CAC & the Coder’s Transition

By Becky DeGrosky, Product Manager at TruCode & Chris Casto, Vice President of Sales & Marketing at Dolbey for Advance for Health Information Professionals

Computer-assisted coding and encoder technology are changing workflow for ICD-10 readiness.

The medical coding career has not experienced a significant change event since the introduction of DRGs in 1984. For decades, processes and technology have remained widely unchanged. ICD-10 is forcing many to reconsider their current workflow and deploy new technology to face the upcoming challenge.

Manual Processes

In an effort to increase productivity, a great deal of the current process requires re-engineering. Coders have been plagued by departmental practices that fly against productivity. Workflow for many is performed manually through spreadsheets of discharge records. The details and documentation required to code an account has been found across multiple applications, each requiring log-in and redundant data entry.

ICD-10 Impact

The expanded code set found in ICD-10 has been appropriately identified as one reason that coders will face greater challenges to complete accounts in a timely fashion. In addition to the sheer number of codes to select from, greater specificity will bring about new medical terms in the documentation and potentially an increased volume of information that will need to be read to complete the job.

All of these challenges will collide once ICD-10 is implemented and without a change in technology, coders face a perfect storm. Computer-assisted coding (CAC) is proving to better equip coders for the challenge. CAC is software that reads the patient chart and, using natural language processing (NLP) and how language relates to medical codes, suggests codes and highlights relevant documentation before the coder has read a single word. This coding head start gives the coder the ability to see the patient story as a coded event right away. Through a process of validation, or rather accepting and deleting suggested codes, coders are faster and, some research suggests, more accurate. “Coders adapt well to the role of auditor”, stated Heather Eminger, CAC product manager at Dolbey Systems. “Many are skeptical at the on-set but the technology proves itself in a matter of hours. Over a short time, coders settle into a cooperative trust with the capability of the software. They begin to understand when the software is likely to be spot on and when further examination is required to produce an accurately coded chart.”

Case Study: CAC Impact on ICD-9 Coding

As mentioned earlier, CAC offers many opportunities to increase coder productivity with both ICD-9 and ICD-10. As compared to the manual workflow and method, CAC offers:

  • A unified workspace that eliminates redundant data entry
  • Suggested codes and linked supporting documentation
  • Smarter workflow based on skill set matched with chart type, DRGs or codes

At a major health system, the Health Information Management Department performed a study on the impact of coder activity before and after having the aid of CAC. The results are in the table below:

With EMR/Encoder
Patient Type Outpatient Test Script
Financial Class Medicaid
Avg. Time to Code  (mm:ss) 01:47
Avg. # of Keystrokes and Clicks 21
Patient Type Inpatient Test Script
Financial Class Self Pay
Avg. Time to Code Complete (mm:ss) 08:39
Avg. # of Keystrokes and Clicks 101


With CAC/Encoder
Patient Type Outpatient Test Script
Financial Class Medicaid
Avg. Time to Code Complete (mm:ss) 00:34
Avg. # of Keystrokes and Clicks 2
Patient Type Inpatient Test Script
Financial Class Self Pay
Avg. Time to Code Complete (mm:ss) 06:34
Avg. # of Keystrokes and Clicks 51

As the results show, coder workflow was positively impacted by adding the technology of CAC. The inpatient results yielded nearly a 50% reduction in keystrokes/clicks and took about 30% less time to code. The outpatient test results are more dramatic. The keystrokes/clicks were reduced to nearly a single click while the time to code was reduced by nearly 70%. This time study shows significant productivity gains by coders using CAC instead of using just the EMR/encoder for coding.

Although this study was performed with ICD-9 instead of ICD-10, a direct impact of the technology can be envisioned with a larger code set, more codes and more specific documentation. The influence of CAC could be even greater than it was when used to code for ICD-9.

CAC and the Encoder

As described before, the role of the coder using CAC is to validate the suggested codes. When validating codes that are not committed to memory, often the coder needs to lean on the capability of the encoder. The encoder plays a role in this validation process and the efficiency that it provides has a direct impact on the productivity results.

The encoder, with or without CAC, essentially provides the capability of (a) finding a code, (b) editing all the codes assigned to the encounter, (c) reviewing coding specific reference material, and (d) grouping and pricing.

Becky DeGrosky, product manager at TruCode, said that “In the realm of computer-assisted coding, the encoder plays a different but vital role. Finding a code is no longer the primary function as CAC has already accomplished that task. The encoder functions of editing and providing applicable references in a visually simple manner take center stage. You also want to provide the coder a simple way to update the code without starting from scratch.”

To achieve maximum efficiency, the encoder features and functionality are integrated as part of the CAC solution. The coder should not have to launch a second application while coding within CAC. All the information required to validate the codes assigned should be part of the CAC display, allowing the coder the ability to drill down using a single click. On the rare occasions when the CAC engine missed a code, or when the code selected needs to be corrected, the encoder should provide an easy to use search where the documentation from the EHR can be copied and pasted directly. The coder should not have to re-launch the edits or the grouper. Assignment of the new DRG should be automatic.

Integration also eliminates an interface, and with many systems involved, one less interface improves efficiency.

Minimized Encoder Steps

The encoder as a tool works best when provided in-line with the CAC workflow. The two options of the process are: (a) codes are suggested by CAC and then preliminary validated codes are examined by the tools of the encoder; or (b) codes are suggested by CAC and simultaneously examined by the encoder with the tools for reference, edits and further look-up. Integration of the functionality allows the coder to use the encoder where it works best for them in the workflow.

As demonstrated in the study, the reduced keystrokes and clicks resulted in increased productivity. Decreasing the steps necessary to validat ethe suggested codes with the encoder also positively impacted the productivity gains.

The encoder methods are quite different depending upon the vendor. With a logic-based approach, the coder answers a series of predetermined questions to arrive at the recommended code. With ICD-9, coders have become accustomed to this effort. With ICD-10, the logic tree expands.

With book-based encoders, the coding books are provided on the screen in an indexed, searchable format that enhances the functionality of a book.

“We asked customers to use their encoder to validate CAC-suggested ICD-10 codes. In a time study, we were able to compare a logic-based encoder and a book-based encoder. The difference in the number of clicks or keystrokes that were required to find a code was quite dramatic” states Eminger. “The book-based version enabled the coder to validate the code with far fewer keyboard and mouse actions.”

Since CAC has already done the work of assigning code and highlighting the corresponding documentation, the historic role of the encoder as the primary resource to arrive at a code is no longer the case. Now the coder is auditing what has already been assigned and using the edits, advice, and references associated with those codes at a much higher level. No longer is the focus of the coder on the individual codes, but on the encounter of codes as a whole. A tool that brings that information to the coder, and doesn’t require the coder to search for it, provides even greater efficiency to the task.

Share Article:

One comment

  1. Was the same chart used to code both scenarios outlined above? The question would is if the same chart was used to code the coder was already familiar with the documentation/code selections which would decrease coding time the second time around.

Comments are closed.

Dolbey Systems, Inc.