Infection as the Primary Discharge Diagnosis
Infection is the second leading cause of hospitalization among dialysis patients. The rate of admission due to infection has not meaningfully improved in recent years, a pattern remarkably consistent across the Census Divisions. The highest rates, among both incident and prevalent dialysis patients, are apparent in the East North Central, Middle Atlantic, and New England areas, while rates tend to be lower in the Mountain and Pacific states. This ordering of rates among the Census Divisions is similar to that seen for primary diagnoses of cardiovascular disease.
Because of widespread use of venous catheters for vascular access during the first year of dialysis, especially at initiation, the rate of admission for infection is markedly higher among incident dialysis patients than in the prevalent population. Unfortunately, between 2003 and 2010, the rate appears to have increased during the first five months of dialysis, especially during the second month. This is concerning. Careful surveillance of infectious complications is needed to assist dialysis providers in addressing this important source of morbidity.
Complicating matters in this domain, just as in the domain of cardiovascular disease, are shifting coding practices likely designed to maximize reimbursement for hospitals. In subsequent pages, we show that hospitalization rates for vascular access infections, including peritonitis, have ostensibly decreased. At this same time, however, hospitalization rates for septicemia have increased. It is important to recognize that Medicare claims may be limited in their capability to distinguish between specific infectious complications among dialysis patients. An assessment of the incidence of all infectious complications is likely necessary both to guide quality improvement and to undergird rating systems, so that improvements are not overstated.
More broadly, it is important to remember these data arise from hospitalization. The burden of infectious complications necessitating inpatient care is important to characterize, but these data fail to represent the burden of infectious complications diagnosed and treated exclusively in the outpatient setting. In future reports we will delve into this area in greater detail, examining outpatient diagnoses of infection, the use of both intravenous and oral antibiotic agents, and the relationship between vascular access technique and the incidence of infection.