invited review

Laboratory Quality Control Based on Risk

Management
James H. Nichols
From the Tufts University School of Medicine, Springfield, MA, USA

Correspondence: James H. Nichols, Ph.D., DABCC, FACB · Professor of Pathology Tufts University School of Medicine Medical Director,
Clinical Chemistry Baystate Health 759 Chestnut Street Springfield, MA 01199, USA · T: +1 413 794-1206 F: +1 413 794-5893 · james.nichols@
baystatehealth.org

Ann Saudi Med 2011; 31(3): 223-228

PMID: **** DOI: 10.4103/0256-4947.81526

Risk management is the systematic application of management policies, procedures, and practices to the
tasks of analyzing, evaluating, controlling and monitoring risk (the effect of uncertainty on objectives).
Clinical laboratories conduct a number of activities that could be considered risk management including
verification of performance of new tests, troubleshooting instrument problems and responding to physi-
cian complaints. Development of a quality control plan for a laboratory test requires a process map of
the testing process with consideration for weak steps in the preanalytic, analytic and postanalytic phases
of testing where there is an increased probability of errors. Control processes that either prevent or im-
prove the detection of errors can be implemented at these weak points in the testing process to enhance
the overall quality of the test result. This manuscript is based on a presentation at the 2nd International
Symposium on Point of Care Testing held at King Faisal Specialist Hospital in Riyadh, Saudi Arabia on
October 12-13, 2010. Risk management principles will be reviewed and progress towards adopting a
new Clinical and Laboratory Standards Institute Guideline for developing laboratory quality control
plans based on risk management will be discussed.

Introduction to Risk and Risk Management familiar in the clinical laboratory, but technical staff

R
isk is defined in ISO31000 as the effect of un- and laboratory directors conduct a number of activities
certainty on objectives, whether positive or nega- that could be considered risk management in the day-
tive.1 For healthcare, risk is generally understood to to-day operations of a laboratory. The performance of
mean the chance of suffering or encountering harm or loss.2 new tests is evaluated before use in patient care, control
So, risk is essentially the potential for harm to occur to a sample failures are investigated for instrument and re-
patient or the possibility of an error that can lead to patient agent problems, and management responds to physician
harm. Risk can be estimated through a combination of the complaints. When incorrect results are reported, the staff
probability of harm and the severity of that harm.3 There must determine and correct the cause, and report the cor-
are two methods to reduce the risk of harm to a patient: rect results. If patients were treated based on incorrect re-
• prevent the error from occurring which averts harm to sults, management must estimate the harm that occurred
the patient, or to the patient and take steps to prevent similar incidents
• detect the error before it can harm the patient. in the future. So, risk management is not a new concept,
Risk management is the identification, assessment, and just a formal description of activities that laboratories are
prioritization of risks followed by coordinated and eco- already doing as part of their quality assurance program
nomical application of resources to minimize, monitor, and to prevent errors and reduce harm to a patient.
control the probability and/or impact of unfortunate events
or to maximize the realization of opportunities.4,5 Risk Sources of Laboratory Error
management is essentially the systematic application of Understanding weaknesses in the testing process is a
management policies, procedures, and practices to the tasks first step to developing a quality control plan based on
of analyzing, evaluating, controlling and monitoring risk.6 risk management. Laboratories should create a process
The terms risk and risk management may seem un- map that outlines all the steps of the testing process

Ann Saudi Med 31(3)  May-June 2011  www.saudiannals.net 223

 invited review LABORATORY QUALITY CONTROL

from physician order to reporting the result. A process phases of testing, outside of the laboratory and beyond
map basically follows the path of the sample from the the supervision of laboratory staff.7 Preanalytic concerns
patient through transportation, receipt and analysis in should include the physician order (is there an order and
the lab to reporting of result. This process map should was it transcribed correctly), was the patient prepared for
include preanalytic, analytic, and postanalytic processes the test (if fasting or withholding medication is required),
required to generate a test result that can be acted on by was the appropriate preservative used to collect the speci-
a clinician. men, was the specimen collected at the expected time,
Weak steps in the testing processes are those that and was the specimen promptly transported to the labo-
have a higher probability of generating an error. These ratory, (protected from freezing or heating)? Reagents,
can be identified through prior experience with similar controls, calibrators and other testing supplies must be
instrumentation or from information collected from the shipped to the laboratory where they may be exposed to
manufacturer and other users about the test and method, conditions (heating and freezing) that could compromise
how the test will be utilized in diagnosing or managing test results. Postanalytical processes should consider how
the patient, the laboratory environment and staff who the test was reported and communicated to the ordering
will perform the test, and local regulatory and accredita- physician since manual transcription is prone to more er-
tion requirements that mandate control over specific as- rors than automated reporting systems using computer
pects of the testing process (Figure 1). This information interfaces. But computers are also not foolproof, and
will be utilized to develop a quality control plan specific instrument interfaces have been known to occasionally
to the device, the laboratory, and the health-care setting report incorrect results to the wrong patients or associate
that reduces the risk of harm to the patient, and meets results with the wrong test due to glitches in the interface
regulatory requirements for quality of testing by the lab- communication. Verbal communication also has the po-
oratory. To identify weaknesses in the testing process that tential to be misunderstood and communication of criti-
could lead to error, laboratories need to acknowledge that cal, life-threatening values can be mixed-up unless the
all medical devices can fail when subjected to the right results are written down and confirmed by read-back to
conditions (environment, operator or device sources of the caller. The prevalence of errors in the preanlytic and
error). Realizing those conditions that may cause device postanalytic phases does not mean that the laboratory
failure and taking steps to protect a testing device from and analytic phase is free of errors. Failure to verify in-
exposure to those conditions is the foundation of a qual- strument performance prior to patient testing, incorrect
ity control plan. maintenance, wrong calibrator setpoints, use of expired
Most errors occur in the preanalytic or postanalytic reagents, aliquotting errors, flawed calculations and dilu-
tion factors can all be sources of analytic error that occur
within the laboratory.
There are, thus, many sources of error to consider
throughout the testing process. Of primary consider-
ation is the impact of the environment, the operator, and
the analysis on the quality of test results. Temperature
can freeze or overheat sensitive reagents and compro-
mise results, but so can humidity, light and even altitude.
Operators can inadvertently misidentify a patient and
associate the labeling of the sample with the wrong pa-
tient. So the optimum laboratory control processes will
be worthless if the specimen is mislabeled with another
patient’s identification. Testing by clinical personnel at
the point of care is more prone to errors than analyses
conducted by experienced laboratory professionals with
training in error recognition and prevention. Analyzers
can fail despite proper operation if incorrect calibrator
factors are programmed or samples are incorrectly ap-
plied, aliquoted or diluted. So, considerations for com-
Figure 1. Process to develop and continually improve a quality control plan. (Obtained
with permission from CLSI. Laboratory Quality Control Based on Risk Management; mon sources of environmental, operator and analyzer
Proposed Guideline. CLSI document EP23-P. Wayne, PA: Clinical and Laboratory error should be considered when developing a quality
Standards Institute; 2010.) control plan.

224 Ann Saudi Med 31(3)  May-June 2011  www.saudiannals.net

LABORATORY QUALITY CONTROL invited review
Control Samples sis can work for low volume patient testing where test
Weak steps in the testing process are sites for control results may not be needed on an immediate basis by the
processes to either prevent errors before they hap- physician. With high volume automation and stat test-
pen or detect errors before they can harm the patient. ing, patient results are continuously released using au-
Historical quality control arose from the industrial set- toverification rules based on periodic analysis of control
ting where factories analyzed samples of a product to samples interspersed with patient specimens. Should a
ensure that the product met specifications and the pro- control sample fail to recover expected results, the labo-
duction line was operating as expected. In the labora- ratory must stop patient testing, correct the problem,
tory, a control sample of known analyte concentration, then reanalyze patient specimens back to when the sys-
sometimes called a “quality control” sample or “QC” tem was reporting acceptable results. Repeat testing can
sample, is analyzed like a patient sample. If the instru- be expensive for both labor and reagent costs. Once the
ment produces a result within an acceptable tolerance of problem is found and corrected, some results may need
the target concentration, then the measurement system to be corrected, and this can lead to physicians ques-
is assumed to be stable and operating as expected. The tioning the quality of the laboratory.
final result from measurement system is the sum of all Control samples, however, are a good means of de-
factors affecting the result including the instrument, re- tecting systematic errors, but perform poorly in detect-
agent, operator and environment. ing random errors. Systematic errors are those that af-
Control samples have a number of advantages and fect every test in a constant and predictable manner.
disadvantages. A control sample is a sample with known Many instruments utilize bottles of liquid reagent to
concentration that consists of a matrix similar to a pa- perform hundreds of tests. These reagents may stay on
tient sample, like plasma, serum or urine. Unfortunately, an instrument for a number of hours to days, so analysis
not all analytes are stable in a biological matrix, like glu- of control samples periodically or with each day of test-
cose in whole blood or blood gases. So, preservatives ing confirms that the reagent is still viable. Analysis of
and other stabilizers must be added to control samples control samples does a good job at detecting errors that
to ensure the analyte recovery and test results are stable affect control samples the same as patient samples, like
over time. The additives that stabilize control samples reagent deterioration, errors in reagent preparation, im-
can change the manner in which some instrumentation proper storage, incorrect operator technique, and wrong
interacts with the specimen, such that control samples pipette or calibration settings. However, if a single pa-
behave differently than patient samples in the same test. tient sample should have a clot or drug that interferes
This non-commutability of control samples prevents with the reagent, the analysis of control samples will not
their use as accuracy based materials in determining be impacted and cannot detect the error. Such random
bias between instruments of different makes and manu- errors affect individual samples in an unpredictable
facturers, unless the sample is certified commutable and fashion, like clots, bubbles and interfering drugs and
accuracy-based. However, stabilized control samples substances. Analysis of control samples does a poor job
do offer a target range for analytes specific to make at detecting random errors.
and model of measurment system that allow their use Optimally, the laboratory needs control processes
in determining ongoing stability of laboratory instru- that function more than periodically. The laboratory
mentation. Control samples can be analyzed each day needs to get to fully automated analyzers that prevent
of testing (or more frequently for high volume testing) errors upfront and provide assured quality with every
and if the test recovers the expected target results, then specimen.
the laboratory staff know that the system is stable and
patient results are acceptable. Other Control Processes
Unfortunately, when control samples fail to recover Newer instrumentation has a variety of control pro-
expected results, something has failed in the test- cesses built-in the device. There are analyzers with elec-
ing process and the laboratory must troubleshoot the tronic controls and system checks performed automati-
source of failure and correct it before patient testing can cally to detect electronic operation as well as reagent
resume. Troubleshooting will work if testing occurs in function. Blood gas analyzers, like the Instrumentation
batches of specimens where results can be held until Laboratories GEM and Radiometer’s ABL80, detect
control results (analyzed with each batch of specimens) baseline sensor signals before and after each specimen.
are examined and compared to expected target concen- If each sensor does not display a characteristic sig-
trations. If the control results are acceptable, then the nal, the flow cells may be blocked by a clot or bubble,
patient results can be released. This type of batch analy- and the system can initiate corrective action flushing,

Ann Saudi Med 31(3)  May-June 2011  www.saudiannals.net 225

 invited review LABORATORY QUALITY CONTROL

back-flushing the flow cell until the sensors return to and the laboratory.8 Information about the function of
expected operation. If not, the analyzer can shut down instrument control processes is needed from the manu-
individual sensors or the entire sensor array until man- facturer to increase the user’s understanding of overall
ual staff intervention. These processes occur with each device quality assurance requirements and so informed
sample to detect certain types of errors, specimen clots decisions can be made regarding which control processes
and bubbles that may block specific sensors specimen are suitable for certain errors in the laboratory setting.8
flow path. Other types of control processes may be en- Some sources of error may be detected automatically by
gineered by the manufacturer into each test card or strip the device and prevented, while others may require the
of unit-use testing devices, like the positive/negative laboratory to do something, like analyze control samples
control area on stool guaiac cards, and the control line periodically and on receipt of reagent shipments, or per-
on pregnancy, rapid strep and drug tests. These built- form specific maintenance. Clear communication of po-
in controls test the viability of the reagents on the test tential sources of error and delineation of laboratory and
(storage and expiration), adequate sample application, manufacturer roles for how to detect and prevent errors
absence of interfering substances (clots, viscous urines, is needed in order to develop a quality control plan.
and drug adulterants), as well as timing and appropri-
ate visual interpretation of test results by the operator Developing a Quality Control Plan
with each test. Still other types of control processes The Clinical and Laboratory Standards Institute (CLSI)
may include barcoding of reagents to prevent use past is developing a guideline; Laboratory Quality Control
the expiration date, device lockouts that prevent opera- Based on Risk Management. This guideline, EP23, de-
tion if control samples have not been analyzed or fail scribes good laboratory practice for developing a qual-
to recover expected target concentration, security codes ity control plan based on manufacturer’s risk mitigation
to prevent inadvertent changes to instrument settings, information, applicable regulatory and accreditation
and disposable pipette tips to prevent sample carryover. requirements, and the individual healthcare and labora-
There are thus a variety of control processes available tory setting. Information collected about the instrument
on different models of laboratory instrumentation, and from the manufacturer, peer literature and other users
each process is intended to reduce the risk of specific of the product is combined with information about the
types of errors. individual healthcare and laboratory setting, and the
No single control process can therefore cover all de- unique regulatory and accreditation requirements and
vices and types of errors. Laboratory devices differ in processed through a risk assessment to develop a labora-
design, technology, function and intended use. Some tory specific quality control plan (Figure 1). This plan
devices have internal checks which are performed au- is an optimized balance of control sample analysis com-
tomatically with every specimen, while the possibility bined with manufacturer engineered control processes
of other errors is reduced through engineering by the in the instrument and laboratory implemented control
manufacturer into the device. For example, the barcod- processes to minimize risk of error and harm to a pa-
ing of expiration date and lot number on each bottle tient when using the instrument for laboratory testing.
of reagent prevents use of reagents past their stamped Once implemented, the QC plan is monitored for con-
expiration date and the requirement of entering lot tinued errors and physician complaints. When trends
number into the instrument reduces the possibil- are apparent, the source of errors is investigated and this
ity of using a lot number whose performance has not new information is processed through a new risk assess-
been previously verified. Barcoding of expiration date ment to determine if changes to the QC plan are needed
and lot number, however, does not verify the stability to maintain risk to a clinically acceptable level. This is
of reagents once they are opened on the instrument. the corrective action and continual improvement cycle
Periodic analysis of control samples may better deter- (Figure 1).
mine open bottle stability. So, the historical analysis of A risk assessment starts with identification of a po-
control samples has provided labs with some degree of tential risk or error (called hazard identification). Once
assurance and analysis of control samples over the past identified, the probability and severity of harm are es-
several decades and will continue to plan an important timated. Take, for example, the risk of an untrained
role in future quality assurance in combination with the operator using a point-of-care testing device. The haz-
built-in controls and on-board chemical and biological ard is “operation by an untrained operator”. The prob-
control processes found in newer devices. ability of harm can be estimated as frequent=once
Developing a quality plan surrounding a laboratory per week, probable=once per month, or remote=once
device requires a partnership between the manufacturer per year or greater. The severity of harm if the device

226 Ann Saudi Med 31(3)  May-June 2011  www.saudiannals.net

LABORATORY QUALITY CONTROL invited review
is run by an untrained operator is unknown, but could must be implemented by the laboratory to reduce the
be serious=injury or impairment requiring medical in- risk to a clinically acceptable level. The sum of all risks
tervention, rather than negligible=inconvenience/dis- identified and control processes to mitigate those risks
comfort or minor=temporary injury or impairment not (manufacturer provided and laboratory implemented)
requiring medical treatment. The risk can be estimated becomes the laboratory’s QC plan specific to this de-
by combining probability of harm with severity of harm vice and laboratory setting. This plan is then checked
in a simple 3×3 matrix to evaluate the clinical accept- against regulatory/accreditation requirements to ensure
ability of the risk (Table 1). More detailed 4×4, 5×5, or conformance with recommendations and signed by the
even 10×10 matrices can be developed to estimate risk, laboratory director as the QC plan for that instrument.
but these necessarily require more granular determina- A single QC plan may cover multiple tests on the same
tions of the exact probability and severity of harm to instrument or refer to multiple instruments of the same
the patient. In risk management literature, the ability to make/model within an institution depending on the
detect an error, detectability, is also factored into the es- clinical application of the specific test results and avail-
timate, but for simplicity, detectability can be assumed to ability of manufacturer control processes on the device.
be zero or worse case scenario. Thus, risk in our simple Once implemented, the QC plan is monitored for
example will depend only on prevention of the error or continuous improvement. Physician complaints, instru-
severity of harm if an error occurs. ment failures, or increasing error trends warrant inves-
In our example, the risk of an untrained operator us- tigation, corrective action and a new risk assessment.
ing the device will depend on the setting. In a central The laboratory should ask if the cause of the error is a
laboratory where the test is performed by medical tech- new risk/hazard not previously considered, or greater
nologists, all who are well supervised and experienced, probability of error or severity of harm than estimated
the probability of this hazard is remote. However, in a in the initial risk assessment. This new information can
point-of-care setting where operation is by a variety of now be factored into the risk analysis to determine if
clinical staff, possibly in a shared department location current control processes are adequate to reduce risk
where there is little supervision, anyone can walk up to to a clinically acceptable level or if additional control
the device and attempt to run a test. In this setting the processes are required. In this manner, the laboratory’s
risk of an untrained operator performing testing is much QC plan defines a strategy for continuous improvement
greater (probable to frequent probability of harm). In and sets benchmarks for monitoring the effectiveness of
both settings, if an error occurs from operation by an the QC plan through the frequency of complaints, in-
untrained operator, the harm could be serious to the pa- strument failures and trends in error rates. The CLSI
tient (depending on the test and how the test is used in EP23 document describing how to develop laboratory
medical management). Combining the probability and QC plans based on risk management is currently in the
severity of harm, we can estimate the risk in a laboratory committee voting stages towards publication as an ap-
setting to be clinically acceptable, but in the point-of- proved guideline.
care setting risk is unacceptable and additional control
measures will be needed to reduce the risk (Table 1). Summary
Now, consider an instrument that has operator lock- Risk management is an industrial process for managing
out features where the operator must enter their identifi- the potential for error, and although the terminology
cation before the instrument will unlock and allow test- may be unfamiliar, the clinical laboratory performs a va-
ing. If only operator identifications on a list of trained riety of risk management activities in day-to-day opera-
operators can unlock the instrument, the probability of tion. The CLSI EP23 guideline simply formalizes the
harm may be reduced to probable or even remote, and process. Newer laboratory instrumentation and point-
thus the risk of using such a device with operator lock- of-care testing devices incorporate a number of control
out features is now clinically acceptable. This example processes, some that rely on traditional analysis of con-
demonstrates how a manufacturer engineered process,
like operator lockout, can reduce the risk of certain er- Table 1. Risk Acceptability Matrix
rors to improve the quality of test results. Severity of Harm
This risk assessment process is repeated for each risk Probability of Harm Negligible Minor Serious
or potential for error identified through weak steps in Frequent Unacceptable Unacceptable Unacceptable
the process map. If the risk of error from manufacturer
Probable Acceptable Unacceptable Unacceptable
recommended and engineered control processes is not
Remote Acceptable Acceptable Acceptable
clinically acceptable, then additional control processes

Ann Saudi Med 31(3)  May-June 2011  www.saudiannals.net 227

 invited review LABORATORY QUALITY CONTROL

trol samples, and others engineered into the device to resents the laboratory’s QC plan, a plan that is scien-
check, monitor or otherwise control specific aspects of tifically supported by the risk assessment process. Once
the instrument operation. Risk management is not a implemented the effectiveness of the QC plan is moni-
means to reduce or eliminate the frequency of analyz- tored through trends in error rates, and when issues
ing control samples, since laboratories must minimally are noted, corrective action is taken, risk is reassessed,
meet manufacturer recommendations and accreditation and the plan is modified as needed to maintain risk to a
agency regulations. Rather risk management helps lab- clinically acceptable level. In this manner, risk manage-
oratories find the optimal balance between traditional ment promotes continuous quality improvement. CLSI
quality control (the analysis of control samples) and EP23 thus translates the industrial principles of risk
other control processes such that each risk in operating management for practical use in the clinical laboratory,
the instrument is rationalized with a control process to and will be a useful in support of the laboratory’s overall
reduce that risk. The sum of all control processes rep- quality management systems.

REFERENCES
1. International Organization for Standardiza- neva, Switzerland. 1999; p10. of risk management to medical devices. ISO,
tion. ISO31000: Risk management – Principles 4. International Organization for Standardiza- Geneva, Switzerland. 2007; p 82.
and guidelines. ISO, Geneva, Switzerland. 2009; tion/International Electrotechnical Commis- 7. Bonini P, Plebani M, Ceriotti F, Rubboli
pp 24. sion. ISO/IEC Guide 73, Risk management – Vo- F. Errors in laboratory medicine. Clin Chem
2. Webster’s Dictionary and Thesaurus. Land- cabulary. ISO, Geneva, Switzerland. 2009; p 15. 2002;48:691-8.
oll, Ashland, OH, USA. 1993. 5. Hubbard D. The Failure of Risk Management: 8. International Organization for Standardiza-
3. International Organization for Standardiza- Why It’s Broken and How to Fix It. John Wiley & tion. ISO15198. Clinical laboratory medicine: In
tion/International Electrotechnical Commis- Sons, Indianapolis, IN, USA. 2009; p. 46. vitro diagnostic medical devices. Validation of
sion. ISO/IEC Guide 51, Safety aspects - Guide- 6. International Organization for Standardiza- user quality control procedures by the manu-
lines for their inclusion in standards. ISO, Ge- tion. ISO14971: Medical devices - Application facturer. ISO, Geneva, Switzerland. 2004; p 10.

228 Ann Saudi Med 31(3)  May-June 2011  www.saudiannals.net