Review Article

Annals of Clinical Biochemistry

2017, Vol. 54(1) 14–19
! The Author(s) 2016
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0004563216669384
acb.sagepub.com

Preanalytical errors in medical laboratories: a review of

the available methodologies of data collection and analysis

Jamie West1, Jennifer Atherton2, Seán J Costelloe3, Ghazaleh Pourmahram4,

Adam Stretton4 and Michael Cornes5

Abstract
Preanalytical errors have previously been shown to contribute a significant proportion of errors in laboratory processes
and contribute to a number of patient safety risks. Accreditation against ISO 15189:2012 requires that laboratory Quality
Management Systems consider the impact of preanalytical processes in areas such as the identification and control of
non-conformances, continual improvement, internal audit and quality indicators. Previous studies have shown that there
is a wide variation in the definition, repertoire and collection methods for preanalytical quality indicators. The
International Federation of Clinical Chemistry Working Group on Laboratory Errors and Patient Safety has defined a
number of quality indicators for the preanalytical stage, and the adoption of harmonized definitions will support inter-
laboratory comparisons and continual improvement. There are a variety of data collection methods, including audit,
manual recording processes, incident reporting mechanisms and laboratory information systems. Quality management
processes such as benchmarking, statistical process control, Pareto analysis and failure mode and effect analysis can be
used to review data and should be incorporated into clinical governance mechanisms. In this paper, The Association for
Clinical Biochemistry and Laboratory Medicine PreAnalytical Specialist Interest Group review the various data collection
methods available. Our recommendation is the use of the laboratory information management systems as a recording
mechanism for preanalytical errors as this provides the easiest and most standardized mechanism of data capture.

Keywords
Preanalytical errors, quality indicators, laboratory errors, patient safety
Accepted: 8th August 2016

Introduction
1
Department of Clinical Biochemistry and Immunology, Peterborough
Preanalytical errors (PAEs) are errors which occur City Hospital, Peterborough, UK
prior to the analytical stage in the total testing process 2
Liverpool Clinical Laboratories, Blood Sciences Department, Aintree
(TTP) and can occur both before and after receipt of University Hospital, Liverpool, UK
3
specimens in the laboratory. They have previously been Derriford Combined Laboratory, Plymouth Hospitals NHS Trust,
Plymouth, Devon, UK
shown to contribute a significant proportion of errors 4
Becton Dickinson Diagnostics, Preanalytical Systems (PAS), Oxford, UK
in laboratory processes (up to 68.2%).1–3 PAEs contrib- 5
Clinical Chemistry Department, New Cross Hospital, Wolverhampton,
ute to a number of patient safety risks, including UK
inappropriate or incorrect therapeutic interventions,
Corresponding author:
unnecessary follow-up investigations and diagnostic Michael Cornes, Department of Clinical Chemistry, New Cross Hospital,
delays, each of which impact on the clinical and eco- Wolverhampton WV10 0QP, UK.
nomical effectiveness of pathology services.4–6 Email: michael.cornes@nhs.net
West et al. 15

Table 1. Mandatory quality indicators of the pre-analytical phase as suggested by IFCC-WG-LEPS.

Step in process Associated quality indicators

Patient identification
Data entry of the request
Sample identification
Sample collection
Storage and transport of samples
Suitability of samples
 Number of requests with errors concerning patient identification (%)
 Number of requests with errors concerning patient identification detected before the
release of results (%)
 Number of requests with errors concerning patient identification detected after the
release of results (%)
 Number of requests with errors concerning test input (%)
 Number of requests with errors concerning test input (missing, %)
 Number of requests with errors concerning test input (added, %)
 Number of requests with errors concerning test input (misinterpreted, %)
 Number of inadequately labelled patient samples (%)
 Number of samples collected with inappropriate sample tube type (%)
 Number of samples collected in inappropriate container (%)
 Number of samples with insufficient sample volume (%)
 Number of damaged sample tubes/containers (%)
 Number of samples transported at an inappropriate time (%)
 Number of samples transported at inappropriate temperature condition (%)
 Number of improperly stored samples (%)
 Number of samples lost/not-received (%)
 Number of samples with inadequate sample anticoagulant ratio (%)
 Number of samples haemolysed (haematology, chemistry, immunology) (%)
 Number of samples clotted (haematology, chemistry) (%)
 Number of lipaemic samples (%)
 Number of unacceptable samples (microbiology) (%)
 Number of contaminated blood cultures (%)

Requirements for accreditation against ISO
15189:2012 (Medical laboratories – Requirements for
quality and competence) dictate that laboratories
should consider preanalytical processes in a number
of areas of the quality management system, including
the identification and control of non-conformances,
continual improvement, internal audit and quality indi-
cators (QIs).7 It states that ‘the laboratory shall estab-
lish QIs to monitor and evaluate performance
throughout critical aspects of pre-examination, exam-
ination and postexamination processes’. In the UK, a
recent report into the quality assurance frameworks
and governance systems highlighted variations in the
processes for error reporting, advising that high levels
of error reporting with low overall error rates is a good
indicator of a quality service.8
In a previous study by the Association for Clinical
Biochemistry and Laboratory Medicine Pre-Analytical
Specialist Interest Group (ACB-SIG-PA), we have
shown that there is a wide variation in the definition,
repertoire and collection methods for preanalytical
QIs.9 Here, we aim to review potential methodologies
for the monitoring of PAEs, including recording
of errors, data collation and review, follow-up data
outputs and integration into a clinical governance
framework. It is hoped that consideration of data col-
lection methods will help to standardize data, allow
development of key performance indicators (KPIs)
and support peer review and continual improvement.
Types of PAEs
A key source of variation identified in the Pathology
Quality Assurance Review8 was the definition of
errors. Standardization in this area underpins the effect-
ive reporting of errors, enabling comparisons between
peers and identification of areas for improvements in the
TTP. These QIs then provide a harmonized platform for
targeted continuous improvement and a means of mea-
suring said improvements.10 The International
Federation of Clinical Chemistry Working Group on
Laboratory Errors and Patient Safety (IFCC-WG-
LEPS) has worked to improve awareness in the field of
laboratory errors and patient safety, developed pilot stu-
dies to monitor error rates and implemented projects to
reduce errors. One element of the project has been to
develop QIs to support the monitoring of critical steps
in the Pathology process. Table 1 details a selection of
harmonized QIs for the preanalytical stage suggested by
the working group to be mandatory to a laboratory’s
quality monitoring processes.11
These suggestions outline a number of common
areas of the preanalytical laboratory processes where
errors can occur, but are not an exhaustive list.
16
Laboratories should consider the risks associated with
each step in the TTP for the services they offer, and
look to ensure that processes are in place to identify
process failures and monitor their rates of incidence.
This project has also triggered the development of
other country-specific pre- and postanalytical external
quality assurance schemes which will further drive the
need to standardize collection methods.12 Further con-
sideration of harmonized definitions of QIs may be
required moving forward. For example, the IFCC-
WG-LEPS defines the presence of haemolysis in
Clinical Chemistry samples as a free haemoglobin con-
centration of >0.5 g/L, while a variety of laboratory
tests are affected by haemolysis at lower and higher
concentrations of free haemoglobin and practice
regarding the handling of haemolysed specimens may
vary between laboratories. Requirements for the label-
ling of samples will also vary between laboratories, with
transfusion laboratories an example where minor errors
will result in rejection, and where the IFCC-WG-LEPS
definition of an inadequately labelled sample as one
which has ‘less than 2 identifiers’ may not be suitable.
As laboratories looked to define and standardize their
own definitions of laboratory errors, it is hoped that
peer comparison schemes will support harmonized
approaches to detect and report laboratory errors.
Recording of errors
In order for laboratories to monitor the error rates for
their services, a robust continuous mechanism of error
collection must be established. When considering the
systems used for highlighting quality failures, it is
important to ensure that the processes are easy to use
and understood by staff in the laboratory tasked with
recording errors. Standardization of processes is
important to ensure the accuracy of the recording
mechanisms and enable effective staff training. To this
end, all error collection procedures must be contained
in a standardized operating procedure which states not
just how to collect data but also collection frequency
and what to do when error frequencies are outside pre-
defined action limits. Below we detail a number of
options for mechanisms to monitor error rates:
Audit
Laboratory quality audit is an essential element of a
laboratory’s quality management system and should
include scheduled audits of process effectiveness. ISO
15189: 2012 states that this should incorporate audits of
preanalytical areas to collect information on the rela-
tive rates of errors. Examples of such audits might be:
an audit of request forms to identify the percentage
which do not have complete information regarding
Annals of Clinical Biochemistry 54(1)
the requester; an audit of specimens to identify the per-
centage which do not have complete (or accurate)
patient information; an audit of ‘booking-in’ processes
to identify the percentage of booking in errors.
While audit forms a crucial part of the laboratory’s
quality processes, the use of audit itself as a tool for
data collection is limited as it is retrospective.
Therefore, audit does not provide a real-time assess-
ment of rates of error incidence, but a survey of error
rates at a particular point in time. Further, audit does
not immediately alert users to the quality issues with
their requests. Finally, audits must be extensive in order
to accurately reflect the true error rate of the labora-
tory, and as such are labour intensive to perform.
Manual recording processes
Systems of recording errors manually by means of
‘quality query reports’ (QQRs) have been described
previously.13 QQRs involve the use of report forms
made available at workstations to records errors.
The process is, again, labour intensive compared with
automated systems of error identification in terms of
both error recording and especially in reviewing data,
and as such risks lower levels of compliance and under-
reporting of errors. Variable frequencies in recorded
errors which may be reflective of variations in compli-
ance the recording process have been anecdotally
observed and would be expected due to human factors.
Such systems also have the disadvantage that they do
not alert the user to quality failures.
Incident reporting
Incident reporting, often using risk management soft-
ware such as Datix (Datix Limited, London, UK),
should be part of any healthcare provider’s clinical gov-
ernance processes. Medical laboratories should have
processes of reporting errors using these systems.
These systems have the advantage of being subject to
host organization governance processes. However,
given the in-depth nature of reporting required, such
systems are not suited to recording large numbers of
lower risk errors.
Use of laboratory information management systems
Laboratory information management systems (LIMS),
primarily used to record receipt of specimens and
report results, have the potential to be used to record
quality errors, report them to users, act as repositories
of error data and also gather data on error rates.
Subclause 5.8.2 of the ISO 15189: 2012 Standard con-
cerns report attributes, and states that the report should
contain ‘comments on sample quality that might
West et al.
compromise examination results’ and ‘comments
regarding sample suitability with respect to accept-
ance/rejection criteria’.7 Many errors will still require
manual identification but the use of the LIMS systems
to record quality failures has the advantage that simple
and standardized processes can be developed which
makes error logging part of the normal sample receipt
process and improves compliance. In the automated
laboratory, some errors have the potential to be both
identified and entered into the LIMS automatically, e.g.
delayed transit, underfilled tubes, wrong specimen type,
etc. LIMS use is both prospective (as opposed to retro-
spective methods described above) in alerting users to
quality failures and retrospective, when regular data
extractions are performed for trend analysis. Setting
up such processes is often labour intensive; however
once in place, there is minimal staff time associated
with collecting and extracting the data.
Suggestions for recording errors via LIMS systems
include the following.
. Use of ‘Report Comment’ fields to record quality
failures and highlight them in the report. For such
systems, a balance between the standardization of
error codes (which improves information gathering
and review) and the flexibility to record errors of a
variety of sources is required.
. Use of a test field in the LIMS system which can be
used to record a number of laboratory errors that
can then be reported to the user (i.e. book in a test
that indicates an error and result this test with a
coded comment to indicate the type of error).
. Use of coded comments in place of results for sam-
ples that cannot be analysed, e.g. HAEM for sam-
ples that cannot be analysed due to haemolysis.
. Keyword searches for comments in reports, e.g.
‘haemolysis’.
The configuration of the recording system will depend
on the configuration of the LIMS system (and poten-
tially other associated operating systems and middle-
ware) and will have to be adapted to allow the
recording of a variety of types of error. Laboratories
should consider the merits of their LIMS systems when
designing systems of recording errors. An ideal process
will allow errors to be easily recorded in the laboratory,
clearly reported to users and support the export and
review of data for continuous improvement.
Data review
The advantage of using LIMS-based systems for the
recording of errors is that it is possible to access real-
time and/or retrospective data for review in a simple
and effective way. Data export processes such as file
17
transfer protocol or database enquiries enable data to
be reviewed quickly and easily. Data should be
reviewed and published to laboratory users periodic-
ally. Trends and patterns should be investigated in con-
junction with the department involved. Data to review
include the following.
. Types of errors – error types (for example as
described in Table 1) can be reviewed for trends and
changes.
. Requesting locations – this can be important in iden-
tifying areas with high error rates for a particular
indicator, such as specimen collection, transport or
haemolysis, which may be a systematic issue relating
to a particular service user. In such cases, training in
specimen collection and/or improvements in logistics
can have a beneficial impact on reducing error rates
and therefore reducing repeat venepunctures and
costs. Reporting to groups of clinical teams (for
example, laboratory service commissioning groups
or hospital clinical teams) may also be useful for
establishing governance processes using laboratory
data.
. Requesting clinicians – error rates relating to
requesting (for example, appropriateness of a
request or completion of a request form)
. Errors by laboratory department – this could indi-
cate an issue with intralab sample processing.
There are a number of quality management tools which
can be utilized in the investigation of laboratory quality
errors, including the following.
. Benchmarking – error rates are compared against
locally or nationally derived benchmarks or against
other laboratories.
. Statistical process control – in the same way that the
laboratory reviews quality control data for outliers,
error rates can be compared against a target mean
and standard deviations to identify statistically sig-
nificant changes.
. Failure mode and effect analysis (FMEA) – FMEA
requires a knowledge of the steps involved in the test-
ing process, target areas of high risk and looks at
methods of adapting processes to reduce failures.14
. Pareto analysis – Pareto diagrams organize elements
in order of frequency of occurrence, enabling the
laboratory to target areas of high risk. Errors can
be grouped as described above, for example, by error
type or location.
Laboratory errors and clinical governance
Once data are established for measuring rates of prea-
nalytical laboratory errors, it is important to
18
incorporate the information into the department’s qual-
ity management system and clinical governance
processes.
KPIs
KPIs are measurands used by organizations to monitor
and assess their effectiveness. In clinical laboratory ser-
vices, a number of indicators have been suggested,15,11
many of which relate to issues regarding PAEs. In the
UK, a recent Pathology Quality Assurance Review8
describes the use of Pathology quality assurance dash-
boards (PQAD) to ensure pathology providers comply
with service specifications and suggests that the dash-
boards should be developed with commissioners of
Pathology services. There have been previous attempts
to develop performance measures for Pathology ser-
vices16 and a UK national project is underway to
develop a dashboard. PQADs can serve the dual pur-
pose of allowing a laboratory to monitor key areas of
its service (which includes elements from the preanaly-
tical phase) and also provide data to users on the effect-
iveness of its service.
Risk management
Clause 4.14.6 of the ISO 15189: 2012 standard7 relates to
risk management and is a core area of the standard. It
advises that the laboratory should ‘evaluate the impact
of work processes and potential failures on examination
results as they affect patient safety, and shall modify
processes to reduce or eliminate the identified risks and
document the decisions and actions taken’. Preanalytical
steps in examination processes, including those that
occur before the sample is received in the laboratory,
are a key source of error in the TTP and risk manage-
ment processes should incorporate such errors.
The ISO Technical Specification 22367:2008
‘Medical Laboratories – reduction of error through
risk management and continual improvement’17 advises
that laboratories should have processes for:
. identifying high-risk processes where the potential
error could lead to a safety risk for patients;
. identifying actual incidents associated with devi-
ations from standard procedures;
. estimating and evaluating the associated risks to
patient safety;
. controlling these risks;
. monitoring the effectiveness of the control
undertaken.
ISO/TS 22367:2008 also recommends assigning actual
(A) and potential (P) scores to the risks associated
with processes, based on the potential impact of the
Annals of Clinical Biochemistry 54(1)
errors on patient safety. Laboratories that fre-
quently monitor rates of laboratory errors in the pre-
analytical phase, and use the information to facilitate
preventative action to improve processes, are in a good
position to monitor the impact of such errors on
the results they report and assess the impact on patient
safety.
Interventions
Efforts to decrease rates of PAEs have been docu-
mented previously but can be difficult. For example,
Kemp et al.18 used posters and screen savers to improve
user awareness of testing protocols but showed no stat-
istically significant change in the frequency of errors.
Salinas et al., however, combined improvement actions
(such as educational programs for nurses, technological
interventions to automate manual steps in phlebotomy
procedure and communication between the laboratory
and peripheral community centres) with monitoring via
a balanced scorecard to show an improvement in PAE
rates over a 10-year period.19
If it is assumed that laboratory testing protocols
across different clinical areas have a consistent rate of
occurrence within a defined variation, an alternative
approach is to target interventions to areas which
have higher error rates, for example higher rates of
sample haemolysis or labelling errors. Here, FMEA
with interventions such as training or process changes
may be more effective. For example, some of the key
causes of sample haemolysis relate to the blood collec-
tion step and overly vigorous collection.20 If areas are
identified where haemolysis rates are higher than
expected, training and communication can be targeted
to reduce this risk.
When targeting interventions to reduce error rates, it
is important to approach issues in a ‘system’-based
manner, as opposed to taking an ‘individual’
approach.21 This approach emphasizes looking at the
processes that a laboratory puts in place, in this case
preanalytical processes such as sample collection and
transport, to identify opportunities for improvement.
Conclusions
PAEs continue to represent the largest source of errors
in the TTP. Harmonized QIs for PAEs have been sug-
gested,11 which allow consistent and efficient reporting
via the laboratory’s quality monitoring processes.
Although it is a challenge to ensure compliance in rec-
ording such errors, the use of the LIMS and/or middle-
ware allows consistent recording of errors, alongside
manual reporting by laboratory staff, and the end users.
Once data have been effectively gathered by the
laboratory, risk analysis using FMEA and reporting
West et al.
of data using PQADs allow judgement about the sever-
ity and incidence of PAEs. Interventions can then be
planned and prioritized, according to the data reported.
Previous studies9 have shown variation in data collec-
tion processes, which will limit the effectiveness of peer
comparisons and as such inhibit opportunities for con-
tinual improvement. The process of data collection
must therefore be standardized to allow comparisons
to benchmarked data which will drive improvement in
this area. Laboratories should carefully consider their
processes to monitor PAEs to ensure that the processes
for obtaining data are effective and areas of risk are
monitored to enable quality improvement in the TTP.
In conclusion, the ACB-PA-WG recommends the
use of the LIMS as a recording mechanism as in the
long term, this provides the easiest and most standar-
dized mechanism of data capture. The ACB-SIG-PA
also recommends that the ability to provide such data
should be a criterion when tendering for new LIMS
systems in order to transfer the responsibility for
developing mechanisms to the experts and produce
standardized data. Other mechanisms are retrospective
and have been shown to under report errors. Data from
KPIs should be available to users and compared with
benchmarked data or against peers.
Acknowledgements
This paper was written by the Association for Clinical Biochemistry and
Laboratory Medicine (ACB) Special Interest Group for the pre-analytical
phase.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the
research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or
publication of this article.
Ethical approval
Not applicable.
Guarantor
MC.
19
Contributorship
All conceived and discussed study. JW wrote the first draft. All edited
manuscript.
References
1. Plebani M and Carraro P. Mistakes in a stat laboratory: type and fre-
quency. Clin Chem 1997; 43: 1348–1351.
2. Plebani M. Errors in clinical laboratories or errors in laboratory medicine.
Clin Chem Lab Med 2006; 44: 750–759.
3. Kalra J. Medical errors: impact on clinical laboratories and other critical
areas. Clin Biochem 2004; 37: 1052–1062.
4. Cadamuro J, Wiedermann H, Mrazek C, et al. The economic burden of
hemolysis. Clin Chem Lab Med 2015; 53: 285–288.
5. Green SF. The cost of poor blood specimen quality and errors in pre-
analytical processes. Clin Biochem 2013; 46: 1175–1179.
6. Jacobs P, Costello S and Beckles M. Cost of haemolysis. Ann Clin Biochem
2012; 49(Pt 4): 412.
7. ISO 15189:2012. Medical laboratories – requirements for quality and com-
petence, 2012.
8. Barnes I. Pathology quality assurance review. Jan 2014. Available at:
https://www.england.nhs.uk/wp-content/uploads/2014/01/path-qa-review.
pdf (accessed September 2016).
9. Cornes MP, Atherton J, Pourmahram G, et al. Monitoring and reporting
of preanalytical errors in laboratory medicine: the UK situation. Ann Clin
Biochem 2016; 53: 279–284.
10. Plebani M. Quality indicators to detect pre-analytical errors in laboratory
testing. Clin Biochem Rev 2012; 33: 85–88.
11. Plebani M, Sciacovelli L, Aita A, et al. Harmonisation of pre-analytical
quality errors. Biochem Med 2014; 24: 105–113.
12. Cornes MP, Church S, van Dongen-Lases E, et al. The role of European
Federation of Clinical Chemistry and Laboratory Medicine Working
Group for Preanalytical Phase in standardization and harmonization of
the preanalytical phase in Europe. Ann Clin Biochem 2016; 53: 539–547.
13. O’Kane MJ, Lynch PLM and McGowan N. The development of a system
for the reporting, classification and grading of quality failures in the clinical
biochemistry. Ann Clin Biochem 2008; 45: 129–134.
14. Agarwal R. Measurement of errors in clinical laboratories. Ind J Clin
Biochem 2013; 28: 227–234.
15. Barth JH. Clinical quality indicators in laboratory medicine. Ann Clin
Biochem 2012; 49: 9–15.
16. Royal College of Pathologists Key Performance Indicators. Proposals for
implementation. Available at: www.rcpath.org/resource-library-homepage/
clinical-effectiveness/key-performance-indicators-kpi.html (accessed 13
August 2016).
17. ISO/TS 22367: 2008. Medical laboratories – reduction of error through risk
management and continual improvement, 2008.
18. Kemp GM, Bird CE and Barth JH. Short-term interventions on wards fail
to reduce preanalytical errors: results of two prospective controlled trials.
Ann Clin Biochem 2012; 49: 166–169.
19. Salinas M, Lopez-Garrigos M, Flores E, et al. Ten years of pre-analytical
monitoring and control: synthetic balance score card indicator. Biochem
Med 2015; 25: 49–56.
20. Carraro P, Servidio G and Plebani M. Hemolyzed specimens: a reason for
rejection or a clinical challenge? Clin Chem 2000; 46: 306–307.
21. Reason J. Human error: models and management. BMJ 2000; 320:
768–770.