SEED Coagulation

Sysmex Educational Enhancement and Development

July 2012

Quality control in coagulation testing

The prothrombin time and activated Prothrombin time reagent
partial thromboplastin time Many of the coagulation reagents and controls that are
As discussed in previous editions, the main indication for required for coagulation testing come in a lyophilised formu-
prothrombin time (PT) (expressed as an international nor- lation. Lyophilisation refers to the process of freeze drying
malized ratio or INR) and activated partial thromboplastin the liquid reagent which removes the liquid component
time (APTT) testing is to monitor the anticoagulant effects and leaves a dry powder behind. As clot based coagulation
of warfarin and heparin respectively. Both drugs have a very testing requires the key components (e. g. tissue factor in
narrow therapeutic window. This means that the patient the thromboplastin reagent Innovin©) to be biologically active
will be adversely affected if the test results fall outside of lyophilisation is necessary to preserve this function. In the
this range – if too high they may bleed or if too low they lyophilised format, the tissue factor is preserved thereby
are at risk of developing a blood clot (Fig. 1). Millions of giving the PT reagent a very long shelf life. Before use, the
patients are on chronic anticoagulant therapy worldwide PT reagent needs to be reconstituted with the exact amount
so the consequences of issuing erroneous test results are of diluent (liquid) as prescribed in the package insert. This
far reaching. reconstitution effectively brings the reagent back to life
and ready to use in testing. However, it must be noted that
In this regard, it is of paramount importance that the PT and once converted back into a liquid, the reagent can only be
APTT results issued by the laboratory are accurate. The only used for a very limited period. This is referred to as the ‘open
way to ensure this is to apply stringent internal quality con- vial stability’. Whilst the expiry times are very long (measured
trol procedures. in months or years), the open vial stability of reagents and
controls is measured in hours or days. It is therefore essential
to ensure that reagents are correctly labeled with the date
Progressive risk therapeutic window Progressive risk and time when reconstitution took place and to ensure that
of clotting ‘safe zone’ of bleeding
of anticoagulation the items are stored appropriately. The individual package
inserts will provide details of the open vial stability under
different storage conditions (Fig. 2).

The lack of adherence to the storage conditions and times

short prolonged
Clotting time in seconds is a major source of error in the coagulation laboratory.

Fig. 1 Diagrammatic representation of the ‘therapeutic window’ of

clotting times for patients on anticoagulant therapy
SEED Coagulation – Quality control in coagulation testing 2
Sysmex Educational Enhancement and Development | July 2012

Storage and Stability

Store at + 2 to +8 °C. At this temperature, the unopened reagent can be used until its expiry date (see vial label).

Stability after reconstitution: + 2 to + 8 °C 10 days (closed vial)

+ 15 to + 25 °C 5 days (closed vial)

+ 37 °C 24 hours (closed vial)

Information about on-board stability is specified in the Reference Guides (Application Sheets) for the different coagulation analyzers.
Do not freeze!
Signs of expiry: Absence of vacuum when opening the vial; reagent is difficult to reconstitute; results are not reproducible.

Fig. 2 Extract of Innovin package insert highlighting storage and stability conditions after reagent reconstitution

Activated partial thromboplastin time reagent If the test method is working well, then the result obtained
In contrast to the thromboplastin reagents used for PT test- for the control sample should be within the specified range
ing, the APTT reagent Actin FS© comes in liquid formulation. of acceptable results – the so-called target values. It is impor-
This is because the activator which initiates the clotting tant to always use a control with a target value that is aligned
reaction is an inert substance that has no inherent biological with what is clinically relevant. In the context of using the
function. It is the charged nature of this substance (ellagic APTT for heparin monitoring, it would be important to have
acid) that reacts with clotting factor XII (the starting point a control with a target value that is within the therapeutic
of the APTT) rather than enzymatic activity. Consequently range. In line with this general principal Sysmex, with its
the reagent is stable as a liquid although the shelf life is reagent partner Siemens, has three levels of controls for
somewhat shorter than the lyophilized reagents. Once PT (INR) and APTT testing (Tab. 1).
opened, the same open vial stability issues apply as exposure
to air initiates a slow decline in the stability of the product. Citrol© Level 1 Normal range
The performance of the product is only guaranteed by the
Citrol© Level 2 Low therapeutic range
manufacturer if the storage times and conditions are strictly
Citrol© Level 3 High therapeutic range
adhered to.
Tab. 1 Control plasmas used for PT (INR) and APTT testing
In contrast calcium chloride is a completely inert chemical
hence it has a very long shelf life included a long stability Control plasma target values
even after opening. Control plasmas for PT and APTT testing may be assayed or
unassayed. A control plasma that is deemed to be ‘assayed’
Control plasmas is supplied with target values. The target values assigned to
As for any laboratory investigation, the accuracy of PT and a control plasma are specific to the reagent and analyser
APTT results must be monitored on a regular basis using used to generate the test result. It is therefore important to
controls. As the PT and APTT essentially measure the reac- ensure the correct target range is used. Unassayed control
tivity of clotting factors, culminating in the conversion of plasmas do not have target values assigned. If a laboratory
fibrinogen to fibrin, the control material has to contain func- chooses to use unassayed control plasmas, they must be
tional clotting factors. However, we know that blood taken aware that they will have to generate their own target range
for coagulation testing should ideally be tested within four by processing the new lot of control plasma on an analyser
hours as the clotting factors start to deteriorate quite rapidly. five times per day for at least five consecutive days.
It therefore follows that controls for clot-based coagulation
testing must be comprised of plasma. As for the PT reagent,
the function of the clotting factors is preserved through
lyophilisation. Likewise, once reconstituted, control plasmas
have the same short viability as do patient samples.
SEED Coagulation – Quality control in coagulation testing 3
Sysmex Educational Enhancement and Development | July 2012

Fig. 3 An extract from the assay data sheet of a normal level of control plasma

The results obtained are then used to calculate the target A second major problem with the use of third party reagents
range using the formula of ‘mean ± 2SD’. During this time is that the protocols on automated analysers are fine-tuned
of establishing such a target range, the daily testing must by the manufacturer for specific reagents which are not
continue to be controlled using the previous lot number of interchangeable. If a laboratory chooses to use reagents
control plasma. In other words, during the days that a new other than those recommended by the manufacturer then
target range is being established, two lots of control plasmas the responsibility of ensuring that the analyser is correctly
must be used. An example of an assay data sheet for a set-up is entirely their own.
control plasma is shown in Fig. 3.
Are all variables that have an influence on test results
Why the use of third party reagents and controls detectable through the use of control plasmas?
is problematic It is important to be aware of the fact that there are several
The use of third party reagents and controls is problematic pre-analytical factors that can give rise to erroneous test
for several reasons. Firstly, the target ranges on the assay results as these factors are sample specific and will not be
data sheet that are supplied with assayed control plasmas detected by the standard internal quality control procedures.
are generally only available for reagents and analysers Control plasmas will only assess the analytical component
belonging to the same manufacturer. If the exact reagent of the testing process. Any error introduced into patient
and analyser (make and model) combination is not listed, sample testing due to sample processing pre-analysis will
then the control plasma is in essence ‘unassayed’ for that not be identified through the use of control plasmas.
combination or analytical system. The laboratory will then
need to establish its own target range. This approach is not How frequently should control plasmas be processed?
recommended unless the laboratory personnel are highly As a general rule of thumb, control plasmas should be
experienced in coagulation testing. Results obtained using processed at least once per 8 hour shift. Both a normal
the same reagent but different models of analyser from the and at least one abnormal control should be included.
same manufacturer are not going to be identical. Differences
are to be expected as the clotting time generated is the How does one determine whether the quality
product of the entire analytical system i.e. the analyser, the control results are acceptable or not?
protocol set up and the reagents. As the therapeutic range One must remember that the purpose of performing internal
for anticoagulation is narrow, any small shifts in clotting quality control is to ensure the reliability of patient results.
times may result in a change in clinical management. In The values obtained for each level of control plasma must
this regard it is imperative that the control target values be compared with the acceptable target range as per the
are specific for the reagent and analyser model in use. data assay sheet that is specific for the lot number of the
SEED Coagulation – Quality control in coagulation testing 4
Sysmex Educational Enhancement and Development | July 2012

control and reagent and analyser combination. The value The first step to be undertaken when troubleshooting internal
must be within the specified range. However, the target quality failures is to judge the anticipated time delay before
range provided is generally rather wide and in practice the the problem is likely to be resolved. This will determine
range of values obtained on a single specific analyser in a whether alternate plans need to be made for patient sample
laboratory from day to day will be much tighter. The degree analysis as no patient can be released until the quality control
of inter-day variation that is considered acceptable will be results are once again shown to be within acceptable perfor-
guided by the CV % of the assay. The CV % is a measure of mance limits.
the precision of the test. If a single value shows a significant
shift from what was obtained on previous days, or a contin- There is no fixed troubleshooting procedure. Each problem
uous trend deviating in the same direction, a possible source needs to be assessed on a case by case basis. A step by step
of error should be sought. Patient results should be withheld approach should be adopted which follows a logical sequence
until the problem has been rectified. of checks. The obvious should be checked first. Judgment
should be prudent taking both cost and time factors into
How would one go about troubleshooting consideration. One should try to gauge whether the problem
failed QC results? is likely to be due to faulty control material, faulty reagents
A number of baseline factors need to be considered when or a faulty analyser. As a rule of thumb, the problem is most
evaluating the possible sources of error: likely to lie with the control material as this is the most labile,
■■ Is an isolated parameter affected or are all followed by the reagents and lastly the analyser which is the
parameters affected? most stable component of the analytical system.
■■ If more than one, do these parameters have a common
testing principle or not; e. g. clot-based tests only or What kind of problems are commonly associated
are chromogenic parameters (e. g. antithrombin) also with the control plasma?
affected? Because the reconstituted control plasma is very labile with
■■ Is only one level of control affected or are all out a short open vial stability (maximum 16 hours for Citrol), the
of range? use of ‘old’ control material is the commonest cause of failed
■■ Is the direction of deviation the same for all levels internal quality control. This is readily preventable if the
and all parameters i. e. do all have values that are laboratory adheres to the practice of labelling each recon-
too high or all are too low? stituted vial with the date, time and signature of the person
■■ Is this a once off event or has the same problem who opened it. All parameters are likely to be affected and
been observed before? usually all levels as in general the person reconstituting the
■■ Has the open vial stability of all reagents been controls for the day (or particular shift) will make up a vial
adhered to? for each level to be used at that time. Another consideration
■■ Has the expiry date of all reagents been adhered to? is the accuracy of pipetting the diluent. An excess of diluent
■■ Does the lot number of the control material used would result in a dilution of clotting factors with prolonged
match that of QC target range? times and too little would give rise to a concentrated sample
■■ Have the reagents and controls been correctly with shorter than expected clotting times. The diluent for
reconstituted? control plasma is distilled water. It is strongly advised that
■■ Was the diluent used for reconstitution the correct one? laboratories use sterile ‘water for injection’ as this comes in
■■ Was the diluent of good quality? a small sized single use sterile packaging. It is however com-
■■ Was there a problem with pipetting? mon practice for laboratories to source the distilled water
■■ Was the reagent adequately mixed and allowed to for control plasma and reagent reconstitution from a local
stand for the appropriate amount of time before use? purification system where sterility, especially during the
■■ Was the analyser correctly calibrated (if applicable)? collection process is not always maintained. The growth
■■ Has analyser maintenance been carried out of fungal elements is relatively common which can inter-
as prescribed? fere with analyser optics and consequently test results. A
simple check would be to make up a fresh vial of control
SEED Coagulation – Quality control in coagulation testing 5
Sysmex Educational Enhancement and Development | July 2012

using sterile water and retesting the sample. Power failures What kind of problems may be due to the analyser?
which lead to intermittent loss of cold chain maintenance The analyser is least likely to be the cause of failed quality
must also be considered as possible sources of error as the control. However there are some simple things that may
open vial stability of Citrol control plasma is reduced from give rise to faulty results. Any spillage of plasma into the
16 hours if kept stoppered and in the fridge at 2 – 8°C to detection wells can interfere with optics. Likewise if the
only 8 hours if the temperature rises to 15°C. intensity of the LED lamps that are the core component of
the optical clot detection system needs to be checked from
What are the commonest problems associated time to time as part of regular analyser maintenance. If this
with reagents? is not done then a drift in results is likely to occur. The
The commonest problem, as for control plasmas, is the lack temperature of the detection wells needs to be at 37 ± 1°C.
of adherence to the maximum open vial stability period and There is an easy self-check on the analysers to confirm that
the correct reconstitution where appropriate. If the reagent the temperature is correct. As clot formation depends on
is the problem then all levels of control plasma would be the biological activity of clotting factors, which function
affected but only those parameters that have a reagent in optimally at body temperature, any small deviation up or
common. For example, if the suspect reagent is Innovin, down can have a significant effect on test results. As each
then the PT (INR) results and extrinsic factors will be affected, of the detection wells has its own LED lamp and temperature
but the APTT and intrinsic factors will be within limits. control, faulty test results are generally restricted to a single
Another common source of error with INR results is the detection well and is thus relatively easily identified.
use of a new lot number or reagent without entering the
new ISI value and newly determined mean control value Take home message
into the method protocol on the analyser. The INR values Internal quality control is an integral process of laboratory
that are produced by the analyser will be based on the testing which is no different for coagulation testing. Always
clotting time of the plasma sample measured with a new ensure that the target values used are appropriate for the
lot of reagents but using the mean control value and ISI that reagent and analyser combination. In the event that the
belong to an earlier lot numbers. Although such differences quality control samples fail, patient results must be withheld,
tend to be small, they can be significant enough to cause and possibly retested once the problem has been resolved.
quality control failure. It is an extremely rare occurrence for There can be multiple sources of error and hence a logical
reagents for which cold chain has been maintained at all troubleshooting approach must be adopted. In general
times, accurately reconstitution and open vial stability limits however, the control plasma is most likely to be the culprit,
adhered to be faulty. followed by the reagents and lastly the analyser.

Compiled by
Dr. Marion Münster

Sysmex Europe GmbH

Bornbarch 1, 22848 Norderstedt, Germany, Phone +49 40 52726-0 · Fax +49 40 52726-100 · [email protected] · www.sysmex-europe.com