Health Information Management

AHIMA study examines CAC impact on coding

HIM-HIPAA Insider, July 29, 2013

Want to receive articles like this one in your inbox? Subscribe to HIM-HIPAA Insider!

by Jaclyn Fitzgerald, Online editor 

A recent study conducted by the AHIMA Foundation in collaboration with the Cleveland Clinic examined the impact of computer-assisted coding (CAC) on coding time, accuracy, and precision. The study revealed that using a credentialed coder with CAC allows for more accurate coding than using CAC alone. The findings were published in the July issue of the Journal of AHIMA.

Participants in the study coded 25 inpatient Cleveland Clinic cases with an average length of stay of 16 days. Six credentialed coders assigned ICD-9-CM procedure and diagnosis codes using CAC and six assigned the codes without the help of CAC. The Cleveland Clinic conducted the first phase of the study several weeks after it implemented the technology, and conducted the second phase six months after implementation. The same 25 cases were used in each phase.

Cleveland Clinic requires coders to participate in a training and evaluation program and demonstrate 95% coding accuracy before permitting them to code inpatient records. Cleveland Clinic compared the assigned codes were compared to their “gold standard,” which is what the clinic deems to be the correct set of codes for each record.

In the study, the use of CAC reduced the time spent on each record by 22%. HIM and coding professionals should consider the use of CAC because data suggests that it could be useful in combating decreased inpatient coder productivity during ICD-10 implementation, according to the Journal of AHIMA article.

The coders using CAC reduced the time spent on coding without reducing quality or accuracy. The findings stated that using CAC without a credentialed coder decreased recall and precision rates. However, diagnostic coding precision and procedure coding recall improved when a coder used CAC. For the purpose of the study, recall was defined as the number of gold standard codes identified by a coder. Similarly, precision was defined as the number of codes in a coder’s set that were in the gold standard set.

The study revealed that the natural language processing (NLP) technology used by CAC systems can improve recall and precision over time. The system coded with greater accuracy in phase two than in phase one because of a process within the NLP known as tuning, according to the article. Over time, NLP tuning allows the system to learn the correct codes for cases.

The results of the study show that HIM professionals need to be active in the CAC process to make sure the systems are used effectively and efficiently, said AHIMA CEO Lynne Thomas Gordon, MBA, RHIA, FACHE, CAE, FAHIMA, in a press release.

For more information, read the findings on the AHIMA website.

Want to receive articles like this one in your inbox? Subscribe to HIM-HIPAA Insider!

Most Popular