For more on healthcare applications of predictive analytics, attend Predictive Analytics World for Healthcare, September 27 – October 1, 2015
But new area of business intelligence does not come without challenges.
Hospitals and health systems can be quite good at leveraging business intelligence tools to better understand their organizations, from analyzing the demographics of the patients they serve to assessing breakdowns of operational expenditures.
But most organizations haven’t yet taken the next step: to predictive analytics.
The CFO needs to drill down and find out what the vendor means by ‘predictive analytics.
Using information technology to discover patterns in large and disparate datasets that point to what the future is likely to bring – and then identifying actions to take – can help health systems improve patient care and stay ahead of issues that could impact their financial status.
Baptist Health in Montgomery, Alabama, is among the early adopters of predictive analytics. A full-service health system comprised of acute care hospitals, a psychiatric hospital, cancer, surgical and imaging centers, employed-physician practices and disease management programs, Baptist is using analytics to drive better patient outcomes, reduce financial risk and improve the revenue cycle – particularly around denials management.
The predictive algorithms in Baptist’s analytics technology makes it possible to alert medical teams to a specific patient’s likelihood of experiencing complications from an illness, acquiring an infection, or experiencing a recurrence or extension of a problem that could lead to readmission.
“Doing more in real-time while the patient is still with the care team is imperative if we are going to learn to educate patients and address them holistically, instead of treating individual disease processes or incidents one at a time,” said Katrina Keefer Belt, Baptist Health’s chief financial officer.
The holistic approach matters as much to putting the patient on the path to continued wellness as it does to a health system’s ability to continue operating on a sound financial footing. The Affordable Care Act can now penalize hospitals for excessive 30-day readmissions for certain health conditions, even as it incentivizes them to keep patients healthy.
The fee-for-service environment under which hospitals have long operated has made it hard for many of them to yet grasp how important predictive analytics will be to “surviving in [this] reformed environment,” said Belt.
Baptist Health is poised to get even more value out of its predictive analytics tool, which is supplied by Suwanee, Georgia-based Jvion. Belt says the health system is evolving its use of the technology to deal with revenue cycle issues related to denials management. Baptist will use the technology to cull through data in search of clinical, technical or administrative error patterns that lead to claims rejections.
“We need to embed triggers in the medical record for caregivers upstream so that we don’t rework claims,” Belt said.
Currently, the quickest adopters of predictive analytics in the hospital sector are large academic and major medical centers, said Jeff Deal, vice president of operations at data science and predictive analytics consultancy Elder Research and [program chair of Predictive Analytics World for Healthcare].
A variety of factors hold other institutions back. Costs are among them, as “the development and application of an advanced analytics program can be an expensive and time-consuming endeavor,” he said. Large medical centers are likely to have more resources to apply to an analytics program than more modestly sized organizations. Deal said academic medical centers are moving into advanced analytics as a matter of mission. “They want and need to be leading on medical innovation.”
On the clinical side, obstacles extend to clinicians’ reservations about the technology’s impact on the practice of medicine. Analytics can encourage a physician to act against his intuition when the data demonstrates that a particular action or decision could result in a better outcome than the action the physician wants to take without that data-driven input, Deal said.
“It is hard for all humans to act against training, experience, and intuition. But it can make a difference when the decision is supported by the sound application of analytic science on good data,” he explained.
There may also be physician resentment if they feel they are being asked by the business office to incorporate advanced analytics into their decision making only to save money. “They need to see the impact on care delivery and health outcomes,” Deal said.
At Baptist Health, Belt said physicians find it a “scary prospect” to have access to information that they otherwise wouldn’t. “Hospitals need to be careful in how they present the information to the care team,” she said. She recommends having the chief medical officer or a well-respected physician help spearhead predictive analytics efforts. Deal agrees that key stakeholders should be engaged early in the process, and advocates for strong leadership from high levels of the organization.
A last piece of advice from both Belt and Deal is to make sure that the IT vendors, consultants or other parties brought on board to help implement predictive analytics really are experts in the technology. “Most solutions that say they are ‘predictive’ are actually after-the-fact reporting, not truly predictive,” cautioned Belt.
“Everyone wants to say they do it,” noted Deal, but in his experience he’s found that they’re not all shooting straight. If a vendor says they do predictive analytics, “the CFO needs to drill down and find out what the vendor means by ‘predictive analytics,’ and what actual experience that vendor has,” he said.
Jennifer Zaino, Healthcare Finance News
Originally published at www.healthcarefinancenews.com
 |
 |
 |
