The DRG dilemma

Physician Executive, May-June, 2004 by C.C. Moreland, David Bloom

This approach is alien to the way physicians are taught to approach clinical problems. Furthermore, to concentrate 15,000 ICD codes and 7,000 CPT codes into 527 DRGs results in a lack of precision in accurately reflecting the severity of each case.

The DRG system is rule-based. The concepts of complexity, severity, complication or comorbid condition (CC) and case mix index are utilized. There are thousands of potential combinations for DRGs that can only be appropriately sorted by a computer. However, the critical choices must be made based upon the patient's medical record as documented by physicians.

Complexity is based upon the resources utilized in treating a case measured by the principal diagnosis or procedure, and a relative weight (RW) is pre-assigned. For example, a newborn has a relative weight of 0.1524 and a heart transplant has a relative weight of 20.2413.

Severity is measured by the secondary diagnoses (number and complexity up to 15 for a given case). The case mix index is measured by adding the relative weights of all the cases treated by a physician or a hospital divided by the total number of cases treated over a given period of time.

Each year on October 1 the relative weight for each DRG is recalculated for Medicare, based on the previous 12 months' data. This is obtained by adding the costs of treating all the cases in a given DRG in the preceding year and by dividing the number of cases in that DRG during the preceding year to obtain the average cost for the DRG translated in relative weights to the fourth digit.

Comorbidity is a condition that exists on admission and increases the length of stay one or more days in 75 percent of the cases. A complication is a condition that develops in the hospital and increases the length of stay one or more days in 75 percent of the cases.

The length of stay allowed by Medicare is based on the average length of stay for a patient with a DRG with a given severity and given comorbidities and complications during the preceding year.

Importance of understanding the DRG system

Using DRG databases, it is possible to compare resource utilization by physicians and hospitals within the constraints of the system. In addition, it is possible to determine expected mortality and morbidity rates. The large size of the Medicare database results in accurate information because of regression to the mean.

The medical record coding staff can only use physician documentation for selecting the DRG for a case. Because of poor documentation in the clinical record, many properly treated cases are coded as less complex and less severe than they actually are. This introduces a significant error in many physician and hospital profiles.

There is a disparity between the proper care actually delivered and the morbidity, mortality and resource usage expected for the reported DRG that is of lesser complexity and severity.

As a result, some physicians and hospitals that are effectively treating patients are misrepresented by the data they generate. They are considered inappropriate utilizers of resources because of low severity levels resulting in high utilization of resources and high mortality and morbidity rates.

It is important to improve the accuracy of the information submitted to the data banks so that correct decisions are made about compliance, reimbursement, pay-for-performance and managed care contracting.

Information in the Medicare data bank is a matter of public record. There are companies that refine and sell these data. With the movement toward the release of physician-specific and hospital-specific information, it is vitally important to ensure that the information is accurate.

Improving data

The data from each encounter a physician has each day with any patient (private pay, managed care, insurance, Medicaid or Medicare) are placed in the Centers for Medicare and Medicaid Services (CMS) databank in Baltimore, Md., under the uniform physician's identification number (UPIN) at billing.

The data allow comparison of clinical efficiency, severity and complexity by ICD 9 codes, CPT codes for diagnoses and procedures, length of stay, resources utilized (cost per discharge) and morbidity and mortality rates.

As the severity and complexity of a case are accurately represented, the resources to care for that case and the length of stay (LOS) are properly accounted for in each group. If documentation does not reflect the severity and complexly correctly and completely, resources used are higher than the average allowed for that case.

This makes the physician appear to be inefficient in appropriately managing a case where the actual data documented reflect that the patient is not severely ill.

This raises the contradiction that the physician is billing for a high-level evaluation and management (E/M) code (and possibly giving a high level of care) but documenting only a low level of care.

Since medical record audits are now becoming more routine, it is essential that physicians become aware of the data accumulated daily in their practices. When audits are done in the hospital or office setting the following simple questions must be answered: