Dr Jonathan Round, PICU Consultant

If you haven’t read it yet, you might want to start with Part 1 – click here

The CVVH circuit

Starting with the circuit (figure 1) used in almost every situation for CVVH, blood is taken out and returned using a large bore double lumen catheter. This could be a temporary cannula sited using a Seldinger technique or a permanent catheter. Its most important attribute is its ability to handle high blood flow rates. The more blood that passes through the filter, the faster its chemistry can be corrected. Flow rates of below 90 ml/min are not much use even for an infant, and a large adolescent will need 200 ml/min. In context, the entire blood volume passes through the filter every 5-10 minutes. This catheter will obviously need to be in a very large vein.

Figure 1: Schematic of CVVH

The blood is then sucked by the blood pump into the circuit. It needs to come from the proximal side (ie the port of the catheter nearest the outside of the body) of the catheter to avoid ‘re-circulation’ which we will get on to later. By definition this is called the ‘arterial’ limb – a nomenclature relic from when dialysis used arterial pressure to drive flow – although these days the blood is a venous as could be. ‘Helpfully’ this access port on the catheter is red. The blood is sucked out of the vein by the blood pump. The speed of the blood pump (also called head speed) defines the flow rate.

A pressure monitor in the pipe from the catheter informs the user of how difficult it is for the pump to suck blood out of the vein. If there is a problem with the catheter or vein, this may become very negative, over -150 mmHg. This is the ‘access pressure’.

After the blood pump the piping now leads to the filter itself. But before it gets there, another port allows dilution fluid to be infused into the circuit. It might seem a bit counter-intuitive to add fluid to the blood, when the overall plan is to get fluid out of it, but this fluid is easy to get out again at the membrane, a little further down the circuit. In Part 1, we looked at why this ‘pre-dilution’ was a good idea above. But by adding fluid before the filter, the blood constituents are diluted, so when fluid is removed at the membrane, it gets it back to the sort of levels of Hb that you would want in blood.

Now the volume and nature of that dilution fluid is crucial to the overall function of CVVH.

Fluid and Electrolyte Management

How CVVH handles volume is simplest to understand. The CVVH circuit is closed: blood comes in, blood goes out; dilution is added, filtrate is removed. So the difference between the dilution and filtrate volumes gives you the difference (or balance) between blood coming out and going back into the body. Both dilution and filtrate volumes are measured at the pumps that push the dilution fluid in and suck out the filtrate. The machine does the hard work and calculates the balance, usually as a running total for a four hour spell. Please note that this balance is not the same as the patient’s overall fluid balance which also has the components of IV or oral fluid intake, urine, stool and insensible losses.

The nature of the dilution fluid is also important. As the aim of CVVH might be to remove K+ ions, lactate or another solute, yet these are in the dilution fluid only to be removed later on at the membrane. The dilution fluid has normal physiological levels of blood solutes (see figure 4), so this dilutes any abnormalities already in the blood. At the filter, fluid and electrolytes are removed as above, leaving the blood passing back into the body with more normal levels. As the blood passes again and again through the CVVH circuit, then blood solute levels trend towards normal. Using a dilution fluid with no K+ at all would correct levels more rapidly, but this runs the risk of overshooting, and is not likely to have practical benefit for the patient.

Coming back to the circuit, as blood comes out of the filter, it will now have had correction both of its overall volume and solutes. Because these changes are small with each pass, they are well tolerated in most cases, but a continuous approach is required to achieve the overall changes needed. The blood then passes into a bubble trap to prevent embolism and then back into the venous limb of the circuit to enter the bloodstream via the distal port of the catheter.

Pressure sensors

There are a couple of other pressure sensors of note, and information from these helps manage the CVVH process. One is before and one after the filter, and the difference between these is the ‘transmembrane pressure’. There is always some resistance to blood passing down the thin membrane tubes in the filter, but as the filter develops microclots inside these tubes, that pressure rises. High transmembrane pressures indicate that the filter is beginning to fail. The pressure after the filter is called the return pressure, and a marker of obstructions to the flow in the catheter as it returns blood to the patient.

If the above narrative suggests that CVVH runs smoothly, that would be false. There are a fair few alarms. There is my guide to them.

Troubleshooting CVVH alarms


CVVH can lead to coagulation through contact of blood with foreign materials, physical damage to red cells in the blood pump and haemoconcentration of blood constituents. Even the pre-dilution process cannot fully prevent this tendency. Even without other factors, children on CVVH may run low platelet counts and raised D-Dimers indicating some coagulation is going on. There is therefore a need for anticoagulation in many cases.

Against this, many very unwell children who are started on CVVH have deranged clotting at the outset, and anticoagulation may max things worse. To manage this, there are several different options and monitoring approaches used. First, avoid harm by deferring active anticoagulation in an unstable patient. However, if the patient is more stable and clotting studies normal, this can be contemplated. The best clue that it is needed is the clotting of the filter. Each filter should last at least 24 hours, and will need replacement at 48 hours. If the filter is becoming clotted before 24 hours, this is a big prompt that anticoagulation is needed.

There are three approaches used – low dose heparin, citrate and prostaglandin. Each has its proponents and detractors, and each PICU will have its favourite. Heparin is best known, but some children will develop Heparin Induced Thrombocytopenia (HIT) and will have some systemic anticoagulation. Citrate functions by binding calcium, blocking coagulation, and theoretically is metabolised before blood re-enters the circulation. However calcium will often be run continuously to counteract this iatrogenic hypocalcaemia. Prostaglandin also has its advocates, but can cause a low grade fever.

Anticoagulation approaches


I have avoided writing about why you might want to start CVVH off until now, mostly because this is pretty simple and one rarely gets to make this call, particularly as a trainee.

As with every intervention, CVVH is used when its not inconsiderable downsides are outweighed by bigger downsides of not using CVVH. Broadly these get into three categories – solutes you don’t want, fluid you don’t want and fringe benefits. And CVVH doesn’t cure the patient – it might stop the patient dying while you do something that does fix them.

Solutes you don’t want

Potassium at levels that cause dysrhythmias (at least 6.5 mmol/L) can be treated with CVVH, and this would be started when other measures aren’t working (beyond this article).

Severe metabolic acidosis, with pH below 7.1 at least can also be corrected by CVVH. The idea is that the acid (eg lactate, or an acid accumulating in a metabolic disorder) is removed at the filter to be replaced by bicarbonate. Always one must ask the question whether the acidosis is going to be a bigger problem than the CVVH.

Hyper-ammonaemia can also accumulate in metabolic conditions and is bad for tissues when at very high levels. It can be filtered (or dialysed) out.

A number of other toxins, especially water soluble poisons that are not protein bound could also be filtered out. This is an unusual indication for CVVH. Alcohols, lithium, aspirin or anticoagulant overdoses could be cleared in this way, but only when life threatening and the body is not clearing the agent using its own systems.

Fluid you don’t want

Lots of children and infants will become fluid overloaded during the course of a critical illness, but mostly this resolves when these patients recover. Sometimes diuretics are useful to manage the consequences, such as pulmonary oedema. Rarely renal issues prevent the diuretics doing their job, and this is where CVVH might be of use. My own preference is to ask whether the patient would develop pulmonary oedema and therefore need ventilation with their fluid overload. If the answer is yes, CVVH might be the answer.

Fringe benefits

There have been a lot of ideas relating to CVVH and marginal gains. This piece is not anywhere near long enough to cover the potential evidence supporting using CVVH in sepsis to remove unhelpful cytokines, but suffice to say CVVH is not routine management of severe sepsis.

Similarly CVVH reduces central temperature and this may be useful in hyperpyrexic situations, such as burns or severe sepsis. There are plenty of other ways to cool a patient that are less invasive, so using CVVH to manage temperature alone would not be smart.

Getting going

Once a decision has been made to start CVVH, decisions have to be made about the access, the circuit, the filter and the ‘prescription’ of CVVH to be used. There are trade offs and compromises to be made: the larger the catheter, the better the flow, but the more issues with clotting; the bigger the filter, the better the potential for haemofiltration, but the greater the haemodynamic disturbance as CVVH is initiated. Simply use the local protocol – every PICU will have one, and if it doesn’t it should not be trying to deliver CVVH.

The whole process has multiple steps, but these can be done in parallel. The catheter needs to be sited and the circuit primed with blood or crystalloid, so it makes sense to do these at the same time. The location of the catheter is important. The best flow is probably from the right internal jugular vein, but this may not be accessible. Remember also that continuous infusions of vasoactive drugs should not be near the catheter, or CVVH will cause CVS instability. Sometimes this requires rewiring and moving of lines around the body.

To start off, when everything is connected as in the diagram (fig. 3), simply turn on the blood pump and circulate blood through the circuit slowly. In unstable children, this may precipitate further wobbles in blood pressure. This may be as the contents of the circuit reenter the bloodstream, because of changes in temperature or blood contents. Once stability is achieved, gradually increase the pump speed to that recommended in the unit protocol for the size of child. At this stage you are not actually doing anything productive for the child, but you are in a position to do so.

Next, start filtration, usually with a neutral balance. How much is expressed in various ways – as ml/kg or as a percentage of the blood flow, which may be easiest to understand. Here, one might start with 5% pre-dilution ie 5% of the blood volume per minute is put into the circuit as ‘replacement fluid’ before the membrane: 5mls/min if the blood pump is running at 100ml/min. As this is currently running a neutral balance, 5 ml/minute is then pulled off as filtrate. CVVH has now begun. However at 5% this is not likely to make a big difference.

Now, slowly increase the pre-dilution percentage to 10 or 15%. You can go higher, according to protocol. If things are still stable, then you can start to remove fluid. Here more filtrate is pulled off than pre-dilution added. Amounts should initially be small, but as this is substituting for renal output, think of similar numbers: 1-4 ml/kg/hour. Getting to this point will take many hours.

By this stage you should be making a significant difference to the electrolytes and over days, to any fluid overload. It may not go so well though, and this is where troubleshooting skills come in.

Not only but also worth a mention


it might at first glance look the same, but the process is very different. But the membrane has much smaller holes and there is no pressure gradient across it. Instead small molecules diffuse across the membrane to balance the concentration gradients each side. Using a dialysate running countercurrent to the blood flow, this is a way of rapidly normalising blood electrolytes, and at the same time removing water from the body. The rapid changes may be too much for critically ill children.

Peritoneal dialysis

PD is nothing to do with CVVH other than that it is another form of renal replacement therapy. Here, using a permanent or temporary catheter inserted into the peritoneal space, dialysis fluid is instilled into the abdominal cavity. Using the semipermeable membrane of the intestinal serosal wall, fluids and electrolytes are exchanged with the dialysis fluid. After a period of time, the fluid is drained out and the process repeated. This is slow, relatively ineffective, and prone to peritoneal infection. However it is low-tech and can provide short or long term support for failed kidneys.


If you made it to the end of this pair of articles, I hope you have a much better understanding of what CVVH is, how it works, and when to use it. It’s pretty simple, but as with many of the simplest things, doing it well is crucial. Ventilation, renal support, sedation, inotropes and other ICU things are in themselves very useful, but can also be very dangerous. Having a team trained and practiced in using these tools is essential to making sure that they are helpful, not harmful.


  • RRT                         Renal replacement therapy
  • CAVH                       Continuous Arterio-Venous Haemofiltration
  • CVVH                      Continuous veno-venous haemofiltration
  • CVVHD                    Continuous veno-venous haemodialysis
  • PD                           Peritoneal dialysis
  • HIT                          Heparin Induced Thrombocytopenia

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.