Just because someone works in a hospital doesn't mean she knows how to navigate the healthcare maze, leverage it to take care of herself, or how to use a blood pressure cuff despite seeing one every day on the job.
Sometimes people forget that healthcare workers are patients too -- and that they also can benefit from predictive analytics in healthcare.
That's where Maiken Himmel comes in. As the employee nurse navigator for Baptist Health Floyd in New Albany, Indiana, she is essentially the case manager for Baptist's employees.
"I help them manage their health and wellness, provide a lot of personalized information, education, resources, utilization, and, bottom line, try to help them manage their health and save them money which, of course, saves the hospital money," she said.
Himmel is using Advanced Plan for Health's (APH) Poindexter risk engine for population health management, which includes data analytics and predictive analytics in healthcare, to help her do her job as nurse navigator; a job created by APH.
"We'll introduce nurse navigator programs where a nurse will start to directly engage the at-risk population within the health plan," Taylor Godbey, executive vice president and chief operating officer at APH, said. "The system itself is designed in a way that we can more efficiently get to actionable data so we can be proactive."
APH's technology is aimed at helping organizations which provide health plans to their workers identify where there is risk among their population.
Godbey said health plan expenditures are one of the biggest costs that any organization -- not just healthcare organizations -- has. "Having insight into what's driving cost is very important," he said.
Dealing with all that data
Taylor GodbeyCOO at Advanced Plan for Health
Godbey explained the technology will find key indicators such as financial, cost, quality, and clinical driver indicators and pull them into one report.
"The bottom line is you can't manage what you can't measure," Himmel said. "Everything that I'm able to measure helps me navigate, not only our employees, but their families as well through the healthcare maze."
Himmel added that APH is real-time, one-stop shopping.
"If I need to get on there any time and want to see, as far as high cost payment, how are we doing from a per member per month standard … did things change from the utilization [for example]," she said. "All of those things I can see."
APH allows Himmel to drill down into the data -- and saves her from having to move from report to report -- and manipulate the data so she can look at exactly what she needs to, whether that's in terms of insurance costs, looking at the diabetic population, who has been going to the ER and how frequently, whether there are any care gaps and more.
"I can put all that information into APH and create my own report … then share that with our senior teams which lets us know what's working, what's not working, what kind of opportunities we have [and] celebrate our successes," Himmel said.
It allows APH's clients to understand what is driving the cost and potentially improve the quality of care by assessing whether better management, better programs or different resources are needed, Godbey said.
In order for APH to effectively make these analyses, the minimum data it needs includes enrollment information, medical claims and pharmacy data. Godbey explained that APH is able to hook up to the EHR; however, it is not necessary in order for the technology to do its job.
APH's risk modeling: Predictive analytics in healthcare
The predictive risk modeling feature of APH's Poindexter allows Himmel to see whether ER visits have increased over the past couple of years and, if so, what can be done to change that.
"You have so many people that aren't managing their health with chronic conditions like they should. Then when something comes up, they go to the ER instead because they either don't have their primary care physician or they haven't seen him in five years," Himmel said. "They're going to the ER. Maybe they're getting admitted because they haven't managed their diabetes and blood pressure, so now they have heart disease."
APH's predictive risk modeling allows Himmel to see, for example, that Baptist has 24 members that have utilized a certain amount of money. The predictive analytics is able to tell her that in the next six months these risks will cost Baptist a certain amount of money.
"That way, I'm able to, hopefully, keep those moderate and low risk claims from becoming high-risk claims," she said.
APH also looks at phenotypes, an organism's observable characteristics or traits, for example, in order to predict risk.
"We're looking at individual characteristics and how those can apply or shift risk for any given disease or behavioral pattern or cost," Godbey said. "When you look at somebody [who is] diabetic, the phenotype using age, gender, location, really any demographic information that we can get our hands on through the enrollment data allows us to dynamically rate risk for that disease."
This allows APH to fine-tune who has the most risk and therefore should be at the top of the list.
"With any program, with any employer or disease management company -- population health, case management -- they all have a limited number of resources that they can apply to an initiative, meaning they might be only able to manage 1% or 5%," Godbey said. "What this system does is it allows them to engage the right 1% or the right 5% so that we're not taking shots in the dark."
While APH's technology helps organizations better manage their at-risk populations and therefore save money, Godbey said, "the real win, the real value in this is that you're able to help the people [who] need it the most."
Mine medical billing and claims data with analytics
Cognitive computing platform helps with medical codes
Improve OR utilization with predictive analytics