Clinica Chimica Acta 404 (2009) 7578

Contents lists available at ScienceDirect

Clinica Chimica Acta

j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c l i n c h i m

FMEA: A model for reducing medical errors

Maria Laura Chiozza a,, Clemente Ponzetti b
a
Quality management Service, University-Hospital of Padua, Italy
b
Gruppo Policlinici di Monza, Monza, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Patient safety is a management issue, in view of the fact that clinical risk management has become an
Received 10 March 2009 important part of hospital management. Failure Mode and Effect Analysis (FMEA) is a proactive technique for
Accepted 10 March 2009 error detection and reduction, rstly introduced within the aerospace industry in the 1960s. Early
Available online 17 March 2009
applications in the health care industry dating back to the 1990s included critical systems in the development
and manufacture of drugs and in the prevention of medication errors in hospitals. In 2008, the Technical
Keywords:
Failure mode and effect analysis
Committee of the International Organization for Standardization (ISO), licensed a technical specication for
Root cause analysis medical laboratories suggesting FMEA as a method for prospective risk analysis of high-risk processes. Here
Risk management we describe the main steps of the FMEA process and review data available on the application of this
Total testing errors technique to laboratory medicine. A signicant reduction of the risk priority number (RPN) was obtained
Laboratory medicine when applying FMEA to blood cross-matching, to clinical chemistry analytes, as well as to point-of-care
Patient safety testing (POCT).
2009 Elsevier B.V. All rights reserved.

1. Introduction clinical laboratory professionals must exploit their knowledge and

Assuring patient safety before an injury occurs is the concern of all
professionals involved in the patient care; patient safety is therefore,
rst and foremost, an issue intrinsically related to professional
identity. However, patient safety is also a cause of immense concern
to all health care systems, because the public has lost its condence in
their ability to provide safe services. It is now particularly clear that
the traditional health care system's reliance on competent people to
do the right thing has not fullled the intended purpose: patients
continue to experience adverse events and medical mishaps occur at
alarmingly high rates [1,2].
Finally, patient safety is also a management issue, in view of the
fact that Clinical Risk Management has become an important part of
hospital management. Reducing the probability of risk in hospitals is
of vital importance in improving the quality of health care, relation-
ships between hospital staff and patients and patient compliance, and
also in limiting malpractice litigation [3].
The report of the Institute of Medicine To err is human: building a
safer health system underlines that health care lags a decade or so
behind other high-risk industries in its attention to ensuring basic
safety [4]. Much of that which needs to be done in order to improve
upon patient safety is already being done in other industries and the
Corresponding author. Quality Management Service, University Hospital of Padua,
Azienda-Ospedaliera di Padova, via Giustiniani 2, 35128 Padova, Italy. Tel.: +39 049
8218338; fax:+39 049 8218351.
E-mail address: marialaura.chiozza@sanita.padova.it (M.L. Chiozza).
tools in order to improve upon patient safety. However, the transfer of
this type of knowledge is not automatic because health cannot be
considered a mere product. A stable, trusting relationship between a
patient and care providers is a factor of critical importance in the cure
or management of an illness [5].
However, the human factor in the provision of health care may be
responsible for some of the safety problems since practitioners are not
computers, their ability to process multiple pieces of often contradictory
information is limited, and of course human errors are often the result of
processes beyond the conscious control of the very professionals who
make errors. Therefore, if we wish to prevent errors in health care we
must understand the human factors causing them [6,7].
Of particular importance is the inner model of the clinical thinking,
and the way in which clinicians cope with the challenges involved in
providing health care, which hinges upon a complex series of interactions
between practitioners and patient, and pieces of technology that help
practitioners make a diagnosis and provide treatment. The challenge
in addressing complexity in health care was recently analyzed from a
clinical view-point, and the ability to cope with complexity has been
identied as is now considered a skill that clinicians must acquire [810].
These human factors must be borne in mind not only when patient
safety problems are mapped, but also when risk analysis tools are chosen
by health managers for their introduction in the health care organisation.
Finally, the management of increasingly complex and innovative
health care processes redenes the competences required from the
process owner as a leader trained to achieve capability and coordination
ability [11].

0009-8981/$ see front matter 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.cca.2009.03.015
76 M.L. Chiozza, C. Ponzetti / Clinica Chimica Acta 404 (2009) 7578
Numerous authors question the difference between reactive (Inci-
dent Reporting, Root Cause Analysis) and proactive (Failure Mode and
Effect Analysis, Socio-Technical Probabilistic Risk Assessment) tools as
methods of choice in risk management. However, while focusing on and
discussing theory differences, they overlook the evidence that fear and
resistance of organisations concerning reactive methods are related to
the fact that these methods are blamed for focusing on events rather
than processes.
The proactive methods are more readily accepted by clinicians
because they call for hope and exploit professional competences
through a positive approach to problems by focusing on the exam-
ination of the entire process, thus anticipating major adverse events
and pre-emptively implementing changes to prevent them from
occurring. For correct risk management, an organisation must
promote the awareness that the human factor cannot be completely
prevented from causing adverse events and that operators must
minimize the chances of making errors [3].
The Joint Commission (JC), formerly the Joint Commission on
Accreditation of Health Care Organization (JCAHO), now requires all
acute care hospitals to perform FMEA regularly (Standard LD 5.2
Accreditation Manual, 2001 Edition) [12] underlining the validity of
the model in reducing medical errors.
Recently the reduction of error through risk management and
continual improvement in medical laboratories has been regulated: in
a specic paper (ISO/TS 22367), issued in 2008, FMEA was accepted as
the method of choice in the identication of potential points of failure
within a process, their effects being determined and actions identied
for mitigating failures. FMEA is also recommended in deciding
whether to introduce a new process in clinical laboratories [13].
2. Failure Mode and Effect Analysis background
The FMEA, which dates back to >30 years ago, was rstly
introduced within the aerospace industry in the 1960s. Unlike other
competing failure preventing methods, FMEA was described in
universally understandable terms by individuals who had limited
technical and/or systems training. This promoted applications that
bridged across companies and industries [14]. In particular, the
automotive industry brought FMEA into the mainstream by adopting
the method as the primary system for error and risk reduction.
Early applications in the health care industry dating back to the
1990s included critical systems in the development and manufacture
of drugs and in the prevention of medication errors in hospitals [15].
Then, in the mid-1990s the Institute for safe Medication Practices
recommended that FMEA should be used to prevent errors occurring in
dispensing medications. More recently, The Veterans Affairs National
Center for Patient Safety implemented a system that modied and
expanded FMEA for its application to health care, the initiative being
named Health Care Failure Mode and Effect Analysis (H-FMEA) [16].
This successful application caught the attention of JCHAO which, in
July 2001, introduced the new leadership standard (L.D. Standard 5.2)
that requires department heads in health care organisations to
perform an FMEA on at least one critical process a year [12].
In 2008, the Technical Committee ISO TC/212, licensed the
technical specication ISO/TS 22367 for medical laboratories using
FMEA as a prospective risk analysis of high-risk processes to identify
needed improvements that will reduce the chance of unintended
adverse event [13].
3. Method description and steps
Admitting that medical errors are inevitable, FMEA places the
primary burden of error prevention, not on individuals, but on the
designers of the systems in which they work [17]. With FMEA, a
selected error-prone process is examined from start to nish, a
multidisciplinary team helping to prioritize the necessary changes.
Analytical methodologies such as FMEA can be very effective in
improving patient safety, but they may also be resource-intensive.
However, process facilitators trained in FMEA methodology can
greatly reduce meeting time requirements and guarantee that all
activities involved are coordinated.
The FMEA process covers the following 5 steps:
1. Choosing a process to be studied;
2. Assembling a multidisciplinary team;
3. Collecting and organizing information on the process studied;
4. Conducting a hazard analysis;
5. Developing and implementing actions and outcomes measures.
3.1. Choosing a process to be studied
The complexity of most patient care processes increases the
likelihood of unexpected and undesirable outcomes. The more steps in
the process and the greater their interdependence, the greater the
chance of error.
The possibility that a failure in a process step will ultimately affect
a patient increases considerably when the time interval between the
process steps is short, and it may not be apparent that an error has
occurred until the next step in the process is well underway. In
particular, whenever the satisfactory completion of a process relies
heavily on human intellectual or physical actions, the likelihood of
failure increases.
During this phase, the organisation identies the process to be
dened as critical, based on the severity of possible harmful events and
the potentially dangerous impact on patient safety. Numerous issues
should be considered when choosing a process for FMEA. In particular,
the processes with a history of adverse patient outcomes or recurrent
vulnerability in the organisation, identied in the literature as being
failure prone, have been identied as sentinel events and reported by
the Joint Commission and other regulatory agencies, and are considered
high risk for litigation claims. Of course the time required to complete
the analysis is reduced greatly by choosing a sub-process.
3.2. Assembling a multidisciplinary team
This step calls for the identication of operators with different
levels and types of training, with specic knowledge and experience of
the process to be studied. An appointed team leader can guide team
members through the process, and can help ensure that team
members complete each step and record the results of FMEA.
3.3. Collecting and organizing information on the process being studied
Team members dene the steps in the process, which is broken
down into sub-processes and straightforward tasks. The process,
described in terms of what actually happens in daily practice rather
than what should happen, is reported in a ow chart. Initially, for
ease of analysis, this should be achieved in the form of an event line of
the steps, after which the team members can create an accurate
diagram of the process as it is being currently performed.
3.4. Conducting a hazard analysis
This, which helps team members make an informed decision
concerning the way in which a safety patient care process is to be
designed, involves ve activities.
3.4.1. Identifying failure modes for each step
A failure mode is anything that could go wrong during completion of
a step in the process, and during the hand-off between steps. Failures
may be due to human error, equipment problems, communication
difculties, missing or misplaced supplies, or any other stumbling block
 M.L. Chiozza, C. Ponzetti / Clinica Chimica Acta 404 (2009) 7578
that might disrupt the seamless ow and ultimate safety of the process.
Ample time should be allowed for team members to identify all potential
failures. Between meetings, additional input can be obtained from staff
involved in the process, but who are not members of the team. The goal
of this activity is to discover everything that may go wrong.
3.4.2. Determining the potential effect of each failure mode
Team members describe failure mode effects in terms of what
those involved in the process, or a patient, might experience if the
failure occurs. Potential failures in the process steps nearest to patient
intervention are more likely to result in patient injury or harm.
Failures that occur earlier on in the process are more likely to result in
process disruption. In the absence of sufcient quality control
mechanisms, however, even failures occurring early on in the process
can affect the patient. If such information in past adverse events is
available for analysis, team members also should take into account the
circumstances surrounding past adverse events.
3.4.3. Ranking the severity of failure mode effects
Although each potential failure has some effect on the efciency of
the process or the outcome, many failures do not result in signicant
harm to a patient. During this activity, FMEA team members estimate
the severity of the effect of each failure on patient outcomes using
various types of numeric ranking systems. The ratings correspond to
the seriousness of the effects of a potential failure mode.
3.4.4. Ranking the probability and detectability of each failure mode
In this step, the team members rst determine the probability that
each failure will actually occur, based on past history and the personal
experience of team members. Then team members determine the
likelihood that each failure will persist throughout the process without
being detected and corrected. A consensus on the probability and
detectability rating scale does not exist, but several examples are
available, including that developed by the VHA [16].
3.4.5. Identifying the areas of greatest concern (critical failure modes)
To avoid the risk of overwhelming the team members by making
them aware that there are too many potential failure modes in the
process, priority should be given to the areas in which process
improvements are needed most. Since a safe process goal of 100% is
unrealistic, this means that FMEA should be used to identify the risks that
exceed acceptable limits. To identify the high priorities for action, the
numeric rating for severity, probability, and detectability are multiplied
so as to achieve a criticality index (CI) score for each failure mode, also
named risk priority number (RPN) or risk probability index index (RPI).
3.5. Developing and implementing actions and outcome measures
Before team members can determine the best way to improve upon
the safety of the process being studied, identifying the root cause(s) of
critical failures is an important step in developing an appropriate action
plan. Traditional root cause analysis methods are used to determine the
underlying cause of each critical failure so that appropriate actions can
be taken [18,19]. Once the root causes of critical failures have been
identied, the team's goal is to eliminate the risk of failures, reduce the
likelihood of failure or mitigate the effects of failure should it affect the
patient. A search of the literature should be made, and benchmarking
procedures used to identify the ways in which other organisations have
made similar processes safer for patients.
Corrective action should be directed at the issues of greatest
concern according to rank ordering by CI or RPN or RPI score. When
the failures potential in the process is understood and the causes
identied, particular process redesign solutions can be selected to
eliminate or reduce the risk of critical failures, essentially through
three types of process improvement strategies. The rst strategy is
designed to eliminate the chance of failure; the second makes it easier
77
for people to do the right thing; the third aims to identify failures
quickly and take appropriate action. The FMEA team members may be
charged with reviewing performance data periodically to assess how
much safer the process has become since implementing the action.
4. FMEA in laboratory medicine
A Google search for FMEA yielded 150,000 hits, a combined search
with the word engineering yielded 40,000 hits while a combined
search with medicine yielded only 3000. Many hits are offers of
manuals, forms, software and training programs for FMEA [20].
While the use of FMEA in engineering is widespread and
sophisticated, in medicine there are relatively few reports on its active
use. When it has been employed in this eld, it appears to have been
benecial and there has been little objection to its use.
Laboratory testing is gaining an increasingly important position in
the diagnostic process, and in monitoring the effects of therapy in
modern clinical medicine. Around two-thirds of important clinical
decisions concerning the admission and discharge of patients from
hospital and the prescription of medicines are based on laboratory test
results [21]. Timely and accurate laboratory test results are the
cornerstone of reliable diagnoses and effective treatments. Therefore,
even a low incidence of laboratory testing errors among the billions of
tests performed every day worldwide might have important public
health and patient safety implications.
In one review of the literature it was found that the majority of
laboratory errors occur in the pre-analytic and post-analytic phases of
the testing process [22,23]; this stresses the complexity of the multiple
steps that affect the delivery of laboratory services. Many mistakes
called laboratory errors are actually due to ineffective communication,
actions by others involved in the testing process or poorly designed
processes that are outside the control of the laboratory. Among the
laboratory-related errors in diagnosis, 50% were a failure to order the
appropriate tests, 32% were a failure to act on the result of tests, and
55% involved avoidable delays in making the diagnosis [24].
While the ISO/TS 22367 document recommends that FMEA should
be used in order to reduce errors and improve patient safety in
laboratory medicine, few data are available on the application of this
technique in clinical laboratories. Woodhouse et al. applied FMEA to
blood cross-matching. According to these authors, multiple checks
and crosschecks were in place for the pre-cross match and actual
transfusion process, but relatively few were used for the process of
cross-matching; they therefore analyzed and redesigned the total
process. Eleven steps were identied in the laboratory cross-matching
process, including such items such as a history check, setting up and
labelling tubes, dropping the reagents and adding patient serum, and
recording reactions. For each of the 11 processes identied, potential
failure modes were developed and relative solutions implemented. By
using these solutions, the possibility of error occurrence was
signicantly reduced and the likelihood of detection was increased,
thus reducing the RPN (risk priority number) from 250 to 40 [25].
Capunzo et al. applied the FMEA technique in a clinical laboratory,
although only a small sample of analytes was evaluated. The study
was, in fact, undertaken while taking into consideration three very
common clinical chemistry tests: glucose, total cholesterol and total
bilirubin. After analyzing the entire testing process, improvement
actions were designed for the following items: a) nonconforming
storage temperature; b) contaminated reagents; and c) contaminated
calibrators. The improvement actions regarding the rst item reduced
the RPI (risk probability index), from 540 to 180. Regarding the second
item, a reduction of the RPI, from 720 to 189, was obtained. For the
third item, a reduction of the RPI, from 128 to 96, was obtained [26].
Nichols et al. applied the FMEA technique to point-of-care testing
(POCT). Errors in POCT are particularly problematic, the test being
conducted by clinical operators at the site of patient care, and imme-
diate medical action being undertaken based on the results prior to
78 M.L. Chiozza, C. Ponzetti / Clinica Chimica Acta 404 (2009) 7578
review by the laboratory. On using the FMEA technique, identication
errors were found to be relatively common. In particular, the Authors
identied as the primary source of system failure the manual entry of
14 digits for each test, ve numbers for operators and nine numbers
for patient account identication. Patient barcoding, introduced to
automate the data entry process, resulted in signicant improvements
in error rates for all laboratory tests performed after an initial
familiarization period [27].
The lessons we can learn from the above experiences in laboratory
medicine are:
a) the total testing process is the unique framework to take into account
for identifying and reducing errors and risk of error in laboratory
medicine;
b) the total testing process is complex, involving procedures, equip-
ment, technology and human skills designed to ensure accurate and
timely diagnosis and treatment decisions;
c) the FMEA technique can be applied only in clinical laboratories that
have already performed a thorough evaluation of all procedures and
processes, particularly if they operate according to the ISO 9000
Quality Management System;
d) preliminary experiences indicate that a signicant reduction can
be achieved in the risk probability index.
5. Conclusions
No prospective risk assessment method, not even FMEA, can
guarantee the absolute safety of health care processes, but risk
assessment can help reduce the occurrence of adverse events when an
analytic technique is used systematically in order to identify and address
failure modes in high-risk processes. FMEA may become the common
standard for measurement and comparison, particularly in laboratory
medicine. In fact, the total testing process is complex, consisting of
numerous steps that are not always taken under the control of
laboratory professionals. Current evidence on the stratication of errors
in laboratory medicine strongly supports the introduction of FMEA for
further reducing error rates, particularly in the extra-analytical steps.
While the rst aim of FMEA is to promote a systematic approach to
ensuring the safety of patient care processes, total cost reduction should
be simultaneously achieved when considering the entire process of
patient care. Both goals, increased patient safety and cost reduction,
were recently achieved after introducing and applying FMEA to the
entire health care system of the Valle D'Aosta, a small region of Italy.
While a detailed description of this experience and results obtained is in
progress, here we would like to stress the point that after one-year
experience a dramatic reduction of the error risk has been obtained both
in hospital and non-hospital health care institutions, including clinical
laboratory and blood bank. Finally, it should be underlined that the gains
derived from risk analysis, using tools such as FMEA, will be short-lived if
the clinical laboratory management fails to visibly support continuous
safety improvement.
References
[1] Lesar TS, Lomaestro BM, Pohl H. Medication-prescribing errors in a teaching
hospital. A 9-year experience. Arch Intern Med 1997;157:156976.
[2] Thomas EJ, Studdert DM, Burstin HR, et al. Incidence and types of adverse events
and negligent care in Utah and Colorado. Med Care 2000;38:26171.
[3] Morelli P, Vinci A, Galetto L, Magon G, Maniaci V, Ban G. FMECA methodology
applied to two pathways in an orthopaedic hospital in Milan. J Prev Med Hyg
2007;48:549.
[4] Kohn KT, Corrigan JM, Donaldson MS. To err is human: building a safer health
system. Washington, DC: National Academy Press; 1999.
[5] Tarrant C, Stokes T, Baker R. Factors associated with patients' trust in their general
practitioner: a cross-sectional survey. Br J Gen Pract 2003;53:798800.
[6] Marx DA, Slonim AD. Assessing patient safety before the injury occurs: an
introduction to sociotechnical probabilistic risk modelling in health care. Qual Saf
Health Care 2003;12:ii338.
[7] Reason J. Human error: models and management. West J Med 2000;172:3936.
[8] Plsek PE, Greenhalgh T. Complexity science: the challenge of complexity in health
care. BMJ 2001;323:6258.
[9] Wilson T, Holt T, Greenhalgh T. Complexity science: complexity and clinical care.
BMJ 2001;22(323):6858.
[10] Fraser SW, Greenhalgh T. Coping with complexity: educating for capability. BMJ
2001;323:799803.
[11] Plsek PE, Wilson T. Complexity, leadership, and management in healthcare
organisations. BMJ 2001;323:7469.
[12] Joint Commission on Accreditation of Health Care Organizations. Performance
improvement. 2002 Comprehensive accreditation manual for hospitals. Oakbrook
Terrace, IL: Joint Commission on Accreditation of Health care Organizations; 2001.
p. 2001.
[13] ISO/TS 22367: 2008. Medical laboratoriesreduction of error through risk
management and continual improvement.
[14] McDermott RE, Mikulak RJ, Beauregard MR. The basis of FMEA. New York: Quality
Resources; 1996.
[15] Duwe B, Fuchs BD, Hansen-Flaschen J. Failure mode and effects analysis
application to critical care medicine. Crit Care Clin 2005;21:2130.
[16] DeRosier J, Stalhandske E, Bagian JP, Nudell T. Using health care failure mode and
effects analysis: the VA national center for patient safety's proactive risk analysis
system. Jt Comm J Qual Improv 2002;28:24867.
[17] Cohen MR, Senders J, Davis NM. Failure mode and effects analysis: a novel
approach to avoiding dangerous medication errors and accidents. Hosp Pharm
1994;29:31930.
[18] Bagian JP, Gosbee J, Lee CZ, Williams L, McKnight SD, Mannos DM. The Veterans Affairs
root cause analysis system in action. Jt Comm J Qual Improv 2002;28:53145.
[19] Neily J, Ogrinc G, Mills P, et al. Using aggregate root cause analysis to improve
patient safety. Jt Comm J Qual Saf 2003;29:4349 381.
[20] Senders JW. FMEA and RCA: the mantras of modern risk management. Qual Saf
Health Care 2004;13:24950.
[21] Carraro P, Plebani M. Process control reduces the laboratory turnaround time. Clin
Chem Lab Med 2002;40:421.
[22] Plebani M, Carraro P. Mistakes in a stat laboratory: types and frequency. Clin Chem
1997;43:34851.
[23] Bonini P, Plebani M, Ceriotti F, Rubboli F. Errors in laboratory medicine. Clin Chem
2002;48:6918.
[24] Leape LL, Brennan TA, Laird N, et al. The nature of adverse events in hospitalized patients.
Results of the Harvard Medical Practice Study II. N Engl J Med 1991;7(324):37784.
[25] Woodhouse S, Burney B, Coste K. To err is human: improving patient safety
through failure mode and effect analysis. Clin Leadersh Manag Rev 2004;18:326.
[26] Capunzo M, Cavallo P, Boccia G, Brunetti L, Pizzuti S. A FMEA clinical laboratory
case study: how to make problems and improvements measurable. Clin Leadersh
Manag Rev 2004;18:3741.
[27] Nichols JH, Bartholomew C, Brunton M, et al. Reducing medical errors through
barcoding at the point of care. Clin Leadersh Manag Rev 2004;18:328334s.