Admissions for heart failure, as well as for the related pathology of cardiomyopathy, are very common among dialysis patients. Explicit designations of heart failure in Medicare claims, however, do not tell the entire story. From a clinical perspective, heart failure and fluid overload can be difficult to distinguish. A diagnosis of heart failure implies that the major area of dysfunction is cardiac in nature, arising from systolic dysfunction, diastolic dysfunction, or both, while a diagnosis of fluid overload suggests that the major problem is volumetric, with cardiac function relatively intact. But these are pathophysiologic principles. The extent to which they extend to coding practices is at least partially dictated by the detailed features of MS-DRGs and the relative rates of reimbursement that accompany alternative arrangements of diagnosis codes. On these two pages we explore admissions for heart failure and cardiomyopathy; on the next two we look at admissions for fluid overload and pleural effusion, a common consequence of fluid overload.
Admissions for a primary discharge diagnosis of heart failure, which are overwhelmingly ascertained from ICD-9-CM diagnosis code series 428, appear to have decreased in frequency during the study era among both incident and prevalent dialysis patients. After the advent of MS-DRGs, rates of admission for a primary or leading secondary diagnosis of heart failure were only modestly higher than corresponding rates based on the primary diagnosis alone. There is substantial geographic variation, particularly among incident dialysis patients, in which first-year admission rates vary by a factor of two between the Census Divisions with highest and lowest rates. Between the annual cohorts of incident dialysis patients from 2003 and 2010, there were improvements in admission rates during every month of the first year of dialysis. Among prevalent patients, there is clear evidence of seasonality — similar to, if not more pronounced than, the seasonality observed in admissions for acute coronary syndrome.
While the rate of hospitalization for heart failure appears to have declined since 2003, analyses displayed in the subsequent pair of pages show that hospitalizations for fluid overload appear to have increased by a roughly commensurate amount since 2006. Collectively, hospitalization for the composite of heart failure and fluid overload appears to have changed very little since 2003. The fact that these rates are unadjusted should certainly be given due consideration, as the gradual aging of the dialysis population has likely increased the underlying risk of heart failure. Nonetheless, these data suggest that increased attention to fluid control is warranted.