Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Research in Social and Administrative Pharmacy

journal homepage: www.elsevier.com/locate/rsap

Error types with use of medication-related technology: A mixed methods T

research study
Shweta R. Shaha, Kimberly A. Galtb,c,∗, Kevin T. Fujic
a
University of Wisconsin Madison School of Pharmacy, Social and Administrative Sciences in Pharmacy1, Department of Pharmacy Sciences and Center for Health Services
Research and Patient Safety, Creighton University School of Pharmacy and Health Professions, Omaha, NE, USA
b
Department of Pharmacy Sciences, Creighton University School of Pharmacy and Health Professions, Omaha, NE, USA
c
Center for Health Services Research and Patient Safety, Creighton University School of Pharmacy and Health Professions, Omaha, NE, USA

ARTICLE INFO ABSTRACT

Keywords: Background: Health information technology has been integrated throughout the medication use process to en-
Automated dispensing cabinets hance safety, quality, and care efficiency. However, technologies have the potential to eliminate or reduce, but
E-prescribing also create some new types of errors.
Health information technology Objective: Assess specific error types before and after the incorporation of two different health information
Medication safety
technologies (HITs), e-prescribing and automated dispensing cabinets (ADCs), into pharmacists’ daily work.
Mixed methods
Patient safety
Methods: A mixed methods design guided use of a pre-existing database of pharmacist survey responses de-
Pharmacists scribing patient safety HIT-related issues in the form of errors prevented and errors observed. In vivo descriptive
text responses were converted into error types. Descriptive analysis was performed to characterize the error
types associated with each HIT.
Results: Four error types were eliminated with the use of e-prescribing, three new error types emerged, and three
error types persisted. With ADC use, four error types were eliminated, three new error types emerged, and three
error types persisted.
Conclusion: Each technology has its own error types, and some persist regardless of HIT use. There is a need to
determine optimal risk reduction approaches for each unique HIT introduced, and design safety practice im-
provement for error types unaffected by the introduction of HIT use.

Introduction reduction and error occurrence observations by pharmacists across

Health information technology (HIT) is believed to be transforma-
tive in remedying problems of inefficiency, preventable medication
errors, and poor care quality.1 Integration of HIT throughout the
medication use process (MUP) has led to the presumption of improved
and safer healthcare.2 Contrary to this expectation, as pharmacists
gradually incorporate and learn to use new HIT in practice, the po-
tential for errors and mishaps continues, resulting in pharmacists who
are unsure about what risks to expect.3 This outcome is consistent with
the widely documented problem of unintended consequences from HIT
adoption and incorporation into daily professional work that has been
captured about electronic health records and their impact.4
When technologies are incorporated into pharmacists’ daily work
specific error types associated with the use of each HIT emerge. A re-
cent study completed by this research team described the error
practice settings related to the use of 13 commonly used technologies in
daily practice. Pharmacists (n = 535) reported observations of error
types reduced or eliminated (n = 1908), as well as, observations of
error types that either continued or were new (n = 1344) once a spe-
cific HIT was introduced into the daily workflow.5 Pharmacists are
trained to be vigilant about risk reduction but do not have the in-
formation necessary to be guided towards implementing appropriate
safety practices specific to the use of each unique HIT. Therefore, it is
important to refine knowledge about safety risks associated with and
unique to each technology being used to devise effective and responsive
safety solutions for daily pharmacy practice.
Objective
The objective of this study was to assess the specific error types

Corresponding author. Center for Health Services Research and Patient Safety, Creighton University, 2500 California Plaza – Boyne Bldg. 123G, Omaha, NE,
∗

68178, USA.
E-mail address: KimberlyGalt@creighton.edu (K.A. Galt).
1
This study was completed at Creighton University for partial fulfillment of the M.S. in Pharmacy Science degree.

https://doi.org/10.1016/j.sapharm.2019.01.010
Received 14 January 2019; Accepted 15 January 2019
1551-7411/ © 2019 Elsevier Inc. All rights reserved.

Please cite this article as: Shweta R. Shah, Kimberly A. Galt and Kevin T. Fuji, Research in Social and Administrative Pharmacy,
https://doi.org/10.1016/j.sapharm.2019.01.010
S.R. Shah, et al.
before and after two different technologies used in the MUP were in-
tegrated into pharmacists’ daily work: e-prescribing and automated
dispensing cabinets (ADCs). These two technologies were chosen be-
cause of their rapid adoption resulting in wide use and integration of
these technologies to all pharmacy practice settings. The aims of this
study were to: a) transform existing qualitative data into categories of
error types, b) count the frequency of these categories, and c) assess the
impact of introducing the technology into daily practice on categories
of error types for the two specific technologies of e-prescribing and
ADCs.
Methods
A sequential transformative mixed methods research design was
used for this study.6 A pre-existing dataset collected and organized by
this study team, the Dyke Anderson Patient Safety Database (DAPSD),
served as the primary source of data.5 The DAPSD was built from a
study in which pharmacists with active licenses practicing in all phar-
macy settings across Nebraska (U.S. state) were surveyed. Respondents
were asked to describe the errors they observed, as well as, the errors
they believed were prevented within a six-month period after using
technologies in their respective pharmacy settings. Pharmacists’ in vivo
qualitative responses to open-ended survey questions were entered into
the DAPSD, resulting in a comprehensive database of qualitatively de-
scribed error-related observations.
In the present study, data transformation techniques were used on
the DAPSD to convert in vivo descriptive text responses into quantifiable
categories of error types for the two technologies of interest: e-pre-
scribing and ADCs.6 First, on-site field observations of e-prescribing and
ADC use processes at two pharmacies selected via convenience sam-
pling (community pharmacy and home care pharmacy) were conducted
from which both themes about safety practices and error types were
generated.6 Second, these observed error types and the National Co-
ordinating Council of Medication Error Reporting and Prevention
(NCCMERP) taxonomy of errors categories were reviewed and com-
pared.7 This evaluation revealed that NCCMERP classifications pro-
vided an incomplete set of descriptors for technology-based error types.
To address this gap, a hybrid of the error types generated by the qua-
litative themes and the existing NCCMERP error types were used to
develop a codebook, which guided the transformation of the qualitative
data in the DAPSD to quantitative data.6 Quantitative analysis of error
types was then performed using Statistical Package for Social Sciences,
version 25, to understand the impact of healthcare-related technologies
on errors.6 Integration of qualitative and quantitative findings provided
a comprehensive understanding of the relationship of errors with in-
troduction of HIT to daily practice. This study was reviewed and ap-
proved by the Creighton University's institutional review board.
Results
While elimination of some error types using each specific HIT was
observed, new error types also emerged, and some error types remained
unaltered regardless of HIT (i.e., the technology appeared to have no
impact on these error types).
With e-prescribing use, there were 4 error types eliminated, 3 new
error types which emerged, and 3 error types that persisted. With ADC
use incorporated into practice, there were 4 error types eliminated, the
emergence of 3 new error types, and 3 error types that persisted. Fig. 1
provides a visual display of the overall findings for the two technologies
within the context of its longitudinal impact. The eliminated error types
are referred to as “error types observed before HIT use” and new error
types as “error types observed after HIT use”. The definition of each
error type category is presented in Table 1. Further details about the
impact of specific HIT use on different error types is described here.
2
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
E-prescribing – error types eliminated
There were 227 eliminated or prevented errors described by 207
pharmacists. These errors encompassed the following error types: leg-
ibility (78%), time delay (6%), knowledge deficit (6%), wrong patient
(5%), handoff (4%), and inaccuracy in drug regimen (1%).
E-prescribing – new error types observed
There were 199 observed errors described by 191 pharmacists
which were categorized into seven error types: inaccuracy in drug re-
gimen (67%), computer system (10%), wrong patient (7%), input error
(3%), time delay (6%), knowledge deficit (6%), and duplicate orders
(2%).
E-prescribing – persistent error types
Four error types: time delay, knowledge deficit, inaccuracy in
dosing regimen, and wrong patient, persisted despite the conversion
from handwritten prescriptions to e-prescribing (i.e., e-prescribing had
no impact on their occurrence).
ADCs – error types eliminated
There were 189 eliminated or prevented errors described by 171
pharmacists. These errors encompassed the following error types: in-
accuracy in drug regimen (57%), dispensing/filling (15%), time delay
(14%), labelling (5%), safety for narcotic use (4%), wrong patient (4%),
and transcription (2%).
ADCs – new error types observed
There were 163 observed errors described by 156 pharmacists
which were categorized into six error types: inaccuracy in drug regimen
(45%), loading error (43%), overrides (7%), wrong patient (2%), time
delay (1%), and computer system error (1%).
ADCs – persistent error types
Time delay, inaccuracy, and wrong patient, are the three error types
that persisted with or without the use of ADCs (i.e. ADCs had no impact
on their occurrence).
Discussion
A key finding from this study highlights the problem of “persistent”
error types that exist irrespective of HIT use, an area of safety not ex-
plicitly identified in the contemporary literature that needs to be ad-
dressed. These error types defy the promise of safer and improved
healthcare attributable to the use of HIT. It is important to study these
errors along with their root causes to understand the major pitfalls in
the design and/or use of each specific technology; to identify areas
within the MUP which are error-prone, and which persist with or
without HIT use.
The results of this study describe the concurrent elimination of some
errors while “new” error types arise due to HIT use from a pharmacist's
perspective.3,8 The pharmacist's viewpoint is new and begins to fill the
detailed knowledge void we have about pharmacy-related HIT specific
technologies, as prior literature does not explicitly discuss this per-
spective. These findings are consistent with the previous findings in the
broader literature that generally discuss new error formation; literature
typically addressing physician or nurses' perspectives on medical error.
It further adds knowledge about two specific HITs, i.e. e-prescribing
and ADCs.
These study results may further assist in differentiating safety risks
inherent to each type of HIT. While not the primary purpose of this
S.R. Shah, et al. Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx

Fig. 1. Error types observed over time with integration of HT Use: E-prescribing and ADCs.

study, this would be an important future study to undertake. We may
consider different factors that might influence safety risk, such as where
in the process of care the error event occurs, the number of patients
potentially impacted by its use (one error to one patient vs. one error to
many patients), and actual function of the technology itself. To illus-
trate, the process of e-prescribing, being integrated into the first step of
the MUP, can cause direct and/or indirect harm to patients. E-pre-
scriptions are specific to an individual, thus harm will generally occur
one patient at a time based upon the individual user of the application.
However, there is an extremely large volume of e-prescriptions making
it a high frequency opportunity. In contrast, accuracy and precision
associated with ADCs is largely dependent upon the accuracy of the
original prescription. Given that medications from one cassette of an
ADC are usually dispensed to many persons, there exists increased
potential for harm to a larger number of patients simultaneously.
Finally, a unique feature about ADCs relative to e-prescribing in the

Table 1
Definition of error types generated on data transformation.
Error type All-inclusive definition for error types observed/prevented

Computer system Software, hardware, application or end-user risk prone design

Duplicate orders Sending same prescription or medication order twice
Dispensing/filling Inaccurate quantity or form of medication used to fill a medication order or prescription
Handoff Transfer of incorrect/inaccurate information specific to a patient from one caregiver to another
Inaccuracy in drug regimen Prescribing or processing wrong drug, wrong dose, wrong timing, wrong route and/or wrong dosage form
Input Wrong entry or typing error
Knowledge deficit Lack of information necessary to safely process a prescription or lack of competence to operate technology or process
Labelling Printed labels with a barcode that can be scanned
Legibility Illegible handwriting or electronically transmitted document (occurs on hand written prescriptions, orders or facsimile)
Narcotic safety Tracking use and distribution of controlled substances
Overrides Skipping the process of scanning while stocking/withdrawing medications from ADCs
Time delay Lag in transmission during prescribing, dispensing manually, and/or delay due to a system breakdown or unavailability of product
Transcription Error in interpreting medication regimen or patient specific information correctly
Wrong patient Misidentification of patients by the prescriber or pharmacist

3
S.R. Shah, et al.
ambulatory care setting is its integration with bar code scanning for
proper drug product identification with the manufacturer's label. This
snapshot of the ways in which just two types of HIT differ, highlights
the importance of both comprehensively understanding the general
impact of all technologies on safety, as well as, the unique impacts on
safety specific to each technology.
Implications for practice
A “technology - type of error” relationship needs to be established
for each HIT, as demonstrated with the examples in this study of e-
prescribing and ADCs. Each HIT must be studied further across
healthcare settings using a framework of classifying errors occurring
due to specific causes inherent in the technology design, operation, and
interface with the human and delivery system.4 This will allow for the
development of tailored solutions matched to specific causes of risk,
which may vary with each HIT used.
Instead of simply devising general guidelines for efficient use of
HIT, greater attention must be paid to both maximizing the safety
benefits resulting from use of HIT while simultaneously addressing
risks. Future studies needed based on the findings and conclusions from
this study include: (a) examination of specific error types uniquely or
predominantly associated with other technologies (smart infusion
pumps, barcode scanners, etc.) on medication errors and patient safety,
(b) pharmacists' views on their patients’ safety using newly introduced
technologies to process and dispense medications, and (c) error event
frequency data based upon HIT-error categories for each specific
technology collected in relationship to the human and delivery system
steps required to care for individuals. With this knowledge, we are
likely to be in a better position to propose strategies/solutions to reduce
risk within pharmacies when technologies are introduced or withdrawn
based upon this newfound knowledge of the safety risks specific to each
technology.
Conclusion
Shifting the paradigm of inquiry to focus on local and specific safety
risks generated by each HIT individually revealed a precise under-
standing of each technology's risks. There are new error types arising
due to use of HIT and persistent error types that exist with or without
use of HIT. This inquiry paradigm shift will help determine the optimal
approach to risk reduction for each unique HIT introduced. This new
knowledge provides the opportunity to reduce risk through focused
strategies of prevention based upon the known specific technology
risks.
Research in Social and Administrative Pharmacy xxx (xxxx) xxx–xxx
Declarations of interest
None.
Funding
This work was supported by the grant “Pharmacists for Patient
Safety” through the Dyke Anderson Patient Safety Grant, Nebraska
Health and Human Services and the Nebraska Office of Rural Health.
The funding source(s) had no involvement in the study design; collec-
tion analysis and interpretation of the data; writing of the report; or the
decision to submit the article for publication.
Acknowledgements
A special thanks to Mr. Ted K. Kaufman, Information Analyst,
Center for Health Services Research and Patient Safety, School of
Pharmacy and Health Professions for his assistance with SPSS software
application use.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.sapharm.2019.01.010.
References
1. National Quality Forum. Identification and prioritization of health IT patient safety
measures. https://www.qualityforum.org/Publications/2016/02/Identification_and_
Prioritization_of_HIT_Patient_Safety_Measures.aspx, Accessed date: 7 November 2018.
2. Bubalo J, Warden BA, Wiegel JJ, et al. Does applying technology throughout the
medication use process improve patient safety with antineoplastics? J Oncol Pharm
Pract. 2018;20:445–460.
3. Oren E, Shaffer ER, Guglielmo BJ. Impact of emerging technologies on medication
errors and adverse drug events. Am J Health Syst Pharm. 2003;60:1447–1458.
4. Harrison MI, Koppel R, Bar-Lev S. Unintended consequences of information technol-
ogies in health care – an interactive sociotechnical analysis. J Am Med Inf Assoc.
2007;14:542–549. https://doi.org/10.1197/jamia.M2384.
5. Galt KA, Fuji KT, Kaufman TK, Shah SR. Health information technology use and pa-
tient safety: study of pharmacists in Nebraska. Pharmacy. 2019;7:7. https://doi.org/
10.3390/pharmacy7010007.
6. Creswell J, Plano Clark V. Designing and Conducting Mixed Methods Research. third ed.
Thousand Oaks, California: Sage Publications Inc; 2007.
7. NCC MERP Taxonomy of Medication Errors. https://www.nccmerp.org/sites/default/
files/taxonomy2001-07-31.pdf. Accessed May 1, 2018.
8. Nadzam DM, Mackles RM. Promoting patient safety: is technology the solution? Jt
Comm J Qual Improv. 2001;27:430–436.