Using DMAIC to Reduce Hospital Insurance Claim Denials and Write-Offs

by bmgi_manager on 12th September 2006

The Challenge

Hospitals lose millions of dollars a year on insurance claims that don't get paid for a variety of reasons. Sometimes the patient's policy is expired, and other times the insurance company has incorrect information about the patient. Sometimes computer hardware or software malfunctions, causing claims to become lost. In still other cases, patients or physician offices provide incorrect information and this causes their claim to be rejected.

This was the issue facing Mercy Medical Center, a 917-bed, non-profit hospital system that employs more than 5,600 people in Des Moines, Iowa. Just one part of that system, outpatient services, was losing revenue on claims that were written off for good by insurance companies.

Mercy Registered Pharmacist Peggy O'Connor earned her Black Belt stripes by tackling this issue during her Six Sigma training. Her main question: Why were certain insurance claims denied and how could Mercy prevent this from happening?

The Process

Using various Six Sigma tools, O'Connor discovered that a very high percentage of insurance claims were denied due to incorrect patient demographic information. Any hospital like Mercy has to deal with upwards of 100 different insurance providers, and they all have their own documentation requirements. For instance, some providers want you to spell out the patient's state of residence while others want it abbreviated. There were a lot of different little requirements like this, or opportunities for error, according to O'Connor. "If you don't have the correct information that the carrier wants, they will deny the claim, and then you have to rework it."

When O'Connor audited the records of about 30 patients, she found that 63 percent of them had been 'written off' due to errors, or failure to comply with insurance company requirements. When a claim is written off, it can never be paid, because the window of time for correcting errors has expired. O'Connor says that these particular records had been 'touched' by the rework process 486 times before they were finally disallowed. That's a lot of rework with no result, and O'Connor's Six Sigma team knew what it had to do: figure out why so many claims were denied and then later had to be written off by the hospital.

"The big problem at first was data," says O'Connor. "What type of data do I need, where is it and how do I get it?"

O'Connor says her teammates were key, and they led her to all kinds of claim data from different parts of the outpatient organization. The team used an affinity process to organize the data into key categories that then became the main branches of a Fishbone Diagram. With this view, the team was able to create a Pareto chart and narrow the problem down to its primary cause: incorrect patient demographic information. That was it; that was the big discovery. While there were many different causes for claims getting stalled, reworked and eventually disallowed, the leverage was in the demographic information part of the claim form.

To confirm, O'Connor audited another 386 charts representing claims that had been denied but not written off. She found that 78.2 percent of the denials were due to the presence of incorrect demographic information, with almost half due to inaccurate insurance information. It all seems so smooth and obvious, but O'Connor assures it was not. "Without the Six Sigma tools, we would never have been able to narrow down to the biggest cause of our problem," she says. "It was a big project, and I would have been lost."

Once the team knew the main cause of its pain, it used other Six Sigma tools to remove it. They used Design of Experiments (DOE) to pinpoint the leverage and benefits of a patient data software system that had been retired. Through the DOE, the team predicted it could save itself a lot of time and errors by bringing the system back into play, which it did. The system automatically loads proper patient information into the claim form system. Also, the team enacted a Failure Mode and Effects Analysis (FMEA) to design a better process for preparing claims more accurately the first time. The FMEA resulted in a new training and career advancement program for all who are involved in the patient data collection and claim process. Now employee training is improved and standardized, and people are more motivated to do a good job. Says O'Connor: "The FMEA was the hardest thing to do because it forces people to stop and think about what the problem really is, as well as what could go wrong in the future. But it's also the most beneficial because it shows how to improve a process from the very beginning of where the problem starts."

Author Name:

Organisation:

Project Objective:

Estimated Cost Benefit: