News: Pilot study finds 63% accuracy in ICD-10 coding

In an effort to understand the effect of ICD-10-CM/PCS coding on the industry, the Health Information and Management Systems Society (HIMSS) and Workgroup for Electronic Data Interchange (WEDI) joined forces to test coders’ efficiency in the new code set.

The ICD- 10 National Pilot Program began on April 12, 2012 and was completed on August 30, 2013. Volunteer AHIMA-approved ICD-10 Trainers from various facilities coded a series of de-duped medical records with an average accuracy rate of just 63%, according to an October report detailing the findings.

Most errors “were functional,” according to the report, either because the case numbers did not match appropriately or other administrative errors, but the study also highlighted some interesting areas for potential targeted improvement including coding additional chest pain specificity. (View the coding efficiency rates by case example online.) Each set of cases is called a “wave” for the purposes of this study.)

Among the two agencies’ recommendations for ICD-10-CM/PCS training and testing were:

• Assigning codes straight from ICD-10 rather that attempting to translate a code from ICD-9 to ICD-10.
• Monitoring DRG shifts to ensure payment accuracy
• Planning for ample time for remediation

“ICD-10-CM/PCS touches almost every aspect of the healthcare continuum. All affected entities should be implementing and testing at this point in time,” the report states. “Create a strategy if planning has not begun. Build a solid structure surrounding internal and external testing, engage trading partners early, identify technical and business needs, and engage the business in developing ‘day in the life’ scenarios.”