ORIGINAL ARTICLE

The Nature, Causes, and Clinical Impact of Errors in the

Clinical Laboratory Testing Process Leading to Diagnostic
Error: A Voluntary Incident Report Analysis
Christel van Moll, MD,* Toine Egberts, PhD,†‡ Cordula Wagner, PhD,§||
Laura Zwaan, PhD,¶ and Maarten ten Berg, PhD**

the adverse events that occurred in patients who died in the

Objectives: Diagnostic errors, that is, missed, delayed, or wrong diagno- hospital were diagnostic errors and that 40% of severe incidents
ses, are a common type of medical errors and preventable iatrogenic harm. in hospitals that were reported to the Dutch Healthcare Inspector-
Errors in the laboratory testing process can lead to diagnostic errors. This ate were related to errors in the diagnostic process.2,3 Diagnostic
retrospective analysis of voluntary incident reports aimed to investigate errors are also internationally considered to be a major patient
the nature, causes, and clinical impact of errors, including diagnostic er- safety problem. The National Academy of Medicine in the
rors, in the clinical laboratory testing process. United States for example concluded in their influential report
Methods: We used a sample of 600 voluntary incident reports concerning on improving diagnosis in health care that most people will expe-
diagnostic testing selected from all incident reports filed at the University rience a diagnostic error at least once in their lifetime.4
Medical Center Utrecht in 2017–2018. From these incident reports, we in- Clinical laboratory testing is a major component of the diag-
cluded all reports concerning the clinical laboratory testing process. For nostic process. Up to 80% of medical decisions are influenced
these incidents, we determined the following: nature: in which phase of by laboratory data.5 Errors in the laboratory testing process can
the testing process the error occurred; cause: human, technical, organiza- lead to diagnostic errors. Gandhi et al6 found for example that in
tional; and clinical impact: the type and severity of the harm to the patient, 37% of diagnostic errors in the ambulatory setting that led to
including diagnostic error. malpractice claims, incorrect interpretation of the laboratory test
Results: Three hundred twenty-seven reports were included in the analysis. result caused the error. Similarly Schiff et al7 found in a survey
In 77.1%, the error occurred in the preanalytical phase, 13.5% in the analyt- among physicians that errors occurred most frequently (44%)
ical phase and 8.0% in the postanalytical phase (1.5% undetermined). Hu- in the diagnostic testing phase (failure to order, report, and
man factors were the most frequent cause (58.7%). Severe clinical impact oc- follow-up laboratory results).
curred relatively more often in the analytical and postanalytical phase, 32% Although the consequences for patients of errors in the labora-
and 28%, respectively, compared with the preanalytical phase (40%). In tory testing process are well known from these studies, there are
195 cases (60%), there was a potential diagnostic error as consequence, no studies in which diagnostic error as an outcome of an error in
mainly a potential delay in the diagnostic process (50.5%). the laboratory testing process was investigated. Such knowledge
Conclusions: Errors in the laboratory testing process often lead to poten- could be useful in developing targeted risk management strategies
tial diagnostic errors. Although prone to incomplete information on causes to effectively improve the safety of diagnostic testing.
and clinical impact, voluntary incident reports are a valuable source for re- The process of clinical laboratory testing can be subdivided
search on diagnostic error related to errors in the clinical laboratory into the preanalytical, analytical and postanalytical phase. The
testing process. preanalytical phase being all the steps occurring before the speci-
Key Words: diagnostic error, patient safety, incident reports, laboratory men is being analyzed, for example, ordering the test, collecting
medicine the specimen, and transport of the specimen. The analytical phase
is the specific analysis of the specimen in the laboratory. The
(J Patient Saf 2023;19: 573–579)
postanalytical phase covers the steps occurring after the specific
specimen has been analyzed, that is, reporting the result to the cli-
D iagnostic errors are a common cause of iatrogenic harm to
patients and are defined as a diagnosis that was wrong, de-
layed, or missed or was communicated with relevant delay to the
nician or patient and interpretation of the result by the clinician.8
Over time, there has been an interesting shift in which phase of
the laboratory testing process most errors occurred. In the 1980s,
patient.1 Recent data from the Netherlands showed that 11% of
most errors occurred in the analytical phase. This changed to the
From the *Department of Internal Medicine, University Medical Center
preanalytical phase 20 years later, when it was estimated in studies
Utrecht; †Utrecht Institute for Pharmaceutical Sciences and Division of with only data on frequencies of type of laboratory errors that 60%
Pharmacoepidemiology and Clinical Pharmacology, Faculty of Science, Utrecht of the errors occur in the preanalytical phase.9–11 Automation, im-
University; ‡Department of Clinical Pharmacy, University Medical Center proved laboratory technology, assay standardization, and better in-
Utrecht; §Netherlands Institute of Health Services Research (NIVEL),
Utrecht; ||Amsterdam Public Health institute (APH), Amsterdam University
ternal and external quality control majorly contributed to quality
Medical Center, Amsterdam, the Netherlands; ¶Erasmus Medical Center, Insti- improvement and reduction of errors in the analytical phase.11
tute of Medical Education Research Rotterdam, Rotterdam, the Netherlands; To prevent patients from harm through diagnostic errors, it is
and **University Medical Center Utrecht, Central Diagnostic Laboratory, important to investigate whether there are strategies to help miti-
Utrecht, The Netherlands.
Correspondence: Maarten ten Berg, PhD, University Medical Center Utrecht,
gate the risk for an error to occur. As Zwaan et al12 recently stated
Central Diagnostic Laboratory, Mailstop G03.550, Heidelberglaan, 100 in their article on the current state of research on diagnostic errors,
3584CX Utrecht, the Netherlands (e‐mail: [email protected]). priority should be given to identification of the main failure points
The authors disclose no conflict of interest. in the diagnostic process and in developing, implementing, and
Copyright © 2023 The Author(s). Published by Wolters Kluwer Health, Inc.
This is an open access article distributed under the Creative Commons Attri-
testing specific interventions. For this type of research, several
bution License 4.0 (CCBY), which permits unrestricted use, distribution, and types of data sources can be used, like claims data, obduction data
reproduction in any medium, provided the original work is properly cited. reports on major incidents, and voluntary incident reports. This

J Patient Saf • Volume 19, Number 8, December 2023 www.journalpatientsafety.com 573

van Moll et al J Patient Saf • Volume 19, Number 8, December 2023

type of reports are an interesting source compared with the others improve the workflow. However, personnel working at the labora-
because it provides an extensive view on the clinical process and tory is being trained to report all incidents as part of daily work.
an insight into the part of incidents that healthcare workers con- The incident reports are used for trend monitoring and im-
sider relevant and that reflect situations that have the potential to provement of workflow and the quality of patient care. Data are re-
become major incidents. corded and are reviewed weekly by independent health-care
Databases with voluntary incident reports have not been widely workers, to check if any high-impact incidents have happened that
used yet for investigating errors in laboratory testing process. We need additional actions. The system can be queried on character-
are familiar with 3 large studies, performed in the United States istics to reports of interest for analysis.
and Canada.13–15 In the first study, an analysis of 37,532 labora-
tory event reports, it was found that most errors occurred in the Data
preanalytical phase (specimen not labeled/mislabeled and im-
For this study, we queried the system for all reports labeled as
proper collection) and that most errors did not cause harm.13 In
diagnostics (regardless of where the reported error occurred) and
the second study, an analysis of 12,278 laboratory related safety
all reports from all other categories for which was reported that
events reported to the British Columbia Patient Safety and Learn-
the error occurred at a clinical laboratory in 2017 and 2018. Of
ing System, comparable results were found. Most incidents con-
all these reports, for convenience purposes (limited time and the
cerned the preanalytical phase (76%), and it was found that the
estimation we did not need all the reports for a representative analy-
majority of incidents (95.9%) resulted in little or no harm and
sis), the first 600 reports (in order of time of reporting) were
0.44% in severe harm.14 In that study, the association between
included. From these we selected reports concerning the clinical
the degree of harm and the phase of testing was investigated. It
laboratory process of clinical genetics, medical microbiology, clini-
was found that postanalytical events contain the highest risk to re-
cal chemistry, pathology, and metabolic diagnostics for the analysis.
sult in severe harm.14 In those studies, it was not specifically
investigated whether the incidents led to (potential) diagnostic er-
rors. The third study from the same group with data from British Outcomes
Columbia Patient Safety and Learning System confirmed the Per report, we determined the nature, the cause, and the clinical
findings from their first study.15 To gain more insight into diag- impact of the error.
nostic errors related to errors in the clinical laboratory testing pro- Before the study, the authors C.M., T.E., and M.B. established a
cess, we conducted a retrospective analysis of voluntary incident 36-step classification scheme for the nature (type of error) of er-
reports in our hospital, a large academic teaching hospital with rors in the clinical laboratory testing process based on previously
high volume diagnostic testing (e.g., approximately 500,000 or- used classifications in order studies (Table 1).10,16,17 All 600 re-
ders for clinical chemistry testing annually) in which we looked ports were read by 2 authors (C.M. and M.B.) and together they
at the phases of the testing process, the type of errors, the causes classified the nature of the incident into the 36 steps of laboratory
and the clinical impact, including potential diagnostic errors, as process. For the classification of the causes, the Eindhoven classi-
well as the relation between these. fication model18 was used. For classification of the clinical
impact, the classification previously used by Graber et al19 for diag-
METHODS nostic error was used. Two authors (C.M. and M.B.) classified the
first 50 cases together, to determine how the different categories
should be classified. Subsequently, C.M. classified the cases alone.
Study Design
In case of doubt, C.M. and M.B. discussed and decided together.
We conducted a retrospective study with reports concerning any As incident reports describe the whole spectrum of incidents
type of incident concerning the clinical laboratory process of clini- (near misses, minor errors with no consequences and diagnostic
cal genetics, medical microbiology, clinical chemistry, pathology, errors, etc.), we defined the outcomes as followed:
and metabolic diagnostics from the voluntary patient safety incident
report system of the University Medical Center Utrecht (UMC • The nature of the error: In which phase of the laboratory testing
Utrecht), a 1024-bed academic teaching hospital in the Netherlands. process the error occurred was analyzed by using both the known
3-phase model of the laboratory testing process (preanalytical, ana-
Voluntary Incident Report System lytical, and postanalytical) and by the 36-step classification scheme
The voluntary patient safety incident report system that was (Table 1).
used at the UMC Utrecht during the study period was the Incident • The cause of the error: Full free-text descriptions of the incident
Management System from the Patient Safety Company (Alkmaar, reports were examined for causes based on the classification
the Netherlands). used in the periodical Dutch report on health-related harm.2
The system was online accessible through the internal Web site Causes were grouped into one of the following 3 main catego-
of the hospital. All personnel of UMC Utrecht can report any in- ries: human factors, technical factors, and organizational factors.
cidents related to patient care including hazards, near misses, and Per report, we only scored the most prominent cause.
adverse events. Reports are filed using a standardized format for • The clinical impact of the error: This includes the severity (in-
what happened where and when. The reporter has to describe the cluding the event resulted in a diagnostic error) and the
nature of the event and the consequences for the patient in free text. consequences of the error. Severity was classified into the fol-
The reporter must assign one of the following categories to the lowing categories: unknown, no harm, distress/anxiety, new
incident: medication, falls, behavior, information and communica- sample taken, potential delay in diagnosis or treatment due to a
tion technology, medical devices, the online patient portal, or other. late test result, a missed diagnosis, and a wrong diagnosis. The
Incidents in the category other are assigned labels by the central in- latter three were considered as a potential diagnostic error, as de-
cident registration of the hospital based on the description of the in- fined by Graber et al.1 The consequences of the error were clas-
cident by the reporter. One of the additional labels is diagnostics sified into the following types: no harm (no negative conse-
that is used for incidents related to any type of diagnostic testing. quences for the patient at all), minor harm (patient inconve-
Incidents will be underreported, as small incidents will happen nience), moderate harm (i.e., delay in surgery schedule, new
and personnel will not be aware that this can be reported to sample taken), major harm (i.e., wrong treatment started based

574 www.journalpatientsafety.com © 2023 The Author(s). Published by Wolters Kluwer Health, Inc.
J Patient Saf • Volume 19, Number 8, December 2023 Errors in the Clinical Laboratory Testing Process

TABLE 1. Nature of Errors: 36 Steps of the Laboratory Testing Process

Main Phases of
Laboratory Process Subprocess Type of Error in Subprocess % of Events (n)
Error in ordering test by clinician Delayed order 0.6 (2)
Wrong order 8.9 (29)
No order 2.4 (8)
Error in providing order to patient Delay in providing the order 0.6 (2)
Wrong patient or wrong form 0.3 (1)
Not providing the order to the patient 0.6 (2)
Error in processing the order Delay in processing the order 0 (0)
Wrongly processing the order 1.5 (5)
Not processing the order 1.2 (4)
Improper collection of specimen Delay in collection 2.4 (8)
Incorrect patient ID 1.5 (5)
Preanalysis Improper specimen 0.6 (2)
Improper container 7.3 (24)
Improper collecting conditions 4.3 (14)
Improper collection of specimen 0.9 (3)
Problems in labeling the container 2.1 (7)
Not collecting the specimen 6.1 (20)
Error in transport of the specimen Delay in transport 4.3 (14)
Improper way of transport 1.2 (4)
Not transporting the specimen 2.1 (7)
Error in preanalytical steps in the laboratory Delay in processing 6,4 (21)
Swapping tests 1.5 (5)
Incorrect patient ID 1.5 (5)
Error in processing 3.1 (10)
Specimen lost 8.0 (26)
Unknown 6,1 (20)
Analytical error 8,0 (26)
Analysis Error in analysis by laboratory specialist. 2,1 (7)
Laboratory device unavailable 1.5 (5)
Unknown 1.8 (6)
Error in reporting result to clinician Delay in reporting 1.2 (4)
Incorrect results reported 3.1 (10)
Not reporting results 0.9 (3)
Not reading the result by clinician 0.6 (2)
Misinterpretation of result by clinician 0.3 (1)
Postanalysis Error in communication of result by Delay in communication 0.3 (1)
clinician to patient
Incorrect result communicated 0.9 (3)
Not communicating the result to the patient 0 (0)
Unknown 0.6 (2)
Unclassifiable 1.5 (5)
Other 1.2 (4)

on wrong laboratory result), disastrous (i.e., death, permanent reports, which were included in the analysis (Fig. 1). A total of
disability, or near life-threatening event), and unknown. 34.5% of these reports (n = 113) were reported by laboratory
workers. Most errors (45.3%, n = 148) occurred in the clinical
RESULTS ward, compared with occurrence in the laboratory setting.
At the UMC Utrecht, a total of 11,827 patient safety incidents For illustration purposes, we provide 3 examples of voluntary
were reported in 2017 and 2018. reports with the data we extracted in Table 2.
Of these, 949 (8%) were labeled as related to diagnostic testing,
including laboratory diagnostics, radiology and function testing Nature of the Error
like electrocardiogram and lung function testing. In Table 1, the distribution of the errors over the phases of the
The first 600 reports in time (of the 949) were screened for be- laboratory testing process is presented. Most errors occurred in
ing related to the laboratory testing process, which resulted in 327 the preanalytical phase, 77.1% (n = 252). Analytical and

© 2023 The Author(s). Published by Wolters Kluwer Health, Inc. www.journalpatientsafety.com 575
van Moll et al J Patient Saf • Volume 19, Number 8, December 2023

(n = 47), a cause could not be determined. Human and organiza-

tional errors were most prevalent in the preanalytical phase. Errors
in the analytical phase were mostly due to technical issues.

Clinical Impact of the Errors

Table 4 presents the distribution of the errors by severity and
the consequences of clinical impact, including diagnostic errors.
In total 195 cases (59.6%), the error in the laboratory testing pro-
cess resulted in diagnostic error. Of these, diagnostic errors with a
highly negative consequence on the patient outcome ([4] major
impact and [5] disastrous impact) accounted for 7.6% (n = 25)
of all laboratory testing error reports. Of the 195 potential diag-
nostic errors, 32.1% concerned misdiagnosis, 17.3% missed
diagnosis, and 50.5% delay diagnostic process.
Table 5 presents the distribution the different phases of the lab-
oratory testing process and the consequences of the errors. Most er-
rors occurred in the preanalytical phase (77.1%). Most errors occur-
ring in the preanalytical phase had no to moderate consequences.
Most errors with severe clinical impact occurred in the analyt-
ical and postanalytical phase, 32% and 28%, respectively. Errors
occurring in the postanalytical phases had major to disastrous con-
sequences in 26.9% of cases.
FIGURE 1. Flow chart selection of incident reports.

postanalytical errors accounted for 13.5% (n = 44) and 8.0% DISCUSSION

(n = 26), respectively. A total of 1.5% of laboratory errors This study provides insights into the nature, causes, and clinical
(n = 5) could not be classified. The top 3 category of error types impact of errors in clinical laboratory testing. To our knowledge,
were improper collection of the specimen (n = 83, 25.4%), errors this is the first study with voluntary incident reports concerning
in preanalytical steps in the laboratory (n = 67, 20.5%), and errors in laboratory testing in which the clinical consequences of
errors in ordering the test by the clinician (n = 39, 11.9%). the error and its association with the type of error were investi-
gated. We found that preanalytical errors, which are most
frequent, have relatively mild consequences for the patients and
Cause of the Errors analytical and postanalytical errors are less frequent but result in
In Table 3, the distribution of the errors by cause category is more harm to patients.
presented. Human factors were the most frequent causes (58.7%, This finding is not unexpected. Preanalytical errors often result
n = 192). Technical causes accounted for 12.5% (n = 41) and or- in a stop or a delay in the test process. Analytical and
ganizational factors accounted for 14.4% (n = 47). In 14.4% postanalytical errors take longer to be detected and result more

TABLE 2. Examples of Incident Reports on Errors in the Laboratory Testing Process (for Illustrative Purposes)

Nature (Phase of the Diagnostic

Incident Description Testing Process) Cause Clinical Impact Error
The laboratory results of this baby of Analysis—analytical error Unidentifiable Wrong diagnosis, major No
04-12-2017 showed a lipase of 858 mmol/L. harm (extra diagnostic test,
The day before the result was 51 mmol/L. wrong therapy started).
Because of a suspected pancreatitis,
hyperhydration was started. On 05-12-2017,
an abdominal ultrasound was performed,
which showed no signs of pancreatitis,
the laboratory results showed a lipase of
41 mmol/L. The results of 04-12-2017
were reanalyzed and showed a lipase
of 28 mmol/L.
Tube for determining the blood type was lost, Preanalytical phase—error in Human Wrong diagnosis, Yes
after arriving at the laboratory. We searched preanalytical steps in the unidentifiable
the laboratory and called the ward, but the tube laboratory—specimen lost.
could not be found. The blood type could not
be determined. The blood must be drawn again.
Wrong leucocyte differentiation results were Postanalysis—error in reporting Human Wrong diagnosis, Yes
reported in this patient. Laboratory worker result to clinician—wrong unidentifiable
2 checked the results of her colleague, results reported
laboratory worker 1, and noticed that
the wrong results were reported in the system.

576 www.journalpatientsafety.com © 2023 The Author(s). Published by Wolters Kluwer Health, Inc.
J Patient Saf • Volume 19, Number 8, December 2023 Errors in the Clinical Laboratory Testing Process

TABLE 3. Causes of Errors in the Laboratory Testing Process

Causes
Human, % (n) Organizational, % (n) Technical, % (n) Unidentifiable, % (n)
Nature of the errors: main phases of the laboratory testing process
Preanalysis 85.4 (164) 85.0 (40) 29.2 (12) 76.6 (36)
Analysis 5.2 (10) 8.5(4) 65.9 (27) 6.4 (3)
Postanalysis 8.8 (17) 6.4(3) 4.9 (2) 8.5 (4)
Unknown 0.5 (1) 0.0 (0) 0.0 (0) 8.5 (4)
Total % (n) 58.7 (192) 14.4 (47) 12.5 (41) 14.4 (47)

often in a wrong diagnostic follow-up and subsequent treatment, preventing these types of errors with a systemic approach different
potentially causing significant harm to the patient. strategies can be chosen, one is to focus on the errors that have
Our finding that most errors occur in the preanalytical testing biggest clinical impact, that is, analytical and postanalytical errors.
phase (77.1%) is consistent with the results from the study by In our opinion, analytical errors, mainly caused by technical er-
Snydman et al13 (81.1%) and the studies of group of Noble rors, are difficult to prevent, because laboratory tests are highly
(69.1%).14,15 Preanalytical events range from 62% to 88% in other developed, controlled, and automated. Analytical errors are
studies in which data from other sources than incident reports mainly causes by technical errors.
were used.6,10,20 Although the number of postanalytical errors is relatively low
In our study, 13.5% of the events are related to the analytical their impact is high compared with preanalytical and analytical er-
phase, and this is consistent with literature in which 11% to rors. Postanalytical processes, especially follow-up of test result,
18% of the events could be related to the analytical phase.20,21 depend strongly on human actions in critical steps. In the past de-
Our number of laboratory errors in the postanalytical phase cades, the risk of lack of follow-up of test results has been identi-
(8.0%) is comparable with the study by Snydman et al,14 in which fied as a major issue in patient safety. A closed-loop system in
5.2% of all errors were postanalytical. which test results are reported to the ordering physician in the
The analysis of the impact of the error on the patient consisted electronic health record is a minimal requirement. However, these
of different measures, the severity, and the consequences. A total systems have to support the physician in such a way there is a min-
of 59.6% of the investigated reports concerned a potential diag- imal change to miss a test result of to forget to follow up the result.
nostic error. A total of 77.1% of the potential diagnostic errors oc- Recent literature shows that follow-up of test results is still a major
curred in the preanalytical phase, but most errors in this phase had health threat to patients. Lack of proper implementation of solu-
no to moderate consequences (88.1%). The rate of potential diagnos- tions and policies is identified as a cause.23
tic errors with a high level of patient impact was low, 7.6% of all lab- For investigating the nature, cause and impact of errors in the
oratory event reports. This rate is in line with literature, in which the laboratory testing process different types of data sources can be
risk of adverse events of inappropriate care due to laboratory errors used. In addition to voluntary incident reports, as we did, one
ranges from 1.7% to 12%.10 Potential diagnostic errors with a high could use databases with medical claims or major safety incidents,
level of patient impact could mostly be related to the analytical and or medical chart review. These data sources will result in different
postanalytical phase. These data demonstrate that the analytical and findings concerning the severity of the impact of the incident. In-
postanalytical phases, in particular the reaction to aberrant laboratory cidents reported as voluntary incident report are less severe and
results, is more important causes of potential adverse outcomes will often concern near misses.24 The value of voluntary incident
for patients, compared with errors in the preanalytical phase.21,22 reports compared with data on major incidents and claims is that
The greater aim of our research is prevention of harm to pa- these reports provided a wider perspective of potential risk in the
tients due to diagnostic error. The results of the current study laboratory testing process where the other two are likely to repre-
can assist in identifying were in the diagnostic process we should sent the tip of the iceberg of incidents.
focus to gain the most impact. Errors in the laboratory testing process Incident report systems are an important tool to detect patient
and their underlying causes are diverse, as are their consequences. In harm and failures in quality of clinical care. Since the release of

TABLE 4. Clinical Impact of Patient Outcome: Number of Errors in the Laboratory Testing Process, Severity Versus Consequences,
Including Diagnostic Errors

Severity of Patient Outcome

No harm Distress New Sample Delay in diagnostic Missed Diagnosis Misdiagnosis Unknown
(%) (%) (%) process (%) (%) (%) (%)
Consequences
No 28 (8.6) 0 (0.0) 0 (0.0) 4 (1.2) 5 (1.5) 20 (6.1) 0 (0.0)
Minor 0 (0.0) 6 (1.8) 56 (17.1) 26 (8.0) 12 (3.7) 10 (3.1) 1 (0.3)
Moderate 0 (0.0) 6 (1.8) 22(6.7) 50 (15.3) 9 (2.8) 17 (5.2) 0 (0.0)
Major 0 (0.0) 0 (0.0) 0 (0.0) 6 (1.8) 4 (1.2) 11 (3.4) 0 (0.0)
Disastrous 0 (0.0) 0 (0.0) 0 (0.0) 3 (0.9) 0 (0.0) 1 (0.3) 0 (0.0)
Unidentifiable 0 (0.0) 1 (0.3) 1 (0.3) 9 (2.8) 4 (1.2) 4 (1.2) 11 (3.4)

© 2023 The Author(s). Published by Wolters Kluwer Health, Inc. www.journalpatientsafety.com 577
van Moll et al J Patient Saf • Volume 19, Number 8, December 2023

TABLE 5. Nature; Main Phases of the Laboratory Testing Process Versus Severity of Patient Outcome

Consequences of Patient Outcome

No, % (n) Minor, % (n) Moderate, % (n) Major, % (n) Disastrous, % (n) Unidentifiable, % (n) Total
Nature: main phases of the laboratory testing process
Preanalysis 17.5 (44) 38.1 (96) 32.5 (82) 3.9 (10) 0.0 (0) 7.9 (20) 77.1 (252)
Analysis 20.5 (9) 18.2 (8) 31.8 (14) 11.4 (5) 6.8 (3) 11.4 (5) 13.5 (44)
Postanalysis 15.4 (4) 19.2 (5) 19.2 (5) 23.1 (6) 3.8 (1) 19.2 (5) 8.0 (26)
Unidentifiable 0 (0) 40.0 (2) 60 (3) 0 (0) 0 (0) 0 (0) 1.5 (5)

the report “To Err is Human,” voluntary reporting systems have REFERENCES
proliferated and are mandatory by, among others, the Joint 1. Graber M. Diagnostic errors in medicine: a case of neglect. Jt Comm J
Commission.25,26 However, it is still challenging to optimize these Qual Patient Saf. 2005;31:106–113.
systems to collect appropriate data and to effectively improve pa-
tient care.25 A successful patient safety learning system must re- 2. Inspectie Gezondheidszorg en Jeugd, Ministerie van Volksgezondheid,
sult in increased awareness of the risk in the work and in improved Welzijn en Sport. In openheid leren van meldingen 2016–2017. Available
at: https://www.igj.nl/publicaties/rapporten/2018/01/25/in-openheid-leren-
risk management. Important aspects of a successful system are sys-
van-meldingen-2016-2017. Accessed November 10, 2022.
tematic data analysis and feedback.26 To maximize the chance to
generate meaningful learning outcomes, it is important to standard- 3. Langelaan M, Broekens , MA, Bruijne MC, et al. Monitor
ize and optimize (e.g., sufficient detail) the classifications used for zorggerelateerde schade 2015/2016: dossieronderzoek bij overleden
type of error, the underlying causes, and clinical impact. With re- patiënten in Nederlandse ziekenhuizen. NIVEL. 2017. Available at: https://
gard to the type of error in the laboratory testing process, our study www.nivel.nl/nl/publicatie/monitor-zorggerelateerde-schade-20152016-
provide a detailed classification, which recently has been imple- dossieronderzoek-bij-overleden-patienten. Accessed November 10, 2022.
mented in the voluntary incident reporting system of our hospital. 4. Balogh EP, Miller BT, Ball JR. Committee on Diagnostic Error in Health
A limitation of our study is lack of detailed data on causes (in Care; Board on Health Care Services; Institute of Medicine; Improving
14.4% undetermined), patient outcomes (9.1% type of harm undeter- Diagnosis in Health Care. Washington, DC: National Academies Press
mined), and information on the patient care setting in incident reports. (US); 2015.
This limited the classification of some of the reports, for exam- 5. O'Kane M. The reporting, classification and grading of quality failures in
ple, the report generally focused on the main cause, and therefore the medical laboratory. Clin Chim Acta. 2009;404:28–31.
it was not possible to analyze multiple causes per incident. Further-
6. Gandhi TK, Kachalia A, Thomas EJ, et al. Missed and delayed diagnoses
more, because the incident reports are voluntary, it is likely that not
in the ambulatory setting: a study of closed malpractice claims. Ann Intern
all laboratory incidents were reported. However, there is no reason Med. 2006;145:488–496.
to believe that diagnostic errors are substantially more or less
underreported than other type of errors. Nevertheless, our data do 7. Schiff GD, Hasan O, Kim S, et al. Diagnostic error in medicine:
not allow to give an estimate of the incidence of laboratory errors. analysis of 583 physician-reported errors. Arch Intern Med. 2009;169:
1881–1887.
8. Plebani M, Laposata M, Lundberg GD. The brain-to-brain loop concept
CONCLUSIONS
for laboratory testing 40 years after its introduction. Am J Clin Pathol.
This study gained insight into the nature, outcomes, and clini- 2011;136:829–833.
cal impact of errors in the laboratory testing process. Most errors
9. Plebani M. The detection and prevention of errors in laboratory medicine.
occur in the preanalytical steps; however, errors in the analytical
Ann Clin Biochem. 2010;47:101–110.
and postanalytical phases have more clinical impact. Errors in lab-
oratory testing often lead to potential diagnostic errors, mostly a 10. Hawkins R. Managing the pre- and post-analytical phases of the total
delay in the diagnostic process. testing process. Ann Lab Med. 2012;32:5–16.
To be even a more useful source for investing safety of diagnostic 11. Plebani M. Exploring the iceberg of errors in laboratory medicine. Clin
testing, voluntary incident reports need to contain more detailed and Chim Acta. 2009;404:16–23.
standardized information, especially on cause and clinical impact.
12. Zwaan L, El-Kareh R, Meyer AND, et al. Advancing diagnostic
Compared with previous estimates most errors still occur in the safety research: results of a systematic research priority setting
preanalytical phase, but we found these errors to have relative low exercise. J Gen Intern Med. 2021;36:2943–2951. doi:10.1007/s11606-020-
impact on the patient outcome. The most harmful events were caused 06428-3.
by human factors and occurred in the analytical and postanalytical
phases. We think that these aspects of clinical laboratory diagnostics 13. Snydman LK, Harubin B, Kumar S. Voluntary electronic reporting of
laboratory errors: an analysis of 37,532 laboratory event reports from 30
currently need the most attention in improving patient safety. To-
health care organizations. Am J Med Qual. 2012;27:147–153.
gether with the new classification of the type of error in the laboratory
testing process presented in this study, classification for cause and 14. Restelli V, Taylor A, Cochrane D, et al. Medical laboratory associated
clinical impact can be implemented in voluntary incident report errors: the 33-month experience of an on-line volunteer Canadian province
systems in hospitals. Wide adoption would provide valuable data wide error reporting system. Diagnosi. 2017;4:79–86.
in guiding efforts in further improving quality of care. 15. Noble MA, Restelli V, Taylor A, et al. Laboratory error reporting rates can
change significantly with year-over-year examination. Diagnosis (Berl).
ACKNOWLEDGMENT 2018;5:15–19.
The authors thank Mariska Hafkamp, Directorate Quality of 16. El-Kareh R, Hasan O, Schiff GD. Use of health information technology to
Care and Patient Safety of the UMC Utrecht, for providing the data. reduce diagnostic errors. BMJ Qual Saf. 2013;22:ii40–ii51.

578 www.journalpatientsafety.com © 2023 The Author(s). Published by Wolters Kluwer Health, Inc.
J Patient Saf • Volume 19, Number 8, December 2023 Errors in the Clinical Laboratory Testing Process

17. Astion ML, Shojania KG, Hamill TR, et al. Classifying laboratory 23. Zimolzak AJ, Shahid U, Giardina TD, et al. Why test results are still
incident reports to identify problems that jeopardize patient safety. Am J getting “lost” to follow-up: a qualitative study of implementation gaps.
Clin Pathol. 2003;120:18–26. J Gen Intern Med. 2022;37:137–144.
18. van Vuuren W, Shea CE, van der Schaaf TW, et al. The development of an 24. Christiaans-Dingelhoff I, Smits M, Zwaan L, et al. To what extent are
incident analysis tool for the medical field. EUT- BDK report. Department adverse events found in patient records reported by patients and healthcare
of Industrial Engineering and Management Science, Technical University professionals via complaints, claims and incident reports? BMC Health
Eindhoven. 1997.85. Serv Res. 2011;11:49.
19. Graber ML, Franklin N, Gordon R. Diagnostic error in internal medicine. 25. Institute of Medicine (US) Committee on Quality of Health Care in
Arch Intern Med. 2005;165:1493–1499. America. To Err is Human: Building a Safer Health System. Kohn LT,
Corrigan JM, Donaldson MS, eds. Washington (DC): National Academies
20. Carraro P, Plebani M. Errors in a stat laboratory: types and frequencies
Press (US); 2000.
10 years later. Clin Chem. 2007;53:1338–1342.
26. Pelletier M, Joint Commission. Accreditation guide for hospitals.
21. Casalino LP, Dunham D, Chin MH, et al. Frequency of failure to inform Oakbrook Terrace Illinois. 2016. Available at: https://www.
patients of clinically significant outpatient test results. Arch Intern Med. jointcommission.org/-/media/deprecated-unorganized/imported-assets/tjc/
2009;169:1123–1129. system-folders/assetmanager/accreditation_guide_hospitals_2011pdf.pdf?
22. Roy CL, Poon EG, Karson AS, et al. Patient safety concerns arising from test db=web&hash=350D19DE3CEF201A9C270B07B7D0FBCD. Accessed
results that return after hospital discharge. Ann Intern Med. 2005;143:121–128. November 10, 2022.

© 2023 The Author(s). Published by Wolters Kluwer Health, Inc. www.journalpatientsafety.com 579