How Integris Health uses data analytics to improve collections

The economy may or may not be on the mend. Healthcare and payment reforms could be a boon or a bust for the healthcare industry's bottom line. Cost containment efforts may or may not save healthcare. But it doesn't really matter whether a hospital or health system is flush with cash or struggling to survive--patient collections has and always will be both a priority and a challenge.

That's certainly true at Integris Health, an integrated system with four hospitals in the Oklahoma City area.

Integris and other organizations are using data analytics and predictive analytics as financial tools. In this exclusive interview, Brent Grimes (right), administrative director of patient account services, tells FierceHealthFinance how Integris is using data analytics to weather shrinking reimbursements and inform its collections strategy.

FierceHealthFinance: What role does technology play in cost control?

Grimes: With reimbursement dropping, technology is the huge driver for us. And to keep and gain market share, you have to be unique in the marketplace. We're working on how to do that. We're going through a change now. We were inpatient focused. That's how we made money--we're a hospital system. Now we're becoming disease-management specialists. There's increased cost with trying to make that move. We're getting paid less. We drive technology to keep costs down and provide a quality outcome.

FHF: How do you use data analytics to enhance collections?

Grimes: We segment collections based on the balance the patient owes and create a probability score using data analytics. We create precise segments based on historical data. We have a 97 percent accuracy rate in predicting whether or not a patient is going to pay.

Based on that, we cater the experience. Do we need an automated dialer campaign? Do I need human intervention with this patient?

I can even customize our collection letters. We have a very nice soft pastel colored paper--pinkish blue. That first encounter, when I'm establishing the balance they owe, I don't want a credit card statement going out. I want a very soft letter. As the account ages, the color changes and become stronger and the language changes to let them know it's getting serious.

FHF: What data points do you use for segmentation?

Grimes: The two most important are past history and the amount of the bill. Even if you have a high credit score, if you owe me $20,000, there's a small likelihood you're going to pay. Conversely, you may have a very low credit score and only owe me $50. There's a chance we can work that out.

We look at life events, too. We continually score and look at those trigger points. And segmentation moves with the economy. Every time a patient comes in the door we talk about patient balances. If the low scores are charity-eligible, we talk about financial assistance.

I'm not just writing it off based on a score--I'm giving them an opportunity to pay it off and work with them if they can't.

FHF: What kind of results have you achieved from data analytics and segmentation?

Grimes: My high end scores have a 97 percent collection rate. All the way down to my segment one, which is the lowest probability of payment, we get 7 percent.

My line changes. My low opportunity has grown--more people get a poverty level score--versus my high end, which has shrunk. We continually monitor that. It's an ebb and flow. As economy and people and healthcare mandates change the responsibility of patients is going up. That's a given in our future--and not to my favor.

Oklahoma is fortunate because we have a large energy market sector and our housing market never hit the ceiling. But 24 percent of our population is uninsured. We have to be good stewards.