SAT-209 The significance of variations in albumin level in the period leading up to dialysis initiation as a predictor of prognosis among the elderly CKD patients

Abstract

The prognosis for elderly patients after the initiation of dialysis is not always favorable. Several epidemiological studies have demonstrated associations between variations in biochemical findings and the prognosis of dialysis patients. The aim of this study was to investigate the significance of variations in albumin level in the period leading up to dialysis initiation among late-stage elderly patients aged 75 or over at the start of hemodialysis. This study was conducted as a single center, retrospective, observational study. The late-stage elderly patients who started hemodialysis at our hospital during the period from April 1, 2011 to March 31, 2014 were included. The yearly trends for serum albumin level and the intra-individual variations in level were examined over a period of two years before the initiation of hemodialysis. The observation period was divided into two phases, the late period (1-12 months immediately prior to starting hemodialysis) and the early period (13-24 months before hemodialysis). Other values, as well as serum albumin level, were compared between the survival group and the death group during and after dialysis initiation. A total of 91 late-stage elderly patients, with an average age of 80.4 (77 - 84.5) years and average duration of dialysis of 34.1 (24 - 77.5) months, were enrolled. Fifteen patients died after dialysis initiation. There were no significant differences in the mean serum albumin values between two groups, the survival group and the death group, in either period. On the other hand, a low albumin level at the time of dialysis initiation was associated with a poor prognosis after dialysis initiation: the survival group = 3.3 (2.2 - 4.3) vs. the death group = 2.7 (2.2 - 3.7) (p < 0.01, Mann-Whitney’s U-test). Variation in albumin level showed an upward trend and accelerated toward dialysis initiation in both groups; and this upward trend was significant in the death group (mean slope: 0.083 ± 0.028/month). The variations in albumin level were significantly larger in the death group than in the survival group in both periods (the late period: the survival group = 0.052 (0.037 - 0.064) vs. the death group = 0.078 (0.046 - 0.10), the early period: the survival group = 0.036 (0.023 - 0.047) vs. the death group = 0.057 (0.031 - 0.074) (p < 0.01, Mann-Whitney’s U-test). Univariate analysis using the values obtained in the late preceding period revealed that the status of activities of daily life (ADL), serum phosphorus level, and variation in albumin level were significant factors related to the prognosis after dialysis initiation. The yearly trend regarding the intra-individual variations in albumin level was considered to be an independent factor related to the prognosis after dialysis initiation among late-stage elderly patients. These results suggest that a future prospective study is required to further investigate the relevant factors.