Skip to main content

Hospital Data Mining Hits Paydirt

 |  By kminich-pourshadi@healthleadersmedia.com  
   November 29, 2010

Perhaps these days the mantra for hospital financial leaders should be based on a familiar seven dwarfs tune used in the Disney classic Snow White: “We dig, dig, dig and we dig, dig, dig, all the day long through. In a mine, where a million diamonds shine.”

How true it is; after all, you’ve likely taken a pick-axe to your budget more than a few times in search of savings, but also in the hopes of finding a hidden gem or two that might offer your facility a glimmer of profitable possibility.

It’s the latter that is infinitely challenging to CFOs; perhaps that’s because in order to uncover the diamonds you need to do more than randomly swing that pick-axe—you need a tool that gives you the precise location of the diamonds among the rocks.

The solution? High-ho, high-ho it’s off to data mine we go!

Data mining, or predictive analytics, can overhaul the conventional hospital billing process, plus offer administrators a much-sought after link between the clinical and the financial. These tools can be used to gather the data needed to support high quality care for a better value while improving the charge recovery process, increasing staff productivity and better controlling costs.

 Without a doubt, there’s money in the data details, though it’s important to realize that not every data mining effort is going to yield billions. Nevertheless, this technology often reveals a few million in unnecessary losses—and these days that can make the difference between red and black for hospital bottom lines.

One area that can benefit from data mining automation is charge recovery; just ask Jason L. Adams, FACHE, vice president of revenue cycle for MultiCare Health System. MultiCare is an integrated health organization made up of four hospitals (Allenmore Hospital, Good Samaritan Hospital, Mary Bridge Children's Hospital and Tacoma General Hospital), numerous primary care and urgent care clinics, multi-specialty centers, and hospice and home health services. This not-for-profit system based in Tacoma, WA has grown steadily over the years and today is the area's largest provider of healthcare services in Southwestern Washington.

Most CFOs know that they and their teams are more than likely missing a few million due to  revenue leaks each year, but determining where to put the plug is a hit or miss process unless you have the necessary data to pinpoint the problem—such was the case at MultiCare. Their steady growth, while positive, was also opening the system up to more opportunities for revenue leaks.

“We were using another tool that was rules-based,” explains Adams. “And it would dump our paid claims into the system and then look at our potential lost charges. So we thought we were doing a pretty good job of tracking this type of problem.”

Still, Multicare wanted to begin to do some predictive analytics modeling, so in early 2010 they decided to invest in a technology by Apollo Data Technologies, a Chicago-based company. Adams explains that the software helped detect anomalies and trends in patient, billing, and clinical data. This data began to show areas where MultiCare was losing revenue as well as ways they could improve their financial performance, and prioritize resources.

The program utilized a similar approach to how retail businesses track their customers charge habits, Adams explains. The program uses advanced statistics to flag errant accounts and claims with a greater degree of accuracy.

“At first I didn’t think we’d find anything [in lost revenue due to charge capture]. We went live in June and three months later we had already realized $1 million in missed charges.”

The analytics uncovered was that nearly 1% of their $1.6 billion annual net revenue was leaking out of their bottom line without anyone realizing it. With the software in place, in less than three months they were able to identify over $1 million in missed charges and by the end of the year they anticipate capturing over $2 million.

Adams speculates that the $2 million figure may have been higher still, but the data actually pointed them in the direction they needed to go in order to nip the problem at the root. Adams and his team were able to see which clinicians were missing opportunities to code additional procedures that had been performed.

For instance, the system flagged specific diagnosis which usually have lab tests associated with them if the lab test codes were missing. In doing so, they were able to capture all the charges associated with a diagnosis and then alert clinicians to be aware of their mistakes.

Plus, Adams and his team can spend less time resolving missing charges, which helps them more efficiently allocate staff resources. Since the charge capture process is no longer as time-intensive, Adams can also allocate clinical staff to higher-value activities. Moreover, the data mining tool is maximizing MultiCare's investment in their EMR because the system links the patient's record with financial information to help the system determine cost/benefit around specific care and service lines.

“This program has helped us create more lean practices and eliminate redundancy for charges,” Adams says. “We’ll always need humans for healthcare delivery, so you’re always going to have variations in the charging practices of these people...With the data mining and predictive modeling we can identify those trends and get to the root cause analysis and make the necessary adjustments quickly.”

Karen Minich-Pourshadi is a Senior Editor with HealthLeaders Media.
Twitter

Tagged Under:


Get the latest on healthcare leadership in your inbox.