Does big data need to be downsized?

Considering the value of personalizing big data down to the patient level
By Michelle Ronan Noteboom
09:40 AM
Small fish big fish

Twice a year my city offers "bulk trash collection" for all those treasures that are neither worthy of donation to Goodwill, nor small enough to fit into a standard trash can. I welcomed the most recent bulk collection as it allowed me to drag to the curb a couple of humongous televisions (circa late 1990s) and massive stereo speakers (probably older than some Healthcare IT News readers).

There used to be a certain prestige when it came to big things: big cars, big computers, and big hair were often a source of envy. And by gosh, who didn't want a Big Wheel?

Today, we covet the small and personal. We love our ultra-thin laptops and TVs. We covet wireless mini-speakers that are both portable and powerful. We are never far from our sleek smart phones that have replaced briefcases filled with bag phones, pagers, cameras, organizers, and luggable computers.  

Yet for all our new-found love for all things small, big data is hot – as it should be, given its potential to help providers improve patient outcomes and cut costs. According to a recent Technavio report, global big data spending in healthcare is predicted to grow at a CAGR of about 42 percent from 2015 to 2019 as providers seek to derive insights from the vast amount of data generated by EHRs, bio-metric devices, social media, and other sources.

However, if you are a consumer or even a physician you probably don't think too often about it, at least in terms of your health. You are likely more focused on a much smaller subset of information that directly impacts your life or your patients' lives and give little regard to the nebulous world of big data.

Thus the challenge is to find ways to downsize big data and make it relevant and of value to individual patients and their physicians.

Consumers today have more opportunities than ever to monitor their own health using wearables and smartphone apps that track exercise, sleep, diet, and other health behaviors. Unfortunately, about one third of users lose interest and abandon their trackers within six months. Meanwhile, physicians are often reluctant to download patient-collected data because they lack tools to disseminate and analyze the information.

Would the value of these bio-metric devices be enhanced if we could leverage big data and provide actionable insights at the patient level? Would more consumers be motivated – and remain committed – to actively monitoring their health? Would it improve physician buy-in?

Consider the potential value of personalizing big data down to the patient level. Ideally we'll one day have a tracker that auto-magically captures all aspects of the wearer's daily diet and exercise habits, monitors vitals, and provides predictions for such things as weight and cholesterol levels over the next 30 days. In other words, the device would not only capture data, but also consider big data insights and provide personalized information that might motivate the user to remain committed to healthy choices.

On the provider side, physicians would be equipped with enhanced tools for disseminating data from individual patients and gleaning insights based on big data analysis. Providers would arguably be more interested in patient-captured data if the information was more than just data points and could aid in the care process.

A few companies are already working on such solutions, most notably IBM Watson Health, which is leveraging cognitive computing capabilities to gain insight from personal health data.

A number of startups are also making inroads, including Cognitive Scale, which offers a solution that delivers patients and caregivers personalized and contextualized insights based on patient behaviors, and Sentrian, which has developed a remote biosensor and machine learning platform for the early detection of patient deterioration.

The problem with big data today is that for the most part it is too big to impact the health of the individual patient. I'm not suggesting we toss big data to the curb, but I do think we need to step up innovation and advance solutions that focus not just on population health but also on the health of one. While we are at it, let's tie some of those enhancements to user-friendly biometric devices in hopes of motivating more patients to remain engaged in healthy behaviors.