VA Coding Pilot Yields Enlightening Coder Productivity Results
A Veterans Health Administration (VHA) coding pilot study brought to light surprising results related to ambulatory and inpatient coding productivity, researchers said during a Tuesday afternoon session at AHIMA’s ICD-10-CM/PCS and Computer-Assisted Coding Summit, taking place in Washington, DC.
Investigators found that coding productivity on ambulatory claims decreased only 6.7 percent, as measured by the time it takes to code a claim, using ICD-10 versus ICD-9. However, when investigators measured loss of productivity in coding inpatient claims, they recorded a 64.5 percent decrease in productivity.
The loss of productivity on inpatient claims can be attributed to challenges in locating documentation for operating room procedures and difficulties in finding specific details needed to assign a procedure code for non-operating room procedures that weren’t documented.
Pilot Study Results
These results were part of a pilot study of 10 VHA sites where coders with varying degrees of experience coded 1,024 ambulatory care records and 382 inpatient records, with the help of coding manuals and an encoder. The pilot’s investigators included Susan Fenton, PhD, RHIA, office of informatics and analytics at the VHA, Shelley Weems, RHIA, CCS, member of the health information management (HIM) program office within the VHA’s office of informatics and analytics, health information governance, and Pamela Heller, RHIA, CCS-P, program director for HIM within the VHA’s office of informatics and analytics, health information governance.
When Fenton—the self-described “data geek” of the group—looked at the data, she said she was excited by some fun findings, such as the breakdown of coder productivity loss by encounter type. For instance, coders were able to code ancillary encounters very quickly, but therapy encounter coding time almost doubled.
“It was very much worth our while to go across cases and look at specialties. That gives you something to work on,” Fenton said, in terms of coaching physicians on documentation improvement.
Unintended Findings from the Pilot
Through having coders code claims using both code sets, investigators were able to make several unexpected observations:
- Coders got excited and wanted to learn ICD-10—their confidence grew over time.
- The benefit of facilitating group discussions among coders cannot be overstated, according to Weems. “Coders being able to talk to each other, it did two things: built collaboration, [and] facilitated each coder’s learning knowledge because they had to explain why they did what they did,” Weems said.
- Using real world claims instead of examples from manuals really provides the best training. “In the real world, you don’t know what the right answer is. Real world cases gave us the best avenue,” Weems said.
- ICD-9 accuracy on cases increased after the study because coders were comparing guidelines between ICD-9 and ICD-10, which reiterated the guidelines.