Cause of death with graft function among renal transplant recipients in an integrated healthcare system.

BACKGROUND: Cardiovascular disease (CVD) is the leading cause of death in renal transplant recipients with a functioning allograft. Modification of CVD risk factors may, therefore, decrease overall mortality in this patient population. We studied renal transplant recipients within an integrated healthcare system (IHS) that uses case management and electronic health records to determine mortality from CVD. METHODS: We retrospectively collected data on all renal transplant recipients over a 10-year period. The primary endpoint was death with graft function (DWGF). Cardiovascular events were used as secondary endpoints. We determined the cause of death and collected laboratory data. The data were analyzed using Student's t test for continuous data, chi square for categorical data, and multivariate logistic regression. Survival was determined using the Kaplan-Meier product-limit method. RESULTS: Death from "other" causes accounted for 29%. This was followed by CVD (24%), infection (16%), and malignancy (12%). The most common "other" causes were diabetes mellitus and end-stage renal disease. Overall, lower hemoglobin, uncontrolled blood pressure, and lower albumin levels were associated with DWGF. There were 184 cardiovascular events in total. Low-density lipid levels were lower in the group with cardiovascular events and DWGF. The use of antihypertensive and antihyperlipidemic agents was similar between the two groups with the exception of diuretics, which were used more often in the DWGF group. CONCLUSIONS: There was a low rate of DWGF because of CVD within this IHS. It is possible that coordinated care within an IHS leads to improved cardiovascular mortality.

Hospitalization rates in daily home hemodialysis versus peritoneal dialysis patients in the United States.

BACKGROUND: Daily hemodialysis (DHD) is associated with improvements in hypertension, left ventricular hypertrophy, mineral metabolism, nutrition, and quality of life, but efficacy is uncertain because of potential selection bias. To reduce the influence of selection bias, we sought to compare hospital admissions for our population of DHD patients with peritoneal dialysis (PD) patients who initiated training during the same period. We also compared our hospital data with the US Renal Data Service database. STUDY DESIGN: Prospective nonrandomized cohort study. SETTING & PARTICIPANTS: 22 (16 male) DHD and 64 (33 male) PD patients who initiated training between March 2003 and September 2007 at our center and remained in our program for at least 6 months. PREDICTORS: Dialysis modality (DHD or PD). OUTCOMES: Number of hospital admissions and length of stay. RESULTS: Median age at initiation of training was 52 years (range, 33 to 76 years) for DHD patients versus 54 years (range, 21 to 82 years) for PD patients (P = 0.5), and median vintage was 23 months (range, 0 to 145 months) for DHD patients versus 0 month (range, 0 to 244 months) for PD patients (P < 0.001). Fifty percent of DHD and 56% of PD patients had a diagnosis of diabetes mellitus (P = 0.8). We observed 27 DHD and 82 PD admissions (0.68 and 0.76 admissions/patient-year, respectively) during the study period (P = 0.5). We also observed 130 DHD and 605 PD hospital days (3.3 and 5.6 days/patient-year, respectively; P < 0.001). LIMITATIONS: Patients were not randomly assigned between the study group and control group; study group was small. CONCLUSIONS: Our study suggests that despite similar patient demographics, patients treated with DHD spend fewer days in the hospital than PD patients in the United States. Although selection bias could partially explain our lower hospitalization rate, other factors, including improvements in blood pressure control, nutrition, and fewer fluctuations in dry weight, probably contributed to the stability of our patients.