ICD-10-PCS Exposed! Using Real Data to Root Out Accuracy and Productivity Issues
By Patty Buttner, RHIA, CDIP, CCS
Are you prepared for the accuracy and productivity challenges you may encounter shortly after October 1, 2015? During a presentation at the Clinical Coding Meeting, which took place on Saturday, Scot Nemchik, vice president of coding compliance and education at IOD, shared the strategies and methods currently in place in his organization to help minimize the impact of ICD10-CM/PCS implementation.
His organization’s priority has been to identify and address the main issues they expect to encounter with ICD-10 coding now to decrease the impact on accuracy and productivity when the new code set is implemented.
Nemchik’s organization began with coder training consisting of two phases. Phase 1 included foundational training, which included anatomy and terminology, 12 hours of ICD-10-CM training, and 19 hours of ICD-10-PCS training, followed by subject matter expert led webinars. Phase 2 involved experiential learning and analytics.
“People learn by doing, [and] experiential learning is best,” Nemchik said. “You can’t monitor what you don’t measure.”
A library of inpatient lessons was created, and subject matter experts coded these lessons in both ICD-9 and ICD-10. The coding work was then reviewed for quality and accuracy.
Coders were given specific DRGs to practice on in the first quarter of 2014. This initial set of coded DRGs established a baseline which trainers used to measure improvement.
Reports generated by the results of these lessons were analyzed and used to identify coding issues that would have the highest impact on reimbursement and productivity. The expectation was that the root operation determination would be a major factor. However, IOD staff were surprised to find out that root operation selection was not a significant problem.
Issues with Accuracy Identified
Analysis of the data revealed some major issues with some PCS codes, but not all. Some of the most common errors included:
Total hip replacement: Coders had difficulty selecting the correct device and qualifier characters.
The solution: Implant device forms were created, along with a policy and procedure for the assignment of the qualifier for cemented or uncemented.
Codes were not being assigned for leads in CRT and Biventricular defibrillator insertions.
The solution: Coders received education surrounding the need to code leads in ICD-10-PCS.
Coronary angiography: Coders struggled with image modality with fluoroscopy vs. plain radiography.
The solution: A contrast key was developed, along with a policy and procedure for imaging.
Selecting the origin and destination for Coronary Artery Bypass Grafts (CABG) was difficult for the coders.
The solution: Additional education was provided on the origin and destination in CABG coding.
Surprisingly, amputations were found to be troublesome. The body part and the qualifier stood out as areas of concern. Coders could not determine the level based on the documentation. This also had an impact on productivity.
The solution: A policy and procedure were created to guide coders in code selection when the level was difficult to determine. This decreased the time spent reviewing the operative report.
The same records were coded again in the second and third quarter of 2014. Analysis revealed significant improvement in coder accuracy in each quarter.
Nemchik briefly discussed the results of ICD-10-CM coding lessons. The data analysis revealed coders used the correct DRG, but were not always on target with the most specific code.
Accuracy Concerns Impact Productivity
Inpatient coders typically coded 1.36 charts per hour using ICD-10, which is a decrease from the average of 1.9 charts per hour. DRG accuracy fell to 78 percent. After a meeting with the coders, it was determined that they were so concerned with productivity that DRG accuracy was impacted. Once the productivity expectation was relaxed, DRG accuracy increased.
Nemchik’s main take away was to have coders code as much as possible in ICD-10 now. The more coders work in ICD-10, the better they will become. Coders should all code the same records for accurate identification of procedures and diagnoses that may be problematic for a given facility.
Also, he recommended providers evaluate the data, and educate and provide tools for coders. Create specific policies and procedures for sorting out coding gray areas, he said. Repeat the process to capture your improvement, Nemchik said, and measure your success.
Identify your organization’s “pain points” and “get through the pain now,” Nemchik recommended.
Catch up on the news and get insights from AHIMA’s 86th annual Convention and Exhibit held September 27-October 2 in San Diego, CA. For a complete list of event coverage on the Journal of AHIMA website, click here.