Machine Learning Times
Machine Learning Times
EXCLUSIVE HIGHLIGHTS
Effective Machine Learning Needs Leadership — Not AI Hype
 Originally published in BigThink, Feb 12, 2024.  Excerpted from The...
Today’s AI Won’t Radically Transform Society, But It’s Already Reshaping Business
 Originally published in Fast Company, Jan 5, 2024. Eric...
Calculating Customer Potential with Share of Wallet
 No question about it: We, as consumers have our...
A University Curriculum Supplement to Teach a Business Framework for ML Deployment
    In 2023, as a visiting analytics professor...
SHARE THIS:

9 years ago
Predictive Medicine Depends on Analytics

 

Regression models, Monte Carlo simulations, and other methods for predicting what’s around the corner have been in use for decades. It’s only recently, though, that advances in information technology have made it possible for predictive tools to access and manipulate big data, and to do so continuously — accelerating the generation of insights, and opening up opportunities to anticipate issues with unprecedented precision. Think of the colleges that are increasingly able to identify students at risk of dropping out and intervene before they do. Or lenders’ enhanced abilities to gauge credit risk. Energy, agriculture, insurance, retail, human resources — no industry is unaffected. But nowhere is the potential of this new era of opportunity more apparent and exciting than it is in health care.

Predictive analytics is fueling a transformation from a focus on the volume of procedures to the value of outcomes. Predictive tools are helping providers — both doctors’ groups and hospitals — assess patients’ risk of contracting a whole host of diseases and conditions. They can come up with individualized regimens by tapping into electronic medical records to identify the types of patients who are most likely to respond to a particular type of therapy. They can pinpoint treatments that sustain health in a more precise way than ever before. And they can identify individuals who are likely to stop benefiting from a specific regimen at a given time. For the volume-to-value paradigm shift in health care, predictive analytics, though rarely visible, is the essential enabler.

Used to its full potential, this is predictive medicine — the ability to integrate and analyze known disease characteristics with a specific patient’s history and health status, and use the resulting insights to change outcomes and inform new directions for life sciences research and development. And in this new arena, the once-clear lines between companies that make drugs and medical devices, providers who diagnosis illnesses and treat patients, and payers who provide the financial support for care are blurring. Actors in this ecosystem are increasingly working together rather than handing off information or tasks to the next entity in a linear process. They are establishing more iterative and interactive connections with each other and with patients. They are collaborating with (sometimes highly unlikely) partners. They’re also sharing risk.

The new business model has yet to solidify, and the leaders have yet to emerge. The positions are there for the taking. But not for long. That’s why pharmaceutical, biotechnology, and medical device companies need to define their relevance in this new health care ecosystem, and soon.

Life sciences companies, for example, might consider staking a claim in the quest to lower hospital admissions and readmissions by working with providers on tailored plans for hospital patients being discharged, based on each individual patient’s propensity to comply with treatment and respond to it.

Consider: Carolinas HealthCare System (CHS), a network of hospitals with more than 900 care locations in North and South Carolina, recently lowered readmission rates by a third by using software from Predixion, a California-based software company. In this particular application, CHS gave its nurses point-of-care information on their patients so that when they were about to be discharged, the nurses could customize clinical interventions based on an individual patient’s predicted risk of readmission. This was a notable success — but what if we combined the insights and resources of a life science company and health provider that were both focused on, say, acute cardiovascular diseases? How much greater would the potential for lowering readmission rates be then?

The options are tangible. Imagine a scenario where a pharmaceutical company marketing a heart failure medication approaches its institutional customers — health systems, hospitals, urgent care centers, and other providers — with a risk-shared value proposition. The contract calls for the provider to use certain evidence-driven predictive analytics tools to define treatment options and possible responses. When predefined treatment goals are attained, both parties contractually benefit. Risk and outcomes, as the scenario plays out, will be managed by predictive models, which are powered by machine learning algorithms that will improve their accuracy rates over time.

Consider another scenario focused on compelling patients to stay the course with treatment. One estimate of the annual cost of medication noncompliance in the United States pegs it at a hefty $289 billion. What if a pharmaceutical company took the lead in conceptualizing and executing a collaborative solution, using predictive analytics to assemble and deliver a package of product and service offerings to motivate patients to stay on track? Think of it as a 360-degree/24-7 support system. It’s not hard to envision; we’re seeing just these sorts of systems popping up — informally, and disconnected from health care providers — with wearable fitness devices that share information among groups of users. With a focus on adhering to treatment, patients, providers, risk bearers, and life science companies would all be beneficiaries.

Medical device companies have begun using predictive analytics and other big data technologies in certain areas of their businesses. For example, consider Minneapolis-based medical device company, Medtronic, which develops diagnostic and intervention devices for cardiac and vascular diseases, diabetes, and neurological and musculoskeletal conditions. Medtronic is using big data and advanced analytics to drive their approach to patient and physician support and manage supply chains.

We’re also seeing some leading pharmaceutical companies — Merck Global Health Innovation, for one — investing and establishing operations to capitalize on new investments in advanced analytics. But this shift shouldn’t just be about capabilities. Life science company executives need to be thinking about what business they want to be in. They need to be thinking about how — and how much — they will develop and integrate predictive analytics capabilities into their services. They need to consider offering services enabled by advanced predictive analytics. And they need to consider business models where partnering to integrate the care that patients receive outside of the walls of provider entities is central to their value proposition.

The variety, velocity, and volume of health care data are allowing predictive analytics solutions to emerge quickly. And while the most visible immediate benefit is cost reduction, the real motivation is a patient-centric business model — one that recognizes that health and care management needs to occur wherever the patient is, not just in hospitals or physician offices. The goal (and it’s within our grasp) is threefold: improve clinical outcomes, enhance patient satisfaction, and drive more value to the entire health care system

By: Jeff Elton, Managing Director, Accenture Strategy and Global Lead of Predictive Health Intelligence & Arda Ural, Senior Manager, Accenture Strategy and Predictive Health Intelligence
Originally published at www.blogs.hbr.org

Leave a Reply