Credit Where Credit is Due
A Lean Six Sigma team at Madigan Army Medical Center generates more than $1 million in extra revenue in one year by improving the coding process. Lessons from this project have been spread throughout Madigan and are now reaching into other Army medical facilities.
Sherry Van Patten’s team estimates that during FY06, Madigan Army Medical Center’s OB/GYN clinic missed out on about $580,000 in revenue because of incomplete or incorrect records of outpatient encounters. Though large, that figure may seem like an underestimate once you learn that improving the process of recording encounters generated an extra million dollars in revenue for the clinic in FY07, despite a reduced patient population.
“At the start of my Green Belt project, we believed that about 21 percent of the clinic’s records were incomplete,” said Van Patten. Her team’s initial goal was to get that number down to 10 percent. But in less than two months, it dropped to an average of 0.06 percent—about six incomplete records for every 10,000 records. That drop means that the clinic can claim credit for more procedures and encounters than it could when the records were incomplete.
The lessons from this project have been spread throughout Madigan and are now reaching into other Army medical facilities. Here’s the story behind the remarkable numbers.
Project Launch
Denise Schultz, a records coder at Madigan, was the first to point out the problem with the high number of incompletely coded outpatient encounters.
“Denise really started the ball rolling,” recalled Van Patten. “She saw that we had a lot of incomplete records across the medical center, and that the OB/GYN clinic was the worst. A group of people including Captain Shawn DeFries, Kathy Pegum, Eileen Kosel, Denise Schultz, Sharon Smith, and Brenda Anthony got interested in looking into the problem.”
At about the same time, Van Patten was nominated to become the first Green Belt for Madigan, which was beginning to use Lean Six Sigma. It seemed a natural fit to formalize the coding efforts as a Green Belt project, and Van Patten started leading the work team, which became known as the Coding & Workload Accounting Improvement (CWAI) project. Major Richard Wilson, chief, patient administration, was the sponsor.
Working the DMAIC Steps
Since this was a Green Belt project, the team followed the standard five-phase DMAIC (define-measure-analyze-improve-control) improvement method built into Lean Six Sigma.
“We had a clear problem and goals, so we breezed through the first phase, Define,” said Van Patten. “We also had a very clear metric we needed to measure: whether a record was complete or incomplete. That information was captured in our database, so the Measure phase went quickly, too.”
The two key activities in these phases were documenting the process and documenting the baseline performance level. Figure 1 shows the process as everyone believed it to be happening originally.
The baseline data for the project is shown in Figure 2. A “real” encounter is one for which all documentation is completed—meaning that information is complete for all encounters. An “inferred” encounter represents work that cannot be fully credited because the records are incomplete.
The most helpful work in these phases was getting out into the clinic to observe the procedures, interview clinic staff, and see if the process steps were occurring as planned.
“We went to every workstation and interviewed the staff for a half-hour or 45 minutes, asking them to walk us through what they did,” said Sharon Smith, lead coder for OB/GYN.
Getting cooperation was no problem for the team. “We had a very strong message,” said Brenda Anthony, a lead outpatient coder. “We explained that we were trying to get them credit for the work they were already doing. We said we knew they were working hard and wanted to make sure that they were receiving their workload credit correctly.”
Through the observation and interviews, the team saw where there was duplication, where people weren’t documenting what they were doing, and so on. The team also discovered that the end-of-day checks were not always completed.
That led them to the conclusion that training was a key issue. Kathy Pegum, chief, Medical Records Branch explained, “Some staff hadn’t been trained at all in using the hospital appointment and documentation database, and in some cases each role had been trained in isolation.” For example, physicians didn’t know how the nurses or aides had been trained in coding procedures, or vice versa, or whether the training for each group matched what the other groups received.
The team’s work got much more complicated once it reached the analyze phase. “We knew from looking at the records and the interviews and observations that there were many, many different causes of incomplete records,” said Van Patten. A data field could have been missed, or a physician didn’t get a co-signature or code something correctly, for example.
These issues came to the forefront when the team worked on a failure modes and effects (FMEA) analysis. FMEA is a tool that helps teams determine the relative importance of different contributors to a problem, and focus their solutions on the problems that have the biggest overall impact.
“We went through the process step-by-step and thought about what went wrong in each step, and the impact that it had on the process.” said Van Patten. “It opened our eyes to the stoppages in the workflow and showed us where training would be most critical.”
Results
The team completed the final Improve and Control phases of DMAIC by developing a 16-point improvement plan that focused on three critical control points: (1) patient check in, (2) documentation of procedures by physicians and nurses, and (3) end-of-day records check. Specific actions addressed:
- Improving documentation/coding compliance
- Initial/retraining for staff
- Clarifying the responsibilities of each role (nurses, techs, clerks)
- Establishing standard business practices
- Assisting in the transition to a paperless system
- Coordinating with other functions for provider’s validation in CHCSI
- Closing out incomplete records with clinic staff
“We learned that the providers didn’t realize just how important it was to code encounters properly,” said Eileen Kosel, assistant chief, medical records. “Complete records are needed so the medical center can figure out how much time we spend on different procedures and allocate resources accordingly.”
The impact of these steps was immediate and dramatic, as shown in Figure 3. The team calculated that the increased revenue in FY07 from improved record keeping was just over $1 million for this one clinic—despite the fact that the redeployment of several brigades lowered the potential patient population substantially. (They don’t like to think about what would have happened to the clinic financially if they didn’t have these improvements in place before the patient volume dropped.)
Expanding the Impact
After this project was completed, the team was asked to take their improvements to other clinics at Madigan. Over the following months, they worked with other clinics in the facility for one to two weeks each, depending on the size of the clinic. Team members knew exactly what to look for in the processes at other clinics because they had studied the OB/GYN process in depth. The team has also helped out facilities in Alaska (Bassett Army Community Hospital) and California (Weed Army Community Hospital and Monterey Medical Clinic).
“There wasn’t one clinic that didn’t have some of the same problems that we had identified in OB/GYN,” commented Pegum. “Some details were different, but the overall pattern was the same.”
Conclusion
Now that their hard work is paying off, what sticks with the team is the challenge of trying to improve a process where so many steps have to work together right. Van Patten remembered that when she was in Green Belt training, the other people were able to create a lot of data charts that helped them focus on one or two specific causes of the problems they were studying. “We didn’t have that pattern here,” she said. “If any one of the steps in the process doesn’t happen right, we don’t have a complete record of the encounter, which could mean lost revenue. We had to tackle the whole process.”
The team is also pleased with the new role that coders have taken on in Madigan and other Army clinics. “We aren’t just seen as coders any more,” said Van Patten. “People now see that we can help with their productivity and efficiency, and making sure they get credit for all their hard work. That’s been very positive.” ♦