Healthcare professionals are increasingly turning to data analytics to support their services; however, inefficient data technology may be holding the industry back.
In a webinar sponsored by MarkLogic and Intel, and hosted by Becker's Hospital Review, MarkLogic Healthcare IT and Payments Executive Bill Gaynor, Owensboro (Ky.) Health CMIO David Danhauer, MD, Senior Vice President at GE Healthcare Camden Group David A. DiLoreto, MD, and Marcee Chmait, who leads innovation at SCAN Health Plan, discussed how both providers and payers can leverage data analytics for improved outcomes.
"Healthcare organizations are having some data challenges," Mr. Gaynor said, noting how legacy software can impede efficiency, security and cost savings. "Business as usual is not gaining the results in enough time. Innovative programs and digital enterprise initiatives require modern and agile data technology."
Healthcare organizations are often focused on bringing together data from a disparate array of sources, many of which present their data in different formats. Today, data scientists spend up to 80 percent of their time preparing, rather than analyzing, data, according to a New York Times article, in part because they lack software to automate the organizing of their varied datasets.
"We're trying to approach 21st century opportunities and challenges with 20th century back-end technology," Dr. Diloreto said.
Enterprise databases can help organizations with this data integration by transforming and delivering data related to a range of sources — such as patients, physicians or products. Although developing an enterprise data strategy can seem daunting, Dr. Danhauer stressed the role it will play in keeping healthcare organizations competitive.
"Data is king," he said. "Those who have the data are going to survive; if you don't have the data, you're not going to survive."
But just having data will not help healthcare organizations make better business decisions. Companies need to be able to leverage all of their data to provide actionable insights, whether combing through EMRs to suggest next steps in patient care or connecting consumers with convenient resources to improve their overall health.
How an enterprise data strategy can help providers
Dr. Danhauer led the successful replacement of Owensboro Health's legacy technology in an effort to alleviate efficiency and quality setbacks associated with the original EMR documentation system.
He sought to identify a new system that would not only improve quality of notes and maintain physician workflow, but would also transform unstructured data, such as voice transcription, into discrete data.
"We wanted some kind of new technology to streamline that process," he said. "To not force our physicians to be data entry clerks, as they've been complaining about over the years."
He ultimately decided to implement an enterprise EMR. This system utilizes computer-assisted physician documentation, which reviews information from a physician note and offers suggestions to the clinician, in an effort to collect more actionable information.
"If they're using our EMR, as they're putting it in, they're now getting cues live in front of them," Dr. Danhauer said. "If they put a voice file in that says 'CHF' [congestive heart failure], they're getting cues right in front of them that say: 'Is that right sided? Left sided? Acute? Chronic?' They're getting those cues to say — please give us more detailed information."
The majority (53 percent) of physicians respond to these suggested documentation improvements, according to Dr. Danhauer, and Owensboro Health has used these insights to develop patient safety programs targeting abdominal aortic aneurysms and lung nodules. The system scans physician notes to identify information related to these two conditions, in order to provide patients with appropriate care as soon as possible.
Although physicians often balk when asked to adjust to a new technology system, Dr. Danhauer found 70 percent of physicians responded favorably toward use of the system. In addition, when he asked providers to review other providers' notes, he saw 83 percent responded favorably to the quality of documentation.
"That quality went through the roof," he said, noting that most of the CAPD suggestions are related to ICD-10. "Our queries out to physicians have significantly decreased for ICD-10. With all of the anticipated backlog we saw with the predictions with ICD-10, we have been current now with coding for the past 6 months."
How an enterprise data strategy can help payers
The healthcare space can be difficult for consumers to navigate. From the payer perspective, data-driven applications can help by addressing patient needs and increasing consumers' access to health and wellness resources with risk-adjusted patient insights.
As an example of a data-driven initiative, Ms. Chmait spoke to how she led the implementation of a consumer-focused care solution at SCAN Health Plan.
The Long Beach, Calif.-based nonprofit health plan connects seniors with a variety of services, from transportation to home care. To personalize this process, SCAN Health Plan has applied graph technology to connect the consumer to his or her healthcare needs.
"The technology that we have decided to use in our human-centered project is graph technology," Ms. Chmait said. This solution — used by companies like Facebook, LinkedIn and Walmart — puts the user at the center of the graph, which then "learns and predicts the needs of the user" through machine learning.
Families and enrollees using SCAN Health Plan can use an app to describe their needs and monitor their experiences. The app, using graph technology, then constructs a user profile to share with primary care physicians. The graph can also link a family's needs to relevant content, such as medical history or related research.
The machine learning algorithm will then use this information, along with the user's location, to determine "best fit" services in his or her area. "Our independent living graph will determine preferences and understand somebody's specific needs," Ms. Chmait said. "It will make recommendations and fulfill services at that very specific need level."
Through replacing legacy technology with NoSQL data platform and advanced analytics, health systems like Owensboro Health are able to improve physician satisfaction and patient care, while healthcare payers like SCAN Health Plan connect consumers to their localized environments. Moving into 2017, an enterprise data strategy will continue to help providers and payers stay competitive, as data insights and personalized care become increasingly important to the healthcare space.
Listen to the webinar recording here. View the webinar slides here.