**2. Convective dose concept—from minimal to optimal dosage**

According to the EUDIAL group recommendations, convective dose is the total volume of ultrafiltration achieved per HDF session (L/session) and summed to week (L/week), standar‐ dized to postdilution convection volume for the various other dilution modalities (i.e., pre‐, mixed, or mid‐HDF), taking differences in the fluid volumes into consideration [22]. This is an easy and clinically relevant surrogate indicator of the convective component. Knowing that only hemodialyzers containing membranes that are highly permeable to both water (Kuf>50) and solutes (sieving coefficient β2‐microglobulin [β2M] >0.6) are used for HDF, it is possible to calculate the convective component of solute clearance. An alternative proposal was to focus more on the biological effect of middle‐molecule removal using biomarkers that reflect the convective action of the HDF. Among uremic toxins, β2M, a 12‐kDa peptide, seems to be the most clinically relevant representative of the middle‐molecule uremic toxins and strongly implicated in both morbidity and mortality of CKD patients.

Following international guidelines, urea Kt/V has been established as the principle dialysis dose quantifier and is regularly used as a quality‐control tool for treatment delivery. To be valid, this approach requires certified methodology (appropriate urea sampling method and time, suitable formula) and appropriate timing of measurements (minimum of once monthly) to capture early deviation in dialysis efficacy. The ionic dialysance (iK) measurement embed‐ ded in some HD monitors is an option. This offers an interesting and cost‐effective alternative that may be performed routinely at each session providing a true continuous quality‐control tool. Using then the simplified concept of iKt, where iK stands for the average ionic dialysance measured by HD monitor and t stands for the duration of the dialysis session, one can estimate the total diffusive dose delivered per session (L/session).

The effective total dialysis dose (i.e., diffusive and convective) delivered to the ESKD patient can be easily assessed. Convective dose as estimated by total ultrafiltered volume per session (L/session) could be considered as a complementary component of the diffusive dose delivered and estimated by iKt (L/session).

A systematic review of studies [retrospective cohorts, prospective randomized controlled trial (RCT)] exploring the relationship of convective dose and patient outcome has shown that survival benefit is observed only when a minimum threshold ultrafiltration volume has been delivered [23]. The critical ultrafiltration volume per session (or per week) required for better patient outcome is between 20 and 22 L/session for a typical European ESKD patient. More recently, it has been shown that total ultrafiltered volume per session (or per week) acted as a continuous variable mimicking a sigmoidal dose‐response curve [24].

The normalization of the convective dose to the patient's anthropometric characteristics has been proposed to match with patient metabolic needs and also to facilitate the generalization of this relationship with different patient profiles (e.g., Asian and American). In fact, such normalization attempts using different scalings (i.e., body weight, height, body surface area, and total body water) have been adopted in different studies, including the Estudio de Supervivencia de Hemodiafiltración On‐Line (ESHOL) [25] and the individual personal data meta‐analysis that was part of the European pooled project [26]. Interestingly, none of these scaling factors have enhanced sensitivity in predicting the relative risk (RR) of mortality. Crude convection volume per session (or per week) or convection volume scaled to body surface area or total body water tend to perform best in predicting survival benefits or mortality risks where HDF is concerned.

### **3. Factors affecting convective dose**

Best clinical practices are essential in daily practice to achieve optimal ultrafiltration flow and the total ultrafiltered volume targeted. Schematically, three basic components need to be considered: patient‐dependent factors, prescription‐dependent factors, and technical or machine‐dependent factors.

#### **3.1. Patient‐dependent factors**

#### *3.1.1. Vascular access*

Following international guidelines, urea Kt/V has been established as the principle dialysis dose quantifier and is regularly used as a quality‐control tool for treatment delivery. To be valid, this approach requires certified methodology (appropriate urea sampling method and time, suitable formula) and appropriate timing of measurements (minimum of once monthly) to capture early deviation in dialysis efficacy. The ionic dialysance (iK) measurement embed‐ ded in some HD monitors is an option. This offers an interesting and cost‐effective alternative that may be performed routinely at each session providing a true continuous quality‐control tool. Using then the simplified concept of iKt, where iK stands for the average ionic dialysance measured by HD monitor and t stands for the duration of the dialysis session, one can estimate

The effective total dialysis dose (i.e., diffusive and convective) delivered to the ESKD patient can be easily assessed. Convective dose as estimated by total ultrafiltered volume per session (L/session) could be considered as a complementary component of the diffusive dose delivered

A systematic review of studies [retrospective cohorts, prospective randomized controlled trial (RCT)] exploring the relationship of convective dose and patient outcome has shown that survival benefit is observed only when a minimum threshold ultrafiltration volume has been delivered [23]. The critical ultrafiltration volume per session (or per week) required for better patient outcome is between 20 and 22 L/session for a typical European ESKD patient. More recently, it has been shown that total ultrafiltered volume per session (or per week) acted as a

The normalization of the convective dose to the patient's anthropometric characteristics has been proposed to match with patient metabolic needs and also to facilitate the generalization of this relationship with different patient profiles (e.g., Asian and American). In fact, such normalization attempts using different scalings (i.e., body weight, height, body surface area, and total body water) have been adopted in different studies, including the Estudio de Supervivencia de Hemodiafiltración On‐Line (ESHOL) [25] and the individual personal data meta‐analysis that was part of the European pooled project [26]. Interestingly, none of these scaling factors have enhanced sensitivity in predicting the relative risk (RR) of mortality. Crude convection volume per session (or per week) or convection volume scaled to body surface area or total body water tend to perform best in predicting survival benefits or mortality risks where

Best clinical practices are essential in daily practice to achieve optimal ultrafiltration flow and the total ultrafiltered volume targeted. Schematically, three basic components need to be considered: patient‐dependent factors, prescription‐dependent factors, and technical or

the total diffusive dose delivered per session (L/session).

continuous variable mimicking a sigmoidal dose‐response curve [24].

and estimated by iKt (L/session).

80 Advances in Hemodiafiltration

HDF is concerned.

machine‐dependent factors.

**3. Factors affecting convective dose**

The achievement of "high" blood flows (>350 mL/min) depends on the type of vascular access (i.e., central venous catheter, arteriovenous fistula, or graft) and on the provision and mainte‐ nance of vascular access patency. Vascular access is defined as inadequate when an extracor‐ poreal blood flow of at least 300 mL/min cannot be reached. Basically, patients treated with online HDF require a vascular access capable of delivering a consistent extracorporeal blood flow between 350 and 400 mL/min or higher. Such extracorporeal blood flows could be reached with large bore tunneled central venous catheters and with arteriovenous fistulas or grafts delivering access flows of ≥500 mL/min. High blood flow is essential to ensure that a sufficient amount of blood is processed during the treatment session. Vascular access flow performances and needle sizes assume a fundamental role in preserving the advantages of applying a high convection volume.

#### *3.1.2. Individual patient criteria*

Hematocrit and protocrit (the volume fraction of plasma proteins that may be calculated as the product of 0.000718 and the total protein concentration of plasma proteins in g/mL [22]) affect negatively the plasma water volume and plasma water flow. High hematocrit (e.g., resulting from anemia management and hemoglobin targets) and high total protein concen‐ tration (e.g., due to the particular nutritional and inflammatory status of the patient) enhance viscosity tremendously, reduce ultrafiltration capacity (filtration fraction), and provide unfavorable conditions for ultrafiltration flow. The additional ultrafiltration applied to achieve a given weight loss is an additive factor that may affect total ultrafiltration. Finally, a dynamic interaction between blood flow (shear rate and shear stress), blood components (hematocrit and protocrit), filtration fraction (ultrafiltration flow to blood flow ratio), and membrane surface (protein layer formation) is crucial to facilitate ultrafiltration flow. Many of these factors are taken into consideration in modern dialysis machine technology.

#### **3.2. Prescription‐dependent factors**

#### *3.2.1. Blood flow*

The effective extracorporeal blood flow delivered is a fundamental determinant of all extrac‐ orporeal cleansing therapies. Among factors that determine the efficiency with which uremic solutes are removed during dialysis, extracorporeal blood flow, whether instantaneous (flow rate) or cumulative (total blood volume processed over the session), is the most critical. Treatment efficiency assessed either in terms of solute clearance or solute mass removal is then dependent on the total blood "processed" within a dialysis session.

High blood flows are critical in HDF, as they have a dual action, one being to maximize the amount of solute removed and the other being to preserve membrane permeability by retarding the formation of a protein layer on the membrane (secondary layer formation). The choice of needle size matters in HDF. Following Poiseuille's law, needle size is a barrier to high blood flow: 15‐gauge needles (optimally 14‐gauge needles) are required to sustain blood flow of ≥350 mL/min at an acceptable pressure regimen.

#### *3.2.2. Choice of hemodiafilter*

The choice of a specifically designed hemodiafilter is important to optimize ultrafiltration flow and prevent dysfunction or alarms occurring on the HDF monitor due to hemorheological changes in the dialyzer. Preferred hemodiafilter features, apart from being equipped with highly permeable membranes, should favor the following to reduce internal convective processes: low blood flow resistance (large lumen diameter of fibers (e.g., 200 μm), short length of dialyzer housing, and increased number of fibers per sectional surface area.

#### *3.2.3. Anticoagulation*

Methods of anticoagulation will not addressed here in detail. However, anticoagulation is required to prevent thrombosis of extracorporeal blood circuit and to ensure a safe and efficient HDF session. Different kinds of antithrombotic agents can be used systematically by intrave‐ nous (IV) injection.

Unfractionated heparin (UH) is administered as an IV bolus dose (30–50 IU/kg) at the start of the HDF session followed by continuous IV infusion (500–700 IU/h). The heparin dosing regimen for HDF does not differ from that in regular HD.

The use of low molecular weight heparin (LMWH) is nowadays favored by many centers because of its ease of use and its better risk profile. In this case, LMWH should be preferably administered IV into the venous line and not into the arterial line to prevent significant loss during the first hemodialyzer passage.

#### *3.2.4. Treatment time prescription*

The prescription of HDF treatment time duration and frequency is usually based on the patient's metabolic needs, extracellular fluid management, and cardiovascular and session tolerance. Increasing both the duration and frequency will facilitate the delivery of a high volume of ultrafiltration. A pragmatic approach is to establish a suitable convective dose for a given patient and to increase the treatment time according to the limitation of effective blood flow delivery.

#### **3.3. Technical or machine‐dependent factors**

#### *3.3.1. Transmembrane pressure management*

Achieving high ultrafiltration rates and targeted ultrafiltration volumes can be challenging and requires the careful management of the transmembrane pressure (TMP) according to the treatment modality selected. In attempting to achieve high ultrafiltration volumes, hemocon‐ centration within the filter commonly results in high TMP, triggering pressure alarms and potentially causing cell damage. Basically, the augmented protein layer formation occurring naturally at the blood‐membrane surface during the course of the HDF session fouls the membrane pores and reduces the membrane's hydraulic permeability. In combination with the increased oncotic pressure (total protein increase) along the fibers, this tends to reduce the ultrafiltration flow. Faced with situations of high TMP increase and/or pressure alarms, nursing staff manually reduce the substitution flow and thus reduce the chances of achieving the targeted ultrafiltration volume.

#### *3.3.2. Machine options*

choice of needle size matters in HDF. Following Poiseuille's law, needle size is a barrier to high blood flow: 15‐gauge needles (optimally 14‐gauge needles) are required to sustain blood flow

The choice of a specifically designed hemodiafilter is important to optimize ultrafiltration flow and prevent dysfunction or alarms occurring on the HDF monitor due to hemorheological changes in the dialyzer. Preferred hemodiafilter features, apart from being equipped with highly permeable membranes, should favor the following to reduce internal convective processes: low blood flow resistance (large lumen diameter of fibers (e.g., 200 μm), short length

Methods of anticoagulation will not addressed here in detail. However, anticoagulation is required to prevent thrombosis of extracorporeal blood circuit and to ensure a safe and efficient HDF session. Different kinds of antithrombotic agents can be used systematically by intrave‐

Unfractionated heparin (UH) is administered as an IV bolus dose (30–50 IU/kg) at the start of the HDF session followed by continuous IV infusion (500–700 IU/h). The heparin dosing

The use of low molecular weight heparin (LMWH) is nowadays favored by many centers because of its ease of use and its better risk profile. In this case, LMWH should be preferably administered IV into the venous line and not into the arterial line to prevent significant loss

The prescription of HDF treatment time duration and frequency is usually based on the patient's metabolic needs, extracellular fluid management, and cardiovascular and session tolerance. Increasing both the duration and frequency will facilitate the delivery of a high volume of ultrafiltration. A pragmatic approach is to establish a suitable convective dose for a given patient and to increase the treatment time according to the limitation of effective blood

Achieving high ultrafiltration rates and targeted ultrafiltration volumes can be challenging and requires the careful management of the transmembrane pressure (TMP) according to the treatment modality selected. In attempting to achieve high ultrafiltration volumes, hemocon‐ centration within the filter commonly results in high TMP, triggering pressure alarms and

of dialyzer housing, and increased number of fibers per sectional surface area.

of ≥350 mL/min at an acceptable pressure regimen.

regimen for HDF does not differ from that in regular HD.

during the first hemodialyzer passage.

**3.3. Technical or machine‐dependent factors**

*3.3.1. Transmembrane pressure management*

*3.2.4. Treatment time prescription*

flow delivery.

*3.2.2. Choice of hemodiafilter*

82 Advances in Hemodiafiltration

*3.2.3. Anticoagulation*

nous (IV) injection.

New features to optimize HDF performances and achieve optimal convective doses are currently available. The adjustment of the substitution mode was the most obvious and primary focus of investigation. When postdilution is problematic, switching to pre‐, mixed, or mid‐dilution mode may be a viable alternative. In all these cases, the targeted convection volume needs to be adjusted for the dilution factor corresponding to the HDF modality chosen (e.g., x2 for predilution HDF or x1.5 for mid‐dilution HDF in manual prescription) to match efficacy with the postdilution mode.

New technical features involving specific software algorithms are currently being implement‐ ed and tested in new online HDF machines. Basically, the idea is to provide an automatic ultrafiltration control system to reduce membrane fouling and ensure maximal ultrafiltration flow considering basic operational conditions of blood flow, hemorheological conditions, and prescription setting. Schematically, the system avoids excessive hemoconcentration by the continuous adaptation of the substitution flow according to changes in blood viscosity within the dialyzer as identified by signal analyses of the pressure pulses transmitted from the peristaltic blood pump. Signal analysis is conducted several times per minute, and the substitution rate is automatically adapted based on pressure pulse attenuation and cross‐ membrane pressure assessment. Using such automatic control systems, it is possible to increase the ultrafiltered volume per session by 10 to 20% without harming the patient, filter, or cells.

#### *3.3.3. Other machine‐related variables*

As mentioned previously, treatment time is one of the factors limiting the increase of "dialysis dose". In the current models of delivering dialysis in dialysis units, based on shifts assigned to nurses with a ratio of one nurse to three to five patients, any further increase of treatment time has an associated cost. Out of the 5.5 to 6 h of the shift length for each single nurse, it is possible to deliver a median of 4 h of treatment to three to five patients. There are two ways to achieve a cost‐neutral increase in treatment time. The first is to reduce the time needed to prepare the dialysis equipment before initiating treatment. Faster disinfection of the dialysis machine is achievable with more simple hydraulic components and/or more effective disin‐ fection processes. Also, the disinfection of the machine surface can be simpler, safer, and faster; this is achievable with especially designed equipment surfaces having as few discontinuities as possible. The second way is to reduce the number and complexity of nurse interventions. This approach may allow nurses to treat more patients, thus decreasing the cost of treatment for each patient per hour. Consequently, it could be possible to increase the time of the shift and the treatment time in a cost‐neutral manner. Is this possible? According to Tsobanelis et al. [27], new dialysis equipment under testing in HDF mode for fistula patients had a 24% reduction in the number of major handling steps compared to the previous dialysis machine from the same company.

Other aspects to be considered when targeting an increase of patient to nurse ratio are those related to safety. The risk of harm has to be reduced as much as possible, and one option is to reduce this risk by the improved design of the equipment. As an example, we can again mention the recent experience of Tsobanelis et al. [28], who reported that the tested new equipment had 27% fewer major process steps and touch points critical to hygiene compared to the current machine version. The authors found it particularly noteworthy that it was possible to avoid disconnection of the arterial line from the arterial needle with the redesigned process of blood reinfusion at the end of the session.

Finally, in the time of green economy, the authors also highlighted that the switch from an infusion line to an integrated infusion port reduced the volume of contaminated waste. The major source of waste relates to the disposables (bloodlines and dialyzers), so that newly designed, integrated disposables can facilitate a reduction of disposable waste, for example, of approximately 0.2 kg/session as reported by Schleser et al. [29], which, given an average of 10,000 treatments yearly delivered by a dialysis unit, translates into 2000 kg less waste. In terms of carbon footprint, one should also consider the consumption of water and energy and how this can be limited, for example, by having a more efficient water treatment system.
