Professional Documents
Culture Documents
Exploring Validation and Verification How They.8
Exploring Validation and Verification How They.8
Exploring Validation and Verification How They.8
John Jacob Barnes, III, MD; Summary Statement: The healthcare simulation (HCS) community recognizes the im-
portance of quality management because many novel simulation devices and techniques
Mojca Remskar Konia, MD, PhD, include some sort of description of how they tested and assured their simulation’s quality.
MACM Verification and validation play a key role in quality management; however, literature
Downloaded from https://journals.lww.com/simulationinhealthcare by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3wX04VDhDA65JyIrQxIAmw8OudmFC7vF5gSn6ug+Sx+k= on 02/09/2020
published on HCS has many different interpretations of what these terms mean and how
to accomplish them. The varied use of these terms leads to varied interpretations of how
verification process is different from validation process. We set out to explore the concepts
of verification and validation in this article by reviewing current psychometric science de-
scription of the concepts and exploring how other communities relevant to HCS, such as
medical device manufacturing, aviation simulation, and the fields of software and engi-
neering, which are building blocks of technology-enhanced HCS, use the terms, with the
focus of trying to clarify the process of verification. We also review current literature avail-
able on verification, as compared with validation in HCS and, finally, offer a working defi-
nition and concept for each of these terms with hopes to facilitate improved communication
within, and with colleagues outside, the HCS community.
(Sim Healthcare 13:356–362, 2018)
Key Words: Verification, validation, healthcare simulation, medical education, quality assurance.
Vol. 13, Number 5, October 2018 © 2018 Society for Simulation in Healthcare 357
Copyright © 2018 by the Society for Simulation in Healthcare. Unauthorized reproduction of this article is prohibited.
industry has is the ability to compare objective information It defines software verification as “objective evidence that de-
from mathematical models and data collected by the electron- sign outputs…meet all of the specified requirements.”21 It
ics of actual aircraft flight with the output data produced by suggests software testing, static and dynamic analysis, and code
simulators to form complex validation techniques.19 inspections as ways of completing this task. The FDA defines
Validation tests are used to compare objectively flight train- software validation in bold text as:
ing device data and airplane data (or other approved reference
data) to assure that they agree within a specified tolerance. “Validation is the confirmation by examination and pro-
Functions tests provide a basis for evaluating flight training vision of objective evidence that software specifications
device capability to perform for a typical training period conform to user needs and intended uses, and that the
and to verify correct operation of the controls, instruments, particular requirements implemented through software
and systems.18 can be consistently fulfilled.”22
A common technique for validation of aviation simula-
tion is to test individual components in a quantitative way in It also recognizes that validation depends on verification
four levels, first individual modules individually, second inter- tasks and recommends testing the functionality of the software
related modules together, third models tested for dynamic re- as a prototype in a simulated use environment and later in lim-
sponse, and finally a pilot using the model.18 ited patient use as key components to ensure validity. The FDA
It is important to note that aviation, based on physics recommends using verification and validation throughout the
principles, lends itself nicely to modeling and simulation. design process.
The principles of aviation are more fundamentally established The pharmaceutical industry, also regulated by the FDA,
and have extensive and accurate mathematical models, which puts a strong emphasis on process validation defined as “the
have been vetted through experimental research over the years. collection and evaluation of data, from the process design stage
Mathematical models based on physics laws and principles can through commercial production, which establishes scientific
accurately predict how an aircraft will behave under variable evidence that a process is capable of consistently delivering
circumstances and actions. This same luxury is not afforded quality product.”24 The FDA details this process in their docu-
to HCS, because human physiology is highly variable within ment “Process Validations: General Principles and Practices”
and between subjects20 and running experimental research accompanied by their current recommendations. Process val-
on human studies is limited by ethical considerations. It is idation breaks down into three stages, with continued process
conceivable that HCS and modeling focusing on specific tasks, verification as the final stage, in which “ongoing assurance is
such as employing an endovascular stent, could approach the gained during routine production that the process remains
precision found in aviation simulation; however, HCS attempting in a state of control (the validated state).”24 Manufacturers of
to emulate a “whole patient,” as in the use of technology- pharmaceutical goods go through a stringent validation pro-
enhanced mannequins, is never likely to achieve the pin- cess proving they “consistently produce drug products…relat-
point accuracy of modeling used in aviation simulation based ing to identity, strength, quality, purity, and potency”24 and
on physics. The differences between simulation in aviation and then must continue to verify that the process does not depart
healthcare are important when discussing verification and from its approved and validated state by collecting and analyz-
validation as the concepts are highly context dependent. ing product and process data. As the pharmaceutical industry
For example, the verification and validation activities for a markets and intends their products for human consumption,
highly mathematical model of aviation physics would rightly the process is scrupulous and overseen by the FDA. Although
look much different than an HCS mannequin used as a HCS for training does not report to the FDA, the broad con-
whole-patient trainer. cepts of validation through scrupulously checking for quality
and reproducibility at various points in production as well as
Verification and Validation in Medicine continued verification are important and practical for applica-
Verification and validation are not foreign concepts to tion to HCS.
medicine or healthcare. As healthcare is an industry with tight The act of validating and verifying biomechanical prod-
regulations and restrictions, overseeing organizations, such as ucts for use in medicine incorporates several similar concepts
the United States FDA, have defined the terms verification used by the nonbiological engineering fields but has its own
and validation and how these concepts apply to determining unique challenges because of their design for and use by the
the safety of products used. Although HCS for training pur- medical community and patients. Hicks et al25 and Stanford
poses is not currently under the umbrella of an overseeing or- University colleagues published recommendations for validat-
ganization, such as the FDA, the industry can learn from the ing and verifying biomechanical products. They recommend
use of these terms in other areas of medicine. verification of computational models and underlying algo-
The FDA publishes guidelines for verification and validation rithms of simulated physiological or physical movement by
of medical devices, now including computational modeling.21–23 comparing with known standards, which they state is more
The FDA relies on making clear definitions and identifying feasible if done individually on components and then on aggre-
broad concepts that can be applied to each unique medical de- gates of components, and validation through comparison of
vice.21 The FDA specifically makes note of how it is common whole model simulation data with patient data from indepen-
to group verification and validation together as “V&V” but dent experiments or other published model data sets. Hicks
specifies it to consider them separate and distinct terms in et al25 mention the limited criterion standard data sets due to
their guidelines for software development in medical devices. difficulty measuring desired output in humans and how there
Vol. 13, Number 5, October 2018 © 2018 Society for Simulation in Healthcare 359
Copyright © 2018 by the Society for Simulation in Healthcare. Unauthorized reproduction of this article is prohibited.
musculoskeletal movement and the underlying established mathe-
matical model of physical movement.25 In this case, the two au-
thors are using different words to describe the similar
mathematical concepts to discuss quality management within
HCS. Gallagher and colleagues,10 who use the word reliability
in their article, caution that readers may misconstrue the word
reliable, because it carries favorable connotations and is syn-
onymous with “good” or “worthwhile” in common speech,
when in scientific writing, reliable means consistent or how a
tool yields the same results when used repeatedly under similar
conditions. Further confounding matters, Cumin et al34 stud-
ied the consistency of physiologic response in various HCSs
and used the term repeatability. It seems that reliability and re-
peatability are not equivocal to verification, rather it may be
more appropriate to consider them individual components
of the larger verification effort. Furthermore, Lampotang36
notes, in an article discussing HCS repeatability, patient
FIGURE 1. Adapted from FDA figure of Design Control Guidance
for Medical Device Manufacturers. A pictorial representation of how variability constitutes an inherent challenge in HCS but also
verification and validation can be incorporated into the design argues that variability within a design framework of HCS
control process (developmental cycle) of HCS as it is in medical “fortuitously relaxes modeling requirements” as repeatabil-
device manufacturing. ity of a simulation should only need to fall within a range of
plausible patient variation that has been predetermined. With
verification and validation into one concept and lack of clear inconsistent use of terms, it is important for authors to state and
separation in the individual activities around verification and describe with clarity what they aim to accomplish and how this
validation. Some studies use the term verification but do not comprises a component of quality control or assurance.
specify what the word means or how it was determined, whereas Sweet et al37 have proposed integration of validation and
other use different terms such as reliability or repeatability verification into the developmental and evaluation process of
when discussing quality management in HCS.9,10,34 The previ- simulation tools. In the article, they describe the technique as
ously mentioned EAES subgroup on evaluating surgical simu- the Center for Research in Education and Simulation Technol-
lators also noted the importance of “reliability” or the “ability ogies (CREST) process, which is based on the earlier “back-
to provide consistent results with minimal errors in measure- ward design principles” of Wiggins and McTighe38 in their
ment” and named test-test reproducibility and internal consis- book on curriculum development. Using backward design
tency as commonly used methods to estimate this.9 Software principles, Sweet et al37 used verification and validation
and engineering fields perform these tests as a form of verifica- techniques throughout the development process and ap-
tion.12,15 In the EAES's work on establishing consensus guide- plied these to early prototypes rather than the final product.
lines for rating simulation equipment available at the time, it The authors noted that using this method helped refine
noted the paucity of data available for verification and decided their models, made verification and validation easier to
to focus their attention solely on validation. Because different complete, and improved their final product. Focus on verifica-
authors in different fields use the words reliability and verifica- tion and validation throughout the developmental process,
tion to describe similar concepts and tasks, it can be confusing. rather than focusing validation and verification efforts after
It seems that verification is the preferred word in software, en- design, is an interesting way to confirm that results are consis-
gineering, and medical device industries for testing and dem- tent (verification) and the individual parts perform their spec-
onstrating test-test reproducibility, internal consistency, ified function (validation).
and accuracy, whereas some HCS publications address sim-
ilar tasks and discuss them under the terms repeatability and Summary
reliability.10,29,30,34,35 Looking at the use of verification and validation in HCS
Gallagher and colleagues10 also discuss how reliability re- provides a glimpse into how these techniques are used and
lates to the medical field. Their article focused on creating val- how broadly defined they are as well as establishes them as
idation systems for evaluating surgical training and assessment unique concepts with interdependence and overlapping of con-
devices. They argued that reliability is a crucial part of a total stituent activities (Fig. 2). As pointed out in articles by
validation system (ie, that it is mathematically consistent, re- Gallagher and colleagues10 and Sweet,35 the educated reader
producible, and dependable) rather than using the term verifi- of scientific publications in this field must be aware of the
cation. Specifically, the group explored split-half method and broad implications and use of the techniques and be careful
test-retest methods to establish reliability in assessment tools.10 to draw their own conclusion of what has been “verified” or
In contrast, Hicks et al25 set out to perform verification in the “validated.” It is again worth noting the importance of using
field of biomechanical devices by determining the accuracy caution when comparing highly context-dependent concepts
of the musculoskeletal models and simulations of movements. such as verification and validation in the distinct fields covered
They defined verification as the determining of agreement in this review. Many of the techniques for verification and val-
between more complex novel computational models of idation in highly scientific or technical fields would not be
practical or have any utility in the evaluation of the applied sci- that we are effectively communicating important ideas for
ence of HCS, but the exploration of these topics helps develop quality management in the rapidly emerging and exciting
and understand purpose of verification and validation in qual- realm of HCS.
ity assurance. Some fields rely on regulatory agencies to pro-
vide strict definitions of and laws governing verification and REFERENCES
validation such as aviation (FAA) and medicine (FDA). We 1. Issenberg SB, McGaghie WC, Petrusa ER, Lee Gordon D, Scalese RJ.
would suggest that this is too stringent for the current state Features and uses of high-fidelity medical simulations that lead to effective
learning: a BEME systematic review. Med Teach 2005;27(1):10–28.
of HCS and the field would rather benefit from further re-
finement of commonly used terms and guidelines for their 2. Rosen KR. The history of medical simulation. J Crit Care 2008;23(2):
157–166.
activities such as verification, validation, and their constituent
elements in a comparable manner to the fundamental sciences 3. Loreiato J, Downing D, Gammon W, et al. The Terminology & Concepts
Working Group. Healthcare Simulation Dictionary. Rockville, MD: Agency
of software and engineering. Efforts have been made to do so for Healthcare Research and Quality; 2016.
as evident by the 2016 publication of a dictionary by Simula- 4. Palaganas JC, Maxworthy J, Epps CA, Mancini ME. Defining Excellence in
tion in Healthcare, which included the terms simulation Simulation Programs. Philadelphia, PA: Lippincott Williams &
reliability and simulation validity. Wilkins; 2014.
Borrowing from those fields, a hybrid definition for veri- 5. Spiegel M, Reynolds PF Jr, Brogan DC. A case study of model context for
fication and validation in HCS could be the following: simulation composability and reusability. Paper presented at: Proceedings
of the 37th Conference on Winter simulation; December 4, 2005.
Verification: the act of confirming the simulator was built
according to its specifications 6. Messick S. Validity of test interpretations and use. ETS Research Report
Series 1990;1:1487–1495.
Validation: the act of confirming the HCS accurately por-
7. American Educational Research Association APA, National Council on
trays what it is modeling and produces the desired outcome Measurement in Education, Joint Committee on Standards for
These definitions differentiate verification and validation Educational and Psychological Testing. Standards for Educational and
and allow for a foundation on which authors can precisely Psychological Testing. Washington, DC: American Educational Research
Association; 2014.
communicate their findings and readers can better understand
the concepts and more effectively draw their own conclusions. 8. Kane MT, Service ET. Validating the interpretations and uses of test scores.
J Educ Meas 2017;50(1):1–73.
We agree that verification and validation should be integral
parts of the design process rather than tasks completed after 9. Carter FJ, Schijven MP, Aggarwal R, et al. Consensus guidelines for
validation of virtual reality surgical simulators. Surg Endosc 2005;
the completions of process, as has been previously suggested. 19(12):1523–1532.
The industries explored, in this article, all stress the impor- 10. Gallagher AG, Ritter EM, Satava RM. Fundamental principles of
tance of integration of verification and validation in the design validation, and reliability: rigorous science for the assessment of surgical
process, as do Sweet and colleagues37 for use in HCS educa- education and training. Surg Endosc 2003;17(10):1525–1529.
tion. Formulating a way to answer the questions from early 11. Oberkampf WL, Trucano TG, Hirsch C. Verification, validation, and
work by Boehm13 with software “Did we build the right prod- predictive capability in computational engineering and physics. Appl Mech
Rev 2004;57(5):345–384.
uct?” and “Did we build the product right?” is much easier and
may more appropriately reflect true validation and verifica- 12. Wallace DR, Fujii RU. Software verification and validation: an overview.
IEEE Softw 1989;6(3):10–17.
tion, before the design process is initiated rather than after
13. Boehm B. Software risk management: principles and practices. IEE Softw
the fact. In the end, the burden of proving a product or tech- 1991;8(1):32–41.
nique as valid or verified falls on the author and the burden
14. Rosson MB, Carroll JM. Usability Engineering: Scenario-Based Development
of drawing those conclusions falls on the reader. Having a clear of Human-Computer Interaction. San Francisco, CA: Morgan
definition of these separate but overlapping concepts ensures Kaufmann; 2002.
Vol. 13, Number 5, October 2018 © 2018 Society for Simulation in Healthcare 361
Copyright © 2018 by the Society for Simulation in Healthcare. Unauthorized reproduction of this article is prohibited.
15. IEEE Draft Guide: Adoption of the Project Management Institute (PMI) 27. Cook DA, Brydges R, Zendejas B, Hamstra SJ, Hatala R.
Standard: A Guide to the Project Management Body of Knowledge (PMBOK Technology-enhanced simulation to assess health professionals: a
Guide)-2008. 4th ed. IEEE; 2011. systematic review of validity evidence, research methods, and reporting
quality. Acad Med 2013;88(6):872–883.
16. Aebersold M. The history of simulation and its impact on the future.
AACN Adv Crit Care 2016;27(1):56–61. 28. Cook DA, Hatala R, Brydges R, et al. Technology-enhanced simulation
for health professions education: a systematic review and meta-analysis.
17. Abrahamson S, Denson JS, Wolf RM. Effectiveness of a simulator in
JAMA 2011;306(9):978–988.
training anesthesiology residents. J Med Educ 1969;44(6):515–519.
18. National Research Council. Appendix F: Validation of Aircraft Flight 29. Devitt JH, Kurrek MM, Cohen MM, et al. Testing the raters: inter-rater
Simulators. In: Webster WC, ed. Shiphandling Simulation: Application to reliability of standardized anaesthesia simulator performance. Can J
Waterway Design. Washington, DC: National Academies Press; 1992. Anaesth 1997;44(9):924–928.
19. Dillard AE. Validation of advanced flight simulators for human-factors 30. Devitt JH, Kurrek MM, Cohen MM, et al. Testing internal consistency and
operational evaluation and training programs. Foundations for Verification construct validity during evaluation of performance in a patient simulator.
and Validation Workshop. John Hopkins University Applied Physics Anesth Analg 1998;86(6):1160–1164.
Laboratory; 2002. 31. Ullah W, Hunter RJ, Finlay M, et al. Validation of a high-fidelity
20. Lipsitz LA, Goldberger AL. Loss of 'complexity' and aging. Potential electrophysiology simulator and development of a proficiency-based
applications of fractals and chaos theory to senescence. JAMA 1992; simulator training program. Simul Healthc 2017;12(1):41–46.
267(13):1806–1809.
32. Morris E, Kesser BW, Peirce-Cottler S, Keeley M. Development and
21. US Food and Drug Administration. Design Control Guidance for Medical validation of a novel ear simulator to teach pneumatic otoscopy. Simul
Device Manufacturers. Rockville, MD: Center for Devices and Radiological Healthc 2012;7(1):22–26.
Health; 1997.
33. Berwick RJ, Mercer SJ, Groom P. Evaluating the fidelity of a novel
22. US Food and Drug Administration. General principles of software part-task trainer for emergency front of neck access training. BMJ Simul
validation. In: Final Guidance for Industry and FDA Staff. Rockville, MD: Technology Enhanced Learn 2017. doi: 10.1136/bmjstel-2017-000208.
Center for Devices and Radiological Health; 2002:6.
34. Cumin D, Chen C, Merry AF. Measuring the repeatability of
23. US Food and Drug Administration. Reporting of computational modeling simulated physiology in simulators. Simul Healthc 2015;10(6):
studies in medical device submissions. In: Guidance for Industry and Food 336–344.
and Drug Administration Staff. Rockville, MD: Center for Devices and
Radiological Health; 2016. 35. Sweet RM. The CREST simulation development process: training the next
generation. J Endourol 2017;31(S1):S69–S75.
24. US Food and Drug Administration. Process validation: general principles
and practices. In: Guidance for Industry. Rockville, MD: Center for Drug 36. Lampotang S. Unlike history, should a simulator not repeat itself? Simul
Evaluation and Research; 2011. Healthc 2015;10(6):331–335.
25. Hicks JL, Uchida TK, Seth A, Rajagopal A, Delp SL. Is my model good 37. Sweet RM, Hananel D, Lawrenz F. A unified approach to validation,
enough? Best practices for verification and validation of musculoskeletal reliability, and education study design for surgical technical skills training.
models and simulations of movement. J Biomech Eng 2015;137(2):020905. Arch Surg 2010;145(2):197–201.
26. Downing SM. Validity: on meaningful interpretation of assessment data. 38. Wiggins G, McTighe J. Understanding by Design. 2nd ed. Alexandria, VA:
Med Educ 2003;37(9):830–837. ACSD; 2005.