FAQ: How to draw the postdialysis blood for dialysis adequacy assessment.

A. Priester-Coary RN and J.T. Daugirdas MD
University of Illinois at Chicago, IL

WHAT IS THE PROPER TECHNIQUE OF DRAWING POSTDIALYSIS BLOOD?
My recommended technique for drawing the postdialysis blood is as follows:
Required equipment: Vacutainer housing with adapter (to allow for needle-free connection to the tubing attached to the dialysis needle and 3 ml sampling tube.

(See Figure)

At the end of dialysis:
1) Set UF rate to zero
2) Reduce blood pump rate to 50-100 ml/min. One may need to manually adjust the venous pressure limits downward as the blood flow rate is reduced to prevent pump shutoff.
3) Exactly 10 seconds later, shut off the blood pump.
4) Clamp tubing attached to arterial needle and disconnect from arterial bloodline.
5) Attach Vacutainer via adapter to arterial needle tubing.
6) Insert sampling tube into vacutainer such that needle inside the vacutainer housing penetrates the sampling tube stopper and withdraw the sample.
7) Remove the saline line from the T-connector and connect the saline tubing to the very end of the arterial bloodline.
8) Return the blood with saline.
9) Clamp the venous bloodline and venous needle tubing. Connect the venous bloodline to the arterial needle tubing. Turn on the blood pump for a few seconds to return the blood in the arterial needle tubing (2.6 ml) to the patient (alternative is to clear this line with a 10 ml saline-filled syringe).
10) Disconnect as per routine.

The details of the method proposed and the reasons for some of the maneuvers are explained below, along with some possible alternatives. The goals of drawing postdialysis blood are:

a) To limit possibility of staff infection via accidental needlestick
b) To assure maximum possible return of blood from the extracorporeal circuit
c) To assure that the post-dialysis sample is not diluted with dialyzer outlet blood in patients with access recirculation
d) To standardize the URR or Kt/V (which is derived primarily from the post/pre BUN ratio).

a) Needlestick injuries.
Approximately 10-30% of hemodialysis patients will be infected with hepatitis C virus. A smaller percentage will be infected with hepatitis B virus or with HIV, and some patients will be infected with unknown or unclassified transmissible infectious agents. There are two easy methods of accessing the blood post-dialysis. The first is to draw blood from the sampling port, which is usually about 7 ml downstream from the arterial end of the bloodline. Because this sampling port must be accessed via a needle, the risk of needlestick injury is present. However, many dialysis centers draw postdialysis blood from this site. The other possible site to draw blood is from the tubing attached to the arterial needle. This tubing usually has a deadspace of about 2.6 ml. After disconnecting the arterial needle tubing from the arterial part of the blood line, the safest method to draw blood via the needle tubing is to use a "vacutainer" with an adapter tubing, enabling needle-free connection to the needle tubing. An alternative method is to use a syringe. However, blood must still be transferred into a collection tube. If one puts a needle onto the syringe to perform this transfer, which requires perforating the top of the collection tube with the needle, the risk of needlestick injury is increased. If one removes the stopper from the collection tube and fills the tube manually, there is still the risk of blood splash. Furthermore, if the sample tube is nearly filled with blood and the stopper is reinserted, the stopper can be pushed out due to the pressure of the air overlying the blood which was compressed as the stopper was inserted. For these reasons, I prefer the vacutainer method strongly.

b) To assure that the maximum amount of blood is returned.
The saline line should be disconnected and inserted into the end of the arterial bloodline, to ensure that all of the blood in the arterial bloodline is returned to the patient. The deadspace in the needle tubing is 2.6 ml. This should also be returned to the patient, either by connecting it for a few seconds to the salaine-filled venous bloodline (after rinseback) and turning on the blood pump briefly as described, of by using a saline-filled syringe.

c) To assure that the post sample is not contaminated with dialyzer outlet blood.
Blood access flow is usually in the range of 1000 ml/min [1] although flow can decrease to below the blood pump rate, at which point access recirculation occurs [2]. In the presence of access recirculation, the BUN concentration of the blood in the arterial line can be reduced substantially. If such admixed blood is sampled, the postdialysis BUN will be underestimated, with a corresponding overestimation of the URR and the Kt/V. The degree of this overestimation of dialysis dose is not trivial. In some patients with severe AR in whom post-dialysis blood was drawn without clearing the arterial line deadspace of admixed blood, the apparent single pool Kt/V of 1.3 translated into an equilibrated Kt/V of only 0.76! [3].

To prevent sampling admixed blood, one needs to clear the deadspace in the arterial needle tubing (and in some cases the arterial bloodline) up to the region where the blood is sampled. As shown in the Figure, if sampling is done from the needle tubing, the deadspace is in the range of 2.6 ml. If sampling is done from the arterial sampling port, the total deadspace (needle plus needle tubing plus arterial bloodline tubing up to the sample port) is approximately 9-10 ml. We measured arterial bloodline deadspace in approximately 25 commonly used blood lines, and found it to average 6.7 ml (range 6.2-7.9 ml). However, some lines may have the arterial sampling port closer to the dialyzer, in which case the deadspace may be increased.

The deadspace can be cleared by reducing blood flow to the range of 50-100 ml (after first reducing the UF rate to zero). At this low blood flow, unless the needles have been reversed or are inappropriately closely spaced, access recirculation will cease, as even the worst accesses deliver flow greater than 100 ml/min. One then makes the assumption that, within several seconds of reducing blood flow, non-admixed blood begins to enter the needle tubing. At a flow rate of 60 ml/min, or about 1 ml/sec, approximately 10 ml of blood will have entered the arterial bloodline after slowing flow. For this reason, we recommend stopping the blood pump after 10 seconds if blood is drawn from the arterial needle tubing (deadspace 2.6 mL), and 15 seconds if blood is drawn from the sampling port (deadspace 9-10 ml). Now one has sequestered the desired sample of blood in the blood tubing. The sample should then be drawn promptly. Usually blood for post-dialysis BUN is drawn in either 3 or 5 ml tubes. These have a draw volume of about 2.5 or 4 ml. When drawing from the needle tubing, it is best to use the smaller tube, as then the blood sample will be limited to that "frozen" in the tubing when the pump was stopped.

A-V gradient:
Within 10-15 seconds of slowing the blood pump, a sharp increase in the BUN concentration of blood at the access begins to occur, which lasts up to 1-2 minutes). This increase is due to the fact that cleared blood is no longer returning to the heart from the dialyzer. Thus, as soon as the blood pump is slowed, the normal A/V urea gradient established during dialysis promptly begins to close. If one is using a draw volume much larger than 3 ml when drawing the sample, and if there is a delay in drawing the sample, some of the blood sample will come from the access and not the "frozen" sample in the tubing. This is also a problem if the sample is being drawn from the arterial sample port using a needle with the blood pump slowed but still running. In this case, there may still be a small artefactual increase in the postdialysis BUN. This increase is usually trivial unless one is comparing the BUN in this "slow flow" sample with one taken at full blood flow just before stopping the blood pump. For these reasons, we recommend waiting no longer than 10 or 15 seconds to stop the pump, and a draw volume of 3 ml or less when drawing from the needle tubing.

There are other methods of clearing the deadspace in the bloodline prior to sampling. One is to clear the deadspace from the 2.6 ml needle tubing with a syringe, by withdrawing blood and setting it aside prior to drawing the sample. This technique has the problem of increased blood wastage, and, if the blood is reinfused (followed by saline reinfusion), the theoretical (albeit miniscule) risk of infection.

d) Standardization of method of drawing postdialysis BUN.
Access recirculation aside, there are two components to the postdialysis urea rebound. Approximately 1/4 to 1/3 of the rebound is due to rapid equilibration of the A/V gradient after dialysis. In fact, with a venous access, rebound should be about 30% less than with an AV access. Because at a K/V of 0.4 Kt/V units per hour, total rebound is about 0.2 Kt/V units [4], one expects about 0.6 units of rebound due to A/V equilibration, and 0.14 Kt/V units due to compartment/flow effects (the magnitude of rebound of course will vary with the K/V, and also with other factors such as cardiac output).

Using the blood drawing technique described, one is sampling arterial post-dialysis blood when sampling from an AV access, and venous post-dialysis blood when sampling from a venous access. Hence when rapid dialysis (0.4 Kt/V units per hour) is being delivered, the (single-pool arterial) Kt/V will be "underestimated" by about 0.6 units when using a venous access (assuming that the dialysis catheter is positioned in such a way to receive a representative sample of mixed venous blood). Either the single pool arterial or single pool venous Kt/V values can be easily corrected to an "equilibrated Kt/V" (Kt/Ve), which is defined as Kt/V based on an equilibrated post-dialysis BUN value. When using an AV access, the equilibrated Kt/V can be estimated as:

Kt/Ve = Kt/Vspa - 0.6 x K/V + 0.03 ref [4]

Using a venous access, once can use the formula:

Kt/Ve = Kt/Vspv - 0.4 x K/V + 0.02 ref [4]

Slowing the blood fow 1-2 minutes:
Some prefer to slow the blood for 1-2 minutes prior to obtaining the postdialysis sample. Slowing the blood for such a long period may increase the risk of clotting (although this has never been reported). Certainly the UF rate should be turned to zero prior to such a long slow flow period. The method has one advantage, in that the blood is sampled after the A/V urea gradient of dialysis has been allowed to dissipate. As the A/V urea gradient can become quite wide in patients with low cardiac output (which may occur transiently at the end of dialysis), there is a theoretical advantage to this method. The resulting Kt/V should be treated as a venous Kt/V, and the second of the above formulas should be used for computing equilibrated Kt/V.

There are some centers which first return the blood and only then draw the sample from the arterial needle tubing. The time elapsing between the end of dialysis and the time when the blood is drawn varies from 2 to 15 minutes, or even longer. The Kt/V and URR values using such techniques will be closer to the equilibrated values. However, as the percent equilibration which will have occurred will vary, one cannot use the above formulas to estimate equilibrated Kt/V if the waiting period is longer than 2 minutes. If an antecedent slow flow period has not been used to clear the deadspace in the needle tubing, waiting for even 15 minutes does not insure that Kt/V will be correct, as recirculated blood might still remain in the needle tubing and form the bulk of the postdialysis sample. An alternative is to clear the needle tubing blood using an additional syringe, withdrawing at least twice the volume of deadspace to the sampling point, but this is cumbersome, and if forgotten, may engender the risk of sampling admixed blood.

In summary, we present a needle-free method of drawing the blood from the needle tubing after reducing flow to 50-100 ml/min for 10 seconds. This method has the advantages of limiting the risk of accidental needlestick, ensuring that the sample is free of admixed blood (in the case of access recirculation), and standardizing the Kt/V.

(June, 1996)

References:

1. Depner TA, et al. . Measurement of access blood flow using ultrasound dilution. ASAIO J 1995;

2. Besarab A, Ross R, Effendi A, Frinak S. The relation of recirculation to access flow. J Am Soc Nephrol 1995; 6:482(abstract).

3. Daugirdas JT, Burke MS, Balter P, Priester-Coary A, Majka T. Screening for extreme postdialysis urea rebound using the Smye technique: identification of patients with access recirculation. Am J Kidney Dis 1996; 27:in press.

4. Daugirdas JT, Schneditz D. Overestimation of hemodialysis dose (delta Kt/V) depends on dialysis efficiency (K/V) by regional blood flow and conventional 2-pool urea kinetic analyses. ASAIO J 1995; 41:M719-M724.

See also, the Hemodialysis AV Access Topic Page




This was the far most exhaustive description of how to draw a blood sample that I have ever seen, but, of course, a very important one. Relating to this subject - I have in vain tried to get some information about when, and perhaps how, the post dialysis sample in the "Core Indicators Project" (http://ns.gamewood.net/forum/core/core.htm) was drawn. Does anyone know?
KG Prütz (kg.prutz@helsingborg.se)
Helsingborg, Skĺne Sweden-Wednesday, July 24, 1996 at 12:03:17 (CDT)


This is really for ask the prof., which my computer has been unable to find or open. Is there any consensus regarding numbing the graft or fistula site prior to needle insertion. Surgeons seems to feel it destroys the graftt, patients often prefer it. Does anyone have any documentation one way or the other?
Marilyn
Green Bay , WI USA-Thursday, October 16, 1997 at 03:00:23 (PDT)