UNC Health Care leverages big data to boost bottom lines

It's always to refreshing to see big data not just being thrown around as a buzzword, but truly being used to save lives and improve bottom lines. That's the case at the University of North Carolina Health Care (UNCHC), a large non-profit healthcare provider in Chapel Hill, N.C., where one doctor is touting data and analytics as "increasingly at the heart of" how his hospitals run.

Growth and consolidation in the UNCHC system saw a massive increase in the amount of data each facility was holding--and about 80 percent of it was unstructured, said Carlton Moore, M.D., associate professor of medicine at UNCHC, in an article in Business Cloud News. Data, he said, now is being used to improve the quality of care and reporting.

"A big part of the accountable care component is stuff like cancer screening. We capture that mostly in a structured format, a checkbox, but a lot of times that doesn't happen--often what happens is patients get cancer screenings at other institutions, and then come to us," Moore said. "We write in the unstructured notes that they've had a screening at another institution and leave the checkbox blank because we aren't providing the service ourselves, so that doesn't get captured by the structured data."

An example of how UNCHC has merited results from big data: They developed a unique algorithm inserted into natural language processing software platform, allowing them to identify, extract and report on abnormal results from pap smears and mammography screenings.

"After reviewing the previous year's mammogram, the radiologist might decide that the small abnormality seen on the current mammogram is unchanged from the previous year and reclassify the mammogram result as benign," Moore said. "By using natural language processing to identify eligible patients receiving screening, we were able to increase our cancer screening numbers by about 10 percent."

And this is just the first case being tested by UNCHC. Moore told Business Cloud News that UNCHC plans to do much more--like analyze medicine compliance and number of clinic visits to predict bad health outcomes.

"In order to provide good care nowadays, you need good data," he said. "You need very good analytics. And that's a huge change from how things were ten or fifteen years ago, or what was even possible then."

In November, CIO highlighted several real-life cases of healthcare organizations that are using big data analytics to improve outcomes and reduce costs in a slideshow. It included examples from Boston-based Partners HealthCare system, which uses big data to connect its financial, operational and clinical analytics systems; Intermountain Healthcare, which mines more than 90 million patient EHRs for outcomes analysis; and Ohio State University's Wexner Medical Center, which uses text message reminder to reduce missed appointments.

So far, though, health payers are more invested in the power of big data and analytics tools than providers, according a recent report from IDC Health Insights.

To learn more:
- read the Business Cloud News article

Related Articles:
3 ways healthcare orgs use big data
Lack of skills, commitment waste power of big data analytics
Big data use: Providers, payers not on the same page
Hurdles to big data use in healthcare more social than technical
Meaningful Use Stage 3 may require social, behavioral data capture