EPO was approved for the treatment of anemia in renal dialysis patients in 1989. In 1991, Medicare changed the way it reimbursed for EPO, and, in response, there was an increase in average EPO dosages from 2793 units to 3200 units in the six month period following the policy change. This change, while meant to improve apparently low EPO dosing, may have had the unintended consequence of maintaining some patients at unnecessarily high hematocrit levels. These patients may experience problems associated with relatively high hematocrits. Thus, while some patients may have received too little EPO prior to the policy change, now they may be receiving too much. Another recent change in EPO reimbursement policy indicates this may be the case. After July 1, 1997 Medicare will reimburse EPO based on a 90-day rolling average hematocrit measurement. If the average is above 36.5%, Medicare will not pay for the EPO. This policy implicitly assumes that high hematocrit levels are directly related to inappropriately high doses of EPO. OBJECTIVE: The purpose of this study is to examine the incidence of complications resulting from abnormally high hematocrits, such as vascular shunt thrombosis, among patients with End-Stage Renal Disease (ESRD). METHODS: This study uses data from the United States Renal Data System (USRDS) and logistic regression to model the incidence of adverse events as a function of hematocrit level while controlling for confounding factors such as age, gender, ethnicity, weight, comorbidities, facility characteristics, dialysis modality, and dialysis history. RESULTS: Preliminary results suggest that the incidence of VST has increased since 1991. CONCLUSION: This study provides Medicare with information concerning the appropriateness of the 36.5% cutoff and the projected impact of the change in reimbursement.