Purpose: To develop disease-specific criteria that differentiate severity levels within each selected chronic disease and to understand costs associated with severity levels.

Methods: A descriptive retrospective analysis using administrative claims from the Medstat Group® 1999-2001 MarketScan® Commercial Claims and Encounters Database. The study identified continuously enrolled adult patients (age ³18 years) diagnosed with any of the 6 diseases of interest—rheumatoid arthritis (RA), inflammatory bowel disease (IBD), chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD), cancer, and congestive heart failure (CHF). Selection of criteria categorizing mild, moderate and severe levels was disease specific utilizing combination of diagnoses (ICD-9 CM), procedures (CPT-4), medications, and durable medical equipment (DME) (HCPCS). RA was primarily by biologics and surgery; IBD by medication, hospitalization, and surgery; COPD by DME; CKD by hospitalization and transplantation; cancer by chemotherapies and metastasis; CHF by concomitant medication and hospitalization. The difference in costs between severity levels within each condition was compared using t-test.

Results: A total of 8,009 RA, 7,200 IBD, 47,837 COPD, 6,093 CKD, 40,857 cancer, and 13,041 CHF patients were identified. Within each disease, patients with moderate/severe conditions incurred much higher average annual healthcare costs than those with mild condition identified by the severity staging, with an exception of CHF population. Moderate/severe RA patients (8%) incurred $29,292 compared $10,687 by patients with mild condition. Moderate/severe IBD patients (10.9%) incurred $37,925 compared to $10,687 by patients with mild level. Moderate/severe COPD patients (8.8%) incurred $47,223 compared to $10,658 by patients with mild condition. Moderate/severe CKD patients (41.2%) incurred $85,145 compared to $24,202 by patients with mild condition. Moderate/severe CHF patients incurred $26,022 compared to $40,650 by patients with mild condition. The difference in costs were statistically significantly different between severity levels (p<0.05).

Conclusions: Disease-specific severity staging that identifies the level of a patient's condition can be applied as an adjustment and/or a predictive parameter in future retrospective or predictive modeling studies. Severity staging criteria should be tailored to different diseases that differentiate severity levels within each disease. The current study developed sets of criteria that would differentiate severity within the six selected chronic diseases. Although patients of moderate/severe level were a smaller proportion of the disease population, the costs incurred were significantly higher with an exception of CHF population in this study.