Healthcare providers are just beginning to figure out how big data, mobile platforms, and integrated software can deliver better care at lower costs, according to speakers at The Economist’s Health Care Forum in Boston.
Talk of using large-scale data analysis to develop customized treatment plans is premature since most healthcare providers are still edging toward joining the big-data movement, said Charlie Schick, IBM’s director of big data, healthcare and life sciences, during a panel discussion at the Tuesday event.
“The reality is hospitals are early on in analysis maturity,” he said. “They’re trying to answer questions required by the government. Big data is a buzzword.”
To receive government reimbursements to defray the cost of electronic health records (EHR) implementations, hospitals must be able to show how they are using IT to improve patient care.
The Cleveland Clinic, a healthcare provider that records 5.1 million patient visits annually, has been aggregating patient, financial and payer data “for a while” and uses that information to deliver better care and reduce costs, said CIO Martin Harris. Physicians are learning how to use the data to determine what care to provide, which is also a metric used in their annual reviews, he said.
The organization is looking ahead to predictive analytics to manage costs. For instance, patient data shows that people cancel surgical procedures during snowstorms. Using this information, the hospital could expect cancelations during inclement weather and better manage operating rooms, which are one of a hospital’s largest cost centers, Harris said.
To that point, just gathering data doesn’t help medicine, said Iya Khalil, co-founder and executive vice president of GNS Healthcare, a Cambridge, Massachusetts, company that analyzes healthcare data to better match patients and treatments.
“To get value, you need to make sense of the data, not only to predict disease trajectory, but how to intervene to get care and make it cost effective,” said Khalil, who added that her company developed an algorithm for health insurance company Aetna that helps identify people at risk for developing metabolic syndrome.
Gathering this data presents its own set of challenges. Managing chronic illnesses—such as asthma—which contribute to between 60 percent and 70 percent of healthcare costs, requires data from a patient’s personal life, said Harris. Patients may be reluctant to capture data from their homes, especially if it requires using a cumbersome device.
“We need to integrate medicine in that setting rather than a physician’s office. The proper device won’t be a medical device, but will be something like a phone,” he said. “They need to see incentives. We need to get them actively engaged in health care.”
Medical workers are also unsure of the potential value in collecting healthcare data.
“People collecting the data need to see value in the data. That’s not happening,” said Schick.
“Physicians should be knowledge workers in the 21st century,” said Harris. “We need data inputted by people who are closest to it.”
In another panel discussion, August Calhoun, vice president of Dell Healthcare and Life Sciences Services, noted that the top reason doctors and nurses say they avoid using technology is because doing so gets in the way of offering care. That’s because hospitals often lack a solid health IT foundation, he added.
“Most hospitals don’t have connected clinical systems,” he said. Instead, data islands exist, so transferring information between systems requires a USB stick and a walk down the hall.
John Halamka, CIO of the Beth Israel Deaconess Medical Center in Boston, warned that financial peril awaits hospitals that don’t use data to change their business model to one that is based on quality of care instead of quantity of tests and procedures.
“My fee-for-service business is dead,” he said. “The days of going to the doctor when you are sick are gone.” The future, he said, is in gathering data from many sources and relying on a team of caregivers to leverage that data to provide care before health problems arise.
Halamka’s future also calls for EHRs with better usability and interoperability.
“We have EHRs designed by people who don’t understand the industry,” he said, which contributes to workflows that aren’t friendly to how doctors work and that, for instance, complicate simple procedures like ordering aspirin.
EHR vendors lack incentives to make their systems work with outside applications. Most have closed APIs (application programming interfaces), making the barriers very high for companies that want to develop applications that work with software from major vendors.
“It will take legislative action to open these legacy systems up,” said Halamka. He likened the current state of EHR interoperability to Apple only allowing applications it developed to work on iPhones.
Despite the walled-garden nature of EHRs, hospitals are inclined to keep existing systems given the substantial sums they invested in the software.
“Ripping that out isn’t possible,” Calhoun said.
While studies have found that EHR adoption remains sluggish, Halamka said that during the past four years it anecdotally seems to him that EHR usage has doubled in the U.S. Adoption hasn’t been even across the country, but progress is being made, he added.
Just trading paper records for their digital counterparts can save hospitals money before quality of care is discussed, Calhoun said. For instance, 30 percent of tests that are ordered are duplications because the tests have already been done. EHRs can identify such duplication and cut wasteful spending.
However, health IT won’t replace every piece of paper.
“Paperless hospitals are as likely as paperless bathrooms,” Halamka said.