Case Study: Transforming Uncompensated Care Via Data Analytics, Part 2

Bad debt cripples healthcare providers, and ever-increasing regulation and governmental scrutiny only serves to exacerbate the problem. But for some, it creates an opportunity to use new technology in innovative ways to transform a constraint into an benefit, not only to patients but also to the provider bottom line.

For James Logsdon, vice president of revenue cycle operations at Texas Health Resources, a chain of 15 hospitals in the Dallas/Fort Worth area, the new regulations and reporting as required under the Patient Protection and Affordable Care Act legislation, rather than prove to be onerous became an opportunity to leverage their data analytic capabilities to create a repeatable business process that “will save money in the revenue cycle, improve patient relations and create a better patient experience.”

In this three-part article, drawn from Logsdon’s presentation at the HFMA Leadership Conference, he describes how his organization employed data analytics to not only fulfill IRS reporting requirements, but also reduce the cost of charity care processing, increase self-pay collections, and shrink the cost-to-collect ratio of Texas Health’s revenue cycle.

(For Part 1, click here)

Part 2: Introducing Presumptive Charity Analytics

“You’re going to write off patient accounts before you even send them a bill?” According to Logsdon, that was a typical CFO reaction to his team’s proposal to introduce data analytics into the charity care processing workflow.

His answer? “Absolutely not,” he says. But by employing data analytics into the revenue cycle, his team can reduce the cost of charity care processing. But he always sends a bill. “I’m going to spend forty-five cents, and then I’m going to do the data analytics,” he says.

The Texas Health team began exploring data analytics two years ago. “We started very, very very conservatively,” Logsdon says, testing the new technology with uninsured patients who visited the emergency room. Texas Health’s payment-to-charge ratio system-wide is 5 percent, Logsdon says; but in the emergency room it plunges to 1.2 percent. “I have a ninety-eight point eight percent chance of being right” that the hospital will collect nothing from the uninsured emergency room patient, Logsdon says.

A percentage of the 98.8 will qualify for charity care; the rest will not, but also will not pay the bill and slide into bad debt. What Logsdon’s team wanted to do is identify those who qualify for charity care but for one reason or another fall between the cracks, for example those who fail to fill out charity care application. He also wanted to do it programmatically and automatically, without requiring human intervention.

For that his team used what is popularly called “presumptive charity analytics,” which enables healthcare providers to use publicly traded data to make charity care determinations “of those who can’t pay or won’t pay.”

Texas Health works with a data analytics vendor. “You can send your patient accounts to a data analytics company and get publicly traded data back, and then use that data to make sound business decisions about one’s ability to pay,” Logsdon says.

Logsdon’s vendor runs Texas Health’s accounts through their scrubbers and returns up to 20 data elements for each account, including:

• Estimated federal poverty limit score;
• Credit score;
• Estimated propensity to pay;
• Estimated propensity to pay medical bills;
• Family size;
• Estimated income.

Of the latter, Logsdon said that he has found that his vendor has been exceptionally accurate.

Logsdon sends his analytics vendor a batch file of patient accounts on Day 4 after the patient’s statement has been created. That’s two days after he has sent a bill. Thirty-six hours later the vendor returns the patient account data with the corresponding 20 data points.

Data for each patient is programmatically compared against the criteria of Texas Health’s charity care policy. “This is important: a patient must meet our charity care guidelines to qualify,” Logsdon says. But even those who are found to qualify are not automatically recorded as charity care. “I let them age a bit,” Logsdon says, until just before the second bill goes out, at which time a charity care determination is made.

The Texas Health team created a separate write-off code for those accounts that have undergone the automated determination process so that the healthcare provider can track the results of the program versus the traditional charity care workflow.

Once the determination is finalized, the system automatically generates what is now called a “charity care determination” form, Logsdon says. “We take all those analytics and spray it on the form.”

The system is completely automated. “I don’t want to have anybody fill it out, I don’t want to have anybody touch it, I don’t want anybody to print it, and I want it stored electronically in my patient accounting system at the account level so anytime we get audited as to why we gave this account charity, we can go into the patient accounting system and say ‘Here it is.’”

According to Logsdon, “Not one single human touched the form, filled the form out, even saw the form.”

Part 1: The Charity Care Challenge

Part 3: The Impact of Data Analytics on Charity Care