Two recent randomized controlled trials (RCT) on online hemodiafiltration (HDF) did not show a treatment effect on patient survival when compared with low- or high-flux hemodialysis. Interestingly, post-hoc (on treatment) analyses from both trials unequivocally showed reduced mortality in the patient group achieving the highest convection volumes. Moreover, a third trial recently found a significant 30% decrease in mortality when HDF was applied with a mean convection volume of 23.7 L per session, which was somewhat higher than the average volumes reached in the aforementioned trials. Altogether, these findings support the concept of a dose-response effect, in which a minimally delivered convection volume is required in order to show a survival benefit.

Hence, the question arises whether high convection volumes are achievable in the majority of patients. The aim of this study is thus to test the following hypothesis: high-volume (>22 liters per treatment) post-dilution on-line hemodiafiltration (HDF) is achievable in the majority (>75%) of patients treated with chronic intermittent hemodialysis. This will be done through the use of a dedicated standardized protocol, in which the three most important determinants of convection volume will be successively optimized: treatment time, blood flow rate and filtration fraction.

Hemodialysis (HD) is the most common used renal replacement therapy worldwide. The main objective is the removal of excess fluid and toxic solutes from the patient. Despite the relative efficiency of modern dialyzers, HD remains inferior to normal kidney function, in part because of the inadequate clearance of so-called 'middle molecules'. In other words, 'uremic toxins' accumulate in chronic HD patients.

The role of middle molecular weight uremic toxins in the pathogenesis of many co-morbid conditions associated with end-stage renal disease is increasingly recognized. Hence, the hypothesis that their enhanced removal could convey a better survival has been proposed. Actually, HDF is the most effective modality in terms of solute removal, because solute transport is achieved by both diffusion (especially for small molecules) and convection (the most important clearance route for middle molecules). Although some convective transport can be achieved with high-flux HD, only in HDF can the amount of filtered volume reach values of 20 liters per session or more.

In line with this thinking, two RCTs comparing HDF to standard hemodialysis with either low- or high-flux membranes were performed. However, treatment assignment did not alter the primary outcome of all-cause mortality in neither of them. Interestingly, post-hoc analyses from both trials unequivocally showed reduced mortality in the patient group achieving the highest convection volumes. Whether these data result from so-called dose-targeting bias (the healthiest patients reaching the highest volumes) cannot be formally excluded, but it is noteworthy that these results remained after correction for known determinants of mortality. Moreover, careful examination of patients' baseline characteristics of the CONTRAST HDF cohort did not reveal a healthier profile among the high convection volume group.

Recently, a third trial found a significant 30% decrease in mortality when HDF was applied with a mean convection volume of 23.7 L per session, which was somewhat higher than the average volumes reached in the aforementioned trials (respectively 20.7 and 19.5 L per session). Altogether, these findings support the concept of a dose-response effect, in which a minimally delivered convection volume is required in order to show a survival benefit.

Hence, the question arises whether high convection volumes are achievable in the majority of patients. In a previous sub-analysis of CONTRAST, it was found that the most important determinants of achieved convection volume were treatment time and blood flow rate. Moreover, it was noted that convection volumes and filtration fraction (defined as the ratio of extracted plasma water flow rate to blood flow rate) differed markedly per participating center, suggesting different practice patterns. Thus, it is plausible that the optimization of these apparently seemingly modifiable factors, on an individual basis through a dedicated standardized protocol, could translate into a higher achieved convection volume.

The aim of this study is to test the following hypothesis: high-volume (>22 liters per treatment) post-dilution on-line hemodiafiltration (HDF) is achievable in the majority (>75%) of patients treated with chronic intermittent hemodialysis.

This study is a prospective observational study, in which all enrolled patients will be treated by post-dilution on-line HDF with the application of a standardized protocol aiming at maximizing the convection volume.

At the beginning of the study, each patient's usual dialysis parameters will serve as starting parameters. In addition, incident HDF patients will start with a filtration fraction of 25 % (or equivalent) on a post-dilution mode.

Then, convection volume will be increased stepwise by successively optimizing the three most important determining factors. First, treatment time will be increased to 4 h, if possible. Second, blood flow will be increased by 50 mL/min per treatment until a value 400 mL/min is reached, provided that pre-specified safety limits are respected. Third, filtration fraction will be increased by 2% per treatment up to a maximum of 33% or the maximally achieved value within safety limits. To take into account different settings between the various dialysis machines used by the participating centers, an easy-to-use conversion table will be provided to the nursing staff, allowing to find the parameter (substitution flow, substitution volume or substitution ratio) corresponding to the desired filtration fraction.

At the end of the step-up protocol, the achieved convection volume will be assessed and compared to the starting value. Additionally, a follow-up period of 8 weeks, in which the maximum values of the targeted parameters will be kept the same, will be observed in order to assess whether the high convective volumes can be maintained for a longer period.

Inclusion Criteria:

• Patients treated with HD or HDF since >6 weeks, 3 times a week;
• Patients able to understand the study procedures;
• Patients willing to provide written informed consent.

Exclusion Criteria:

• Current age < 18 years;
• Severe incompliance to dialysis procedure and accompanying prescriptions (frequency and duration of dialysis treatment and fluid restriction);
• Life expectancy < 3 months due to non-renal disease.

Patients treated with hemodialysis or hemodiafiltration recruited at dialysis units in 6-7 hospitals in the Netherlands