This past January, Forrester Research analyst Mike Gaultieri did some soothsaying. “My prediction: TIME magazine will name big data its 2013 person of the year,” he wrote.
We’ll see about that. But it’s undoubtedly true that big data and analytics – despite the challenges they pose – are starting to make their mark on healthcare. It’s not just the huge academic medical centers doing groundbreaking initiatives with genomics and pop health, either. Smaller hospitals are starting to show interest in what they can do with their own troves of clinical and financial information.
It’s been said that healthcare is perhaps 15 years behind other industries when it comes to making smart use of data analytics. That’s not very surprising: After all, it’s pretty hard to do much data mining on a basement full of dusty paper records.
But now that electronic health records are up and running in hospitals and physician practices nationwide, patient health data has begun flowing in earnest.
That’s to say nothing of an explosion of new consumer technologies: mobile medical devices, ever-more ubiquitous smartphones loaded up with health and wellness apps, even a new so-called “smart fork” that aims to help hungry people lose weight by logging data about how long it took to eat and the number of bites they take per minute.
It can be a lot to handle. The lucky thing is that just as all this data is coming at us – and more every day – just as the need for it is more pressing than ever. Big changes in care delivery and reimbursement models require lots of data to be studied and understood to accurately gauge the effectiveness and efficiency of patient care.
“If I’m at risk, I want to start doing a better job to make sure I have the data to manage it,” says John McDaniel, national practice leader, healthcare provider market at Sunnyvale, Calif.-based storage and data management firm NetApp.
A recent survey from NetApp polled healthcare decision-makers about their willingness and readiness to deal with big data.
Asked how important clinical analytics are to their organizations, 56 percent of respondents said it was of utmost importance; just 6 percent ranked it least important.
“Are you having a hard time assessing the future impact of Big Data?” asked another question. Sixty percent of respondents said they were.
So they know it’s important, but many providers seem unsure just how to approach the opportunities afforded by big data.
At a recent industry event in Atlanta, health IT professionals from area hospitals weighed in on the promises and pitfalls of big data.
One pointed out how the growth of unstructured data has been “explosive,” especially since 2005, and noted that it’s “almost impossible from an infrastructure perspective” to manage the deluge.
Another clinician mentioned the challenges of convincing clinical staff of the merits of capturing more data: explaining why the need is a challenge “on a continuing basis,” she said, since time is at such a premium.
Still another mentioned that her facility was migrating over to a new IT system, and so the challenge there was dealing with legacy data. How long to keep it? How to access it? How to report on it? New radiology modalities were adding another level of complication, she said.
But another hospital employee saw the necessity of having lots of data to swim in, however daunting it may be.
“What I’m finding is that there are a lot of hidden patterns,” he said. “Conclusions that can be drawn when we don’t know the questions yet to ask. Relational connections and statistical connections. In order to be able to find those patterns, we need the big collection of data, all at once. However you segment it and slice it, if you can’t span everything, then something is missing.”
So the interest is there. But the hurdles are many, especially for small providers. Data sets are getting so big and so complicated, so fast, that it’s getting harder to capture, store, search, share, visualize and analyze them using the technologies most hospitals have on hand.
The fact that 80 percent of data is unstructured is a sticking point. “It’s scary when you think about how much is in email form, text message form, video form,” said McDaniel. “It’s hard to do something with data unless you can figure out how to tag it.”
Privacy and security are similarly pressing. And, of course, storage will continue to be an ongoing challenge.
“Overall storage and archiving volume requirements for U.S. medical imaging data will cross the one exabyte mark by 2016,” wrote Frost & Sullivan analyst Nadim Daher, “which marks medical imaging’s definitive entry into Big Data territory.” (For a bit of perspective, it’s been estimated that every word ever spoken by humans in history would take up some five exabytes of data.)
“Part of the challenge is getting everybody on the same page,” says McDaniel. “When we talk about big data, we’re not just talking about images, which is huge, but we’re talking about genomic information, my entire health history, a lot of components that a physician or care provider is going to need for a lot of reasons.”
For now at least, not everyone is on the same page. Not yet. If healthcare is lagging other industries when it comes to making hay with big data, smaller hospitals have a way to go before they’re at the level of the UPMCs and Intermountains of the world.
One recent survey found that fully 100 percent of health organizations said they’re collecting and managing more information today than two years ago. The good news was that 87 percent of those organizations reported deploying software to help them leverage that information to make better decisions.
The bad news? Few were satisfied with their overall preparedness for all that new data: 40 percent of respondents graded themselves a D or an F.
But the opportunities available for smart use of data analytics are obvious. One study showed that $300 billion could be saved annually by sniffing out administrative inefficiencies and improving care coordination.
But no one expects small hospitals to move mountains with the insights gleaned from their data just yet. Low-hanging fruit will be OK for now.
“My advice is to start with a finite, measurable plan and to constantly re-evaluate,” says Mical DeBrow, practice director for outcomes at Siemens Medical Solutions. “Start with a small project that demonstrates effectiveness of change.” n
