B-type natriuretic peptide and interdialytic fluid retention are independent and incremental predictors of mortality in hemodialysis patients.

AIM: Management of fluid homeostasis remains a major challenge in hemodialysis patients. We aimed to establish whether the cardiac strain marker B-type natriuretic peptide (BNP) could help to identify hypervolemic patients at increased risk of death. METHODS: BNP levels were determined before dialysis in the entire HD population at our institution (n = 57). IDWG and BNP were stratified above or below 1.5 kg or the median value, respectively. All patients were prospectively followed for 35 months. The influence of IDWG and BNP on mortality was assessed with a Cox proportional hazards model, adjusted for each other, as well as for demographics, comorbidities, cardiac function, residual diuresis, dialysis duration and efficiency and complications of renal failure. RESULTS: Median BNP was 303 (135 - 692) and 21 (36%) patients displayed an average IDWG below 1.5 kg. During follow up a total of 25 (44%) patients died, 5 (26%) in the low IDWG group and 20 (53%) in the high IDWG group (adjusted hazard ratio (adjusted HR) 5.31 95% CI (1.47 - 19.1), p = 0.011). In the low BNP group 7 (25%) patients died and in the high BNP Group 18 (62%) patients died (adjusted HR 3.53 95 CI (1.37 - 9.09), p = 0.009). When both factors were considered simultaneously, patients with low BNP and low IDWG had an 11 times lower risk of death compared to patients with high BNP and high IDWG (HR. 0.08 95% CI (0.01 - 0.6129, p = 0.015). CONCLUSIONS: BNP and IDWG are independent and incremental predictors of mortality in HD patients. These findings suggest that BNP guided fluid management could improve survival in these patients.

Patients' representations of their end-stage renal disease: relation with mortality.

BACKGROUND: Self-regulation theory explains how patients' illness perceptions influence self-management behaviour (e.g. via adherence to treatment). Following these assumptions, we explored whether illness perceptions of ESRD-patients are related to mortality rates. METHODS: Illness perceptions of 182 patients participating in the NECOSAD-2 study in the period between December 2004 and June 2005 were assessed. Cox proportional hazard models were used to estimate whether subsequent all-cause mortality could be attributed to illness perception dimensions. RESULTS: One-third of the participants had died at the end of the follow-up. Mortality rates were higher among patients who believed that their treatment was less effective in controlling their disease (perceived treatment control; RR = 0.71, P = 0.028). This effect remained stable after adjusting for sociodemographic and clinical variables (RR = 0.65, P = 0.015). CONCLUSIONS: If we consider risk factors for mortality, we tend to rely on clinical parameters rather than on patients' representations of their illness. Nevertheless, results from the current exploration may suggest that addressing patients' personal beliefs regarding the effectiveness of treatment can provide a powerful tool for predicting and perhaps even enhancing survival.