Introduction: Hemodiafiltration is currently one of the most effective techniques of extra-renal purification but results in an increase of albumin loss in dialysate. We aimed to determine the factors associated with albumin loss during post-dilution hemodiafiltration, compare an "automatic" mode of infusate flow control versus a "manual" control, and assess the potential nutritional impact.
Methods: This prospective observational study included all hemodialysis patients in our institution who underwent post-dilution hemodiafiltration 3 times a week on a Fresenius 5008 for at least 2 months. At each session, albumin content was measured in a representative effluent dialysate volume. The automatic mode of the Fresenius 5008 was used for automatic infusate flow control.
Results: In all, 18 patients (mean age 60.7 ± 15 years) underwent 85 post-dilution hemodiafiltration sessions. The mean albumin loss was 3134 ± 2450 mg/session. Albumin loss was significantly affected by infusate flow, infusate volume, transmembrane pressure and ultrafiltration volume. The loss was greater with Toraysulfone and FX 1000 membranes rather than FX 80 or FX 100 membranes. With AutoSub rather than manual control, infusate flow was greater (P<.001), transmembrane pressure was higher (P = .004), and the albumin loss was greater (P = .010). However, there was no correlation between albumin loss and nutritional variables.
Conclusions: Albumin loss during post-dilution hemodiafiltration was correlated with increased transmembrane pressure and infusate flow, especially AutoSub flow control, and type of membrane. However, this loss, when moderate, did not seem to affect nutritional aspects and should not limit the use of hemodiafiltration.