Many of the promises of Big Data are being felt in the healthcare profession as real-time processing and data analytics is allowing for faster and more comprehensive decision-making and actions on the part of the medical field.
The field is slowly maturing as industry-specific Big Data software and consulting services come to market, but there is still a long way to go before the market can be considered mature. Some things, like patient privacy issues, are still evolving but progress toward a solution is being made.
One aspect of the impact of Big Data is coming courtesy of a major change in how health care is done, and it pre-dates Big Data by a few years. The push toward Electronic Medical Records (EMR), which began around the turn of the century, is really starting to pay off in ways not expected at the time.
“That meant the amount of data grows exponentially. Now we’re at a state with a lot of data in the system. We’re able to make better decisions and make near real-time activities, whether it’s clinical or revenue, where in before could not do it,” said David Chou, CIO and CDO of Children's Mercy Hospital in Kansas City, Missouri.
When the market crashed in 2008, there was a big stimulus program with lots of incentives for hospitals to be electronic, said Chou. Health care it was on the receiving end of it and because of that, it brought in much more data. “Now we’re able to utilize the data to make better decision,” he said.
Healthcare Big Data Analytics
Analytics is being used in health care to help predict outcomes of actions taken by physicians. For example, what would be the outcome of a surgery for a condition, based on patient data points like age, relative health, existing conditions, and so forth.
“Big Data is helping us move toward precision medicine, trying to predict the outcome. By predicting outcomes we can get ahead of the curve. We can predict a need for major surgery in a few months based on a test. Test results may predict a symptom or outcome so you can get ahead of it to prevent that outcome or heal quicker or not get into needing surgery,” said Chou.
“One of the biggest things I’ve seen change is the power of being in the ballpark. It’s more of the idea of approximation. We don’t necessarily have to wait for the normal timeline of events to occur to be sure it’s x, y, or z. We can tell well in advance of traditional indicators in health care. That’s only possible in Big Data,” said Damian Mingle, Chief Data Scientist with Intermedix, a provider of analytics tools to healthcare providers.
In leveraging Big Data, you would start to know at the point of admission to the hospital. When you can add variables like demographic and historical data to make small demographic data large and crunch that using machine learning, that could remove hours of time for starting treatment, said Mingle.
In some cases, there can be as many as 2,200 variables around local and national population health statistics that are compared to a presenting patient’s initial symptoms. A patient who has been to a region with disease infections might be flagged, or a patient presenting with symptoms common in one area might be flagged for potential testing.
“The Big Data and machine learning technologies enable the clinical setting to be aware of potential risk factors that they would not normally have seen,” said Justin Schaper, senior vice president of analytics at Intermedix. “Say at a broader level, community or facility level, using what the machine has learned on the large scale of hundreds of thousands of patients coming through the door that this patient has a higher probability of of a condition. Before it might not be discovered until well into the lab work.”
Big Data Benefits for Healthcare
The market for Big Data health care software is enjoying a boom, to the point that the industry has a wealth of choices in front of it. One health system that had so many vendors approaching them that there was chaos, according to Alex Coren, chief innovation officer and co-founder of Wambi, a health care and recognition platform.
“Patient engagement analytics are a hot topic but aren't for everyone. They're for specialists in the data arena. So it seems like the data is getting into hands but lacking in widespread use and being conveyed in a meaningful way,” she said. In addition to patient engagement software, she is also seeing some “impressive” scheduling systems for ensuring continuity of care.
Another bit of good news for doctors is that the tools around Big Data are becoming more sophisticated and it’s no longer necessary for a data scientist to hand code very sophisticated models that take months to put together, said Schaper.
“We have used third-party tools to make multiple models in a short period time,” he said. “The ability to be productive has increased and hardware costs have come down.”
Big Data and Healthcare Considerations
The biggest challenge facing Big Data in health care is not data or software or data scientists, but getting doctors to enter their documentation. If a physician does not document notes in real time after seeing patient then you won’t get the information on the patient in real time.
“They hate data entry,” said Chou. “A lot of doctors hate the fact that there is an EMR now. They could write a lot faster. And they are becoming data entry clerks because of this mandate. That’s the case with people not trained properly. The lack of training creates additional burden. That is a big dissatisfier for a lot of physicians.”
He added that the younger generation of doctors who grew up with tech are better about it. “You see a big divide between people who use a system well and those who do not. A lot of that is generational but it’s not 100% of the case,” said Chou.
Schaper said you’d be surprised how many doctors dabble in Big Data. “We actually do have a few client physicians very engaged in this kind of work,” he said.
At the hospital level, most hospitals don’t have big teams of data scientists, said Schaper. Oftentimes, it’s just a Business Intelligence analyst who made a move into Big Data or a software engineer who takes up the role. Not surprising, larger healthcare systems that have very sophisticated data science and analytics departments, he added.
The Future of Big Data in Healthcare
The cloud is the wave of the future for most industries, but Health Care has been slower to embrace the cloud due to strict privacy compliance laws like HIPAA. Mingle said there are some health care-specific vendors like Tigertext, but that Amazon is the go to vendor thanks to Amazon Shield, which securely transmits data back and forth between its data centers and medical offices.
Mingle believes we are going to see a change from quantitative data to qualitative data. As it matures people will be more happy with qualitative data. Instead of numeric values you get categorical stuff, such as moving patients into a sub cluster. So instead of national trend lines, you see more precision clustering of patients.
There will also be more social listening in health care data. “If we notice information on Twitter or Facebook that might have a correlation with what we see in patient volume spikes, we can act on it. More health care will reach outside hospital walls to include that data,” he said.
Coren said the medical profession is moving to a value-based model where hospitals are paid based on how healthy the patient is, rather than how much treatment they need. “The old system was the more you treated the patient, the more money you made. The value system means the healthier your patients are, the more money you make. So it’s important to get analytics to make data driven decisions more than ever,” she said.