E-Book

QUALITY CONTROL

Improving Laboratory Performance Through Quality Control
Five simple steps for QC success

Q
How Often is Right for QC?
Ask the Right Questions to get the Right Answers

Complete QC solutions for results you can trust

How Often is Right for QC?

It is widely accepted that laboratories should perform QC at least every day of patient testing. However, is this
adequate for every assay and for every laboratory? Is running QC once per day really sufficient? What is the
“right” frequency for running QC samples in your laboratory?

ISO 15189 regulations don’t state a recommended QC frequency but they do recommend that
“Quality Control materials shall be periodically examined with a frequency that is based on
the stability of the procedure and the risk of harm to the patient from an erroneous result”.
ISO 15189 understands that differing tests and situations will require differing QC frequencies.

So how do you use this advice to work out the correct QC frequency for each assay in your lab?
There are various factors in line with ISO regulations that you need to consider when deciding an appropriate QC frequency.
Deciding an appropriate QC frequency, starts with asking the right questions.

1Which assays are more Which tests

stable compared to others? 2are higher risk
and have a
higher impact
if an erroneous

result is
released?

4What is 3How many patient

the time samples are you running in
between QC between QC evaluations?
evaluations?

It is of great importance that laboratories ask all of these questions in order to determine the “right” QC frequency. If you
2 ask the ‘right’ questions, you’ll get the ‘right’ answers.

1. Which assays are more stable compared to others?

Some assays naturally perform better than others, giving consistently better results and rarely producing any errors. On the other
hand, some assays perform inconsistently, having a higher rate of error and much lower stability.
It’s important that laboratories can recognise which assays are more stable and consistent in comparison to others and ensure that
they are running QC at an appropriate frequency, in order to detect any clinically significant errors.
By utilising an EQA scheme and/or peer group reporting programme, method validation and peer performance comparison can be
monitored, helping laboratories to assess their precision over time and easily identify which tests generally perform better.

Identify any unstable assays and increase QC frequency for those assays.

3

2. Which tests are higher risk and have a higher impact if an erroneous result is released?

It’s important that you run QC more frequently for higher risk tests. With higher risk tests there is a greater possibility of harm to
the patient, therefore it’s of utmost importance that the results released are both accurate and reliable.
Any tests that have the following characteristics should be considered high risk and QC should be run more frequently in these
instances.
• A test where there could be a detrimental consequence, should the wrong test results be released
• A test that supports the clinician’s decision in isolation
• A test that is acted upon immediately
• A test that is performed on a specimen that is difficult/painful to collect

3. How many patient samples are you running in between QC evaluations?

This is an important factor that must be taken into consideration when deciding QC frequency. Imagine there are two labs “Lab A”
and “Lab B”, both of which perform QC every morning.That’s sufficient, correct?
Now consider that ‘Lab A’ tests 10 patient samples a day, whereas Lab B tests 1000 patient samples a day. Is performing QC once per
day still sufficient for each lab? Say an error occurred in the test system after 50% of the patient samples had been tested, both labs
would not recognise that they had a QC failure until the next day. Arguably, this could be problematic for both labs as erroneous
patient results may have been released.
In both cases, they will have to re-evaluate the patient samples from the last successful quality control event as recommended by
ISO 15189.
Potentially Lab B will have to repeat 1000 patient samples, costing a significant wastage of both time and resources. Ideally patient
samples should be run in batches, starting and ending with a QC evaluation, perhaps running QC every 50 or 100 patient samples.
This will save time, save money and most importantly this will reduce risk of harm to the patient.

4

4. What is the time between QC evaluations?

Let’s consider another scenario, both Lab A and Lab B now decide to change their QC strategy. Instead of running QC once per
day, they now decide to run it every 100 patient samples. Great news, correct? This perhaps is good news for Lab B, as they will
now be running QC more frequently, reducing risk for their patients. However, this is not so good for Lab A. Let’s say that an error
occurs after 50 patient samples have been run. For Lab B, they will detect the problem straight away on day 1 and will be able to
investigate the problem preventing the release of erroneous patient results.
For Lab A, the error will have occurred on day 5 of their patient testing, but the problem would not be recognised until day 10.
This could spell disaster for any laboratory, with the release of potentially erroneous results causing misdiagnosis, incurring cost and
resulting in a negative impact on patient care.
Therefore, it’s of utmost importance that you consider the time between QC evaluations in addition to the number of patient
samples being tested. Keep the time between QC evaluations shorter than the time needed to undertake any corrective action in
the case of an erroneous result.This is a good rule of thumb to ensure you select an appropriate QC frequency.

It’s good to take a holistic approach and take all factors into consideration.

Conclusion

Unfortunately there is no straightforward answer to how frequently you should run QC. However, if you
ask the right questions, you’ll reach the right answer. Make sure you are running QC more frequently for
high risk and unstable tests, that you start and end patient testing with a QC evaluation and you make
the time between QC evaluations shorter than the time needed to take corrective action in the case of
an erroneous result.
Finally QC samples should also be tested before and after any event that has the potential to adversely
affect the testing process e.g. change of reagent batch, instrument maintenance and calibration.  Testing
prior to the event provides confidence that patient results since the last successful QC check are reliable.
Testing QC samples immediately after the event ensures the test system is in control prior to running
more patient samples.  In the case of an unplanned event QC testing should still be performed immediately
after the event to assure the testing process is operating correctly before continuing with patient testing.

Acusera Advisor*

Designed for use with the Acusera range of third party controls, our Acusera 24•7 software will help your lab monitor and interpret
your QC data. Acusera Advisor is a tool that automatically recommends QC multi-rules and optimal frequency for each individual
test used in your lab.

Employing Acusera Advisor in your lab will reduce false rejections and unnecessary troubleshooting, therefore saving time and
money on expensive repeat tests.

Recommendations are based on normalised OPSpec charts. Once performance limits have been defined, the software will determine
the %CV and %Bias.These are then used to calculate the normalised operating point. A normalised OPSpec chart is then used to
select the appropriate QC strategy.

* not available in USA
5

QUALITY CONTROL

Randox Laboratories Ltd, 55 Diamond Road, Crumlin, County Antrim, BT29 4QY, United Kingdom
+44 (0) 28 9442 2413 +44 (0) 28 9445 2912 [email protected] randoxqc.com

True third party controls offering complete test menu consolidation
Uniquely combining more than 100 parameters conveniently in a single control,
laboratories can significantly reduce costs and consolidate without compromising
on quality. As true third party controls, unbiased performance assessment with any
instrument or method is guaranteed.

24•7

Online QC software with real-time peer group statistics
Compatible for use with the Acusera range of third party controls, the Acusera
24•7 software is designed to help laboratories monitor and interpret their QC data.
Access to an impressive range of features including interactive charts and real time
peer group data generated from our extensive database of laboratory participants,
ensures Acusera 24•7 is the most comprehensive package available.

The largest global EQA scheme with over 31,000 lab participants
Comprising over 360 routine and esoteric parameters in 24 comprehensive and
flexible EQA programmes, RIQAS is designed to cover all areas of clinical testing.
Each programme benefits from a wide range of concentrations, frequent reporting
and comprehensive yet user-friendly reports.

find out more

Information correct at time of print. Randox Laboratories Ltd is a subsidiary of Randox Holdings Limited a company registered within Northern Ireland with company number N.I. 614690. VAT Registered Number: GB SEP15
151 6827 08. Product availability may vary from country to country. Please contact your local Randox representative for information. Products may be for Research Use Only and not for use in diagnostic procedures
in the USA. 6

QUALITY CONTROL

Improving Laboratory Performance Through Quality Control
Five simple steps for QC success

Q

Designing an Appropriate QC Procedure for Your Laboratory

Complete QC solutions for results you can trust

The Importance of a well-designed QC Procedure

It is easy to get caught up in an abundance of QC statistics and forget the fundamental reason why
QC exists in the first instance. QC is about detecting errors and ensuring that the results you
produce are accurate and reliable. All QC procedures should focus on reducing the risk of harm
to the patient.We are not examining statistics; we are examining real patients, real results and real
lives. Around 70% of all medical decisions are based on laboratory results, which is why it is of
utmost importance that each and every laboratory, has a well-designed QC procedure in place.

An effective QC strategy is not as complicated as you might think, it does not
require you to become an ‘advanced mathematician’. If you stick to these five simple
steps you can be assured that you are releasing quality QC results.

1Identify the Quality 2Choose Good Quality
Control Materials
Specifications for the test

5Ensure you are 5 Simple Steps
for QC Success
able to recognise
and troubleshoot 3Start and End
an out-of-control
Patient testing with
event a QC evaluation

2 4Ensure you know

the characteristics
of good QC results

1. Identify the Quality Specifications for the test

• Not every test performs the same
• Not every test needs to perform the same
• Give poor-performers and high risk tests more QC attention
Not every test in the laboratory performs the same way. Some tests for example are the well-known, well liked, laboratory “good-
performers, always showing excellent recovery rates. However, other tests such as the laboratory “poor-performers”, show a lot
more inconsistencies in performance, and can often be slightly more problematic and frustrating.

“It’s important that you know which tests are your well performing tests
and which are your poor performing tests.”

You can identify which tests generally perform better, through monitoring your precision over time, via method validation and
comparing your performance to peers through either an EQA scheme or through a peer-group reporting software. Using this
differentiation, afford your “poor-performer” QC tests, a little more QC attention. Run QC more frequently and monitor the results
closely, to ensure that they are both accurate and reliable.
It is also important to remember that not every test needs to perform the same. Some tests, compared to others, have a little more
scope for error before they have an impact on patient care. It is therefore important to ensure that you identify the important
“high risk tests”. Any tests that have one or more of the following characteristics, should be considered high risk:

• A test where there could be a detrimental consequence, should the wrong result be released
• A test that supports the clinicians decision in isolation
• A test that is acted upon immediately
• A test that is performed on a specimen that is difficult/painful to collect
Additionally, make sure you take note of the CLIA analytical quality requirements, to identify the quality specifications of particular
tests. For any test that you identify as high risk, make sure you run QC more frequently and play close attention to your QC results.

3

2. Choose Good Quality Control Materials

The effectiveness of any QC procedure is underpinned by the quality of the control materials that you choose to use. ISO 15189
recommends that the

“laboratory shall use quality control materials that react to the
examining system in a manner as close as possible to patient samples”.

It is important to note that differently manufactured QC material, will not perform in the same way. It is of utmost importance that
you choose the best quality material available on the market, as you may otherwise be putting patients at risk.

Look out for the following qualities in your QC material:
• Have accurately assigned values; the narrower the control ranges used, the more efficient you will be at assessing your

performance. Unfortunately, many manufacturers assign their QC values based on only a handful of results, meaning that control
ranges are far too wide to monitor performance effectively. Choose materials that have assigned values based on thousands of
laboratory results.

• QC material that is “as close as possible to a patient samples”; as recommended by ISO 15189; In order to behave
and react like a patient sample, the material needs to be in a matrix similar to a patient’s as this will allow the best assessment of
performance. Look out for materials that are described as 100% or fully human and look out for lyophilised products that are free
from any artificial materials and preservatives. Many manufacturers describe their material as “human-based”, as these materials
contain components of animal origin and will not be “as close to a patient sample” as possible. This is particularly important
for anti-body based tests. Make sure you check your QC kit insert to ensure you are using 100% human material and that your
lyophilised material is free from preservatives.

• A format suitable for use; QC materials come in lyophilised, liquid-frozen and liquid ready-to-use formats. Lyophilised materials
are great for hotter climates. Liquid frozen and liquid ready-to-use samples, eliminate the chance of reconstitution errors, with
liquid ready-to-use samples requiring the least effort for preparation, saving you time and making them
ideal for point of care testing.

• Are they third party materials; ISO 15189 recommends the “use of
independent third party control materials”.Third party materials offer many
advantages, enabling an unbiased assessment of performance that can be
used on all instrument platforms.

• Cover the full clinical range; ISO 15189 recommends that the
“laboratory should choose concentrations of control materials,
wherever possible, especially at or near clinical decision values,
which ensure the validity of the decisions made”.

• Match their stability claims; Many manufacturers, in an
attempt to make their material ‘look better’, give inaccurate
stability claims and often give different stability claims for different
analytes in the material. Don’t forget that a QC material is only as
stable as its least stable analyte. Make sure your QC manufacturer
is being honest about their stability claims and avoid those who
make different stability claims for different analytes in the same
sample.

4

3. Start and End Patient testing with a QC evaluation

Make sure you are testing patient samples in “batches” and are starting and ending patient sampling with a QC evaluation. If you
only run QC every morning and one morning you find that your QC is out of range, you may be unaware as to whether the
malfunction occurred the previous day or not.Your staff will have to troubleshoot, without knowing if the previous days results have
been affected or not, by re-running samples, perhaps unnecessarily.

Running QC at the end of the day, assures you that that day’s results are not affected, saving you a lot of time and money! If your
lab runs in 24hr operation, make the time between QC evaluations shorter than the time needed to correct erroneous results, as
this is the most cost effective strategy for your lab. Additionally, ensure that you know the number of patient results between QC
evaluations, as this will help you ascertain the number of patients at risk, should an out-of control event occur.

4. Ensure you know the characteristics of good QC results

In order to recognise a “bad” result, it is important that you are able to recognise the qualities of a “good” set of QC results. In an
ideal world, the characteristics of good/normal QC results on a Levey-Jennings chart should;

• Fluctuate randomly around the mean for comparison. There should be an equal number of results above the mean
as below the mean.

• Very rarely exceed 3SD; 3 in every 1000 results should be outside 3SD
• On the whole be within 1SD.
• On Occasion exceed 2SD; It is normal for 1/20 results to exceed 2SD

It is important to keep these normal characteristics in mind when examining your QC results, to ensure you are not falsely rejecting
acceptable results. Ensure you are maintaining a record of your QC results, to evaluate over time, the norm for your laboratory.

There are various QC management software packages available on the market, that make monitoring your QC results easy. Some
QC management software packages also have functionality that allows you to compare your results to other laboratories, using the
same system as yourself. Knowing that you are comparing well to others, provides you with confidence that your results are in fact
good/normal. Look out for software that allows:

• Multiple instrument registrations
• Results to be entered online, anywhere, any time.
• Real time monitoring of peer group data. Ideally peer group data should be updated daily.
• Interactive charts such as Levey-Jennings and Histogram charts
• Different user level accounts so that lab managers can keep trace of the entering and acceptance of results.
• User defined multi-rules, to be applied to QC results, so that results can automatically be rejected or accepted.

68.26% +3 +3
95.46% +2 +2
99.73% +1 +1
MM
-1s -1s
-2s -2s
-3s -3s

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

5

5. Ensure you are able to recognise and troubleshoot an out-of-control event

• The most effective way to recognise errors is to use QC multi-rules
• Participate in a ISO 17043 accredited EQA scheme to help recognise errors
• Make sure you estimate the magnitude and size of the out of control event, before you correct it

There are various approaches for recognising possible out-of-control events. Some labs use a single rule and other labs opt to use
QC multi-rules. It is important for whatever rule(s) you use, that you are able to identify errors quickly and effectively, without
falsely rejected results, wasting time and money. It is also important that you choose assayed QC material that has tight ranges or
conversely assign the ranges yourself to ensure you have effective error detection.

Recognising out-of-control QC results using a Single rule

A laboratory favourite is the 1 result outside 2SD rule.This is a great rule for alarming possible out-of-control events, however, you
must be careful as it does have a high false rejection rate. Remember, it is normal for 1/20 results to be outside 2SD! It is important
that you also consider the strategy you use when you have a result outside 2SD. Avoid the “repeat, repeat, repeat… got lucky!”
tactic.This is not an effective error detection method, as if 2/3 results exceed 2SD you may have a problem at hand that is worth
investigating further. In short, be careful when using 1 result outside 2SD rule and ensure that if you do repeat, you do so only
once!

Another single rule that is sometimes used, is when 1 result is outside 3SD.This rule has a very low false rejection rate. Remember
only 3 in 1000 results should be outside 3SD. However, this is not the best rule for sensitive error detection. This rule ideally
shouldn’t be used in isolation.

Recognising out-of-control QC results using Multi-rules

Using a combination of multi-rules is the most effective way of recognising out-of-control events. Applying multi-rules means that
you will have a high rate of error detection, whilst maintaining a low rate of false rejection.This means you will run less unnecessary
repeats and waste less time carrying out unnecessary troubleshooting, in turn saving you money.

As a general rule, apply more multi-rules to poor-performer tests and high risk tests.With stable, good-performer, you can use less
multi-rules.The diagram below summarises a variety of commonly used multi-rules.

Control
Data

No ACCEPT RUN
1 IN - CONTROL
2S

Yes No No No No
1 2R 4 10
3S 2S 2s 1S X

Yes Yes Yes Yes Yes

OUT - OF - CONTROL REJECT RUN

6

Participate in an Accredited EQA Scheme
QC alone isn’t sufficient for detecting errors and so you should participate in an IS0 17043 accredited EQA scheme. ISO 15189
states that the “laboratory should participate in interlaboratory comparison programmes”. Not only will EQA help you detect
errors, it will provide you, the physicians and patients you supply results to, confidence and evidence that you are releasing accurate
patient results.
Troubleshooting Out-of-Control Events
Make sure you estimate the magnitude and size of the out of control event before you correct it. It’s a good idea to take note and
monitor your average patient mean or test a known patient sample.That way, you can measure the extent of the problem and the
effect the out-of-control event has had on patient results. Measuring the direction and magnitude of the shift in results, can help you
decide whether any clinical significant errors may have occurred and whether or not you need to repeat the patient results.

When an out-of-control event has occurred, ISO 15189 requires laboratories to
“evaluate the results from patient samples that were examined
after the last QC event”.

Ensure you know how many samples were run from the last QC event and do not release any patient results until the problem has
been rectified.

Conclusion

Armed with these 5 simple steps, you can be assured that you are on the way to producing
an effective and simple QC strategy, reducing the risk of harm to your patients. Remember
that you do not need to become a statistician to design effective QC procedures.
Improvement in your QC procedures simply requires the desire and determination to
take incremental steps. With each step you take, the next step becomes easier and you
will soon reach your goal of QC success!

7

QUALITY CONTROL

Randox Laboratories Limited, 55 Diamond Road, Crumlin, County Antrim, BT29 4QY, United Kingdom
+44 (0) 28 9442 2413 +44 (0) 28 9445 2912 [email protected] randoxqc.com

True third party controls offering complete test menu consolidation
Uniquely combining more than 100 parameters conveniently in a single control,
laboratories can significantly reduce costs and consolidate without compromising
on quality. As true third party controls, unbiased performance assessment with any
instrument or method is guaranteed.

24•7

Online QC software with real-time peer group statistics
Compatible for use with the Acusera range of third party controls, the Acusera
24•7 software is designed to help laboratories monitor and interpret their QC data.
Access to an impressive range of features including interactive charts and real time
peer group data generated from our extensive database of laboratory participants,
ensures Acusera 24•7 is the most comprehensive package available.

The largest global EQA scheme with over 31,000 lab participants
Comprising over 360 routine and esoteric parameters in 24 comprehensive and
flexible EQA programmes, RIQAS is designed to cover all areas of clinical testing.
Each programme benefits from a wide range of concentrations, frequent reporting
and comprehensive yet users friendly reports.

find out more

Information correct at time of print. Randox Laboratories Limited is a subsidiary of Randox Holdings Limited a company registered within Northern Ireland with company number N.I. 614690. VAT Registered MAR15

8 Number: GB 151 6827 08. Product availability may vary from country to country. Please contact your local Randox representative for information. Products may be for Research Use Only and not for use in diagnostic
procedures in the USA.

QUALITY CONTROL

Improving Laboratory Performance Through Quality Control
Five simple steps for QC success

Q
Which QC is the Right QC?
Ask the Right Questions to get the Right Answers

Complete QC solutions for results you can trust

Which QC is the Right QC?

Anyone working in a laboratory will understand the importance of quality control
(QC). QC is the process used to detect errors in order to ensure the reliability and
accuracy of patient test results. Without QC, laboratory errors could go unnoticed
potentially resulting in misdiagnosis and inappropriate or delayed treatment, all of
which could be life-threatening for the patient.

When running internal QC, laboratories need to be assured of the accuracy of the results produced and, to ensure this, they
need to have confidence in the QC materials they are using.

ISO15189:2012 states that:
“The laboratory shall use quality control materials that react to the
examining system in a manner as close as possible to the patient samples.
The laboratory shall choose concentrations of control materials,
wherever possible, especially at or near clinical decision values, which
ensure the validity of decisions made. Use of independent third
party control materials should be considered, either instead of, or in
addition to, any control materials supplied by the reagent or instrument
manufacturer.”

2

This document explores the different types of quality control products available and draws conclusions about what best meets
ISO15189:2012 requirements.
QC material is tested by laboratory staff in the same way a patient sample is. The results are then used to validate whether the test
system is operating within the pre-defined specifications. If QC results fall outside the pre-defined limits this infers that there may
be a problem with the test system or equipment, that could lead to the release of inaccurate patient results.
QC products are patient-like materials usually made from human serum, urine, or in some cases, spinal fluid. Control products can
be liquid or freeze-dried (lyophilised) and are either assayed, where target values and ranges for the test results are pre-defined by
the control manufacturer, or unassayed, where the laboratory has to define their own target values.
The validity of QC results largely hinges on the reliability of the QC material in use. When deciding on which type of control
product to trust it is important to be aware of the different types of material available on the market.

3

There are three types of control material available:

Dependent Semi-Dependent Independent or
Control Control Third Party Control

1 2 3

1 Dependent Control

Dependent controls refer to any control material that has been produced by the instrument or
reagent manufacturer for use on a specific test system. Such controls are often manufactured
from the same raw materials as the calibrator making them less sensitive to subtle changes in
performance.

As dependent controls are generally optimised for use with manufacturer’s test system, these
controls can mask weaknesses, and therefore, are increasingly considered less effective than
completely independent controls.

2 Semi-Dependent Control

Semi-dependent control material although produced independently of the instrument or reagent
is often supplied or recommended by the instrument/reagent manufacturer. It is this manufacturing
relationship between the two that requires close scrutiny when considering if these controls are
fit-for-purpose.

In many cases although the control material is not produced by the instrument manufacturer it is
produced according to their exact specifications and therefore, optimised to work with a specific
platform. For example, if an instrument has known precision issues at lower concentrations,
control material can be manufactured at higher levels, thus masking any issues.

As semi-dependent controls usually carry product labelling that is different from the instrument
or reagent some laboratories may unwittingly be using semi-dependent controls. In doing so,
weaknesses and errors may go unnoticed.

4

3 Independent or Third Party Control

Independent or third party quality control material has not been designed or optimised for use
with any instrument, kit or method. This complete independence enables the quality control
material to closely mirror the performance of patient samples, and in doing so, provide an
unbiased, independent assessment of analytical performance across multiple platforms.

This is a fact that is recognised by regulatory bodies and standards around the world, and is corroborated in ISO15189.
The Clinical and Laboratory Standards Institute (CLSI) sums it up in the following statement:

“Quality control materials should be different from the calibrator
materials to ensure that the QC procedure provides an independent
assessment of the measurement procedure’s performance in its entirety,

including the procedure for calibration of the measurement.”

Independent Third Party Controls Versus Dependent/Semi-Dependent Controls

True third party controls are manufactured Some manufacturers use the same raw material to
independently meaning they give totally unbiased manufacture both controls and calibrators, making the
results. controls less sensitive to performance changes.

The leading third party control manufacturers assign It is not uncommon for instrument/reagent
target values using data from thousands of independent manufacturers to assign their QC values using only
laboratories resulting in genuinely independent, multi- a limited number of results generated on their own
method, and multi-analyser data.The target values are, instruments, using their own reagents and calibrators.
as a result, more accurate and reliable. This can result in perceived accuracy, unrealistic wide
ranges and batch to batch variability.The same is true
for semi-dependent controls.

Some third party controls offer an extended shelf life of Laboratories using an ‘in-kit’ control will receive a new
up to four years.This enables long term QC monitoring lot of QC with each new batch of reagents meaning
and the detection of shifts upon change of reagent they are constantly changing lot number and do not
batch.This can help laboratories save time and money have the benefit of long term QC monitoring.
due to fewer lot crossovers.

5

Case Studies

The following real-life laboratory case studies highlight the importance of third party controls and how they can be used to help
identify instrument, reagent and procedural errors. If errors remain undetected they could result in incorrect patient results,
misdiagnosis, unnecessary treatment, increased costs and may even be potentially life threatening for the patient.
For more information on the benefits of third party controls visit randoxqc.com.

Identifying Lot-to-Lot Variations with Third Party Controls
A laboratory in the UK contacted Randox Technical Services, reporting higher than expected QC results for Thyroglobulin. When
using a third party control (Acusera Immunoassay Premium Plus Control) the results were four times higher on their main analyser
compared to other systems. However, when they ran the instrument manufacturer’s control alongside the third party control it did
not show the same problem.
After reviewing the EQA data, the Technical Services team confirmed there was a significant difference in results between the
instruments, and set about helping the laboratory troubleshoot. After an exhaustive review of procedures and processes, the
customer contacted the instrument manufacturer, who advised of a positive bias with several batches of reagent, including the batch
the laboratory was using.
Conclusion:
By using a third party control the laboratory was able to detect a shift in results after changing reagent batch that the
instrument manufacturer’s control did not.

Overcoming Instrument Errors with Third Party Controls
A laboratory using a third party chemistry control contacted Randox Technical Support after observing a consistent negative bias
for ALT. The instrument manufacturer advised the laboratory that it was a control problem and recommended they use a control
from another manufacturer.
Randox investigated the problem and demonstrated that patient results were also being wrongly reported low. When Randox
ran the test using the control recommended by the instrument manufacturer it failed to highlight the problem. The instrument
manufacturer later recommended a wash stage to the customer which eliminated the carryover interference which was causing the
low ALT results.
Conclusion:
The customer overcame this problem by programming in an extra wash cycle. However, had the lab used the QC
recommended by the instrument manufacturer they would never have identified the problem in the first instance and
would have continued to report inaccurate patient results leading to possible patient misdiagnosis.

Identifying Procedural Errors using Third Party Controls
A laboratory, using the Randox Immunoassay Premium Plus Control, reported that their QC results for Sex Hormone Binding
Globulin (SHBG) were much higher than the target value provided.When the lab ran the reagent manufacturer’s control the results
were within the supplied ranges.
After reviewing the reagent kit insert the Randox Technical Services team noticed that the instrument the laboratory was using
automatically diluted patient samples but not controls. Lab staff were unaware of this and had not been diluting the controls.
Conclusion:
By using third party controls, a procedural error was identified that the reagent manufacturer’s control was masking.
Good laboratory practice is when a quality control material is run in the same way as a patient sample.

6

Acusera Third Party Controls

The Acusera range of true third party controls is extensive, covering over 390 routine and esoteric parameters. All controls are
designed to deliver unbiased performance assessment with any instrument or method.

Accurate Target Values Consolidation

Randox controls have the most accurately With Randox multi-analyte QC, you can
assigned target values and ranges on the dramatically reduce the number of controls
market. This is made possible through access used in your laboratory and save money by
to a unique database of over 31,000 laboratory replacing multiple instrument specific controls

par ticipants. with a single multi-analyser product.

Highest Quality Material Unrivalled Stability

Randox controls are the most commutable Randox lyophilised controls have a shelf life
available providing a matrix that reacts to the of up to four years enabling long term QC
examining system in the same manner as a monitoring. The extended working stability of
patient sample. Our lyophilised controls contain Acusera controls means they meet and often
no added preservatives or stabilisers which can
exceed the stability claims quoted.
cause interference.
7

QUALITY CONTROL

Randox Laboratories Ltd, 55 Diamond Road, Crumlin, County Antrim, BT29 4QY, United Kingdom
+44 (0) 28 9442 2413 +44 (0) 28 9445 2912 [email protected] randoxqc.com

True third party controls offering complete test menu consolidation
Uniquely combining more than 100 parameters conveniently in a single control,
laboratories can significantly reduce costs and consolidate without compromising
on quality. As true third party controls, unbiased performance assessment with any
instrument or method is guaranteed.

24•7

Online QC software with real-time peer group statistics
Compatible for use with the Acusera range of third party controls, the Acusera
24•7 software is designed to help laboratories monitor and interpret their QC data.
Access to an impressive range of features including interactive charts and real time
peer group data generated from our extensive database of laboratory participants,
ensures Acusera 24•7 is the most comprehensive package available.

The largest global EQA scheme with over 31,000 lab participants
Comprising over 360 routine and esoteric parameters in 24 comprehensive and
flexible EQA programmes, RIQAS is designed to cover all areas of clinical testing.
Each programme benefits from a wide range of concentrations, frequent reporting
and comprehensive yet user-friendly reports.

find out more

Information correct at time of print. Randox Laboratories Ltd is a subsidiary of Randox Holdings Limited a company registered within Northern Ireland with company number N.I. 614690. VAT Registered Number: GB JUL15

8 151 6827 08. Product availability may vary from country to country. Please contact your local Randox representative for information. Products may be for Research Use Only and not for use in diagnostic procedures
in the USA.

QUALITY CONTROL

Improving Laboratory Performance Through Quality Control
Five simple steps for QC success

Q
Troubleshooting QC Problems:
Your QC has failed, what do you do next?

Complete QC solutions for results you can trust

Your QC has failed, what do you do next?

So you ran QC this morning and realised that one of your analytes has been flagged
as “out-of-control”, what do you do next? Do you ignore the warning and continue
patient testing, repeat the control until it’s within range or do you halt patient
testing and investigate the source of the error?

When it comes to troubleshooting QC errors, unfortunately there is no easy path to take. However, it’s important that you have
standard operating procedures in place, outlining what to do in the event of an out-of-control error. Errors occur in laboratories all
over the world. A lab with effective troubleshooting procedures in place will still have errors but will be able to detect them, quickly
reducing their impact and reducing the risk of wasting both time and money.

Although there is perhaps no correct way to go about troubleshooting, here are some helpful tips that
your laboratory can use in order to ensure it has effective troubleshooting procedures in place.

1

Put the problem
into perspective
before you begin
troubleshooting

Implement Review your 2
corrective Levey-Jennings

4 actions, check the charts to
effectiveness of the understand the
corrective actions type of error that
and document the has occurred
solution

Relate the type of
error to possible

causes

3

2

1. Put the problem into perspective before you begin troubleshooting

• Outline what you recognise as an out-of-control event
• It’s important to estimate the magnitude and size of the out-of-control event before you

attempt to correct it.
Using QC multi-rules is a great way to ensure sensitive error detection, whilst keeping the false rejection rate low. Make sure you
outline what you recognise as an out-of-control event that warrants corrective action to take place. In the event that a series of
QC multi-rules have been broken, you should halt patient testing immediately until the problem has been rectified. In the event that
only a single rule has been broken, you should repeat the control ensuring you do so only once.
It’s important to estimate the magnitude and size of the out-of-control event before you attempt to correct it. It’s a good idea to
monitor your average patient mean or test a known patient sample.That way, you can measure the extent of the problem and the
effect the out-of-control event has had on patient results.
Measuring the direction and magnitude of the shift in results, can help you decide whether any clinically significant errors may have
occurred and whether or not you need to repeat patient results.

When an out-of-control event has occurred, ISO 15189 requires laboratories to
“evaluate the results from patient samples that were examined
after the last QC event”.

Ensure you know how many samples were run from the last QC event and do not release any patient
results until the problem has been rectified.

3

2. Review your Levey-Jennings charts to understand the type of error that has occurred

Random Errors There are Systematic Errors
two types
of analytical

errors:

Both of these types of errors can be recognised on a standard Levey-Jennings chart and by using QC multi-rules. Identifying the type
of error will help you relate the error to a possible cause.

Systematic Error

Systematic errors create a characteristic bias and can be identified by a change in the mean of control values. The change in the
mean may be either a gradual trend or an abrupt shift in control results.

>+3 >+3

Standard Deviation +2 +2 Standard Deviation

Shift

+1 +1

Trend

00

-1 -1

-2 -2

<-3 <-3
1 2 3 4 5 6 7 8 9 10 12 13 14 15 16 17 18 19 20 1 2 3 4 5 6 7 8 9 10 12 13 14 15 16 17 18 19 20
Run Number Run Number

Random Error

Random errors vary in an unpredictable manner with regard to magnitude and sign.There will always be a degree of random error
associated with any set of QC results. There is an acceptable and unacceptable amount of random error. Using QC multi-rules can
help decide when the amount of random error in your test system has become unacceptable. It is acceptable for 1 in 20 results to
be outside 2SD, anymore than this is an unacceptable amount of random error.

Acceptable amount of random error Unacceptable amount of random error

>+3

+2

+1

Standard Deviation 0

-1

-2

<-3

1 2 3 4 5 6 7 8 9 10 12 13 14 15 16 17 18 19 20 1 2 3 4 5 6 7 8 9 10 12 13 14 15 16 17 18 19 20

Run Number

4

Using QC Multi-rules to Identify Systematic and Random Error
You can use QC multi-rules to differentiate between systematic and random error.

+3sStandard Deviation Standard Deviation +3s

+2s 12s Rule +2s

+1s +1s

M M

-1s -1s
-2s 13s Rule
-2s
-3s
-3s 1 2 3 4 5 6 7 8 9 10
1 2 3 4 5 6 7 8 9 10
Run Number
Run Number
Rule 13s:
Rule 12s: Any QC result outside +/-3s breaks this rule.This rule identifies
This is a warning rule that is broken when a single control unacceptable random error or possibly the beginning of a large
observation is outside the +/-2s limits.This rule warns that either systematic error. If this rule is broken, troubleshooting should
a random or systematic error is present in the test system. If occur to investigate the source of the error.
no other unacceptable results are apparent in the test system,
it must be assumed that this is simply a random error and no
further troubleshooting action is necessary.

Standard Deviation+3s +3sStandard Deviation

+2s 22s Rule +2s
+1s
+1s
M
M
-1s -1s R4s Rule

-2s -2s

-3s -3s
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

Run Number Run Number

Rule 22s: Rule R4s:
This rule is broken when two results are greater than 2s on the If there is at least a 4s difference between control values within
same side of the mean. It is indicative of systematic error. It a single run, this rule is violated and can represent random error.
could indicate the start of a trend or a shift in QC results.

+3s +3s

+2s +2s

Standard Deviation +1s 41s Rule Standard Deviation +1s
M 31s Rule
7x Rule
-1s
M
-2s
-1s
-3s
1 2 3 4 5 6 7 8 9 10 -2s

Run Number -3s
1 2 3 4 5 6 7 8 9 10
Rule 31s:
When three respective control results are greater than 1s and Run Number
on the same side of the mean.
Rule 41s: Rule 7X, 8X, 9X 10X and 12X:
When four results are greater than 1s and on the same side These rules are broken when 7, 8, 9, 10 or 12 QC results are
of the mean. Both these rules are indicative of systematic error on the same side of the mean. This is indicative of a systematic
but don’t necessarily require you to reject the analytical run. As error. Again, these rules don’t necessarily require you to reject
long as this error is not clinically significant there is no need to the analytical run. As long as the error is not clinically significant,
troubleshoot this small amount of error. However, this analytical there is no need to carry out any troubleshooting. However, this
bias can be eliminated by performing calibration or instrument analytical bias can be eliminated by performing calibration or
maintenance. instrument maintenance.

5

3. Relate the type of error to possible causes

Ask yourself questions in order of likely relevance and review the most common solutions to the problem. Dependent on whether
you have identified a systematic or random error in your system, this can help you determine the possible root cause of the error.
See below for some common causes of systematic and random errors.

Systematic Error: Shift

Change or failure in light source 1 6 Change in reagent formulation
Major instrument maintenance 2 7 New reagent lot
Failure in sampling system 3 8 Recent calibration
Change in temperature 4 9 Change in calibration lot
Failure in reagent dispense system 5

Systematic Error: Trend

Slowly deteriorating reagent 1 3 Change in instrument temperature
or control material 2 4 Deteriorating lamp or filter

Calibration shift

Random Error

Bubbles in reagent/sample syringes 1 3 Clog in pipette
2 4 Power supply fluctuations
Improperly mixed/dissolved
reagent/control

6

4. Implement corrective actions, check the effectiveness of the corrective actions and document
the solution

Once you have identified possible causes of the out-of-control event, implement any necessary corrective action. It’s important
that you implement only one change at a time and monitor the improvement of that change on your QC and patient results. It’s
important that you document the solution and learn from your previous laboratory failures. Put procedures in place to prevent any
errors from reoccurring. Remember that a single unacceptable result is most likely due to random error. In this instance re-run the
sample, if the result of repeat analysis is acceptable then corrective actions is not required. If the issue persists, investigate possible
sources of systematic error. See below for some suggested actions that may help to resolve systematic errors:

Perform
instrument
maintenance

Recalibrate
instrument

Perform
staff training

Prepare Review
fresh reagent
/ sample
reagents storage e.g.
and re-run
Check
sample pipettes

In conclusion, make sure you have effective troubleshooting procedures in place. Keeping these few tips in
mind will help assure your laboratory is on the right track when it comes to troubleshooting QC failures.

7

QUALITY CONTROL

Randox Laboratories Ltd, 55 Diamond Road, Crumlin, County Antrim, BT29 4QY, United Kingdom
+44 (0) 28 9442 2413 +44 (0) 28 9445 2912 [email protected] randoxqc.com

True third party controls offering complete test menu consolidation
Uniquely combining more than 100 parameters conveniently in a single control,
laboratories can significantly reduce costs and consolidate without compromising
on quality. As true third party controls, unbiased performance assessment with any
instrument or method is guaranteed.

24•7

Online QC software with real-time peer group statistics
Compatible for use with the Acusera range of third party controls, the Acusera
24•7 software is designed to help laboratories monitor and interpret their QC data.
Access to an impressive range of features including interactive charts and real time
peer group data generated from our extensive database of laboratory participants,
ensures Acusera 24•7 is the most comprehensive package available.

The largest global EQA scheme with over 31,000 lab participants
Comprising over 360 routine and esoteric parameters in 24 comprehensive and
flexible EQA programmes, RIQAS is designed to cover all areas of clinical testing.
Each programme benefits from a wide range of concentrations, frequent reporting
and comprehensive yet user-friendly reports.

find out more

Information correct at time of print. Randox Laboratories Ltd is a subsidiary of Randox Holdings Limited a company registered within Northern Ireland with company number N.I. 614690. VAT Registered Number: GB JUL15

8 151 6827 08. Product availability may vary from country to country. Please contact your local Randox representative for information. Products may be for Research Use Only and not for use in diagnostic procedures
in the USA.

QUALITY CONTROL

Improving Laboratory Performance Through Quality Control
The five simple steps for QC success

Q
2. The Role of EQA in Quality Control

Complete QC solutions for results you can trust

The Role of EQA in QC

Quality Control (QC) is a vital process for laboratories to monitor the accuracy
and precision of patient sample testing. Without QC, laboratory errors could
go undetected potentially resulting in misdiagnosis and inappropriate or delayed
treatment, all of which could be life threatening for the patient.

Taking an holistic view

While Internal Quality Control (IQC) does a vital job it is only part of the overall quality management picture. For example, IQC
doesn’t address calibration issues, instrument systematic errors or wide ‘acceptable’ limits. Equally, while IQC effectively monitors the
reproducibility or precision of laboratory testing, it doesn’t enable laboratories to monitor bias or accuracy. Accuracy refers to how
close the result generated by the laboratory is to the true value of the sample, in order to effectively monitor accuracy laboratories
must participate in a high quality External Quality Assessment (EQA) programme.

External Quality Assessment (EQA), also known as Proficiency Testing (PT)
provides:
“a system of objectively checking laboratory results by means of an
external agency” (World Health Organisation, 1981).

Participation in an EQA
programme is recommended
for all laboratories, indeed ISO

15189 states:
‘the laboratory
shall participate in
interlaboratory comparisons
such as those organised by
external quality assessment

schemes’.

It is important that laboratories participate
in an ISO 17043 accredited EQA scheme
in order to verify the reliability of their
results, IQC alone is not sufficient.

2

Real life examples

The following real life example highlights the importance of participating in a reputable EQA scheme in addition to daily IQC, in
order to ensure laboratory quality, and prevent the release of incorrect patient results.
This laboratory was running the instrument manufacturers QC for their procalcitonin assay. Figure 1 shows that the lab’s results
were always within 1SD and showed no apparent bias. According to these results the laboratory has excellent precision and
accuracy. However, when the lab participated in an EQA scheme (fig 2.) they discovered this was not the case.

Fig 1: Graph of Internal Quality Control

1 2 3 4 5 6 7 8 9 10 12 13 14 15 16 17 18 19 20

Fig 2: Graph of External Quaility Assessment Results

Figure 2 shows the results the lab received from their participation in an EQA scheme. The Levey-Jennings chart on the report
clearly shows erractic performance which led to the identification of an instrument fault, that the IQC failed to detect.

3

How does EQA work?

EQA provides a means of periodically assessing laboratory performance in comparison with other laboratories using the same
method and instrument. Unlike IQC, EQA provides an effective method of monitoring a laboratory’s bias or accuracy through the
analysis of samples of unknown concentrations otherwise known as ‘blind samples’. It also enables assessment of the complete
testing process including pre-analytical, analytical and post analytical phases.

After analysis, results are submitted to the scheme organiser for statistical evaluation. Laboratories will subsequently receive a report
comparing their performance to that of other participants in the programme.

REGISTER 2

1 Register by completing enrolment You will receive a
document and submit at least 3 set of numbered
weeks before the cycle starts samples for the
cycle along with a
username/password
for RIQAS.Net

Analyse the sample on the 3 QUALITY CONTROL
recommended date
1

4 Enter and submit 36.000 mg/dl 5 Receive report
results online 84.800 IU/ml within 72 hours and
before the final 132.000 ng/ml
deadline review to assess
performance

Certificate of Participation 6

<Name> Certificates will be awarded
providing that more than half
<Laboratory/Hospital>
of the results are returned
Stephen Doherty
abc RIQAS Manager PR-281 APR15

Randox International Quality Assessment Scheme
Randox Laboratories Ltd., 55 Diamond Road, Crumlin, Co. Antrim, BT29 4QY, United Kingdom

4

Benefits of EQA

EQA plays an essential role in assuring laboratory quality by supporting daily IQC. It facilitates interlaboratory performance
comparison, bringing greater standardisation in diagnostic testing.

It helps laboratories EQA has
detect equipment a number
failures, identify of functions:
reagent problems
and review staff
competency

It maintains and
improves the analytical

quality of laboratory
tests

It improves It compares different
interlaboratory analytical methods
agreement and helps
raise standards Initiates and
evaluates corrective
Provides an
objective view actions
of test system
performance
that IQC alone
cannot provide

In short, participation in an EQA scheme will give labs greater confidence (and evidence) that the patient results they are reporting
are reliable and accurate. Quality results will reduce time and labour costs, and most importantly provide accurate patient diagnosis
and treatment.

5

What should a laboratory look for when choosing an EQA scheme?

Firstly, make sure the scheme is inspected and accredited to international standards of best practice. It should be accredited to ISO/
IEC 17043:2010, which outlines the general requirements for proficiency testing.Accreditation to this level gives scheme participants
confidence that the scheme is fit for purpose.
The frequency of programme reporting is also important, the
more frequent the better. Some international schemes offer both
bi-weekly and monthly reporting enabling identification of errors
sooner.
It is vital that EQA is performed regularly. Many EQA schemes
operate on a quarterly basis or even less frequently, under these
circumstances laboratories are unable to identify errors in their
test system until they receive their next EQA report which
could be several months away.
In such instances laboratories are faced with the following
questions; when did the error occur and how many patient
samples have potentially been affected?
Some schemes can turn around reports within 24-72 hours,
this allows corrective action to be implemented immediately with
minimal effect on laboratory operations.

“External quality assessment programmes should as far as possible, provide
clinically relevant challenges that mimic patient samples”

ISO15189

A good EQA scheme will have a large number of international participants – the larger the peer group the better.
This ensures an extensive database of results for many analytical methods and increases statistical validity. By participating in an
international scheme (as opposed to a national scheme) laboratories have access to peer data giving a truly global representation
of manufacturer’s instruments and diagnostic kits.

6

As part of your EQA scheme you will be provided with ‘blind’ samples to be analysed as if they were patient samples.Analytes should
be provided in a realistic range of concentrations, allowing assessment of normal and abnormal ranges, both analytical and clinical
decision levels should be covered. It’s important that samples should be stable, but free from interfering preservatives or stabilisers,
and offer a matrix consistent with a human sample. Labs should be confident that EQA material mirrors that of patient samples.
EQA programmes should offer flexibility to customise the programme to suit the laboratory’s unique requirements. Some schemes
offer effective consolidation of programmes through the provision of multi-analyte samples.This allows laboratories to significantly
reduce the number of EQA programmes needed to cover their assay range, thereby increasing productivity and efficiency while still
meeting their EQA needs.
Perhaps the main value of participating in an EQA scheme lies in the report generated by the scheme provider. Not only should
reports be comprehensive covering a wide range of statistics and charts it’s also important they are user-friendly and easy to
interpret. Look out for reports that provide a means of visually assessing your laboratory’s performance over time.

Supplementary EQA schemes

If your existing EQA scheme does not fully meet your
laboratory needs you could benefit from supplementing
your existing scheme with individual programmes
from another international quality assessment scheme.
Completely changing EQA providers can be a tricky
proposition for some laboratories, particularly if you are
part of a lab chain.
By supplementing your existing scheme with individual
programmes from another EQA scheme you could benefit
in a number of ways, for example greater programme
frequency, access to less routine programmes and multi-
analyte programmes.

Conclusion

By looking further than your EQA scheme, you may be able to fulfill your complete EQA
needs by covering your full assay range, increasing productivity and ultimately saving
your laboratory time and money.

7

QUALITY CONTROL

Randox Laboratories Ltd, 55 Diamond Road, Crumlin, County Antrim, BT29 4QY, United Kingdom
+44 (0) 28 9442 2413 +44 (0) 28 9445 2912 [email protected] randoxqc.com

True third party controls offering complete test menu consolidation
Uniquely combining more than 100 parameters conveniently in a single control,
laboratories can significantly reduce costs and consolidate without compromising
on quality. As true third party controls, unbiased performance assessment with any
instrument or method is guaranteed.

24•7

Online QC software with real-time peer group statistics
Compatible for use with the Acusera range of third party controls, the Acusera
24•7 software is designed to help laboratories monitor and interpret their QC data.
Access to an impressive range of features including interactive charts and real-time
peer group data generated from our extensive database of laboratory participants,
ensures Acusera 24•7 is the most comprehensive package available.

The largest global EQA scheme with over 31,000 lab participants
Comprising over 360 routine and esoteric parameters in 24 comprehensive and
flexible EQA programmes, RIQAS is designed to cover all areas of clinical testing.
Each programme benefits from a wide range of concentrations, frequent reporting
and comprehensive yet user-friendly reports.

find out more

Information correct at time of print. Randox Laboratories Ltd is a subsidiary of Randox Holdings Limited a company registered within Northern Ireland with company number N.I. 614690. VAT Registered Number: GB JUN15

8 151 6827 08. Product availability may vary from country to country. Please contact your local Randox representative for information. Products may be for Research Use Only and not for use in diagnostic procedures
in the USA.