RESUMO
INTRODUCTION: Prior observational studies conducted in the hemodialysis population have suggested a reverse association between dialysis-unit blood pressure (BP) and mortality. The present study aimed to investigate the prognostic association of home versus dialysis-unit BP with all-cause mortality in hemodialysis patients. METHODS: At baseline, 146 patients receiving maintenance hemodialysis underwent assessment of their BP with the following methods: (i) 2-week averaged routine predialysis and postdialysis BP measurements; (ii) home BP monitoring for 1 week that included duplicate morning and evening BP measurements with the use of validated devices. RESULTS: Over a median follow-up period of 38 months (interquartile range [IQR]: 22-54), 44 patients (31.1%) died. In Kaplan-Meier curves, predialysis and postdialysis systolic BP (SBP) was not associated with all-cause mortality, while home SBP appeared to be of prognostic significance (log rank p = 0.029). After stratifying patients into quartiles, all-cause mortality was lowest when home SBP was ranging from 128.1 to 136.8 mmHg (quartile 2). In univariate Cox regression analysis, using quartile 2 as a referent category, the risk of all-cause mortality was 3.32-fold higher in quartile 1, 1.53-fold higher in quartile 3 and 3.25-fold higher in quartile 4. The risk-association remained unchanged after adjustment for several confounding factors (adjusted hazard ratio: 4.79, 1.79, 3.63 for quartiles 1, 3, and 4 of home systolic BP, respectively). CONCLUSION: Our findings suggest that among hemodialysis patients, 1-week averaged home SBP is independently associated with all-cause mortality. In sharp contrast, SBP recorded either before or after dialysis over 2 weeks is not prognostically informative.
RESUMO
INTRODUCTION: Assessment of dry-weight among patients on dialysis is challenging in the absence of reliable markers to define fluid overload (FO). This study aimed to explore the value of two simple clinical signs, pedal edema, and crackles at pulmonary auscultation, in diagnosing hypervolemia, using bioimpendence spectroscopy (BIS) as reference standard. METHODS: In a cohort of 107 asymptomatic dialysis patients, FO was assessed with physical examination and BIS shortly before the mid-week dialysis session. Patients were also asked to perform home blood pressure (BP) monitoring with a validated, automatic device (HEM-705, Omron, Healthcare) for 1 week in order to determine their BP outside of dialysis. FINDINGS: Patients within the high tertile of predialysis FO had longer dialysis vintage, lower serum albumin and higher home systolic BP, despite the more aggressive treatment with a higher average number of antihypertensives daily. In receiver-operating-characteristic (ROC) curve analysis, pedal edema (area under curve [AUC]: 0.534; 95% confidence interval [CI]: 0.416-0.651) and pulmonary crackles (AUC: 0.551; 95% CI: 0.432-0.671) had limited accuracy in detecting excess predialysis FO > 2.2 L. The agreement of pedal edema (k-coefficient: 0.065) and pulmonary crackles (k-coefficient: 0.122) with BIS-derived FO was poor. In multivariate linear regression analysis, longer dialysis vintage (ß: 0.306, p < 0.001) and higher home systolic BP (ß: 0.287, p < 0.01) were the two factors that were associated with predialysis FO. CONCLUSIONS: This study showed that among asymptomatic dialysis patients, pedal edema and pulmonary crackles in physical examination had limited discriminatory power in detection of FO, as assessed with the method of BIS.
