The evaluation was based on a review of existing publications obtained from public sources and potential vendors. Specific evaluation criteria included:

• Internal consistency with a stated definition of "disease severity" concept, such as:
• disease progression
• likelihood of death
• likelihood of high inpatient expenses
• likelihood of lengthy hospital stay
• burden of disease (e.g., presence of comorbidities)
• Focus on all patients rather than a single payer or population
• Medical meaningfulness, that is, physicians were involved in system development
• Wide use by organizations such as hospitals, payers, peer review and accreditation organizations.
• Relevant research published in the public domain on the development, performance, and applications of the severity algorithm
• Value for cost
• Vendor willingness to let the product be disseminated in the HCUP restricted access public release databases (i.e., the NIS and KID)

Results of the Evaluation

The evaluation results suggested that in addition to Medicare DRGs and AHRQ Clinical Classification System (CCS) which are already disseminated with HCUP data, severity measures from four systems should be added to the database -- Disease Staging,² the AHRQ comorbidity measures,³ All Patient Refined (APR) DRGs,⁴ and All Payer Severity-adjusted (APS) DRGs.⁵ These systems are sufficiently differentiated in their technical properties making them unlikely substitutes for one another in any potential application (see Table 1). Each system has a relative advantage over the others depending on the purpose of the analysis.

Disease Staging

Disease Staging and the AHRQ comorbidity measures represent clinically based systems. Disease Staging is a product of Medstat. In Disease Staging, severity is defined as the likelihood of death or organ failure resulting from disease progression and independent of the treatment process.² Disease progression is measured using four stages of increasing complexity and substages within Stages 1, 2, and 3. Stages and substages are defined specific to each disease category following clinical criteria:

Stage 1 – no complications or problems of minimal severity
Stage 2 – problems limited to a single organ or system; significantly increased risk of complications
Stage 3 – multiple site involvement; generalized systemic involvement; poor prognosis
Stage 4 – death

Disease Staging uses information on diagnoses, gender, Cesarean section procedure codes, and discharge status to assign cases to disease categories (version 4 of the software includes 606 categories) and to measure disease-specific severity. Major operating room procedures are not used in assigning disease categories, thus use of procedures does not figure into the assignment of severity (as with DRG-based systems). However, to improve the value of Disease Staging in explaining costs, length of stay, and mortality, Medstat developed expense, stay and mortality predictive scales based on Medicare DRGs, using the DRG medical/surgical splits. Patient age, gender, and admission and discharge status are also used in these models. The following measures are included in the NIS and KID, based on Disease Staging software: disease categories and stages for the principal diagnosis and predictive scales for mortality, length of stay, and total charges.

AHRQ Comorbidity Measures

The AHRQ comorbidity measures identify coexisting medical conditions that are not directly related to the principal diagnosis, or the main reason for admission, and are likely to have originated prior to the hospital stay.³ These comorbidities can make a hospital stay more expensive and complicated. The AHRQ comorbidity measures were developed originally as one of the HCUP tools. Complete documentation on the comorbidity measures is available on the HCUP User Support Website under Tools & Software (http://www.hcup-us.ahrq.gov/toolssoftware/comorbidity/comorbidity.jsp).

APR-DRGs and APS-DRGs

APR-DRGs and APS-DRGs are DRG-based severity measurement systems. A brief review of different DRG classification systems helps shed light on the rationale for selecting these two particular systems. DRG-based systems, as illustrated in Figure 1, include:

- Medicare DRGs
- Refined DRGs (R-DRGs)
- All-Patient DRGs (AP-DRGs)
- Refined DRGs (R-DRGs)
- All-Patient DRGs (AP-DRGs)
- All-Patient Refined DRGs (APR-DRGs)
- Severity-adjusted DRGs (S-DRGs), andv - All-Payer Severity-adjusted DRGs (APS-DRGs)

There are two major limitations in using Medicare DRGs for severity adjustment. First, there is limited adjustment for severity of illness.⁶ Principal diagnoses and procedures are stratified into categories based on the presence of a substantial complication or comorbidity (CC) in secondary diagnoses. The CC list includes about 3,000 diagnosis codes for diverse conditions that range from major acute illness to less severe chronic conditions. As a result, DRG categories are unable to sufficiently account for the differential effects of these secondary diagnoses on resource use. Second, Medicare DRGs have limited categories for conditions specific to the nonelderly, particularly for neonates,⁷ for whom there are only seven broadly-defined DRGs.

Developed during the mid to late 1980s, Refined DRGs (R-DRGs) and All-Patient DRGs (AP-DRGs) represented the first modifications of Medicare DRGs that attempted to account for severity of illness. Both systems address the limitations of DRGs through refinement of the CC list and the DRG categories for neonates.

With R-DRGs, developed by researchers at Yale University, the CC list for secondary diagnoses was specified separately for each DRG and the DRGs for neonates were revised to incorporate birthweight.⁶ AP-DRGs, first implemented in New York State, expanded Medicare DRGs to include neonatal, obstetric, and other conditions typical to the under-65 population.⁴ Based on research conducted by the National Association of Children’s Hospitals and Related Institutions (NACHRI), neonates were stratified by age, birthweight, and procedure.⁷ The assignment of CCs for secondary diagnosis was determined at the MDC level, rather than the DRG level.

AP-DRGs formed the basis for All-Patient Refined DRG (APR-DRGs) which were developed by 3M Health Information Systems in the early 1990s.⁴ APR-DRGs added severity of illness and risk of mortality subclasses for each base APR-DRG. In determining the severity level, 3M not only revised the CC list to accommodate the non-Medicare population but also incorporated principal diagnosis, age, interactions of multiple secondary diagnoses, and combinations of non-operating procedures with principal diagnosis. The severity of illness and risk of mortality subclasses have levels of 1 to 4, indicating minor, moderate, major, and extreme, respectively. Based on these enhancements, APR-DRGs represented a significant improvement over both R-DRGs and AP-DRGs.

In 1994, the Center for Medicare and Medicaid Services (CMS, formerly HCFA) developed Severity-adjusted DRGs (S-DRGs) which incorporated aspects of R-DRGs and AP-DRGs.⁸ The major enhancement of S-DRGs lies in the revision of the CC list. Instead of defining the CC list by DRG (like the R-DRG system) or by MDC (like the AP-DRG system), researchers at CMS evaluated the presence of each ICD-9-CM diagnosis code as a secondary diagnosis to determine its effect on resource use. Based on this analysis, each code was assigned as a major CC (MCC), CC, or non-CC. Two concerns with S-DRGs are: 1) no refinement for neonatal groups, making this system inapplicable to an all-payer population; and 2) inconsistent severity categories across DRGs.

To address these limitations, HSS Inc. introduced All-Payer Severity-adjusted DRG (APS-DRGs) in 1995.⁹ As an expansion of S-DRGs, APS-DRGs are intended to be used for all hospitalized patients. Neonatal categories are defined by birthweight, diagnoses, and disposition. APS-DRGs also extend the exclusion criteria inherited from Medicare DRGs and S-DRGs to eliminate certain secondary diagnoses from the CC and MCC lists because those conditions are associated with the principal diagnosis. The severity categories are standardized across DRGs with 0, 1, and 2 denoting without CC, with CC, and with MCC, respectively.

In summary, APR-DRGs and APS-DRGs are built upon prior DRG-based systems with substantial refinements in severity specification and neonatal categories. These two systems differ from one another in their trajectory of development, clinical logic, severity classification structure, and level of complexity.

APR-DRGs (version 20) include 316 base disease categories. The following measures are included for APR-DRGs: base APR-DRG, severity of illness subclass, and risk of mortality subclass within each base APR-DRG.

APS-DRGs (version 20) consist of 375 base disease categories called Consolidated DRGs (CDRGs). The APS-DRG group number is represented by the CDRG category (XXX) plus the one-digit severity class (Y). Weights developed using the NIS for mortality, length of stay, and total charges are also included.

Documentation

Documentation for all systems is distributed with the NIS. For the three proprietary systems (Disease Staging, APR-DRGs, and APS-DRGs) abbreviated documents outlining the measures and their recommended use are included with the NIS. For users who wish to understand the specific assignment of cases to individual disease categories/DRGs and levels of severity, full documentation that provides the complete clinical and coding logic will be available from the vendors at a special reduced price. Contact information for each product is provided below.

References

1. Hornbrook MC. 2003. Disease Severity Adjustment Software Evaluation for the Healthcare Cost and Utilization Project. A report submitted to AHRQ.



2. Disease Staging Software User Guide. 2001. Ann Arbor, MI: Medstat Group, Inc.



3. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Medical Care 1998;36(1):8-27.



4. All Patient Refined Diagnosis Related Groups (APR-DRGs): Methodology Overview. 1998. Wallingford, CT: 3M Health Information Systems



5. Definitions Manual for All-Payer Severity-adjusted DRG (APS-DRGs) Assignment. 2003. Germantown, MD: HSS, Inc.



6. Freeman JL, Fetter RB, Park H, Schneider KC, Lichtenstein JL, Hughes JS, Bauman WA, Duncan CC, Freeman DH Jr, Palmer GR. Diagnosis-related group refinement with diagnosis- and procedure-specific comorbidities and complications. Med Care 1995;33(8):806-27



7. Muldoon JH. Structure and performance of different DRG classification systems for neonatal medicine. Pediatrics 1999;103(1 Suppl E):302-18



8. Edwards N, Honemann D, Burley D, Navarro M. Refinement of the Medicare diagnosis-related groups to incorporate a measure of severity. Health Care Financ Rev 1994;16(2):45-64



9. Leary RS, Johantgen ME, Farley D, Forthman MT, Wooster LD. All-payer severity-adjusted diagnosis-related groups: a uniform method to severity-adjust discharge data. Top Health Inf Manage 1997;17(3):60-71