Background: Large convective volumes are recommended for online haemodiafiltration (HDF) to maximize solute removal. There has been little systematic evaluation of factors that determine convective volumes in routine clinical practice.
Methods: In the present study, potential patient- and treatment-related determinants of convective volume were analysed in 235 consecutive patients on post-dilution HDF using multivariable linear regression models. All patients (age 64 +/- 14 years; 61% male) participated in the ongoing CONvective TRAnsport STudy (CONTRAST). Additionally, differences in convective volumes between dialysers were evaluated.
Results: The mean convective volume was 19.4 +/- 4.0 L (+/-SD) per treatment, with a large variation between the participating centres (centre means ranging from 13.4 +/- 0.9 L to 24.5 +/- 0.12 L, +/- SE). The mean filtration fraction of the blood flow was 25.9 +/- 3.6. In the multivariable analysis, factors that were significantly related to convective volume were haematocrit [inversely, regression coefficient (B) = -1.4 +/- 0.4 L per 10%], serum albumin (positively, B = 1.0 +/- 0.4 L per 10 g/L), blood flow rate (positively, B = 0.4 +/- 0.04 L per 10 mL/min) and treatment time (positively, B = 5.1 +/- 0.4 L/h). In addition, significant differences between dialysers were observed, likely explained by different operational conditions.
Conclusions: Apart from increasing the treatment time and blood flow rate, convective volumes could be optimized by increasing the filtration fraction in each individual, provided that transmembrane pressures are well within safe limits. The precise role of dialyser characteristics on maximal achievable convective volumes in clinical practice is a topic for further research.