Maryland’s Experience: A Closer Look at the Impact of APR-DRGs

There is much to learn from Maryland's hospitals, which for more than a year have been using the All-Patient Refined (APR-DRG) severity-adjustment system from 3M Health Information Systems. APR-DRGs form the foundation of the system that the Centers for Medicare & Medicaid Services (CMS) initially proposed to use in its nationwide roll-out of a severity-adjustment system.

Although CMS is now evaluating alternative severity-adjustment systems (including Ingenix's APS-DRG system), APR-DRGs remain in contention for CMS' selection. Therefore, there is value in evaluating the experience that Maryland's hospitals have had with the system. Specifically, health information managers there have made several compliance-related observations:

• The complexity of the APR system and the complicated interactions between multiple diagnoses and other factors, make it difficult to analyze case-mix changes or to evaluate the accuracy of DRG assignment.
• Coding professionals with APR experience indicate that the complexity of the system makes it nearly impossible to determine why similar cases are classified differently and to identify errors. Most professional coders have a very good understanding of how different diagnosis codes affect particular DRGs. This knowledge is then used when a similar case is coded. The APR-DRG system, however, contains logic that changes for each secondary diagnosis code, depending upon which principal diagnosis code is assigned, and which other diagnosis codes are assigned.
• It appears that there is a "more is better" mentality to coding, in which the coders have learned to assign codes for "anything and everything" documented in the record, in an effort to ensure capture of all conditions that may affect severity. But this method of coding may actually violate official coding guidelines, especially in scenarios indicating that certain codes, such as those for signs and symptoms, should not be reported in the presence of a related specific disease process. There is an inappropriate incentive to assign these additional codes if the APR-DRG's severity level is affected.
• From a productivity standpoint, APR-DRGs have been very expensive. Some Maryland hospitals, particularly the large facilities, have reported that productivity results are half the pre-APR implementation numbers. Additionally, these hospitals have tripled their coding staffs to ensure appropriate monitoring and feedback. It's clear that many hospitals around the country will not have the ability to adjust their staffing levels to this degree. Management and compliance staff must then ensure that the coding staff receives as much education related to the severity-adjusted system as possible.

Regardless of the final system selected by CMS, there will definitely be a learning curve for the coding staff and the staff members who are involved in making sure that the clinical documentation is as complete and accurate as possible. Many facilities have implemented documentation improvement programs during the last several years, but these have been based on the current CMS DRGs. As a result, many of the specific guidelines developed for clinicians will now be outdated as a new system is put in place. There will be a necessary lag time between system implementation and the time when most coders are comfortable with the nuances of the system and have recovered some of the downturn in productivity.

Any change to the DRG system will require additional training for coding staffs. However, the experience of Maryland's hospitals with the APR-DRG system seems to indicate that the design of the system creates some unnecessary challenges that could be overcome with a more streamlined system. What are your thoughts?