetched. Ditto for those researchers, investors and visionaries embracing the field of precision medicine. But what is precision medicine and how will it change our lives?

The terms precision medicine and personalized medicine are often used interchangeably to describe medical care that has been tailored to a particular based on molecular profiling. However, genetic analysis alone doesn’t tell the whole story. Diseases are multi-modal, but not yet studied this way.

In the midst of this are the simple truths of medicine – patients simply want to feel better. Providers want to prescribe the right treatments for their patients. Payors want to help patients avoid disease or to treat disease as early and as cost-effectively as possible. Precision medicine promises to fulfil all of these ambitions by understanding diseases more thoroughly and being able to bring that knowledge to bear in the context of each patient to:

  • accurately predict which patients will do best on which medications
  • identify which patients are at risk of adverse drug:drug or drug:disease or drug:food interactions
  • predict which lifestyle or treatment plans would enable patients to optimize their health and wellness pre- or post-diagnosis

Why it’s Been out of Reach – Until Now

The vast majority of diseases are polygenic. The vast majority of people at the end stage of their lives have been diagnosed with more than one chronic disease and are typically consuming more than three medications. The complexity of this polygenic – polypharmaceutical – comorbidity problem is huge and it is, indeed, a very tough nut to crack.

Twenty years ago, conventional wisdom was that The Human Genome Project would reveal all by decoding our DNA sequence. In contrast, the Project exposed how little we understand about the complexity of disease and the influence of genetics, environment and lifestyle. The simple hypothesis of one-gene-one-disease has since been thoroughly debunked and we are now in agreement that how a patient responds to treatment will vary based on a multitude of factors. Historically, we’ve studied each factor one at a time in isolation which isn’t representative of how disease progresses, but we haven’t had the capability to analyze all of those factors in parallel – until now.

It’s certainly not for a lack of data. We currently have two Exabytes (that’s two million Terabytes) of healthcare data. As we push new imaging and molecular profiling technologies and wearable/IoT sensors further into the mainstream of healthcare, the volume of data we generate will continue to grow exponentially. Generating data is only one (largely solved) part of the problem. The bigger challenge is how do you analyze it all at the resolution and complexity at which disease actually manifests itself? That’s a multi-factorial hyper-combinatorial nightmare even for the world’s biggest supercomputers.

In addition to clinical history, genomic data, epigenetic effects, environmental factors, phenotype, behavioral and lifestyle habits like smoking and drinking must all be analyzed together to assess disease risk, progression and treatment response. This requires having tools that can straddle extreme variety in structured and unstructured data types from MRI scans to gene sequences to blood pressure readings. And, to be meaningful, all of these data types need to be analyzed in parallel across large populations to identify patterns that yield insights on how to treat disease at the level of personalized medicine. This has been beyond our computational capacity and collective know-how but is now within our reach.

So What’s Next?

Disease charities, researchers, data scientists, clinicians and hospitals are joining forces to tackle the thorny problem of disease using highly innovative computational biology analytics approaches. We need all of these skills to create and analyze multi-dimensional datasets. The goal is to identify the hyper-combinatorial features that drive disease processes across populations so that we may determine disease risk and therapy response for individual patients. We then need to take this knowledge and build decision support systems that can be used by clinicians and patients to cut through the complexity to inform choices about treatments, diet and lifestyle to optimize an individual patient’s health.

One such effort, GIRO Health, is an open initiative bringing together disease experts, clinical imaging and genomics solutions plus a unique approach to complex data analytics which leverages the power of artificial intelligence (A.I.). Experts are invited to join RowAnalytics, The MGH/HST Martinos Center for Biomedical Imaging, Envision Genomics, IBM and others to lend their expertise, tools and funding to engage in wholistic studies of disease. The initiative is taking aim first at neurodegenerative diseases such as Alzheimer’s, ASD (Autism Spectrum Disorder) and ALS (motor neuron disease). Cancers and other diseases will be studied as the GIRO Health open initiative in precision medicine expands.

Genomic, imaging, clinical, phenotypic and lifestyle data will be collected for large populations and analyzed in parallel to develop personalized plans for treatment. Point of care solutions will empower providers, caregivers and patients to make better informed decisions – at the personal level of healthcare. By working together, the promise of precision medicine may be realized sooner than we thought.

By | 2018-01-05T14:10:53+00:00 January 5th, 2018|blog|1 Comment

About the Author:

Loralyn is the team translator with nearly 20 years of experience listening to what scientists need and to what software developers say they can build, liaising between the two groups to drive product development and messaging so that both sides get what they need – and want. Loralyn has held a variety of roles within the life sciences anchored in market development for analytics and ‘omics technologies. She brings a combination of marketing, alliance management, sales and innovation to the company. Her specialty is connecting the dots aligning needs to products to people with a personal mission to help others and upholds the philosophy that better tools = better health.

One Comment

  1. Clinical Decision Support February 5, 2018 at 6:52 am - Reply

    Thanks for sharing the valuable information. The ED clinical team can be greatly helped by electronic health record (EHR) tools with embedded or add-on clinical decision support (CDS) that supports both patient safety as well as throughput efficiency to reduce such risks. CDS should be used in the hospital ED because it helps to reduce the risk of medication errors, reduce misdiagnoses, provide the entire care team with consistent, reliable information, and Improve efficiency and patient throughput.

Leave A Comment

Solve : *
13 − 1 =