Objective The National Kidney Foundation (NKF) Kidney Disease Outcome Quality Initiative (KDOQI) guidelines have recommended the use of arteriovenous fistula (AVF) at the initiation of dialysis. However, there are significant differences in the dialysis environments of Japan and the United States, and there are few people who receive hemodialysis via a central venous catheter (CVC) in Japan. The aim of the present study was to examine the association between the type of vascular access at the initiation of dialysis and the incidence of mortality in Japan. Methods This study was a prospective, multicenter, cohort study. The data was collected by the Aichi Cohort study of Prognosis in Patients newly initiated into dialysis (AICOPP) in which 18 Japanese tertiary care centers participated. The present study enrolled 1,524 patients who were newly introduced to dialysis (the patients started maintenance dialysis between October 2011 and September 2013). After excluding 183 patients with missing data, 1,341 patients were enrolled. The Cox proportional hazards model was used to evaluate mortality based on the type of vascular access. The types of vascular access were divided into four categories: AVF, arteriovenous graft (AVG), CVC changed to AVF during the course (CAVF), CVC changed to AVG during the course (CAVG). Results A multivariate analysis revealed that AVG, CAVF and CAVG were associated with a higher risk of mortality in comparison to AVF [hazard ratio (HR), 1.60; p=0.048; HR, 2.26; p= 0.003; and HR, 2.45; p=0.001, respectively]. Conclusion The research proved that the survival rate among patients in whom hemodialysis was initiated with AVF was significantly higher than that in patients in whom hemodialysis was initiated with AVG or CVC.

Background: Serum albumin is a marker of nourishment and inflammation. Although hypoalbuminemia in hemodialysis patients is reported as a risk factor for poor prognosis, few studies describe its effects on infectious diseases specifically. This study aimed to examine the relationship between the serum albumin level on admission and infection-related in-hospital death among hemodialysis patients. Methods: This was a multicenter retrospective observational study that was undertaken in Japan. We reviewed the medical records of 507 hemodialysis patients aged > 18 years, whose blood cultures were obtained based on suspicion of infectious disease, and who were managed at seven Japanese tertiary dialysis units from August 2011 to July 2013. The outcome measure was infection-related in-hospital death. Multivariate logistic regression models adjusted for age, sex, the dialysis vintage, diabetes mellitus, bacteremia, and log C-reactive protein levels were used for the statistical analysis. Results: Four hundred patients were analyzed and allocated to three groups based on their serum albumin levels: marked hypoalbuminemia (< 2.5 mg/dL), mild hypoalbuminemia (≤ 2.5-< 3.5 mg/dL), and normal albumin levels (≤ 3.5 mg/dL). The infection-related in-hospital death rates were 22.9% (n = 11), 12.5% (n = 25), and 4.6% (n = 7), respectively. The multivariate logistic regression models determined that a low serum albumin level was an independent risk factor for infection-related in-hospital death (odds ratio 0.35, 95% confidence interval 0.18-0.66). Conclusions: A low serum albumin level strongly predicts infection-related in-hospital death in hemodialysis patients hospitalized on suspicion of infection. Like those with bacteremia or diabetes mellitus, hemodialysis patients with hypoalbuminemia require careful management of their infections.

Multiple studies have shown that Staphylococcus aureus bacteremia (SAB) has been a major cause of death in hemodialysis patients. We examined whether SAB is a risk for mortality among chronic hemodialysis patients in Japan where the standard vascular access is arteriovenous fistula (AVF). This was a multicenter, retrospective study of maintenance hemodialysis patients with bloodstream infection (BSI) from 2011 to 2013 at tertiary care centers in Japan. The endpoint was hospital mortality. Our cohort contained 32 SAB cases (14 MRSA and 18 MSSA) and 42 non-SAB cases. Hospital mortality was higher among SAB cases than non-SAB cases (46.9% vs. 23.8%, P = 0.038). In patients with BSI, SAB was significantly associated with hospital mortality after adjustment for potential confounders, including type of vascular access (OR 3.26). S. aureus was the leading cause of BSI and hospital mortality among this cohort. Therefore, initial empiric treatment should cover for S. aureus.

A 72-year-old woman with a history of type 2 diabetes mellitus was brought to the ER with metforminassociated lactic acidosis. She received continuous hemofiltration and hemodialysis, but the laboratory analyses showed no improvement. She died 11 hours after admission. Metformin is minimally bound to proteins and is readily dialyzable, but a prolonged period of dialysis is required, because metformin has a very large distribution volume and is distributed to multiple compartments. The peak blood metformin level was 432 mg/L in this case, which is one of the highest metformin concentrations ever reported, and eight hours of hemodialysis were not sufficient to reduce the serum level.