CLINICAL ASPECTS

FAILURE MODES AND EFFECTS ANALYSIS (FMEA) –

AN ASSESSMENT TOOL FOR RISK MANAGEMENT IN
CLINICAL LABORATORIES

REMONA ELIZA DAVID1, MINODORA DOBREANU2

1,2
University of Medicine and Pharmacy Tîrgu-Mureș

Keywords: risk Abstract: The present paper aims to be an example of the clinical laboratory risk assessment, though it
management, pre- does not include all the possible risks. This paper presents the evaluation technique FMEA (failure
analytical errors, modes and effects analysis) as a tool for risk management and quality improvement of the clinical
FMEA laboratory analyses. The purpose of FMEA is to aid the clinical laboratory in raising awareness and in
identification of the possible hazardous situations of a testing system. Once the hazardous event has
been identified, the risk can be estimated, analyzed and treated. Using the standard CLSI EP18-A
guidelines, the table FMEA has been laid and, thus the errors from the pre-analytical process,
especially, have been ranked according to criticality.

This paper aims to be an example of risk assessment in ISO 15189 as a system for reducing laboratory errors and
a clinical laboratory, but without the inclusion of all possible improving patient safety by applying the principles of risk
risks. This example does not apply to all proceedings management, with reference to pre-examination, examination
encountered in a clinical laboratory and should not be confused and post-examination processes. ISO/TS 22367:2008 proposes a
with a customized quality control plan of a laboratory, which is methodology for identifying clinical laboratory errors that would
the next step of risk assessment. be avoided with the application of ISO 15189.(3)
For risk assessment, the laboratory collects EP18-A2 is used as a guideline for risk management
information from several different sources: activities describing different techniques to identify and control
• accreditation requirements; laboratory errors.(4)
• information about the measurement system provided by the Information about the measurement system provided
manufacturer; by the manufacturer
• information about the laboratory’s particular environment; A clinical laboratory has the manual of the equipment
• information about health and clinical applications of test available, which contains a number of requirements referring to
results.(1) conditions and environmental precautions established by the
Accreditation requirements manufacturer (e.g. the storage conditions should be between 2 –
Laboratories purchase standards that provide 8° C). The users’ guide comprises a section with possible error
guidelines to be followed: [1] SR EN ISO 15189: 2013, messages that provides information about the equipment or the
“Medical Laboratories - Particular requirements for quality and testing process that can be affected or involved.(5)
competence”; [2] ISO / TS 22367: 2008 “Medical laboratories - Manufacturers determine and describe hemolysis, jaundice,
Reduction of error through risk management and continual lipemia (HIL) indices and alert indices for clinical laboratory
improvement”; [3] Clinical and Laboratory Standards Institute equipment, which use an automatic system for the detection of
(CLSI). “EP18-A2: Risk Management Techniques to Identify HIL.(6)
and Control Laboratory Error Sources”. Reagent product inserts can include sections such as:
SR EN ISO 15189: 2013, section 4.14.6: “The “Limitations”, “Warnings and precautions”, with reference to
laboratory must evaluate the impact of work processes and certain drugs, food or chemicals that can influence test
potential failures on examination results as they affect patient results.(1)
safety, and must modify processes to reduce or eliminate the Information about the laboratory’s particular
identified risks, with records to be kept of decisions and action environment
taken”.(2) The clinical laboratory establishes training
SR EN ISO 15189: 2013, section 4.14.7: “The programmes for staff members by covering different aspects of
laboratory must establish, monitor and periodically review the system.
quality indicators for critical aspects of pre-examination, Information about health and clinical applications of
examination, and post-examination processes. The monitoring test results
processes of indicators will be planned, which include setting Information regarding the usefulness of a test is
objectives, methodology, interpretation, limits, plan of action provided by physicians, which is very important for the clinical
and duration of the measurements. The laboratory must laboratory. Thus, feedback is provided about the performance
periodically evaluate whether or not the indicators meet their limits of the test based on the clinical experience.
effectiveness.”(2) All this information can be used to create a map of the
ISO/TS 22367:2008 characterizes the application of process or a “fishbone diagram”. A map of the process classifies

1
Corresponding author: Remona Eliza David, Str. Păltiniș, Nr. 21-23, Cod 540520, Tîrgu-Mureş, România, E-mail: elizaremona@yahoo.com, Phone:
+40740 168508
Article received on 29.08.2015 and accepted for publication on 30.10.2015
ACTA MEDICA TRANSILVANICA December 2015;20(4):130-134
AMT, vol. 20, no. 4, 2015, p. 130
 CLINICAL ASPECTS

the measurement process in several stages/ components, thus errors identified by the manufacturer may affect the clinical
facilitating the identification of errors which may present the laboratory. Laboratory staff examine whether there are other
source of major risks to patients.(1,5) possible errors, and whether the existing control measures are
Aspects that should be considered when mapping a adequate.(9) In cases where control measures do not reduce the
process for an automatic measuring system: risk to a clinically acceptable level, additional control measures
• training and competence of the operator; are sought that will be implemented later. It is important to
• shipment and storage of reagents and calibrators; include pre-analytical and post-analytical processes in FMEA. A
• assessment of the acceptability of specimens; clinical laboratory may show reluctance to apply FMEA,
• initiation of equipment; because the staff responsible for the development and
• calibration of equipment; implementation of this technique requires experience, and this
• operation of equipment; also imposes multidisciplinary teamwork.
• review and validation of test results.(1,5,7) FMEA needs detailed information for each
phase/component of the system, thus making possible to analyze
Pursuing the process map we can identify possible
errors which belong to different phases/ components of the the modalities in which each phase/component can fail. The
process. It is practical to group possible errors into five major required information is not always readily available.
categories (samples, operators, reagents, environmental The followings should be identified for each
conditions and measuring system), and then represent these in a phase/component of the process:
“fishbone diagram” (a great visual tool that conveys several • the manner how failure/errors can occur;
information subsequently used in a FMEA table).(8) The list of • mechanisms that can produce these errors;
possible errors will be represented in one of the columns of the • effects that can occur after the error is produced;
FMEA table. It is important to take into account that errors can • the severity of effects by deciding if the error is acceptable
also occur in the pre-examination, examination and post- or unacceptable from the clinical point of view;
examination process.(3) • methods to detect failures/errors.(8)
Risk evaluation is a comprehensive process of risk FMEA is a process that is finalized under the form of a
identification, analysis and assessment. ER EN 31010: 2011; table containing the list of possible errors that can occur during
“Risk management - Risk assessment techniques” provides the pre-analytical, analytical and post-analytical processes. The
guidance on selection and application of systematic techniques table can contain information provided by both the manufacturer
for risk assessment. and clinical laboratory describing the importance of each error,
The first step in risk assessment is to identify possible which could be the source of incorrect results or delays in
errors and their causes. Risk management aims at preventing turnaround time.
situations in which errors may occur, so that incorrect results are FMEA working teams should be as conservative as
not reported to clinicians, thus, preventing to cause harm to possible in deciding what is and what is not an error, as a way of
patients. Based on identified errors, a personalized quality prioritizing errors. CLSI EP18-A2 uses numerical quantification
control plan (QCP) is compiled by using the process map or called “criticality” (called “risk” in SR EN ISO 14971:2011).
“fishbone diagram”.(5) According to the ISO/IEC GUIDE 51:2014, “risk is a
The purpose of this paper is to present the FMEA combination of the probability of occurrence of harm/error and
evaluation technique as a tool for risk management and quality the severity of that harm”.
improvement in a clinical laboratory. CLSI EP18-A2; “Risk
Management Techniques to Identify and Control Laboratory Table no. 1. The definition of risk (9,10,11)
Error Sources; Approved Guideline - Second Edition”; 2009 Risk is the the probability of harm and the severity
describes the elements of FMEA (Failure Modes and Effects combination of occurrence of that harm.
Analysis)/ FTA (Fault Tree Analysis) and FRACAS (Failure of
The probability “Harm is According to
Reporting and Corrective Action System) risk assessment of occurrence injury or SR EN ISO
techniques.(4) includes damage to the 14971: 2011,
FMEA aims at helping equipment and reagents exposure to a health of severity is the
manufacturers, as well as clinical laboratories to raise awareness hazardous people, or measure of
and identify possible hazardous situations associated to a testing situation, the damage to possible
system. Once a hazard situation is identified, the risk can be occurrence of a property or consequences
evaluated. In case of critical errors, control measures are hazardous event the of a hazard.
implemented to reduce risks. and the environment.
possibility to ”
FMEA is used and applied in the following situations:
prevent or limit CLSI EP23-
• to increase customer satisfaction (physicians and/or damage. A:2011 says
patients); that damage
• to identify human errors and their effects; to property or
• when the decision to introduce a new product or process in the
the clinical laboratory is taken; environment
is not
• to establish methods of control for newly introduced
considered
processes; harmful
• for existing processes when their goals of improvement is unless
modified in order to ensure greater safety for patients; directly harm
• in case of error analysis in existing processes for people.
continuous quality improvement.(8) How to compile a FMEA table:
FMEA analyzes the situation of a clinical laboratory The columns have been suggested by CLSI E18-A2.
before a measurement system is acquired and subsequently For each of them it has been described the way to fill it in and its
implemented. This allows users to check whether potential specific characteristics.

AMT, vol. 20, no. 4, 2015, p. 131

 CLINICAL ASPECTS

The FMEA table comprises the following After errors are identified in a clinical laboratory, it is
columns:(12) necessary to establish the probability of occurrence for each of
1. Column “phase/component of the process” them. The clinical laboratory should establish a scale for
The “phase/ component of the process” column is assessing the probability of occurrence of the cause. The scale
useful in identifying the phase/component of the process to that is used by our laboratory has been 1 to 5 (5 is more likely
which most of the errors/failures correspond. In order to than 1) and the correlation between the probability of occurrence
maintain the phase/ component of the process a flow diagram of and each number on the scale will be established. Alternatively,
the process should be compiled (e.g. “fishbone diagram” where one of the scales of SR EN ISO 14971: 2011, Annex D can be
errors/failures are grouped into five major categories - samples, used.(9)
operators, reagents, laboratory environment and measuring
system). Table no. 3. Probability of harm (1 – 5) example from our
2. Column “the main sources of errors/failure” laboratory and example according to SR EN ISO 14971:
The “main sources of errors/failure” column is used to 2011, Annex D
list the potential failure modes. Terms Evaluation Practical Example
3. Column “the main causes of potential errors/failure modes” example from according to SR
Using a list of causes of errors, it should be mentioned our laboratory EN ISO 14971:
which errors will be monitored. 2011, Annex D
Frequent 5 Once a day ≥10-3
4. Column “the consequence of errors”
Probable 4 2 – 10 a week <10-3 and ≥10-4
Any error can possibly trigger a cascade of events. For Occasionally 3 Every week <10-4 and ≥10-5
example: concerning the consequence of incorrect laboratory Isolated 2 Once a month <10-5 and ≥10-6
results on the health care of a patient, it should be taken into Improbable 1 Once a year <10-6
consideration what happens after the clinician receives the Note: the clinical laboratory must determine the significance of each term on the scale, this may
report of the medical analysis. The clinical laboratory ought to vary for different tests.

evaluate the severity of the error. For our laboratory it has been 8. Column “detection methods”
decided to use a scale from 1-5. Detection methods do not prevent errors but the
Regarding the health care of the patient, two outcomes occurrence of their effects. It is necessary to raise awareness of
are possible for the above example: the importance of the detection phase because this is similar to
• an incorrect result that may or may not lead to an incorrect the accuracy of the diagnosis of a medical test. Detectability is
medical decision; expressed as the probability of the implemented control process
by the laboratory to detect or prevent an error and can be
• delayed test result, which may or may not affect a patient’s
quantified by using a scale of 1 to 5. The clinical laboratory will
health care.
decide whether detectability is taken into account to estimate
5. Column “evaluated” (yes/no)
criticality.
When the FMEA table presents a generic list of
possible errors, the laboratory should select the errors that
Table no. 4. Scale of detectability (1 – 5)
correspond to the analyzed testing system by adding new issued
Evaluation Practical example
errors or marking with “No” those errors which are not 5 Ineffective control
evaluated. 4 It is unlikely that control measures detect errors
6. Column “the severity of harm” 3 Control measures may or may not detect errors
In a clinical laboratory, the consequences of errors can 2 Control measures almost always detected errors
be as follows: an incorrect test result, a result received late by 1 Control measures can detect errors
the clinician or lack of the requested test result. These situations 9. Column “criticality”
can influence the health care of a patient as a test result can lead For FMEA, criticality is the process of severity
to misdiagnosis which subsequently may be followed by referring to the probability of occurrence of errors. Criticality is
inadequate treatment or lack of appropriate treatment. For each synonymous with risk. The probability of occurrence of an error
possible error, the laboratory must evaluate the severity of harm, and the severity of harm are both descriptive criteria, therefore it
which can be the consequence of this, using a scale with a is difficult to draw up a procedure that combines both of them.
number of levels necessary to cover the range of possible EN ISO 14971: 2011 proposes the outline of risk matrix as
degrees of severity. Too many levels may not result in an follows: the probability levels of error occurrence are
accurate and objective assessment of the severity of harm. This represented on the Y axis, while the severity levels of harm are
estimation requires the decision of the medical laboratory in represented on X axis. Each cell in the table indicates whether
collaboration with clinicians who use these test results. CLSI the risk of the evaluated error is acceptable or unacceptable from
EP23-A, Section 6.3.2.2, describes the severity of harm using the clinical point of view.(4,5) The clinical laboratory should
semi quantitative scale levels of severity of 1 to 5, as suggested correlate the scale levels with the numerical value of criticality
in SR EN ISO 14971: 2011.(5,9) obtained by multiplying the probability level of the occurrence
of the error, the level of the severity of harm and /or
Table no. 2. Scale of severity (1 – 5) detectability. High criticality numbers must be followed by
Evaluati Description of SR EN ISO specific quality control, and then the assessment of residual risk
Terms
on 14971:2011
of harm to decide whether this is clinically acceptable. If the
Catastrophic 5 Deceased patient
Life threatening, permanent harm/
residual risk is still not clinically acceptable, the laboratory must
Critic 4 identify additional measures to reduce the risk. This process is
injuries
Injuries that require medical repeated until the residual risk has been reduced to the clinically
Severe 3 acceptable level. In case the risk was considered “unacceptable”
intervention
Temporary lesion which does not this means that, the adopted control measures are not adequate
Minor 2
require medical intervention to keep the risk at a low level. The interpretation of the results of
Negligible 1 Temporary discomfort the risk matrix should be performed by both clinical laboratory
7. Column “likelihood of occurrence” and clinicians to determine if they are applicable. For example,
AMT, vol. 20, no. 4, 2015, p. 132
 CLINICAL ASPECTS

in a large laboratory, a frequent minor error can be considered QMS06-A3: Quality Management System: Continual
acceptable, because in the context of thousands of results this is Improvement; Approved Guideline-Third Edition, 2011).(13)
admissible. But a serious error that occurs occasionally cannot The laboratory should not choose only those quality indicators
be accepted. that are aimed for laboratory processes but also for the processes
10. Column “prevention” that extend beyond the laboratory and those that they
Preventive measures involve modifying a encounter.(14)
phase/component of the process. If the error is prevented, the Various organizations, regulations, standards and/or
consequence of the error is eliminated. contracts may influence the laboratory’s choice of quality
11. Column “measures of continuous improvement” indicators. The laboratory shall monitor all aspects of the testing
Control measures (prevention, detection or process (pre-examination, examination and post-examination) in
improvement) implemented by the laboratory to minimize the addition to the general administrative processes. The indicators
risk usually cannot change the severity of the errors, they can are defined by each laboratory, but it needs to cover all aspects
only reduce the probability of occurrence of errors. mentioned.(14) It would be ideal to be able to monitor all
12. Column “evaluation - Six Sigma indicator” laboratory processes, but it is not practical too. Risk
SR EN ISO 15189: 2013N, Section 4.14.7, says that “a management tools can be used to select specific indicators of
clinical laboratory should establish quality indicators for certain critical components that can be effectively implemented
systematically monitoring and evaluating the centre’s and monitored by a medical laboratory (CLSI EP18-A2).(4)
contribution to patient care”. According to the same standard, Laboratory management should be sure that the selected
Section 3.19, the quality indicator is defined as “the degree to indicators allow to measure a wide variety of non-conformity. In
which a set of inherent characteristics fulfils particular order to assess the quality of each indicator selected by the
requirements”. Measurements could be expressed as percentage, laboratory, a worksheet can be compiled in which the
number of defects per million opportunities (DPM) or on the Six recommended actions, the limitations of these actions as well as
Sigma scale. For the selected indicators, the Six Sigma value 4 – the activities of the quality control plan are mentioned.(14)
5 shows that the assessed laboratory processes are well Table no. 5 is an example of the proposed FMEA to
supervised and stable over time. Developing measurement our laboratory, which was outlined by taking into account the
principles is a strategic point of a clinical laboratory. Without indications of the CLSI EP18-A2 standard, and where probable
measurement performance cannot be evaluated objectively. The errors from pre-analytical process were described according to
performance of the process should be an important part of the the internal and external context of the laboratory.
program of continuous improvement in the laboratory (CLSI

Table no. 5. Example of FMEA

Phase/ component of the

Main causes of potential

Measures of continuous
Measures of detection
Main sources of error

Evaluated (yes/no)

Six Sigma quality

Severity of harm
Effect of error

Probability of

improvement
failure mode

occurrence

Evaluation
Prevention
Criticality

indicator
process

Sample Sampling Incorrect Results from Yes 2 4 The operator must 24 Revision of the tagging Staff training ---
results due to another confirm the patient (ID process.
incorrect patient and name) by inserting Retraining staff
identification his/her birthday or
data personal ID number in
the device before testing
is initiated.
3
Lack of request The results Yes 3 4 Staff with experience 24 The barcode system Limiting the access 5.2
or incorrect are useless and appropriate training ensures the proper of specialized
interpretation for patient can confirm two IDs for registration of requests. personnel to file
of the request care a patient before testing. Errors due to barcodes are registration by using
for laboratory 3 rare but due to manual password to enter the
analysis registration of requests, computer system.
errors are more common.
Presentation Sample Sampling Yes 1 2 5 10 Retraining staff Staff training 4.7
of samples collected in must be
unsuitable repeated
vacutainer/
additives
Inadequate Delayed Yes 2 4 The detector identifies 16 Retraining staff involved in Monitoring the 4.4
volume diagnosis the samples with low sampling (recommendation tendency of accepting
and volume. The sensor for coagulation and samples with low
treatment detects a meniscus hematology tests, where volume by tracking
waiting for a generating electricity, volume is a criterion for the frequency of error
second thus ensuring that the rejecting the sample). messages of the
sampling volume is at an Visual examination can measurement system
with appropriate level. identify low volume and personnel
adequate Otherwise, a system samples before they are training.
volume. error occurs and the placed in the analyzer
result will not be (recommendation for
released. biochemical tests).
2
Conformity of Hemolysis Delayed Yes 3 4 1 12 Personnel training on HIL Automatic 4
samples diagnosis sampling techniques, detection system. The
and respectively, on the development of a
treatment interference of hemolysis study to establish the
waiting for a with some biochemical limits of interference
second parameters. of HIL on
sampling. quantitative
determinations.

AMT, vol. 20, no. 4, 2015, p. 133

 CLINICAL ASPECTS
Lipemia Delayed Yes 3 4 1 12 Training personnel and HIL Automatic 3.9
diagnosis informing patients about detection system. The
and the importance of blood development of a
treatment collection for medical study to establish the
waiting for a analysis in the morning, limits of interference
second after fasting for 8-10 hours. of HIL on
sampling. quantitative
determinations.
Icterus Interference Yes 3 4 1 12 --- HIL Automatic 4.5
on the detection system. The
determinatio development of a
n of certain study to establish the
analytes limits of interference
of HIL on
quantitative
determinations.
Blood Incorrect Yes 4 2 Micro-clots cannot be 16 Implementation of a pre- Staff training 4.5
coagulation results detected examination procedure,
2 which assumes the
assessment of quality
sampling and personnel
training for this procedure.
The assessment of the
frequency of system errors
caused by the presence of
blood clots and retraining
staff responsible for blood
sampling.
Testing staff responsible
for sampling and
processing of samples.
Reagents/ Reagents Incorrect Incorrect Yes 3 3 Testing QC samples that 27 Reagent storage condition Continuous ---
calibrators results due to results were received by the recommended by the monitoring of
/ controls damage to laboratory on the manufacturer is between temperature by using
reagents during occasion of another 2˚C - 8˚C. alarm systems for
transport transport than reagent There is no certainty that refrigerators in the
transport. these conditions were laboratory.
Damaged QC samples maintained during
can lead to the incorrect transportation.
assessment of the It is recommended to check
integrity of the reagent. the temperature for cool
3 boxes during transportation
(thermogram).
Calibrators Incorrect Incorrect No 1 1 The user’s manual 1 The manufacturer Continuous ---
results due to results released by the recommends the monitoring of
damaged manufacturer describes a verification of the temperature by using
calibration process of internal calibration performance by alarm systems for
materials control that monitors analysing the control refrigerators in the
calibrator absorbance samples. laboratory.
values, ensuring that the More frequent analysis of
value of the controls can provide the
measurement for laboratory assurance that
calibration is between the system is stable and
predefined acceptable calibration is not diverted.
limits.
1

REFERENCES evaluare a riscurilor. Acest standard este identic cu

1. Clinical and Laboratory Standards Institute (CLSI). EP23- standardul european EN 31010:2010; 2011.
A-WB: A Practical Guide for Laboratory Quality Control 9. International Organization for Standardization. ER EN ISO
Based on Risk Management USA; 2011. 14971:2009. Dispozitive medicale. Aplicarea
2. International Organization for Standardization. ISO managementului de risc la dispozitivele medicale; 2011.
15189:2012. Medical laboratories – Requirements for 10. Asociația de Standardizare din România (ASRO). SR Ghid
quality and competence; 2012. ISO 73:2010. Managementul riscului – Vocabular. Acest
3. International Organization for Standardization. ISO/TS standard este identic cu ghidul internațional ISO GUIDE
22367:2008. Medical laboratories – Reduction of error 73:2009; 2010.
through risk management and continual. Geneva. Published 11. International Organization for Standardization ISO/IEC
in Switzerland; 2008. GUIDE 51:2014. Safety aspects – Guidelines for their
4. Clinical and Laboratory Standards Institute (CLSI). EP18- inclusion in standards. Geneva. Published in Switzerland;
A2: Risk Management Techniques to Identify and Control 2014.
Laboratory Error Sources; Approved Guideline – Second 12. Clinical and Laboratory Standards Institute (CLSI). EP18-
Edition. USA; 2009. A2: Risk Management Techniques to Identify and Control
5. Clinical and Laboratory Standards Institute (CLSI). EP23- Laboratory Error Sources; Approved Guideline – Second
ATM: Laboratory Quality Control Based on Risk Edition. USA; 2009.
Management; Approved Guideline. USA; 2011. 13. Clinical and Laboratory Standards Institute (CLSI).
6. Clinical and Laboratory Standards Institute (CLSI). C56-A: QMS06-A3: Quality Management System: Continual
Hemolysis, Icterus, and Lipemia/Turbidity Indices as Improvement; Approved Guideline-Third Edition; 2011.
Indicators of Interference in Clinical Laboratory Analysis; 14. Gimenez-Marin A, Rivas-Ruiz F, Perez-Hidalgo MM,
Approved Guideline. USA; 2012. Molina-Mendoza P. Pre-analytical errors management in
7. Clinical and Laboratory Standards Institute (CLSI). EP23- the clinical laboratory: a five-year study. Biochemia
A-WS: A Sample Form for Laboratory Quality Control Medica. 2014;24(2): 248-257.
Based on Risk Management. USA; 2011.
8. Asociația de Standardizare din România (ASRO). SR EN
31010, CEI/ISO 31010. Managementul riscului. Tehnici de

AMT, vol. 20, no. 4, 2015, p. 134