ASSESSING THE EFFECT OF VARIED DESIGN ELEMENTS ON INFORMATION
PROCESSING IN MEDICAL DEVICE LABELS
By
Do Chan Seo

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Packaging – Doctor of Philosophy
2014

ABSTRACT
ASSESSING THE EFFECT OF VARIED DESIGN ELEMENTS ON INFORMATION
PROCESSING IN MEDICAL DEVICE LABELS
By
Do Chan Seo
The labeling of medical devices plays a very important role in communicating
product information to healthcare providers so that the device is used safely and effectively.
Cai (2012) identified various labeling issues as problematic in medical devices. Specifically:
small font sizes, poor color contrasts, absence of latex status or sterility status, and varied
locations for three pieces of critical information (latex status, sterility status and expiration
dating).
Research proposed herein is comprised of three experimental parts: (1) a
benchmarking study to verify (or refute) Cai’s findings, (2) a study investigating how design
strategies impact early stages of information processing using a change detection
methodology; this portion also evaluates symbol comprehension, and (3) a forced choice
task which enumerates the effect of design elements on the correct selection of a device
and time to select the same. The four design factors which were evaluated are: grouping of
critical information, boxing of critical information, symbol presence/absence and color
coding.
The key findings from our benchmarking study support Cai’s conclusion (2012) that
the three pieces of critical information were scattered throughout medical device labels, and
that their font sizes were relatively smaller than those of the product name. Legibility testing
also bolstered Cai’s findings: all three pieces of information deemed critical to the safe and
effective use of a medical device were indicated to be significantly less legible (α=0.05) than
product name and brand name when 20 commercial labels were tested by 99 participants.

Attentive behaviors of participants were evaluated measuring the proportion of the
sample that successfully detected changes to stimulus prior to time out (60 seconds) and
time to successful change detection. Participants detected changes significantly faster when
the three pieces of critical information were boxed than when they were unboxed, in both
grouped (p=0.0086) and ungrouped (p<0.0001) formats. Color-coded designs enabled
participants to detect changes significantly faster than non-color-coded design, in both
grouped (p<0.0001) and ungrouped (p<0.0001) formats. In addition, the 3-way interaction
term of Boxing x Symbol x Color was found to be significant (p=0.0323). Though grouping
enhanced performance in treatments with colors, it slowed performance in the boxed
conditions.
When comprehension rates of symbols were evaluated, only 6 out of 38 symbols in
the internationally recognized standard, AAMI/ANSI/ISO 15223: 2007 A1: 2008, were
classified as successful. Three symbols from the same standard were categorized as
“critically confusing" for participants: they were not only misunderstood, but, in fact,
interpreted to have the opposite meaning of what was intended.
When subjects were asked to identify a product with a particular feature (e.g.
containing latex) as quickly as possible, three design effects, namely, Color (p<0.0001),
Grouping (p=0.0104), and Symbol (p<0.0001) decreased time to selection. Grouping
information in one location, the presence of symbols and color coding showed significantly
higher probability rates and less time to correct device selection when compared with the
two commercial labels (α=0.05).
Our work indicates that medical device manufacturers should seriously consider
employing these design elements to develop a standard labeling format for critical
information. Further, policy changes regarding stand-alone graphical representation on
medical device labels should be carefully considered prior to their implementation.

Copyright by
DO CHAN SEO
2014

Special thanks to Dr. Laura Bix for all the advice
given during my Ph. D program

v

ACKNOWLEDGEMENTS

The author wishes to express appreciation to the Blue Cross Blue Shield of
Michigan Foundation for the award of an Investigator Initiated Grant which supported
this work.

Further, the author wishes to express appreciation to the sponsors and

attendees of the Healthcare Packaging Immersion Experience (HcPIE); proceeds from
this event also supported the research presented herein. Additionally, the author
wishes to acknowledge CR Bard and Teleflex that provided several cases of indwelling
urinary catheters for our benchmarking study, free of charge on short notice. The
author also wishes to acknowledge the Association of Surgical Technologists (AST)
support for subject recruitment through their 2013 Instructors Forum at Savannah,
Georgia and 2014 Annual National Conference at Denver, Colorado.
The author would like to thank Dr. Laura Bix for being a wonderful,
understanding advisor for this research.

It has been a great honor to work with you in

this medical device research area.
The author would also like to extend a deep thank-you to Dr. Mark Becker who
guided this research to right directions regarding the experimental designs of change
detection and forced choice task methods.

And, the author wishes to acknowledge Dr.

Diana Twede and Dr. Susan Selke for your tremendous advice during my Ph. D
program.
The author would like to thank Healthcare, Universal Design, Biomechanics
(HUB) members (Tony Trier, Cory Wilson, Lanqing Liu, James Richardson, Audrey
Wilson, Adams Watts, Eric Seeley) for your big support to recruit subjects and to

vi

conduct research at Savannah and Denver.

It was a long trip to drive to Savannah,

GA and Denver, CO, but was the great fun driving which should be kept in my good
memory book.
Finally, the author would like to express a special thank-you to my wife
(Seungeun Kim) for your patience and commitment to devote your life to our family.
Thank God for being with me all the times.

vii

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... xii
LIST OF FIGURES ......................................................................................................... xv
CHAPTER 1 INTRODUCTION .........................................................................................1
1.1 Significance of this research...........................................................................1
1.1.1 Labeling matters .................................................................................1
1.1.2 End-user VOCs ..................................................................................3
1.1.2.1 Latex status .........................................................................3
1.1.2.2 Expiration dating .................................................................4
1.1.2.3 Sterility status......................................................................4
1.1.2.4 Summary.............................................................................5
1.2 Research goals ..............................................................................................5
CHAPTER 2 LITERATURE REVIEW ...............................................................................7
2.1 Medical Device labeling regulations ...............................................................7
2.1.1 Misbranding ......................................................................................7
2.1.2 General labeling requirements ..........................................................7
2.1.3 Special labeling requirements ......................................................... 10
2.2 End-users’ voices on labeling ....................................................................... 13
2.3 Significance of latex status ........................................................................... 16
2.3.1 Manufacturing process of latex-containing medical devices….. ...... 16
2.3.2 Latex allergy types .......................................................................... 16
2.3.3 Risks associated with latex allergy ................................................. 18
2.3.4 FDA countermeasure actions ......................................................... 19
2.4 Significance of sterility status ....................................................................... 21
2.4.1 Packaging functions and materials ….. ........................................... 21
2.4.2 Sterilization methods....................................................................... 22
2.4.3 Health care-associated Infections (HAIs)........................................ 23
2.4.4 Surgical microbial contamination .................................................... 25
2.5 Significance of expiration dating ….. ............................................................ 26
2.6 Information Processing ….. .......................................................................... 27
2.7 Labeling Noticeability .................................................................................... 29
2.7.1 Change Detection .......................................................................... 29
2.7.2 Eye tracking ................................................................................... 34
2.8 Text Legibility................................................................................................ 40
2.9 Symbol Comprehension ............................................................................... 43
2.9.1 Global standards ….. ...................................................................... 43
2.9.2 Comprehension evaluations ........................................................... 47
2.10 Forced Choice Task ................................................................................... 49
viii

CHAPTER 3 BENCHMARKING EXISTING, COMMERCIAL LABELING FOR
INDWELLING, URINARY CATHETERS ........................................................................ 51
3.1 Objective ...................................................................................................... 51
3.2 Methodology ................................................................................................ 52
3.2.1 Placement of labeling information.................................................. 52
3.2.2 Measurement of leading, kerning and type size, and color contrast
evaluation ............................................................................................... 53
3.2.2.1 Equipment ...................................................................... 53
3.2.2.2 Materials and Methods ................................................... 54
3.2.3 Symbol evaluation ......................................................................... 56
3.2.3.1 Originating standard ....................................................... 56
3.2.3.2 Presence/absence (with and/or without text) ................. 57
3.2.3.3 Symbol size.................................................................... 57
3.2.3.4 Color contrast of symbols............................................... 57
3.2.4 Legibility of texts ............................................................................ 57
3.2.4.1 Subjects ......................................................................... 57
3.2.4.2 Equipment ...................................................................... 58
3.2.4.3 Materials and methods ................................................... 59
3.3 Results ......................................................................................................... 63
3.3.1 Placement of labeling information.................................................. 63
3.3.2 Measurement of leading, kerning and type size, and color contrast
evaluation ............................................................................................... 65
3.3.3 Symbol evaluation ......................................................................... 68
3.3.4 Legibility ........................................................................................ 72
3.4 Discussion and Conclusions......................................................................... 76
CHAPTER 4 DESIGN EFFECTS (BOXING, GROUPING, SYMBOL AND COLOR
CODING) ON EARLY STAGES OF THE INFORMATION PROCESSING MODEL
USING CHANGE DETECTION ...................................................................................... 78
4.1 Objective & Hypothesis ................................................................................ 78
4.2 Methodology ................................................................................................ 79
4.2.1 Subjects ......................................................................................... 79
4.2.2 Equipment and Software: Change Detection ................................. 80
4.2.3 Materials and Method: Efficacy of boxing, grouping, symbol and
color........................................................................................................ 85
4.2.4 Materials and Method: Comprehension of symbols ....................... 94
4.2.4.1 Stimulus materials .......................................................... 94
4.2.4.2 Procedure ...................................................................... 94
4.2.4.3 Categorization ................................................................ 95
4.3 Results ......................................................................................................... 96
4.3.1 Subject demographics ................................................................... 96
4.3.2 Descriptive statistics on questionnaire evaluation ......................... 98
4.3.2.1 Years of experience ....................................................... 98
4.3.2.2 Employment settings ...................................................... 99
4.3.2.3 Position & role .............................................................. 100
ix

4.3.2.4 Critical pieces of labeling information ........................... 100
4.3.2.5 Critical labeling problems ............................................. 101
4.3.2.6 Medical errors due to labeling issues ........................... 102
4.3.2.7 Recommendations on labeling designs........................ 102
4.3.3 Statistical analysis on Change Detection ..................................... 103
4.3.3.1 Binary Variable – Change Detected (Yes/No) .............. 103
4.3.3.2 Continuous Variable – Time to detect change
(milliseconds) .............................................................. 106
4.3.3.2.1 Significant 2-way interaction terms: Grouping x
Boxing ................................................................ 107
4.3.3.2.2 Significant 2-way interaction terms: Color x
Grouping ............................................................. 108
4.3.3.2.3 Significant 2-way interaction terms: Grouping x
Symbol ............................................................... 109
4.3.3.2.4 Significant 3-way interactions: Boxing x Symbol x
Color ................................................................... 110
4.3.4 Percentage statistics on symbol comprehension evaluation ....... 112
4.4 Discussion .................................................................................................. 119
4.4.1 Change Detection ........................................................................ 120
4.4.2 Symbol Comprehension .............................................................. 122
CHAPTER 5 DESIGN FEATURES (BOXING, GROUPING, SYMBOL AND COLOR
CODING) INFLUENCE ON INFORMATION PROCESSING DURING A FORCED
CHOICE TASK ........................................................................................................... 124
5.1 Objective & Hypothesis .............................................................................. 124
5.2 Methodology .............................................................................................. 124
5.2.1 Subjects ....................................................................................... 124
5.2.2 Materials and Method .................................................................. 125
5.3 Results ....................................................................................................... 134
5.3.1 Subject demographics ................................................................. 134
5.3.2 Descriptive statistics on questionnaire evaluation ....................... 137
5.3.2.1 Years of experience ..................................................... 137
5.3.2.2 Employment settings .................................................... 138
5.3.2.3 Position & role .............................................................. 138
5.3.2.4 Critical pieces of labeling information ........................... 139
5.3.2.5 Critical labeling problems ............................................. 139
5.3.2.6 Medical errors due to labeling issues ........................... 140
5.3.2.7 Recommendations on labeling designs........................ 141
5.3.3 Statistical analysis on Forced Choice Tasks ............................... 142
5.3.3.1 Binary Variable – correct choice (Yes/No) ................... 142
5.3.3.2 Continuous Variable: Time taken to make a correct
choice (milliseconds) ................................................. 144
5.3.3.2.1 Significant main terms: Grouping, Symbol and
Color ................................................................. 145
5.3.3.3 Pairwise comparisons between optimal (grouped +
symbol present + color-coded) label and commercial
x

labels ......................................................................... 148
5.3.3.3.1 Binary Variable: Probability of correct choice
(Yes/No) ........................................................... 148
5.3.3.3.2 Continuous Variable: Time taken to make a correct
choice (milliseconds) ....................................... 149
5.4 Discussion .................................................................................................. 149
CHAPTER 6 CONCLUSIONS, AND LIMITATIONS & FUTURE WORK .................... 153
6.1 Conclusions ............................................................................................... 153
6.2 Limitations & Future work .......................................................................... 155
APPENDICES .............................................................................................................. 158
APPENDIX 1. Proposed Rules of the FDA on medical device labeling ........................ 159
APPENDIX 2. Misbranding (specified by Section 502, Federal Food, Drug and Cosmetic
Act)….. ......................................................................................................................... 163
APPENDIX 3. Latex glove manufacturing process ....................................................... 164
APPENDIX 4. Stimulus materials for Legibility test ...................................................... 165
APPENDIX 5. Data collection sheet: Legibility test ...................................................... 170
APPENDIX 6. Data collection sheet: Comprehension test ........................................... 172
APPENDIX 7. Research questionnaire form: Change Detection/Forced Choice Task
test ............................................................................................................................... 179
APPENDIX 8. Consent form: Legibility test .................................................................. 182
APPENDIX 9. Consent form: Change Detection/Comprehension tests ....................... 185
APPENDIX 10. Consent form: Forced Choice Task test .............................................. 188
APPENDIX 11. Recruitment flyer: Legibility test ........................................................... 191
APPENDIX 12. Recruitment flyer: Change Detection/Comprehension tests ................ 192
APPENDIX 13. Recruitment flyer: Forced Choice Task test......................................... 193
APPENDIX 14. Critical pieces of labeling information (Change Detection) .................. 194
APPENDIX 15. Critical labeling problems (Change Detection) .................................... 195
APPENDIX 16. Critical pieces of labeling information (Forced Choice Task) ............... 196
APPENDIX 17. Critical labeling problems (Forced Choice Task) ................................. 197
APPENDIX 18. Benchmarking study labels.................................................................. 198
APPENDIX 19. Placement of critical information .......................................................... 205
APPENDIX 20. Evaluation of text size of critical information ........................................ 206
APPENDIX 21. Evaluation of text leading of critical information ................................... 207
APPENDIX 22. Evaluation of text color contrast of critical information ......................... 208
APPENDIX 23. Evaluation of symbol size of critical information .................................. 209
APPENDIX 24. Evaluation of symbol color contrast of critical information ................... 210
APPENDIX 25. Evaluation of originating symbol standard of critical information ......... 211
APPENDIX 26. Evaluation of presence/absence of symbols for critical information .... 212
BIBLIOGRAPHY .......................................................................................................... 213

xi

LIST OF TABLES

Table 1. Information critical to the safe and effective use of medical devices as identified
by Cai ....................................................................................................................... 3
Table 2. General labeling requirements for medical devices (CFR Title 21, 801, 2013) . 8
Table 3. Adequate directions for use (CFR Title 21, 801.5, 2013) .................................. 9
Table 4. Failures of prominence or conspicuousness (CFR Title21, 801.15, 2013) ...... 10
Table 5. Latex rubber formations (CFR Title 21, 801.437, 2013) .................................. 11
Table 6. Labeling information on sterilization (CFR Title 21, 820, 2013) ....................... 12
Table 7. Medical device package considerations by nurses ((Bustchli, 2008)............... 13
Table 8. Challenges associated with the labeling of critical information and suggested
information (Cai, 2012) ........................................................................................... 15
Table 9. Medical devices that commonly contain latex (Adapted from Alwilda et al., 2003)
............................................................................................................................... 17
Table 10. Articles addressing risks associated with latex allergy .................................. 19
Table 11. Healthcare-associated infections in U.S. hospital during 2011(CDC, 2012a) 24
Table 12. Change detection studies relating to packaging or labels ............................. 34
Table 13. Eye tracking studies relating to packaging or labels ...................................... 39
Table 14. Terminology definition of letter and design elements on a label .................... 41
Table 15. Symbols comparison among international and US standards ....................... 46
Table 16. Labeling information for legibility evaluation .................................................. 63
Table 17. Experiment combinations of Change Detection............................................. 86
Table 18. Color coding formats: Change Detection ....................................................... 87
Table 19. Matrix Chart of Change Detection Trials ....................................................... 88
Table 20. Critical pieces of labeling information with top 5 out of 14 response groups 101
Table 21. Critical labeling problems with top 5 out of 10 response groups.................. 102

xii

Table 22. Medical errors participants experienced due to labeling issues................... 102
Table 23. Suggested recommendations to resolve labeling problems ........................ 103
Table 24. Percentages of each category response on medical device symbols ......... 113
Table 25. Percentage of correct response category on medical device symbols: Means
and Upper & Lower Confidence Limits at 95% confidence level .......................... 115
Table 26. Passing symbols in comprehension: Means and Upper & Lower Confidence
Limits at 95% confidence level ............................................................................. 116
Table 27. Symbols at less than 10% percentage in comprehension: Means and Upper &
Lower Confidence Limits at 95% confidence level ............................................... 118
Table 28. “Critically confusing symbols”: Means and Upper & Lower Confidence Limits
at 95% confidence level ....................................................................................... 119
Table 29. Color coding formats: Forced Choice Task ................................................. 126
Table 30. Matrix chart of Forced Choice Tasks: Newly developed labels ................... 128
Table 31. Matrix chart of Forced Choice Tasks: Commercial labels............................ 128
Table 32. Questions of Forced Choice Tasks ............................................................. 129
Table 33. Critical pieces of labeling information with top 5 out of 15 response groups 139
Table 34. Critical labeling problems with top 5 out of 10 response groups.................. 140
Table 35. Medical errors participants experienced due to labeling issues................... 141
Table 36. Suggested recommendations to resolve labeling problems ........................ 141
Table 37. Key benefits through UDI implementation ................................................... 160
Table 38. Misbranding items ....................................................................................... 163
Table 39. Latex glove manufacturing process (Zalglaniczny, 2001; Yunginger, 1998) 164
Table 40. Legibility data sheet ..................................................................................... 171
Table 41. Critical pieces of labeling information (Change Detection) .......................... 194
Table 42. Critical labeling problems (Change Detection)…. ........................................ 195
Table 43. Critical pieces of labeling information (Forced Choice Task) ....................... 196
Table 44. Critical labeling problems (Forced Choice Task) ......................................... 197
xiii

Table 45. Placement of critical information.................................................................. 205
Table 46. Text size of critical information .................................................................... 206
Table 47. Text leading of critical information ............................................................... 207
Table 48. Text color contrast of critical information ..................................................... 208
Table 49. Symbol size of critical information ............................................................... 209
Table 50. Symbol color contrast of critical information ................................................ 210
Table 51. Originating symbols standard of critical information .................................... 211
Table 52. Presence/absence of symbols for critical information .................................. 212

xiv

LIST OF FIGURES

Figure 1. Diagram of Information Processing Model ..................................................... 29
Figure 2. Change Detection Image Cycle...................................................................... 31
Figure 3. Locations of labeling information .................................................................... 53
Figure 4. Optical comparator ......................................................................................... 54
Figure 5. Leading of typefaces ...................................................................................... 55
Figure 6. Kerning of typefaces ...................................................................................... 55
Figure 7. Relative x-heights of typefaces ...................................................................... 56
Figure 8. Symbol to fill square ....................................................................................... 57
Figure 9. The Lockhart Legibility Instrument (LLI) ......................................................... 59
Figure 10. Visual acuity card ......................................................................................... 60
Figure 11. Pseudo-iso chromatic plates for testing color perception ............................. 61
Figure 12. Legibility test procedure ............................................................................... 62
Figure 13. Placement findings in regard to four pieces of critical information................ 65
Figure 14. Leading measurement results in regard to three pieces of critical information
............................................................................................................................... 66
Figure 15. x-height analysis results in regard to four pieces of critical information ........ 67
Figure 16. Findings on text color contrast in regard to four pieces of critical information
............................................................................................................................... 68
Figure 17. Non-standard symbols found from the benchmarking study ........................ 69
Figure 18. Symbol presence/absence findings in regard to three pieces of critical
information ............................................................................................................. 70
Figure 19. Symbol measurement results in regard to three pieces of critical information
............................................................................................................................... 71
Figure 20. Findings on symbol color contrasts in regard to three pieces of critical
information ............................................................................................................. 72
xv

Figure 21. Demographics information of subjects for gender, race, native language and
education (highest level achieved) ......................................................................... 73
Figure 22. Subject characteristics on visual acuity, color blindness and health literacy 74
Figure 23. Estimated least square means (LSM) of degree of rotation on legibility of the
five pieces of labeling information with estimated upper and lower limits. Letters
indicate statistical significance at α = 0.05. ............................................................ 76
Figure 24. Sequence of Change Detection Images ....................................................... 81
Figure 25. The same label appears with sections in different locations. Location was
randomized across subjects. .................................................................................. 84
Figure 26. Change Detection Trials: Image Type A ...................................................... 89
Figure 27. Change Detection Trials: Image Type B ...................................................... 91
Figure 28. Change Detection Trials: Image Type C ...................................................... 93
Figure 29. Age of participants ....................................................................................... 96
Figure 30. Demographics information (%) of participants on gender, ethnicity, native
language and education (highest level achieved) .................................................. 97
Figure 31. Subject characteristics (%) for visual acuity, color blindness and health
literacy .................................................................................................................... 98
Figure 32. Experience in years ...................................................................................... 99
Figure 33. Employment settings of participants (%) ...................................................... 99
Figure 34. Position & role of participants (%) .............................................................. 100
Figure 35. The effect of ‘Grouping’ on the probability of successful change detection:
Estimated least square means (LSM) with estimated upper and lower limits. Letters
indicate statistical significance at α=0.05. ............................................................ 105
Figure 36. The effect of ‘Color’ on the probability of successful change detection:
Estimated least square means (LSM) with estimated upper and lower limits. Letters
indicate statistical significance at α=0.05. ............................................................ 105
Figure 37. The effect of ‘Grouping’ and ‘Boxing’ formats on Time to detect change:
Estimated least square means (LSM) with estimated upper and lower limits. Letters
indicate statistical significance at α=0.05. ............................................................ 108

xvi

Figure 38. The effect of ‘Grouping’ and ‘Color-coding’ formats on Time to detect change:
Estimated least square means (LSM) with estimated upper and lower limit bars.
Letters indicate statistical significance at α=0.05. ................................................ 109
Figure 39. The effect of ‘Grouping’ and ‘Symbol’ formats on Time to detect change:
Estimated least square means (LSM) with estimated upper and lower limits. Letters
indicate statistical significance at α=0.05. ............................................................ 110
Figure 40. The effect of ‘Boxing’, ‘Symbol’ and ‘Color’ formats on Time to detect change:
Estimated least square means (LSM) with estimated upper and lower limits. Letters
indicate statistical significance at α=0.05. ............................................................ 111
Figure 41. Test cell # 26 from Table 30 (Latex vs. Latex free information).................. 130
Figure 42. Test cell # 25 from Table 30 (Latex vs. Latex free information).................. 131
Figure 43. Commercial label A (Latex vs. Latex free information) ............................... 132
Figure 44. Commercial label B (Latex vs. Latex free information) ............................... 133
Figure 45. Age of participants ..................................................................................... 134
Figure 46. Demographics information (%) of participants on gender, ethnicity, native
language and education (highest level achieved) ................................................ 135
Figure 47. Subject characteristics (%) on visual acuity, color blindness and health
literacy .................................................................................................................. 136
Figure 48. Experience in years .................................................................................... 137
Figure 49. Employment settings of participants (%) .................................................... 138
Figure 50. Position & role of participants (%) .............................................................. 138
Figure 51. The effect of ‘Boxing’ on Probability of correct choice: Estimated least square
means (LSM) with estimated upper and lower limits. Differing letters indicate
statistical significance at α=0.05........................................................................... 143
Figure 52. The effect of ‘Symbol’ on Probability of correct choice: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters
indicate statistical significance at α=0.05. ............................................................ 144
Figure 53. The effect of ‘Grouping’ on Time to make a correct choice: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters
indicate statistical significance at α=0.05. ............................................................ 146

xvii

Figure 54. The effect of ‘Symbol’ on Time to make a correct choice: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters
indicate statistical significance at α=0.05. ............................................................ 147
Figure 55. The effect of ‘Color’ on Time to make a correct choice: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters
indicate statistical significance at α=0.05. ............................................................ 147
Figure 56. The effect of ‘Grouped + Symbol presence + Color-coded’ design on
Probability of correct choice, compared to commercial label designs: Estimated
least square means (LSM) with estimated upper and lower limits. Differing letters
indicate statistical significance at α=0.05. ............................................................ 148
Figure 57. The effect of ‘Grouped + Symbol presence + Color-coded’ design on Time to
make a correct choice, compared to commercial label designs: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters
indicate statistical significance at α=0.05. ............................................................ 149
Figure 58. Legibility stimulus materials for Brand A .................................................... 165
Figure 59. Legibility stimulus materials for Brand B .................................................... 166
Figure 60. Legibility stimulus materials for Brand C and D .......................................... 167
Figure 61. Legibility stimulus materials for Brand E .................................................... 168
Figure 62. Legibility stimulus materials for Brand F ..................................................... 169
Figure 63. Medical device symbols ............................................................................. 172
Figure 64. Recruitment flyer: Legibility test ................................................................. 191
Figure 65. Recruitment flyer: Change Detection/Comprehension tests ....................... 192
Figure 66. Recruitment flyer: Forced Choice Task test ............................................... 193
Figure 67. Benchmarking study labels ........................................................................ 198

xviii

CHAPTER 1
INTRODUCTION
1.1 Significance of this research
1.1.1 Labeling matters
General labeling requirements for medical devices that are sold in the US can
be found in the Code of Federal Regulations, specifically, 21 CFR 801.

At minimum,

device manufacturers must prominently label:
•

The name and place of business (21 CFR 801.1)

•

The intended use of the device (21 CFR 801.4) and

•

Adequate directions for use (21 CFR 801.5).

Beyond this, there are labeling requirements related to the specifics of the devices
themselves, such as those containing latex (21 CFR 801.437) or that are delivered in a
sterile state.

In September of 2013, the FDA published a final rule mandating the

presence of a unique device identifier (UDI) pertaining to most medical devices sold in
US commerce. Two items in the final rule, the device identifier and the production
identifier, have the potential to force device manufacturers to revise their labeling.
device identifier references information specific to the model.

The

The production identifier

requires one or more of the following items be presented in both plain-text format that
can be easily readable by patients and health care professionals, and a format that can
be read by a barcode scanner or an Automatic Identification Data Capture (AIDC)
technology (FDA, UDI Final Rule, 2013):
1

• The lot or batch within which a device was manufactured;
• The serial number of a specific device;
• The expiration date of a specific device;
• The date a specific device was manufactured;
• The distinct identification code required by 21 CFR 1271. 290 (c) for a
human cell tissue, or cellular and tissue-based products (HCT/P) regulated
as a device.
The likely revision of the vast majority of labels within the device industry
presents an opportunity to update not only the information dictated by the final rule, but
to optimize the label content so that it can be readily read and understood by healthcare
providers.
In April 2013, another proposed rule regarding medical device labeling was
issued by FDA.

It is composed of two important changes: (1) to allow for the inclusion

of stand-alone graphical representation of information or symbols which are established
as part of a standard developed by a nationally or internationally recognized standards
development organization (SDO) and accompanied by a symbols glossary, and (2) to
authorize the use of the symbol statement “Rx only” on the labeling of prescription
devices (FDA, Use of certain symbols, Proposed rule, 2013).

Its primary purpose is to

make medical device labeling more user-friendly by replacing small, difficult-to-read text
with pictorial information, and to harmonize the labeling information of US and foreign
regulatory bodies (e.g. European Commission). If this rule is enacted, it will benefit

2

medical device manufacturers in several ways: e.g. increased space utilization and a
single labeling system for multiple markets (e.g. US, EU countries)
1.1.2 End-user VOCs (Voice Of Customers)
Cai (2012) conducted a series of seven focus groups comprised of operating
room personnel (primarily perioperative nurses and surgical technologies) with the goal
of identifying the most prevalent problems associated with medical device packaging.
Two major problems emerged as critical: device labeling and difficulty associated with
sterile presentation of devices (Cai, 2012).
As groups drilled down into the intricacies regarding device labeling, a central
theme was that non-critical information hindered healthcare personnel from finding
critical information easily.

Four pieces of information were repeatedly indicated as

critical to patient care and problematic in the labeling of commercial devices at present
(see Table 1).
Table 1. Information critical to the safe and effective use of
medical devices as identified by Cai
No.

Critical information

1

Latex status

2

Expiration dating

3

Sterility status

4

Product name

1.1.2.1 Latex status
Study participants reported that a lack of information regarding latex status
causes confusion; for instance, when information regarding latex is absent, many
respondents reported uncertainty regarding whether or not latex was of concern (Cai,
3

2012).
CFR 801.437 dictates that devices containing natural rubber or synthetic rubber
are required to have the following caution statement on the outside package, container
or wrapper: “Contains or presence of natural rubber latex which may cause allergic
reactions” (CFR 801.437). Focus group results suggest, however, that some confusion
still exists regarding the current approach.
1.1.2.2 Expiration dating
For sterile medical devices, packages have the unique function of creating a
sterile barrier system (SBS) that maintains device sterility as the product traverses
distribution (Philchik, 2003).

Many times the expiration date is a direct function of the

stability date available for the seal, rather than the device itself (Philchik, 2003).

As

such, upon expiration, the integrity of a package, and therefore the sterility of the device,
is no longer guaranteed. Expiration dating that can be noticed and understood is
paramount to the safety of the contents within.
Focus group participants reported a lack of standard location, poor contrast and
small font sizes as problematic for expiration dating (Cai, 2012).

The UDI final rule

mandates standardized content as, year, month and day (e.g., 2013-09-30) so as to
ensure that dates are unambiguous and understood by users clearly (FDA, UDI Final
rule, 2013), but placement and textual formatting requirements are not addressed.
1.1.2.3 Sterility status
As with latex status labeling, focus group participants indicated that an absence
of information regarding sterility caused confusion regarding the status of the device
within the package (Cai, 2012).

Reported problems included difficulty in reading
4

sterility status information due to relatively small font sizes and inconsistency in
formatting (Cai, 2012).
1.1.2.4 Summary
The issues of expiration dating and confusion regarding sterility status both have
the potential to contribute to hospital/healthcare acquired infections (HAIs), a noted
problem for healthcare. HAIs are caused by a wide variety of common and usual
bacteria, fungi and viruses during the course of receiving medical care (CDC, 2012b).
Researchers have estimated that 1.7 million of these infections occurred in 2002 and, of
these, 274,098 were Surgical Site Infections (SSI), 16% of the total.

This same group

reported 424,060 HAIs were Urinary Tract Infections (UTIs), approximately 25% of the
total (Klevens, 2007).
1.2 Research goals
The goal for this research was to develop a new labeling system for medical
devices, which facilitates timely and efficacious processing of three of the pieces of
critical information (sterility status, latex status and expiration dating) Cai identified as
crucial to the safe and effective use of devices, but at present, reported as problematic
(Cai, 2012).
It is hypothesized that a standard location and format of information
deemed critical to care will benefit healthcare providers during most stages of
information processing.

To test this hypothesis, we chose to use the labeling of

indwelling urinary catheters as a model product. The perioperative use of indwelling
urinary catheters has become routine practice in orthopedic surgery services (Wells,
2004; Skelly, 1992; Michelson, 2004).

Indwelling catheters were chosen as our model
5

for several reasons, including their price, widespread use, association with Urinary Tract
Infections (UTIs) and because they may or may not contain latex.

Six brands of

indwelling catheters: Bard, Teleflex, Amsino, Dynarex, Covidien and Kendal, were
collected for the benchmarking portion of this research to verify the problematic labeling
conditions reported by Cai (2012).

Findings from the benchmarking study were used

to create labels for mock brands of indwelling catheters to objectively test the effect of
varied design factors (specifically: standard location, the use of a graphic box and
symbol presence or absence, and color-coding) on varied stages of information
processing (e.g. attention, comprehension, and, ultimately, choice).

Further, our

creations were compared with existing systems to objectively evaluate the same.

6

CHAPTER 2
LITERATURE REVIEW
2.1 Medical Device labeling regulations
2.1.1 Misbranding
Misbranding is defined in the Federal Food, Drug and Cosmetic Act (FFDCA) as:
“labeling that is false, misleading or incorrect in some detail.”

Products that are

misbranded cannot legally move through interstate commerce in the US. Section 502
of the FFDCA provides a listing of ways that products regulated by the FDA can be
misbranded (see Appendix 2).

Eleven out of the 23 listed violations relate to the

labeling of a medical device (FDA, Labeling Requirements-Misbranding, 2013).

It is

imperative that medical device manufacturers familiarize themselves with labeling
requirements in order to create products that can be legally distributed in interstate
commerce in the United States.
2.1.2 General labeling requirements
The labeling requirements for medical devices which are sold in USA are defined
in the Code of Federal Regulations (CFR) Title 21. Specifically, part 801 of 21 CFR
describes the general labeling provisions of commercial medical devices as addressed
in Table 2.

7

Table 2. General labeling requirements for medical devices
(CFR Title 21, 801, 2013)
Section
Contents
Name and place of business of
Part 801.1
manufacturer, packer or distributor
Part 801.4
Meaning of intended uses
Part 801.5
Adequate directions for use
Part 801.6
Misleading statements
Part 801. 15
Prominence of required label statements
Spanish-language version of certain
Part 801.16
required statements
If a medical device is being sold in the US, it must contain information regarding
the place where it was manufactured, packed or distributed as well as the name of
business ownership for manufacturing, packing or distribution (CFR Title 21, 801.1,
2013). This information allows medical device users to identify a contact in case of
questions or problems associated with a purchased medical device. Further, if
violation(s) is (are) present, the FDA may contact the business unit shown on the
labeling of a specific medical device for appropriate legal actions.
Part 801.4 addresses the meaning of ‘intended uses’ as follows:
“The words “intended uses” refer to the objective intent of the persons legally
responsible for the labeling of devices. The intent is determined by such
persons’ expressions or may be shown by the circumstances surrounding the
distribution of the article.”
“Intended uses” are communicated to users by legally responsible persons or
representatives for a specific medical device through varied printed information,
including labeling, advertising matter, and written statements.
The directions for use specified on the labeling of medical devices should be
adequate for a layperson to understand safe product use (CFR Title 21, 801.5, 2013).
The required contents for the directions for use are listed in Table 3.
8

#
1
2
3
4
5
6
7

Table 3. Adequate directions for use
(CFR Title 21, 801.5, 2013)
Contents
Statements for all conditions, purposes, or uses for which such device is
intended
Quantity of dose, including usual quantities for each of the uses for which it
is intended
Frequency of administration or application
Duration of administration or application
Time of administration or application
Route or method of administration or application
Preparation for use

To meet the listed requirements in Table 3, the labeling of commercial medical devices
should contain what specific purpose it is intended for, how it is used in a safe manner
along with route or method of administration or application, and preparation for use (if
applicable), as well as specific conditions to be avoided.

Mandatory directions for use

require bold statements and medical device symbols intended for the purpose of
warning or cautioning.

Recommendations for appropriate handling, and the quantity of

dose per use also have to be addressed, as well as how often, long or what specific
time it has to be used.
Misleading statements render a device misbranded (CFR Title 21, 801.6, 2013).
Furthermore, insufficient prominence of legally required information also constitutes a
case of misbranding (Federal Food, Drug, and Cosmetic Act, 2013).

As such, all

wording requirements, statements and other information should be visually prominent or
conspicuous (CFR Title 21, 801.15, 2013). The prominence or conspicuousness of the
required labeling information may fail due to the conditions addressed in Table 4.

9

No.
1
2

3
4
5
6
7

Table 4. Failures of prominence or conspicuousness
(CFR Title 21, 801.15, 2013)
Contents
The failure of the required information to appear on the part or panel of
the label which is presented or displayed under customary conditions of
purchase
The failure of the required information to appear on two or more parts or
panels of the label, each of which has sufficient space therefore, and each
of which is so designed as to render it likely to be, under customary
conditions of purchase, the part or panel displayed
The failure of the label to extend over the area of the container or
package available for such extension, so as to provide sufficient label
space for the prominent spacing of the required information
Insufficient label space for the required information, resulting from any
word, statement, design or device which is not required by or under
authority of the act
Insufficiency of the label space for the required information, resulting from
the use of label space to give materially greater conspicuousness to any
other information not in association with the required information
Smallness or style of type in which the required information appears,
insufficient background contrast, obscuring designs or vignettes, or
crowding with other written, printed, or graphic matter
Insufficiency of the label space for the required information, resulting from
the use of label space for any representation in a foreign language

To sum, it is recommended that medical device manufacturers use sufficient
label space when supplying all required information, considering font size, spacing and
background contrast, as well as appropriate placement so as to avoid insufficient
prominence or conspicuousness

.

If a device is sold in the commonwealth of Puerto Rico, where Spanish is the
predominant language, a label written solely in Spanish is acceptable (CFR Title 21,
801.16, 2013).
2.1.3 Special labeling requirements
CFR Title 21, Subpart H addresses the special requirements of Specific Devices.
10

Among several special requirements in subpart H, relevant to the study presented
herein is the requirement, “User labeling for devices that contain natural rubber” (CFR
Title 21, 801.437, 2013).

It is intended to protect medical device users from any

potential risks associated with natural latex proteins, which may cause anaphylactic
reactions. Natural latex proteins refer to the latex formations listed in Table 5, as well
as the synthetic rubber that contains natural rubber in its formulation.

Type
Natural rubber latex

Dry natural rubber

Table 5. Latex rubber formations
(CFR Title 21, 801.437, 2013)
Definition
Rubber that is produced by the natural rubber latex
process that involves the use of natural latex in a
concentrated colloidal suspension. Products are formed
from natural rubber latex by dipping, extruding, or coating.
Rubber that is produced by the dry natural rubber process
that involves the use of coagulated natural latex in the
form of dried or milled sheets. Products are formed from
dry natural rubber by compression molding, extrusion, or
by converting the sheets into a solution for dipping.

Devices that contain these substances should be labeled prominently in a legible
manner with one of the requisite warning statements in bold print on the principal
display of the device packaging, the outside package, containers or wrapper, and the
immediate device package, container or wrapper (CFR Title 21, 801.437, 2013).
Acceptable warning statements include the following:

• Natural rubber latex: “Caution: This Product Contains Natural Rubber Latex
Which May Cause Allergic Reactions.”

• Dry natural rubber: “This Product Contains Dry Natural Rubber.”
The Quality System Regulation (CFR Title 21, 820) emphasizes the importance
of labeling sterility status of medical devices.
11

If only components of a medical device

are sterilized, appropriate labeling is required to indicate the sterilized parts. For
example, a device that was partially sterilized might be labeled with a statement like:
“Caution: Only the fluid path of the set is sterile and non-pyrogenic. Do not use
in a sterile or aseptic area without proper precautions.” (CFR Title 21, 820, 2013)
If a kit contains some mixed components with regard to sterility status (i.e. some sterile,
others not), it may not be stated (or implied) that all contents are sterile (CFR Title 21,
820, 2013).
Other requisite statements regarding sterility have to do with special conditions.
For instance, in cases where user sterilization is required, or re-sterilization is required
prior to reuse, further information is required on the labeling.

Table 6 describes the

required information for both cases.
Table 6. Labeling information on sterilization
(CFR Title 21, 820, 2013)
Type
Required information
Sterilization by the user Special cleaning methods required
before use
Re-sterilization
Changes in the physical characteristics of the device that
may result from reprocessing which affect its safety,
effectiveness, or performance; and the limited number of
times for resterilization and reuse that can be done
without affecting the safety or effectiveness of the device
In cases where re-sterilization or reuse of a medical device is not appropriate,
manufacturers should include information warning against such behaviors.
In-vitro diagnostic devices (IVD) must include expiration dating information as
part of the labeling because some components of the in-vitro diagnostic may contain a
battery or diagnostic reagent, which has limited use life (CFR Title 21, 809.10).

12

2.2 End-users’ voices on labeling
Neid conducted a survey to gain better understanding of what nurses face, to
get feedback, and, ultimately, to make packaging easier for end users (Butschli, 2008).
The survey included a ranking activity, in which nurses ranked eight different medical
device package considerations on a scale of one to eight, with one being the most
important (Butschli, 2008).

The top two considerations were “easily read text/font

labeling” and “speed of opening package” (see Table 7).

Most votes converged on

those top two considerations, while the last four considerations received only single
votes (Butschli, 2008).
Table 7. Medical device package considerations by nurses
(Bustchli, 2008)
Ranking
Packaging considerations
1
Easily read text/font labeling
2
Speed of opening package
3
Manufacturer’s instructions for use provided via Web
4
Manufacturer’s instructions for use in every package
5
Smallest possible package
6
Color-coded labeling
7
Consistent package sizes
8
Least amount of packaging waste
Cai (2012) conducted seven-focus groups with perioperative personnel
(primarily nurses and surgical technologists) to investigate their needs regarding
medical device packaging.

Qualitative data was converted to quantitative data using a

process called “content analysis” (Neuendorf, 2002).

Focus group discussions were

transcribed and broken into “thought units” which were organized using a coding
scheme and enumerated to build inferences by analyzing the frequency of common
themes.
13

Seven hundred and ninety-five thought units were enumerated in the category
“packaging issues”.

68.4 % of these were sub-categorized as “opening & aseptic

presentation” while “identification” accounted for 22.6% and “packaging waste” 8.6%
(Cai, 2012).

Nurses reported that they did not have enough time to scan and read the

labeling in its entirety due to time pressures.

Focus group participants recommended

“packages that nurses don’t need to read” and the “presence of critical information in a
format that can be quickly identified and read” (Cai, 2012).
Specific to “packages that nurses don’t need to read”, there were several
recommendations by survey participants (Cai, 2012):

• Transparent packaging to allow quick identification of contents
• Diagrams to indicate size and shape (if transparent packaging is not
available)

• Color coding systems (with the caveat request that they be consistent and
universal)

• Different opening features (i.e. certain package structures reserved for use
with sterile devices only)
Most participants reported difficulty in identifying critical information on packages
and that non-critical information interfered with accessibility of critical information (Cai,
2012).

They identified four pieces of critical information: “expiration dating”, “latex

status”, “sterility status” and “product name”. Challenges (and solutions proposed)
relating to identifying and reading critical information quickly are presented in Table 8
(Cai, 2012).
14

Table 8. Challenges associated with the labeling of critical information and
suggested solutions (Cai, 2012)
Solutions
Challenges
Non-critical information gets in the way making it
Get all the wanted information
harder to find the wanted information
together, highlight the critical
information
Expiration
No standard location
Standardize a location for this
dating
information
Light colors
Make it dark and black or bold,
bright color
Small font size
Use bigger font size
Latex status Lack of any information regarding
latex status causes confusion
regarding its presence or absence
Latex-free info not provided
Sterility
The sterility information printed on The outer package should have
information the inner package
the information
for double
Small font size
Use bigger font size
barrier
Wrong highlighting of sterility for
Use circle and slash
unsterile item
Cai (2012) categorized suggestions for improvement into 3 actionable items:
•

Single location and standardized placement
9 To gather the critical information in one location
9 Standardized location for expiration date

• Noticeable text
9 Bolded, bright or color-contrasted expiration date
9 Circled, slashed sterility information (symbol use)
9 Bigger font size for expiration date, sterility information

• Presence of critical information
9 To present latex status information
9 To present sterility status information on all packages (outer and
inner)
15

These actionable items have the potential to be critical factors in the development of
labels which facilitate timely and efficacious processing of the critical information.
2.3 Significance of latex status
2.3.1 Manufacturing process of latex-containing medical devices
The term “latex” is familiar to most healthcare professionals as a potential
catalyst for allergic reactions.

The main source of latex is the sap of commercially

grown rubber trees, Hevea brasilensis (Zaglaniczny, 2001; Kam, 1997; White, 1996).
Ammonia and sulfite are added as chemical preservatives while the sap is extracted
from the rubber trees (Zaglaniczny, 2001; Virant, 1996).

Several additives such as

compounding agents, emulsifiers, stiffeners, etc. are added to improve the rubber’s
structure quality in processing (Zaglaniczny, 2001; Virant, 1996). Items are processed
from rubber into molds for products such as gloves, balloons, and condoms
(Zaglaniczny, 200; Cheng, 2000). The process of making latex-rubber containing gloves
is described in Appendix 3.
2.3.2 Latex allergy types
Latex is used for fabrication of several functional medical devices, many of which
are listed in Table 9. The allergy symptoms from latex-containing medical devices are
caused by a response of the human immune system against foreign proteins (Alwilda et
al., 2003). Hypersensitive responses to latex are classified into Type I immunoglobulin
E (IgE) or Type IV cell-mediated response (Alwida et al., 2003). Type II and III allergy
responses are not associated with latex rubber. Type I latex reactions are initiated by
IgE antibodies which are produced against water-soluble proteins remaining in natural
latex products (Alwida et al., 2003).

These reactions may be generated within minutes
16

after exposure to the latex allergen.

The severe symptoms caused from the type I

latex reactions are temporary, rapid constriction of bronchial smooth muscles, increased
vascular permeability, and dilation of postcapillary venules. Type IV latex reactions
may be triggered by chemicals used as accelerants and antioxidants during the
manufacturing process.

Type IV hypersensitivity reaction can occur 24 to 72 hours

after exposure to the latex allergen (Alwida et al., 2003).

Expected symptoms from

type IV latex reactions are pruritis, erythema, and vesicles or blister at the point of
contact (Alwilda et al., 2003).
Table 9. Medical devices that commonly contain latex
(Adapted from Alwilda et al., 2003)
27. Injection ports
1. Ace bandages
28. Intravenous meditation pumps
2. Adhesive tape
29. Multidose/single-use vial tops
3. Anesthesia masks
30. Nasogastric tubes
4. Bandages
31. Operating room masks, hats, and
5. Bath mats
shoe covers
6. Bite blocks
32. Oral and nasal airways
7. Blood pressure cuffs
33. Orthopedic appliances
8. Bulb syringes
34. Protective sheets
9. Catheters
35. Pulse oximeter
10. Colostomy pouches
36. Reflex hammers
11. Crutch pads
37. Respirators
12. Dental dams
38. Spacers for inhaled medication
13. Dentures
39. Stethoscopes
14. Disposable gloves
40. Stretcher mattresses
15. Electrode pads
41. Suction catheters
16. Endotracheal tubes
42. Surgical gowns and drapes
17. Enema kits
43. Surgical lifts
18. Feminine sanitary pads
44. Surgical masks
19. Fluid-circulating warming blankets
45. Syringes
20. Foam pillows
21. Gastroscopy tubes
46. Tape
47. Tourniquets
22. Goggles
48. Tympanometers
23. Hot water bottles
49. Vascular stockings
24. Identification bands
50. Wheelchair cushions and tires
25. Incontinence pads
51. Wound drains
26. Incubators
17

2.3.3 Risks associated with latex allergy
Risks caused by the latex allergen have been investigated by several
researchers in various medical areas (see Table 10). According to an FDA
announcement regarding latex allergies, more than 1,000 cases were reported between
1988 and 1992, and an additional 500 cases by early 1996 (Dillard, 1992; Kellett, 1997).
It was estimated that 3% to 17% of healthcare workers had allergic reactions as a result
of their exposure to latex (Bowyer, 1998). This percent increased to 24% of healthcare
workers who were atopic (i.e. those with a tendency toward multiple allergic conditions)
(Bowyer, 1998).

Patients who were atopic had higher risks to latex allergy than the

general population (Bowyer, 1998). Children who were atopic or required frequent
surgical interventions were more likely to have a latex allergy (Queiroz, 2009).
Workers involved in the manufacture of latex products were also at a high risk for latex
allergy (Bowyer, 1998).

18

Table 10. Articles addressing risks associated with latex allergy
Dillard et
al. (1992)

-

Between 1988 and 1992, the FDA was informed of 1,133 allergic
reactions that had occurred due to 30 different medical products
made of latex.
408 involved reactions to latex examination gloves and 77 were to
latex surgical gloves.

Kellett
(1997)

-

500 reports of latex allergy and seven more deaths (six associated
with barium enemas and one with latex gloves) were reported
additionally by early 1996.

Bowyer
(1998)

-

Estimates for healthcare workers reveal as few as 3% and up to
17% as having latex allergy.
This increases to 24% of healthcare workers who are atopic
(having a hereditary tendency for immediate Type I allergic
reactions).
Approximately 7.5% of surgeons and 5.5% of theatre nurses have
a latex allergy.
Patients who have pre-existing allergies are more likely to develop
latex allergy. Atopic patients are more prone to latex allergy. 60
to 80% of latex allergic patients are atopic as opposed to 20% of
the general population.
Occupationally exposed people such as those involved in the
manufacture of latex products are at high risk of latex allergy.
Those with occupational exposure have a 2.9-17.0% chance of
latex allergy, whilst research into those working in latex glove
manufacture plants found 11% of workers with latex allergy.
Allergic or immediate hypersensitivity reactions to latex have been
reported in children with increasing frequency in the past.
Children’s subpopulations at particular risk include: atopics,
individuals with spina bifida, children undergoing surgical
procedures during the neonatal period and individuals who
required frequent surgical instrumentations.

-

-

Queiroz
(2009)

-

2.3.4 FDA Countermeasure actions
The FDA has taken several actions to inform the public of latex allergy risks and
to revise related regulations to minimize potential risks.
actions taken by the agency (Farnham et al., 2002):

19

Farnham et al. list five specific

• Medical device reporting/MedWatch data and the emergence of
natural latex allergy: In March 1991, the FDA issued a Medical Alert to
the medical community to inform healthcare workers of the problem of
natural latex allergy, to make recommendations for patient care and
advice, and to request health professionals report adverse reactions to
natural rubber in medical devices.

• User labeling rule for devices containing natural rubber: In September
1998, the rule “User Labeling for Devices That Contains Natural Rubber”
became effective.

This rule requires medical device labeling to disclose

the presence of natural rubber in medical devices and device packaging
when present. There are two types of natural latex rubbers of concern:
Natural Rubber Latex (NRL) and Dry Natural Rubber (DNR). The
following statement is required for medical device/packaging containing
one of those latex rubbers:
For DNR: “This Product Contains Dry Natural Rubber”
For NRL: “Caution: This Product Contains Natural Rubber Latex Which
May Cause Allergic Reactions.”

• Good manufacturing practices for devices containing natural rubber:
In October 1997, the FDA issued the Quality System Regulation final rule.
This rule requires the removal of “manufacturing material” from the
finished product when manufacturing has the potential to affect product
quality.

Water-soluble natural rubber proteins are defined as
20

manufacturing material in 21 CFR Part 820.3(p).

As such, current Good

Manufacturing Practices (CGMP) require device manufacturers to remove
such soluble proteins to the extent possible and to document this removal.

• Standard activities: The FDA has been involved in the creation of
standards regarding medical devices that contain latex. Specifically, FDA
scientists participated in the development of standard test methods for
quantification of Natural Latex proteins.
2.4 Significance of sterility status
2.4.1 Packaging functions and materials
Medical devices that contact a patient’s blood or other internal tissues should be
sterile until their package is opened for medical treatment (Sherman, 1998). As such,
the vital function of a medical device package is to keep the contents sterile. Three
basic elements of package design have been indicated as crucial indicators for ensuring
sterility maintenance (Pilchik, 2003):

• Seal strength: The property to hold the sealed components of the package
together

• Seal integrity: The property associated with the seal being of sufficient
quality to prevent microorganisms from penetrating through the seal

• Package integrity: The property to ensure that the entire package is free
from defects that can allow penetration of microorganisms.

21

Two broad categories of packaging materials are commonly used in the
manufacture of medical devices; selection of the appropriate material(s) is frequently
dictated by the sterilization process which will be used (Pilchik, 2003):

• Porous packaging materials: Tyvek (a polymeric fiber strand distributed in
multiple layers to produce a flat sheet stock) and paper (medical grade) to
allow for gas sterilization methods

• Nonporous packaging materials: Polymeric films (used individually or in
combinations through lamination, co-extrusion, or coating) and foils (used in
combinations with polymeric components to increase the oxygen and water
vapor resistance) to allow for other sterilization methods
2.4.2 Sterilization methods
The sterilization of a medical device is defined as the process by which
anticipated levels of microbial contaminants in a load of items are exposed to a specific
number of decimal reduction values (D-values, time or dose to kill 90% of the organisms
at a given set of conditions) for the sterilant being utilized (Sherman, 1998). The
probability of a survivor per item (PSI) is generally less than 10-3 for topical products,
and less than 10-6 for implantable or blood-contacting items (Sherman, 1998).

When

selecting the appropriate sterilization method, design factors to be considered include:
product materials, product design, packaging, marketing requirements, current
manufacturing and sterilization capabilities, and process economics (Sherman, 1998).
The available methods of sterilization for packaged medical devices are (Sherman,
1998):

22

• Steam under pressure: Saturated steam under pressure is the most
practical and dependable agent for sterilization of heat-tolerant medical
supplies and packaging.

• Dry heat sterilization: Dry heat is transferred by means of convection and
conduction to sterilize medical items.

This method requires longer

sterilization and higher temperatures than does moist heat.

• Gaseous sterilization: Ethylene oxide (ETO) and propylene oxide (PO) are
generally used for gaseous sterilization. ETO is the most commonly used
gaseous sterilant for sterilization of medical items and instrumentation.

• Ionizing radiation sterilization: Absorption of high-energy radiation (gamma
and electron beam radiation) by organic matter causes chemical changes in
the material. Unlike the ETO sterilization, ionizing radiation does not impart
toxicity to plastic materials, but may change their color and stability.

• Gas plasma sterilization: “Plasma” is an ionized, or partially ionized gas
which contacts the surface of devices for sterilization. This method was
developed to reduce and/or to eliminate the dependence on ETO as a
sterilant for moisture and heat labile items.

The commercially available

plasma sterilization systems are the Plazlyte system and the Sterrad system.
2.4.3 Healthcare-associated infections (HAIs)
Healthcare-associated infections (i.e. an infection that a patient acquires during
the course of receiving treatment for other conditions in a health care setting) are
caused by a wide variety of common and usual bacteria, fungi and viruses during the
23

course of receiving medical care (CDC, 2012b).
Researchers have reported that 38,785 of these infections occurred in 2011
(see Table 11). The numbers of infections noted were 18,113 for “Central-line
associated bloodstream infection, 14,315 for “Catheter-associated urinary tract
infection”, and 6,357 for “Surgical site infections”, and “Central-line associated
bloodstream infections” comprised 47%, “Urinary Tract Infections”, 37%, and “Surgical
site infections”, 16% (CDC, 2012a).
Table 11. Healthcare-associated infections in U.S. hospitals during 2011
(CDC, 2012a)
Type of
Category
# of infections
Percent (%)
infection
Intensive Care
10,134
Central-line
Units(ICUs)
associated
bloodstream
Wards
5,781
infections
Non-intensive Care
2,198
(CLABSI)
Units (NICUs)
Sub-total
18,113
47%
Intensive Care Units
Catheter(ICUs)
associated
urinary tract
Wards
infections
Sub-total
(CAUTI)
Surgical site
Combined SCIP
infections (SSI)
procedures
HAIs
Total
(Observed)

8,925
5,390
14,315

37%

6,357

16%

38,785

100%

The estimate of average attributable costs ($ base year) per patient in “Centralline associated bloodstream infections”,

“Urinary Tract Infections (UTI)” and “Surgical

Site Infections (SSI)” were $29,116, $1,007 and $34,670, respectively, in 2007 (CDC,
2009).

24

2.4.4 Surgical microbial contamination
Sources of surgical microbial contamination may be either resident flora (i.e.
endogenous microorganisms) or transient flora (i.e. exogenous microorganisms)
(Hopper, et al., 2010).

Resident flora are bacteria or microorganisms considered to be

permanent residents of the skin and are not readily removed by hand washing (AORN,
2010a; Hopper et al., 2010).

Transient flora are bacteria and microorganisms that

colonize the superficial layers of the skin and are easily removed by hand washing or
use of a hand rub agent, and they are easily transmitted from patients and inanimate
surfaces to other locations (AORN, 2010a; Hopper et. al, 2010).

One potential cause

of exogenous microbial contamination is a break in sterile technique (Hopper et al.,
2010).

Sterile technique is defined by the Association of peri-Operative Registered

Nurses (AORN) as “methods by which contamination by microorganisms is prevented.”  
Aseptic transfer, or transfer of the device to the sterile field without contaminating it, is
paramount.

To ensure asepsis, the following appropriate, preventive actions are

recommended by the Association of perioperative nurses (AORN, 2010b):

• A properly designed Sterile Barrier System (facilitates sterilization, maintains
sterility throughout distribution, assists in verification of sterility maintenance
and enables aseptic transfer)

• Correct sterilization processing
• Maintenance of seals throughout the distribution process
• Verification of the sterile barrier system’s integrity by personnel, and
• Aseptic transfer to the sterile field.
25

In addition, the Association of Surgical Technologists (AST) has recommended
their members use specific standards of practice for creating the sterile field.

The

following key actions are recommended to maintain aseptic technique in opening sterile
device packages (AST Standards, 2011):
•

Placement of sterile items on clean, dry surfaces

•

Verification of external chemical indicator or integrator, integrity of packaging,
and expiration date prior to opening

•

Establishment of an appropriate routine for opening sterile items
o Sequence of opening sterile items: backtable pack, basin set, small
wrapped items (e.g. sterile towel pack) and peel pack items
o Opening of gown and gloves on a separate flat surface
o Flipping small wrapped items, peel packs and suture packs onto the
sterile field using aseptic technique
o Not allowed to flip heavy or difficult items onto the sterile field
o Opening sterile items in a grouping manner for establishment of a
logical, sequential, and efficient routine (e.g. sharp items, drapes on
each designated area of the back table on the sterile field, etc.)

2.5 Significance of expiration dating
The shelf life of medical devices is determined by multiple factors, including:
bio-burden (both in the air and on the surfaces of sterile packaged products and
packaging materials), seal strength, distribution stresses, airflows, personnel traffic
patterns, storage location, temperature, pressure, humidity, and bio-barrier properties of

26

packaging materials (1998, Sherman). Generally, the expiration date is a direct
function of the stability date available for the seal, rather than the device (Pilchik, 2003).
Most medical devices have an expiration date as part of their labeling.

The UDI

final rule (2013) requires standardized content format, in the form of year, month and
day (e.g. 2013-09-30) so as to ensure that dates are unambiguous and understood by
users clearly (FDA, UDI Final rule, 2013). But, multiple labeling problems regarding
expiration dating were identified from Cai’s research (2012): lack of a standardized
location, the use of poor contrast, and small font size.

These problems are likely to

cause equivocal expiration dating of medical devices, and have the potential to result in
increasing the number of devices used beyond a point where sterility is guaranteed.
2.6 Information Processing
The significance of these three pieces of critical information (latex presence,
sterility status and expiration dating) has clear ramifications for health.

As such, the

clear communication of this information is paramount at the point of use.

Thus, a

review of one theoretical frame to assess information processing is germane to the
development of useful labeling systems.
Commonly cited models of information processing (Rousseau, 1998; Dejoy, 1991)
suggest that for information to be effective, five steps of interaction must occur between
the message recipient (in this case, a healthcare provider) and the message.

These

are:
•

Step 1: Exposure (absence of needed information can be problematic)

•

Step 2: Perception (the user must take the message in through one of the
five senses)
27

•

Step 3: Encodation (the external signal from the environment must be
converted into an internal one that can be processed by the cognitive
system)

•

Step 4: Comprehension (messages that are beyond the reading level of
the individual or symbols that are confusing are problematic)

•

Step 5: Action (the physical systems perform the desired and appropriate
action)

Success or failure at each processing step is directly influenced by four broad inputs
(the user, i.e. message recipient; the context of interaction; the task to be accomplished;
and the design of the product/package system) (de la Fuete, 2013; see Figure 1).
Specifically, “Context” refers to the environment of interaction; “Package” means a
physical object to contain the product; “User” relates to the individual interacting with the
package.

As a given task is accomplished (e.g. selection of the appropriate product),

the user goes through the 5 steps of the information processing model to take an action.
When the task is accomplished, the state of things is altered, and the next task may
begin, and the information processing sequence begins anew.

28

Figure
F
1. Diagram
D
of Information
I
n Processing Model

2.7 Labe
eling Notic
ceability
Early
E
stage
es of informa
ation proce
essing (expo
osure, percception) invvolve attentiion.
In recen
nt years, tec
chniques co
ommonly us
sed to mea
asure the atttentive beh
haviors of
people have
h
been applied
a
to labeling and
d packaging
g.
2.7.1
2
Change Detection
Change
C
dete
ection is a technique
t
that is frequ
uently used for evaluatting the
attention
nal prioritiza
ation people
e give varying compon
nents within
n a scene ((Goldstein,
2007). It is comm
monly referre
ed to as a “flicker
“
taskk”. During each flicke
er trial, a control
240ms) con
ntinuously alternates
a
with
w the testt image (24
40ms) (the ccontrol onlyy
image (2
slightly altered)
a
with
h a brief, grray screen interleaving
g between the two (80
0ms) as sho
own
29

in Figure 2.
(1997).

The techniques and its timings were originally developed by Rensink et al.

This sequence image-blank-test-blank loops until the participant presses the

space bar, indicating that they have found the alternation, or until they time out.

30

First screen: Control image appears for 240ms.

Second screen: Gray image appears for 80ms.

Third screen: Test image appears for 240ms.

Changed part
Fourth screen: Gray image for 80ms (Returning back to the control image if a subject
does not press the computer’s space bar).

Figure 2. Change Detection Image Cycle

31

The change detection methodology has been applied on a limited basis to
objectively evaluate the visual salience of varied elements of labeling and packaging
(Table 12).
Gaschler et al. (2009) used change detection to study how individuals attend to
and process newly introduced formats for food labels. Tested elements included text
and graphic formats of four pieces of nutrition information (fat content, “best before”
date, recycling information and organic status).

Changes to “organic” product

information were detected significantly faster than other information on the food labels,
while changes in the “fat content” information were detected significantly slower than all
other types of product information (p < 0.01). Additionally, there was a correlation in
change detection time and age for “organic” and “recycling” information; older subjects
detected changes significantly slower than young subjects (p < 0.01).
DeHenau (2010) evaluated the effect of TALL-Man lettering in differentiation of
look-alike, sound-alike drug names.
names in uppercase letters.

TALL-Man lettering is a practice of writing drug

Eight pair images of drug names were presented to

subjects in two formats (TALL-Man vs. traditional).

Participants were able to decipher

a doppelganger faster when the changes occurred in TALL-Man pairs when compared
with the same in traditional lettering (p<0.0001).

The effect was particularly

pronounced for nurses.
Bix et al. (2010) evaluated the prominence of different label elements on a
beverage container. Six portions of labels were evaluated: the manufacturer name, the
product name, text within a warning dot in three colors, and presence/absence of the
warning dot.

The time required to detect changes to the manufacturer’s name was
32

significantly longer than for any of the other label elements (p<0.0001).

Changes to

the warning dot with red text were located marginally faster than the warning printed in
black (p=0.0566).

In addition, an effect of the location of the change was noted,

suggesting that subjects tended toward a standard scanning pattern across the stimulus.
Sundar (2013) evaluated the effect of design, specifically color and facial icons,
on participant’s attention to nutrition information.

Front Of Pack (FOP) nutrition labels

of varied design were compared with nutritional information conveyed through traditional
labeling (in the form of Nutrition Facts Panels, or NFPs).

Changes to the FOPs were

more likely to be successfully detected than those to the NFPs (Nutrition Fact Panels)
(p<0.0001); when detected successfully, researchers noted a significant effect of color
(p<0.0001).

33

Table 12. Change detection studies relating to packaging or labels
Title
Change detection
for new food labels
(Gaschler et al.,
2009)

Objective/Stimuli
Objective: To investigate the person-variables (e.g. age of
participants) associated with lower change detection latencies
for specific food-related information/To evaluate whether the
change detection task is useful for studying how individuals
attend to and process formats and contents of food labels
Stimuli: A label with the fat content, best-before date, recycling,
and organic information/ A design in which the content (general,
organic and health) of the product information was presented in
text and graphic format
Applying change Objective: To qualify TALL Man lettering as a method to
detection to test
differentiate look-alike sound-alike drug names/To evaluate the
the noticeability of change flicker method so that it can possibly be used in future
components of
labeling studies
medical labels
Stimuli: Eight pairs of look-alike, sound-alike names in two
(DeHenau, 2010) formats (TALL Man vs. traditional)
The use of change Objective: To develop change detection software and
detection as a
methodology for label use/To compare the relative prominence
method of
of different label elements on a beverage container
objectively
Stimuli: A beverage container to have 6 label elements: the
evaluating labels manufacturer name, the product name, and a warning dot with
(Bix et al., 2010) text in three colors
Investigating the Objective: To evaluate the effect of color and facial icons on the
effect of color and ability of a nutrition FOP to attract attention
icon on information Stimuli: 3 factorial designs (Color vs. No color, Text vs. Facial
processing
Icon, and Healthy vs. Unhealthy) with 3 brands of cereal/24 FOP
behaviors related (Front Of Pack) trials and 24 NFP (Nutrition Facts Panel) trials
to Front-Ofregarding 4 nutrients: ‘FAT’, ‘SATFAT’, ‘SUGARS’ and ‘SALT’
Package nutrition
labels (Sundar,
2013)
2.7.2 Eye tracking
Fixations and saccades are the two main components of eye movements
(Buswell, 1935).

Fixations describe the status of the still eyes at a certain point of a

stimulus, lasting 200-500 milliseconds, while saccades are quick eye movements,
lasting 20-40 milliseconds (Rayner, 1998). The pattern of fixations and saccades are
called a scan path (Noton & Stark, 1971). Usually, eye trackers record this pattern of
34

fixation and saccades (Wedel, 2008).

Most commercial eye trackers use an infrared

corneal reflection methodology to measure the distance and angle of the reflection of
infrared light from the center of the pupil to determine the point of fixation of the person
(Young and Sheena, 1975).
In recent years, limited academic research has evaluated early stage
information processing with packaging and labeling as the stimulus material (Table 13).
Bix et al. (2009) evaluated the prominence of warnings (one which indicated the
lack of a child resistant feature and another alerting consumers to the presence of
tamper evident features) on OTC pain relievers using a head-mounted optics ASL eye
tracker. Other label elements examined were brand name, claim statement and drug
facts for comparison purposes.

Research participants spent less time on the zone of a

warning indicating that the packaging had no child resistant features than other label
elements (p< 0.05). Results of a free recall test subsequent to the eye tracking study
suggested further that subjects recalled regulatory information (e.g. warnings for
“Alcohol”, “CR (Child Resistant)”, “Child statement”, and “TE (Tamper Evidence)”)
significantly less frequently than marketing information (p< 0.05). Specifically, the most
frequently recalled elements were brand name, indications and package color.

Text

legibility was evaluated using a Lockhart legibility instrument. Rotation of the
instrument’s handle correlated with rotation of the first of a pair of polarizing filters in
series.

Greater angles of rotation related to more light: as such, items that were more

difficult to read required a greater degree of filter rotation than those that were readily
deciphered.

The child resistant warning and a tamper evident warning, both of which

are required by law to be prominent or conspicuous, required significantly more rotation
35

than any of the other elements tested, suggesting that these two warnings were
relatively less legible than the others tested (p < 0.05).
Oh (2010) measured the relative prominence of the traditional format versus an
altered format for nutrition labels, using a Pan Tilt ASL eye tracker. The Nutrition Facts
Panels (NFPs) of nine cereals were presented with two types of format: iconic face
versus text only for three nutrients (sodium, sugar and fat).

The iconic face format

showed improved prominence in all three dependent variables: total time in zone (p <
0.0001), probability of noticing in zone (p< 0.0001) and number of hits to the zone (p<
0.0001).
Graham et al. (2011) evaluated the visual attention of research participants to
the Nutrition Facts label under a simulated grocery shopping exercise, using an Eye
Link 1000 eye tracker. Sixty-four foods were presented for purchase decisions: “would
buy”, “would not buy”, or “not applicable”.

Each individual food included three images

on a computer screen: the food’s price and description, a photograph of the food and an
ingredient list, and a Nutrition Facts Label. Researchers indicated that the label
components located on the top of the label were viewed more frequently than those on
its bottom (p<0.05), and labels located in the center of the computer screen were
viewed more frequently than those on its sides (p<0.05).
Herpen et al. (2011) evaluated the effect of Front-of-Pack nutrition labels of
cereal products, using a remote eye tracker manufactured by SMI.

There were 6

cereal boxes, composed of three different nutrition labeling schemes: logo, multiple
traffic-light (MTL) label, and nutrition table.

It was reported that participants were less

likely to attend to nutrition tables than to logos (p<0.01) and attention to logos was
36

marginally higher than attention to MTL labels (p=0.067).

However, the total amount of

time spent on labels was longer for the nutrition table than the logo (p<0.05), and the
average dwell time on the logo was lower than the average dwell time on either the MTL
label (p<0.001) or the nutrition table (p<0.01).
Hurley et al. (2012) evaluated the effect of an amount of product visible through
the primary display panel on consumer attention and purchase decision in the category
of grill ware.

There were 4 levels of visible product exposure tested: 0%, 40%, 90%

and 100%; packages were positioned on the shelves of a fully immersive simulated
shopping environment while eye movements were tracked using a Tobii glasses eye
tracking device.

It was found that the packaging with the greatest product exposure

was chosen more than the other packaging configurations and the 0% visible product
received significantly fewer fixations, a slower time to first fixation and lower total
fixation durations than the other 3 configurations (p’s<0.01).
Hurley et al. (2013) also conducted research to determine whether consumers
preferred a public label product versus a private label product in terms of product
purchase and visual attention, using a mobile eye tracking system manufactured by
Tobii.

Researchers suggested that the public label product was preferred in purchase

decisions and received more visual attention (fixation time) than the private label
product (p<0.05).
Bix et al. (2013) evaluated the effect of color contrast on participants’ attentive
behaviors and perceptions of fresh produce, using an ASL Pan Tilt eye tracker.    Six
different types of produce (red apples, oranges, lemons, green apples, purple onions
and white onions) were photographed with four different colored mesh bags, resulting in
37

four contrast treatments: the same (as the produce), complementary, complementaryanalogous and analogous.

Researchers concluded that the same color or analogous

color treatment inspired more visual attention as measured by the number of visual
fixations (p<0.001) and fixation time (p<0.001) than complementary or complementaryanalogous treatment. A post-hoc survey of purchase intention, quality and visual
appeal as measured with a Likert scale was done, and odds ratio estimates were
conducted for statistical analysis. The same color or analogous color treatment was
perceived by participants as higher quality, more visually appealing and garnered a
higher level of purchase intention than other color treatments (α=0.05).
Sundar (2013) evaluated the effect of Front-Of-Pack (FOP) nutrition labeling on
the attentive behaviors of participants viewing breakfast cereals and crackers, using a
head-mounted ASL eye tracker. There were two label types evaluated: Nutrition Facts
Panel (NFP) only and FOP +NFP.

It was reported that there was evidence of a

significant effect of label type (NFP only or FOP + NFP) for analyzed response variables:
probability of noticing nutrition information on the package (p<0.0013), time to first
fixation of nutrition information (p<0.0001) and total time spent on nutrition information
(p=0.0032).

These findings suggested that when the FOP was present, participants

were more likely to fixate the nutritional information, hit nutritional information faster, and
spend longer time viewing the information.
Gomes et al. (2013) evaluated the effect of shelf presence of full body graphic
labels versus partial body graphic labels on plastic beverage bottles, using a mobile eye
tracking system manufactured by Tobii.

There were 12 beverage bottles, composed of

six different flavors and two different labels for each flavor.
38

It was reported that there

was evidence of a significant effect of label type (full body vs. partial body) for analyzed
response variables: the total number of fixations (p<0.05), and visit count (p<0.01).
These findings suggested that the partial body labels had more within-AOI (Area-OfInterest) fixations than the full body labels, and participants returned to looking at partial
labels more often than full body labels.
Table 13. Eye tracking studies relating to packaging or labels
Title
Examining the
conspicuousness and
prominence of two
required warnings on
OTC pain relievers
(Bix et al., 2009)
Measuring the relative
prominence of graphic
symbols vs. text for
nutrition labels using eye
tracking
(Oh, 2010)
Eye-Tracking evidence
that consumers
preferentially view
prominently positioned
nutrition information
(Graham, Robert, &
Jeffery, 2011)
Front-of-pack nutrition
labels. Their effect on
attention and choices
when consumers have
varying goals and time
constraints (Herpen et al.,
2011)

Objective/Stimuli
Objective: To evaluate the prominence of warnings on
child-resistant and potential product tampering (Time
spent viewing the warnings compared with other areas of
the label, recall and legibility ability)
Stimuli: 4 kinds of OTC pain relievers to include Brand
name, claim statement, child resistant and tamperevidence warnings
Objective: To examine the attentive behaviors of
subjects when viewing 9 cereals with nutrition information
presented in the traditional format and an altered format.
Stimuli: Nutrition Facts Panels(NFPs) of Cereals: Text
only (commercially available) NFPs vs. Icons inserted
NFPs
Objective: To track the visual attention of individuals
making simulated food-purchasing decisions to assess
Nutrition Facts label viewing
Stimuli: Three images on a computer screen: food’s
price and description, a photograph of the food and
ingredient list, and Nutrition Facts Label
Objective: To examine consumer attention to and use of
three different nutrition labeling schemes (logo, multiple
traffic-light label, and nutrition table)
Stimuli: 6 cereal boxes with three labeling schemes

39

The effect of modifying
structure to display
product versus graphical
representation on
packaging
(Hurley, Galvarino,
Thackston, Ouzts, &
Pham, 2012)
Effects of private and
public label branding on
consumer purchase
patterns (Hurley, Ouzts,
Fischer, & Gomes, 2013)
The effect of color
contrast on consumers’
attentive behaviors and
perceptions of Fresh
Produce
(Bix, Seo, & Sundar,
2013)
Investigating the effect of
color and icon on
information processing
behaviors related to
Front-Of-Package
nutrition labels (Sundar,
2013)
The effect of full body
versus Partial body
graphic labelling on
beverage packaging
(Gomes et al., 2014)

Table 13. (Cont’d)
Objective: To investigate whether the amount of physical
product visible from the primary display panel of a
package has an effect on consumer attention and
purchase decision in the category of grill ware
Stimuli: Three similar products with four distinct package
structures varying the amount of visible product exposure
(0%, 40%, 90% and 100%)
Objective: To determine whether consumers prefer a
public label product versus a private label product
Stimuli: 2 boxes of cookies: public brand product
(General Mills Cookies Crisp) versus private brand
product (Southern Home Kookies)
Objective: To identify the impact of simultaneous color
contrast (i.e. the produce viewed through a mesh bag) on
attentive behaviors as measured by eye tracking and
perceived quality, visual appeal and purchase intention as
measured with a Likert scale
Stimuli: Six different types of produce (red apples,
oranges, lemons, green apples, purple onions and white
onions) were photographed with four differently colored
mesh treatments
Objective: To evaluate whether an FOP (Front-Of-Pack)
encourages attention to the more detailed NFP (Nutrition
Fact Panel), or acts as an informational short-cut, thereby
reducing attention to the traditional NFP.
Stimuli: 8 packages with 3 factorial design: FOP vs. No
FOP, Healthy icons vs. Unhealthy icons, and Breakfast
cereal vs. Crackers
Objective: To evaluate the shelf presence of full body
graphic labels versus partial body graphic labels on
plastic beverage bottles
Stimuli: 12 beverage bottles, composed of six different
flavors and two different labels for each flavor

2.8 Text Legibility
After labeling information is exposed to and noticed by users, next interactions
between a user and a package are perception and encodation (see Figure 1). During

40

these interactions, the proper text legibility of labeling information on the package is
crucial for users to maintain successful information processing.
A label consists of textual and design elements that are used to communicate
information to users, and much of this information is required by regulators.

Text

designs of the label can be varied to improve its legibility.
“Legibility is the overall goal in a complex system of interrelated elements (letter
weight, letter compression, counter form shape, stress, type style, type size,
message layout, leading, kerning, ink, substrate, and printing process) that come
together to create a message” (Bix, 2001).
Textual elements include font size (e.g. x-height), shape of ascenders and descenders,
typeface design, counter form shapes, line spacing, color contrast, the use of serifs or
san-serifs, stroke weight, kerning and leading (Bix, 2001).

Terminology definitions

relating to letter elements are described in Table 14.
Table 14. Terminology definition of letter elements on a label
Terms
Definition
Font
x-height
Ascenders
Descenders
Typeface
Family
Counter forms
Serifs/San-serifs
Stroke weight
Kerning
Leading

One size of one particular typeface
The height of the body of the lowercase letter
Any portion of the letter extending above the x-height
Any portion of the letter falling below the x-height
Full range of type of the same design
All the typefaces of the related designs
Negative spaces within letters
Serifs: Terminal strokes that are short cross lines at the end of
main stroke
San-serifs: Serifs are not present.
Variations in stroke thickness
Negative spacing between letters
Amount of space between lines of type/Distance between two
baselines of letters

41

x-height is the height of the body of the lowercase letter, and it is a better
indicator to show legibility of the letter because the font size varies, depending on the
typeface (Craig, 1980).
“The size of a given font is based on the now-antiquated system of setting metal
type. Metal type setting was the system used when letterpress, a type of relief
printing, was the only way to print text. In letter press printing, each letter is
raised from the surface of a metal block. The block is referred to as the body:
the printing surface (the letter) is referred to as the face. Type size is based on
the size of the block from which the letter is carved and is not directly related to
the height of the letter” (Craig, 1980).
Serifs (presence of terminal strokes) have pros and cons in legibility. They
contribute to improvement of legibility by combining separate letters into word-wholes
horizontally (Perles, 1977); however, some researchers suggest that their strokes can
cause visual distractions (Garcia, 1981).
Letter spacing is considered as an important design element in legibility.
Kerning (negative spacing between letters) may reduce legibility dramatically (Pettit,
2000).

And, the optimal amount of leading (distance between two baselines of letters)

depends on the letter elements as well as the message design (Becker et al., 1970).
Another important factor affecting legibility is the color contrast between the
letter and its background.

It has been suggested that messages printed in black on

white result in better legibility than other color contrasts (Sorg, 1985; Sundar, 2009).
Bix (1998) conducted research to evaluate the effect of age on legibility readings
as measured with the Lockhart Legibility Instrument (LLI). In the Lockhart Legibility
Instrument, Greater angles of rotation of the first of a pair of polarizing filters in series
related to more light.

That is, items that were more difficult to read required a greater

42

degree of filter rotation than those that were readily deciphered. There were 4 age
groups (21-35, 36-50, 51-65 and 66-80) in this study. It was found that there was
significant difference in legibility indices among age groups; specifically the older age
group had a higher legibility index requiring a great degree of rotation of a polarizing
filter in series than their younger counterparts.
2.9 Symbol Comprehension
Along with texts being clearly legible, symbol comprehension is an important
factor in communication of required labeling information to users as intended, for safe,
effective use of medical devices.
Medical device symbols are used as an effective way to convey information in
different languages because of their benefits such as high visual impact, less space
occupation and information independent of language (Davies et al., 1998; Wolff &
Wogalter, 1998; Perry, 2003; Liu et al., 2005).

Notwithstanding these advantages,

there are potential risks of medical errors if those symbols are not interpreted correctly.
2.9.1 Global standards
The international standard for medical device symbols is ISO 15223-1 2007:
“Medical devices – Symbols to be used with medical device labels, labeling and
information to be supplied.”

The standard is comprised of recognized symbols which

convey information considered by regulatory authorities to be essential for the safe and
proper use of medical devices.

These symbols, which have precise meanings defined

by the standard, are intended to reduce confusion and delays that can result from
labeling in multiple languages.
43

Symbol use is widespread in the European Union, largely in response to the EU
Commission’s 1993 Medical Directive, which requires that text must be presented in
multiple languages in order to be accessible to providers in multiple countries (EU
Council Directive 93/42/EEC, 1993). Recognizing challenges related to label space,
the EU Directive also indicates “where appropriate this information should take the form
of symbols,” and that symbols should conform to harmonized standards (EU Council
Directive 93/42/EEC, 1993).

When harmonized standards do not exist, symbols must

be described in accompanying documentation.

The harmonized symbol standard

recognized by the European Union on medical device symbols is EN 980 2008:
“Symbols for use in the labeling of medical devices”.

Its main purposes are to reduce

the need for multiple translations of words into national languages, to simplify labeling
wherever possible, and to prevent divergent symbols intended to convey the same
information in Europe.
With a few exceptions (e.g. FDA, Use of symbols on labels and in labeling in
Vitro Diagnostic devices intended for professional use, 2004), US regulations indicate
that graphics, pictures or symbols that represent required information on medical
devices must be accompanied by explanatory English text adjacent to the symbol.
However, FDA published a proposed rule involving the use of standardized symbols for
medical devices. If enacted, the rule would: (1) allow for the inclusion of stand-alone
graphical representation of information or symbols, provided they are part of a standard
developed by a nationally or internationally recognized standards development
organization (SDO) and accompanied by a symbols glossary, and (2) authorize the use
of the symbols statement “Rx only” on the labeling of prescription (FDA, Use of certain
44

symbols, Proposed rule, 2013). The harmonized symbol standard recognized by the
US is an American National Standard, ANSI/AAMI/ISO 15223-1: 2007: “Medical
devices-Symbols to be used with medical device labels, labeling and information to be
supplied on medical device symbols.” The primary intentions of this proposed rule are
to make medical device labeling more user-friendly by replacing small, difficult-to-read
text with pictorial information and to harmonize the labeling information of US and
foreign regulatory bodies (e.g. European Commission).
ISO 15223 contains 31 basic symbols for medical devices (Table 15). Since
ANSI/AAMI/ISO 15223: 2007 adopted ISO 15223, its symbols are the same as those of
ISO 15223 in terms of symbol image and symbol meaning.
added in Amendment 1: 2008 for ANSI/AAMI/ISO 15223.

But, 7 symbols were
The added symbols are

“Sampling site”, “Fluid path”, “Non-pyrogenic”, “Contains or presence of natural rubber
latex”, “Drops per milliliter”, “Liquid filter with pore size” and “One-way valve”.
EN 980 has 32 symbols for medical devices (Table 15). The EN 980 symbols
not included in the ISO 15223 document are: “Manufacturer”, “Authorized representative
in the European community”, “Sufficient for”, “For IVD performance evaluation only”,
“Contains or presence of natural rubber latex”, and “Sterile fluid path”.

The five

symbols which are included in ISO 15223 not present in EN 980 are: “Fragile, handle
with care”, “Protect from heat and radioactive sources”, “Patient number”, “Humidity
limitation” and “Atmosphere pressure limitation”.
The FDA guidance (Use of symbols on labels and in labeling of In Vitro
Diagnostic Devices intended for Professional use, 2004) contains fewer symbols (see
Table 15) than ISO 15223. The symbols which are not included in the FDA guidance
45

are “Fragile, Handle with care”, “Keep away from sunlight”, “Protect from heat and
radioactive sources”, “Keep away from rain”, “Do not resterilize”, “ Non-sterile”, “Do not
use if package is damaged”, “Patient number”, “Humidity limitation”, and “Atmosphere
limitation”.
Table 15. Symbols comparison among international and US standards
No

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

Symbol Description

ISO
15223

EN 980

ANSI/AAMI/
ISO 15223

FDA
Guidance

√
√
√
√

√
√
√
√

√
√
√
√

√
√
√
√

√
√
√

√

√
√
√

√
√
√
√
√
√
√
√
√
√
√
√
√
√

√
√
√
√
√
√
√
√
√
√
√
√
√
√

√
√
√
√
√
√
√
√
√
√
√
√
√
√

√
√
√
√
√
√
√
√

√
√
√
√
√

√
√
√
√
√
√
√
√

Biological risks
Do not re-use
Consult instructions for use
Caution, consult accompanying
documents
Fragile, handle with care
Keep away from sunlight
Protect from heat and radioactive
sources
Keep away from rain/Keep dry
Lower limit of temperature
Upper limit of temperature
Temperature limitation
Use by date
Date of manufacture
Batch code
Catalog number
Serial number
Control
Negative control
Positive control
Sterile
Sterilized using aseptic processing
techniques
Sterilized using ethylene oxide
Sterilized using irradiation
Sterilized using steam or dry heat
Do not resterilize
Non-sterile
Do not use if package is damaged
In Vitro Diagnostic medical device
Patient number
46

√

√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√

√

30
31
32
33
34
35
36
37
38
39
40
41
42
43

Table 15. (Cont’d)
Humidity limitation
√
Atmosphere pressure limitation
√
Manufacturer

√
√
√

√

Authorized representative in the
European community
Sufficient for/Contains sufficient for
<n> tests

√

√

√

√

For IVD performance evaluation
only
Contains or presence of natural
rubber latex
Do not use if package is damaged
Sterile fluid path
Sampling site
Non-pyrogenic
Drops per milliliter
Liquid filter with pore size
One-way valve
Total # of symbols

√

√

√
√
√

31

32

√
√ 
√ 
√ 
√ 
√
√
38

25

2.9.2 Comprehension evaluations
ISO 9186-1: 2007 is recognized as the international standard to assess the
comprehensibility of graphical symbols. During the comprehension test, each
respondent is presented with the question: “What do you think this symbol means?”
Responses are categorized as one of the five standard categories: 1, 2a, 2b, 3 or 4
(specified in ISO 9186-2007):
•

1: Correct

•

2a: Wrong, 2b: Wrong and the response given is the opposite of the
intended meaning

•

3: The response given is “Don’t know”

•

4: No response is given
47

The responses which belong to “category 1” are considered as a correct answer; the
other responses are considered incorrect.
Liu et al. (2004) evaluated 16 symbols used in Intensive Care Units (ICU) for
comprehension, using methods prescribed by ISO 9186.

Twenty healthcare workers in

Germany and 13 healthcare workers in China participated in this study.

Of the 16

symbols evaluated, only of half symbols in Germany and four symbols in China reached
the 67% criterion (specified by ISO 3864), and only 3 out of 16 symbols would be
accepted if the 85% criterion (specified by ANSI Z535.3) was considered.
Hermans et al. (2011) evaluated the comprehensibility of 18 symbols used for In
Vitro Diagnostic Devices (especially, Rapid Diagnostic Test (RDT) kits) in accordance
with methods described in ISO 9186.

Two conditions were tested: stand-alone

symbols and symbols presented in context (i.e. a color photograph of a malaria RDT kit
package).

Study participants were health care workers from four international settings

(Belgium, Cambodia, Cuba and Congo). The comprehension level of the participants
was not satisfactory for most of the tested symbols, based on the 67% criterion
(specified by ISO 3864).

The symbols which received fewer than 10% correct

responses were: “Do not reuse”, “In vitro diagnostic medical device”, “Sufficient for”,
“Date of manufacture”, “Authorized representative in EC”, and “Do not use if package is
damaged”.
Our review of the literature suggests a scarcity of work investigating
comprehension of medical device symbols, with all identified publications coming from
outside of the US; the limited research consistently finds poor comprehension rates
related to medical device symbols.
48

2.10 Forced Choice Task
As described previously in the information processing model (Figure 1), a
sequence of 5 interactions (i.e. exposure, perception, encodation, comprehension and
action) between a user and a package continues until the user takes an action.

Thus,

it is interesting to assess how the user comprehends labeling information through these
interaction steps on varied designs of labeling.
A forced choice task is a commonly used tool to evaluate psychological
concepts such as perception, recognition or decision making (McKenzie et al., 2001).
This method has been applied to labeling-related research on a limited basis.
Filik et al. (2006) used a forced choice task methodology to investigate the effect
of tall man letters on perceptual similarity. Participants were given a “same/different”
judgment task, in which they were presented with a pair of drug names on a computer
screen and had to indicate, as quickly and accurately as possible, whether the two
names were the same or different.

The task was chosen to represent a situation in

which people were faced with similarly named products that were placed next to each
other on a shelf.

This forced task study was composed of experiment 1 and

experiment 2, using the same test procedure.

The difference between the two

experiments was that the following instruction was provided prior to experiment 2, but it
was not given prior to experiment 1:
“Tall man letters are used in an attempt to make similar names less confusable with
each other.”
In experiment 1, there was no evidence of an effect of letter style on response time to
indicate a same or different name. However, in experiment 2 (p<0.01), there was
49

evidence of a significant effect of letter style, suggesting that shorter response time for
the tall man name pairs was taken than for the lowercase names.

Another experiment

was conducted, using a recognition memory task to assess the effect of color.

Names

were presented either in lowercase or in tall man letters, and they consisted of black
text alone or of black-and-red text. During this recognition memory task, there were
two phases: a study phase and a test phase. In the study phase, 5 names were
presented; in the test phase, 10 names presented were composed of the 5 names
(previously provided during the study phase) and 5 distractor names.

Participants

were asked to indicate, for each name on the test list, whether or not it had appeared in
the study list.

This study reported fewer overall errors for names containing tall man

letters than for names in lowercase (p<0.05), and no evidence of a significant effect on
color for the overall number of errors (p>0.05).
Borgmeier and Westenhoefer (2009) investigated which signpost food label
format enabled consumers best to differentiate healthier products from less healthy
ones, using a forced task choice methodology.

Participants were given a

“heathier/less healthy” judgment task, in which they were presented with 28 food pairs
in 5 different nutrition label formats: (1) a simple “healthy choice” tick, (2) a multiple
traffic light label, (3) a monochrome Guideline Daily Amount (GDA) label, (4) a colored
GDA label and (5) a “no nutrition label” condition. There was evidence of a significant
effect of different nutrition label formats on the average number of correct choices for
each subject (p<0.001). The traffic light label yielded the highest average of 24.8
correct choices of the 28 pairs, and the “no nutrition label” condition was associated with
the lowest average of correct choices (20.2 of 28 pairs).
50

CHAPTER 3
BENCHMARKING EXISTING, COMMERCIAL LABELING FOR
INDWELLING, URINARY CATHETERS
3.1 Objective – Benchmark existing, commercial packages for indwelling urinary
catheters to verify (or refute) the reports of difficulty with four pieces of critical
information
There are three sub-objectives for the first experiment:
•

Sub-Objective 1 - Characterize the placement of four pieces of labeling
information (sterility status, latex status, expiration dating and product name)
on packages of commercially available indwelling, urinary catheters

• Sub-Objective 2 - Objectively evaluate the text of four pieces of critical
labeling information
Measure:
o leading
o kerning
o color contrast
o type size
•

Sub-Objective 3 - Objectively evaluate the symbols present on the labeling of
several brands of indwelling urinary catheters, the product being used as a
model in this study
Identify:
o originating standard (where possible)
o presence/absence (with and/or without text)
51

o symbol size
o color contrast
•

Sub-Objective 4 - Objectively evaluate the comparative legibility of five pieces
of labeling information on packages of commercially available indwelling
urinary catheters

3.2 Methodology
3.2.1 Placement of labeling information
The placement of four pieces of information (sterility status, latex status,
expiration dating and product name) was evaluated using the nomenclature presented
in Figure 3. The locations of the labeling information were classified in binary fashion
(as present or absent) in all 4 sectors of the lidding web for the commercially available
catheters (six brands) included.

52

Full lab
bel view

1

3

2

Section 1

Sectio
on 2

Section3

Sectio
on 4

4

Figure 3. Location
ns of labelin
ng informattion

3.2.2 Measu
urement off leading, kerning
k
and
d type size
e, and colo
or contrastt
evaluation
3.2.2.1 Equipment
Leading,
L
ke
erning and letter
l
size were
w
measu
ured and re
ecorded using a Bridge
eport
optical comparator
c
(Bridgeporrt, CT) in su
urface illum ination mod
de.

A testt label was

placed on
o the front moving pla
ate (Up/Dow
wn and Rig
ght/Left) with a holding
g fixture made
from pollystyrene. Once the targeted
t
lab
bel text wass focused, adjusting th
he control b
bar
53

of the fro
ont plate, th
he surface illumination
n mode dep
picted the te
ext on the sscreen where it
was mea
asured (Fig
gure 4).

Figure 4. Optical
O
com
mparator
3.2.2.2 Mate
erials and Methods
Leading
L
is the
t distance
e between baselines o
of the succe
essive liness as presen
nted
in Figure
e 5.

54

Figure
F
5. Le
eading of tyypefaces
Kerning
K
is negative spa
acing betwe
een charactters in a texxt line (Figu
ure 6).

Figure
F
6. Kerning of tyypefaces
s measured
d to charactterize the tyype size.
x-height was

Research h
has suggessted

eight, the height of the
e body of a lowercase letter (speccifically, the
e x; see Figure
that x-he
7), is a better
b
indica
ator of legib
bility than ty
ype size, which is base
ed on the h
height of a
letter-pre
ess block frrom the now
w-antiquate
ed system o
of type setting (Bix, Lo
ockhart, Selke,
Cardoso
o, & Olejnik
k, 2003).

55

Figure
e 7. Relativ
ve x-heightss of typefacces
Texts
T
were further
f
charracterized by
b recording
g the color contrasts w
with which tthey
were pre
esented (e.g. black (te
ext)/white (b
background
d)).
3.2.3 Symbo
ol evaluation
3.2.3.1 Orig
ginating sta
andard
Two
T
globally
y recognize
ed symbol standards
s
fo
or medical d
devices are
e ISO 1522
23
(Medical Devices – Symbols to
t be used with
w medica
al device la
abels, labeliing and
informattion to be su
upplied) an
nd EN 980 (Graphical
(
S
Symbols fo
or use in the
e labeling o
of
medical device). In 2007, the
e Associatio
on for the A
Advanceme
ent of Mediccal
entation (AA
AMI) publis
shed a standard for symbol use w
with medica
al devices,
Instrume
harmoniizing with IS
SO 15223. This new standard rrecognized additional ssymbols in
2008. Herein, sym
mbols on co
ommercial labels
l
were
e characteriized as “sta
andard or nond” symbols,, based on AAMI/ANSI/ISO 1522
23: 2007 A1
1: 2008 (see
e Table 15)).
standard

56

3.2.3.2 Pres
sence/abse
ence (with and/or witthout text)
The
T status of critical in
nformation was
w recorded as: sym
mbol/text, text only and
d
symbol only.
o
3.2.3.3 Sym
mbol size
Using
U
an op
ptical comp
parator man
nufactured b
by Bridgepo
ort (Bridgep
port, CT) in
n the
surface illumination
n mode, the
e size of sym
mbols used
d for three p
pieces of crritical
y status, lattex status and
a expiratiion dating) was measu
ured in widtth
informattion (sterility
and heig
ght, and rec
corded by drawing
d
a dimensioned
d square (F
Figure 8) ass indicated by
ISO 918
86-1 2007.

Figure
F
8. Symbol to filll square
or contrastt of symbo
ols
3.2.3.4 Colo
Symbols
S
we
ere further characteriz
zed by reco
ording the co
olor contrassts with wh
hich
they werre presente
ed (e.g. blac
ck (text)/wh
hite (backgrround)).
3.2.4 Legibiility of textts
3.2.4.1 Subjjects
Ninety
N
nine subjects we
ere recruite
ed from PKG
G 101 and PKG 330 cclasses at th
he
School of
o Packagin
ng, Michigan State University and
d tested usiing procedu
ures approvved
under IR
RB # 13-698
8.

Prior to
o this legibility experim
ment, a rese
earch conse
ent form wa
as

provided
d to subjectts to acquire
e their agre
eement for planned ressearch activities (see
57

Appendix 8). Three pre-tests: visual acuity, color blindness and health literacy, were
conducted to characterize participants.

A questionnaire was given to subjects to

collect their demographic information (see Appendix 5).
3.2.4.2 Equipment
The Lockhart Legibility Instrument (East Lansing, MI), or LLI was used to
evaluate the relative legibility of the label information for various catheter brands (Figure
9).

Participants recruited from within and around the University were instructed to

rotate the hand wheel of the LLI until the first point that they could “easily read” the label
information. Rotation of the hand wheel rotates a single polarizing filter of the LLI, and
in series, its second filter is held in place. The more the filter is rotated, the more light
is allowed through. The polarizing filter can be rotated to a total of 90○ of rotation (total
light).

As such, information that requires a larger degree of rotation is expected to be

more difficult for a participant to decipher than information that requires a lesser degree
of rotation. Within subjects comparisons were made to derive the relative legibility of
the 3 critical pieces of information (sterility status, latex status and expiration dating),
compared to other elements of the label, such as the brand and product information.

58

Figure 9. The Lockha
art Legibilityy Instrumen
nt (LLI)
3.2.4.3 Mate
erials and Methods
Prior
P
to the legibility ex
xperiment, three pre-te
ests were u
used to cha
aracterize te
est
participa
ants.

Thes
se were: vis
sual acuity, color blind
dness and h
health litera
acy.

A visual acu
uity card from Bernell/v
vision traini ng productss (Mishawa
aka, IN) was
utilized to
t measure
e near pointt visual acuity (see Fig
gure 10).

T
The approxximate16-inch

distance
e between a subject’s eyes and th
he visual accuity card w
was kept un
nder genera
al
room illu
umination during
d
the visual acuity
y test, and tthe score of visual acu
uity per sub
bject
was reco
orded as 20
0/20, 20/30
0, 20/40, etc
c.

59

Figure 10. Visual acuity card

Pseudo-isochromatic plates from Richmond products (Boca Raton, Florida) were
used to evaluate the color perception of subjects (see Figure 11).

If a subject

responded correctly to 10 or more out of 14 test plates, the color vision status of the
subject was considered as “normal”: otherwise, if 5 or more incorrect response were
given, the subject was considered “at risk for color blindness”.

60

Figure 11. Pseudo-isochromatic plates for testing color perception

Subjects’ health literacy status was conducted using the Rapid Estimate of Adult
Literacy in Medicine-Reduced (REALM-R) (Bass III, Wilson, & Griffith, 2003).

This

card consisted of 11 words: Fat, Flu, Pill, Allergic, Jaundice, Anemia, Fatigue, Directed,
Colitis, Constipation, and Osteoporosis.
the card aloud.

Subjects were asked to read 11 words from

The first 3 words: Fat, Flu and Pill, serve as an acclimation period,

which were not tallied for a final score.
didn’t know the given words.

Subjects were instructed to say “blank” if they

If less than 6 out of 9 words (excluding the first 3 words)

were pronounced correctly, a subject was reported as “at risk” for poor health literacy.
For each commercial package, five pieces of information (brand name, product
name, sterility status, latex status and expiration dating: see Table 16) were further
evaluated according to procedures prescribed by ASTM D7298-06, “Standard Test
61

Method of Comparative Legibility by Mea
ans of a Po
olarizing Filtter” with 99 subjects
d from withiin and arou
und the Univ
versity.
recruited

The detailed
d procedure
e of the legibility

test is ex
xplained in Figure 12.

gure 12. Legibility test procedure
Fig

Each
E
particip
pant condu
ucted 30 tria
als (six bran
nds x five p
pieces of infformation)
using the legibility instrument
i
(see Appen
ndix 4).
actex” to miitigate any run order effect.
e
“Proc Fa

Trials were rrandomized
d using the SAS
The
e response variable wa
as measure
ed

as degre
ees of rotattion of the polarizing
p
filter in the le
egibility insttrument.

T
The respon
nse

variable was modeled using a general lin
near mixed model fitted with the M
Mixed
procedu
ure of SAS software.
s

62

Table 16. La
abeling info
ormation forr legibility evaluation
Inform
mation item
ms

Symbol
S

Con
ntents (Tex
xt)

Bra
and name

NA

BARDEX
X I.C. (e.g. Bard)

Prod
duct name

NA

Anti--Infective Fo
oley Cathetter(e.g. Barrd)

Lattex status

Cauttion: This p
product Con
ntains Naturral
Rubb
ber Latex W
Which May C
Cause Allerrgic
R
Reactions.

Sterrility status

Ster ile: Unless packaging is opened or
d
damaged

Expirration dating
g

USE BY YYYY-M
MM

3.3 Results
Twenty
T
cath
heters (six brands) we
ere characte
erized.

Th
hose cathetter label

images are presented in Appe
endix 18. Details from
m this bencchmarking sstudy are
d in Append
dices 19 to 26.
reported
3.3.1
3
Place
ement of la
abeling info
ormation
The
T produc
ct name (e.g
g. ‘Foley Ca
atheter’), w
was primarilyy located in
n section 1
(Figure 13).

Fourtteen out of the 20 cath
heters had their produ
uct name in section 1; the

rest of th
hem, in sec
ction 4.

In addition, various sub--product na
ames such as ‘Pediatric’,

‘All silico
one’, ‘Size (e.g.
(
66 Fr.)’, ‘Volume (e.g. 5 cc)’’, etc. were present ass
supplem
mentary info
ormation. Ninety
N
five % of the in vestigated catheters ccontained th
heir
sub-prod
duct information in sec
ction 4.
The
T require
ed latex warrning, “Cau
ution: This p
product con
ntains naturral rubber la
atex
which may
m cause allergic
a
reac
ctions”, was
s commonlyy located in
n sections 1 and 4 (Fig
gure
13).

Tw
wo out of th
he 20 cathe
eters did not have any latex-relate
ed informattion. As su
uch,
63

they were not included in percentage calculations of placement frequencies.
The common text regarding sterility status was “Sterile unless packaging is
opened or damaged”. The top two locations for this information were sections 1 and 2
(Figure 13). Six out of the 20 catheters had different locations for their text and symbol
regarding sterility status; locations used for text were considered for percentage
calculations.

Most of the investigated catheters included their expiration date in

section 4 (Figure 13). One had no information regarding the expiration date.

Thus, it

was excluded from percentage calculation. A second was excluded from the
calculation totals because information regarding the expiration date was contained on
the reverse side of the packaging.
None of 20 catheters studied had text and symbols for all four pieces of critical
information in one location; 40% of them used 2 locations and the majority (60 percent)
used 3 locations to display the information.

64

(18)

100%

100%
%
90%
%
80%
%

(14)

70%

70%
%
60%
%

(6)

(6)
(2)

440%

(2)

(2)

11%
%
11
1%

20%
%

50%

33%

30%

30%
%
10%
%

(8)

45%

50%
%
40%
%

(10)

(8)

10%
0%
%

0%0%

0% 0%
0 0%

0%
%
Produ
uct name
Section 1
S

Latex staatus
Section
n 2

Steerility statu
us
Secction 3

Exp. dating

Section 4
 

Note) Values in
n parenthesiss represent a  number of frrequencies 

Fiigure 13. Placement
P
findings
f
in regard
r
to fo
our pieces o
of critical infformation
3.2.2 Measu
urement off leading, kerning
k
and
d type size
e, and colo
or contrastt
evaluation
e distance between
b
lines of type ((see Figure
e 5) was alsso measure
ed
Leading, the
ntaining crittical text. In instance
es where an
ny of the criitical
for lines of type con
eading wass not measu
ured.
informattion was priinted in a single line, le

Ave
erage, minim
mum

and max
ximum valu
ues of leadin
ng measure
ements are
e presented
d in Figure 1
14.

Leadin
ng

surround
ding the pro
oduct name
e tended to be larger t han for the other criticcal elementts.

65

The leadings surrounding expiration dating was not measurable because it was always
presented in a single line.
Unit: mm

12.0
Max.

11.3

Avg.

10.0

Min.

8.0
7.3

6.0

4.0

3.6
2.8

2.7

2.0

2.0

2.9
2.5
2.4

0.0
Product name

Latex status

Sterility status

Figure 14. Leading measurement results in regard to three pieces of critical
information
The x-height of text used for critical information for each of the 20 packages was
measured to characterize the type size.

When text was present in only capital letters,

the height of the capitals was used. Average, minimum and maximum values of xheight measurements are presented in Figure 15.

66

Unit: mm

7.0
Max.

6.0

6.1
Avg.

5.0

Min.

4.0

3.9

3.0

2.9

2.0
1.0

2.2

2.0
1.4

1.1

1.5

1.7
1.2

1.4

1.1

0.0
Product name

Latex status

Sterility status

Exp. Dating

Figure 15. x-height analysis results in regard to four pieces of critical information

Text color contrast was also observed and recorded for the four pieces of critical
information. Of the 20 catheters evaluated, two did not have the text associated with
latex status and three had the latex status symbol only (Appendix 22).

Regarding the

sterility status, two out of them had the sterility status symbol only (Appendix 22).

One

out of them did not have its expiration date. Catheters that had symbols only and did
not have critical information were not counted for percentage calculations (Appendix 22).
Detailed findings of the color contrast study are presented in Figure 16.

67

100.0%
90.0%
80.0%

10.0%

15.8%

(2)

15.0%

33.3%

(3)

(5)

70.0%
60.0%

35.0%

(3)

55.6%

(10)

5.5%

(1)

(7)

50.0%

(5)

33.3%

40.0%

84.2%

(16)

30.0%
20.0%

40.0%

(8)

(5)

33.4%

38.9%

(7)

10.0%
0.0%
Product name
Green/White

Latex status
Black/white

Sterility status
Blue/White

Exp. dating
White/Blue
 

Note) Values in parenthesis represent a number of frequencies

Figure 16. Findings on text color contrast in regard to four pieces of critical
information
3.3.3 Symbol evaluation
Most of the symbols investigated originated from AAMI/ANSI/ISO 15223: 2007
A1: 2008 (Medical Devices – Symbols to be used with medical device labels, labeling
and information to be supplied).

Three latex-free symbols and one symbol related to

expiration dating present are not part of the recognized standard, AAMI/ANSI/ISO
15223: 2007 A1: 2008.

These non-standard symbols are presented in Figure 17.

68

Figure 17
7. Non-standard symbo
ols found frrom the ben
nchmarking
g study
Symbols
S
we
ere evaluatted on whetther or not they were p
present on a label with
h
supplem
mentary textts, in a text only forma
at or in a symbol only fformat.

Off the 20

catheterrs evaluated
d, two did not
n have the
e text or symbol assocciated with latex statuss
(Append
dix 26).

One out of th
hem did nott have its exxpiration da
ate informa
ation.

Catheters that did not
n have crritical inform
mation were
e not counte
ed for perce
entage
calculatiions (Appen
ndix 26). The
T percen
ntage of ‘syymbol/text’, ‘text only’ a
and ‘symbo
ol
only’ for three piece
es of critica
al informatio
on are pressented in Fig
gure 18.

69

100.0%
16.7%

(3)

10.0%

(2)

40.0%

(8)

10.5%

(2)

89.5%

(17)

80.0%
60.0%
66.6%

(12)

40.0%
20.0%
0.0%

50.0%

(3)

(10)

16.7%

Latex status
Symbol/Text

Sterility status
Text only

Exp. dating

Symbol only

Note) Values in parenthesis represent a number of frequencies

Figure 18. Symbol presence/absence findings in regard to three pieces of critical
information
Average, minimum and maximum values of symbol measurements for three
pieces of critical information are presented in Figure 19.

70

Unit: mm
m

35.0
M
Max.

30.0

30.4

A
Avg.

25.0

M
Min.

20.0

18.9

15.0
10.0
5.0

14.6
9.2

8.3

8.8

8.4

5.0
2.1

3.5

7.2

7.77

6.8
4
4.7
3.2

5.9
4.7
3.3

0.0
Width
h

Wid
dth

W
Width

Height

Height

Height

S
mea
asurement results in re
egard to thrree pieces of critical
Figure 19. Symbol
infformation
The
T color co
ontrasts of symbols
s
forr three piecces of critica
al information were
evaluate
ed.

Of the 20 cathete
ers evaluate
ed, six had the latex sttatus symb
bol, twelve the

sterility status
s
symb
bol and sev
venteen the
e expiration dating sym
mbol (Appen
ndix 24).
Catheters that did not
n have sy
ymbols asso
ociated with
h each piecce of critica
al informatio
on
were not counted fo
or percenta
age calculattions.
Figure 20.
2

71

The
e detailed fiindings are
e presented in

100.0%
80.0%

33.3%

(2)

60.0%

16.7%

(1)

25.0%

(3)

16.7%

(2)

40.0%
50.0%

20.0%

58.3%

(3)

5.9%

(1)

94.1%

(16)

(7)

0.0%
Latex
Black/white

Sterile
Green/White

Exp. Dating
Blue/White

Note) Values in parenthesis represent a number of frequencies

Figure 20. Findings on symbol color contrasts in regard to three pieces of critical
information

3.3.4 Legibility
Ninety-nine subjects were recruited from PKG 101 and PKG 330 classes at the
School of Packaging, Michigan State University.
students and faculty from those two classes.
female.

The subject group consisted of

Of the 99 participants, 54 were male; 45,

The average age of participants was 20 years old (ranging from 18 to 57,

median: 19).

More details on the subject demographics information are presented in

Figure 21.

72

Ethnicity (%)

Gender (%)
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%

100.0%
80.0%
60.0%

54.5%
45.5%

40.0%

81.6%

12.2%
1.0%

3.1%

1.1%

1.0%

20.0%
0.0%
Male

Female

Education

Native language
100.0%

89.9%

100.0%

80.0%

97.0%

80.0%

60.0%

60.0%

40.0%

40.0%

20.0%

10.1%

20.0%

2.0%

1.0%

Bachelor's
degree

Master's
degree

0.0%

0.0%
English

High school
degree

Others

Figure 21. Demographics information of subjects for gender, ethnicity, native language
and education (highest level achieved)
Three pre-tests regarding visual acuity, color blindness and health literacy were
conducted prior to the legibility experiment.
presented in Figure 22.

73

Details of the pre-test results are

Visual acuity
100.0%
80.0%

68.7%

60.0%
40.0%

30.3%

20.0%
1.0%
0.0%
20/20

20/30

20/40

Color blindness
100.0%

Health literacy

98.0%
100.0%

97.0%

80.0%
80.0%
60.0%

60.0%

40.0%

40.0%

20.0%

20.0%
3.0%

2.0%
0.0%

0.0%
Normal

At risk

Healthy

At risk

Figure 22. Subject characteristics on visual acuity, color blindness and health literacy

Legibility index readings were collected from a total of 99 subjects. The
collected data were analyzed, using a general linear mixed model of the statistical
software SAS 9.3 (SAS Ins., Cary, NC). The response variable (degrees of rotation)
was modeled as a function of the fixed effects of labeling information (5 pieces).
Gender, age, ethnicity, education level, visual acuity, health literacy, color blindness,
74

inside-light level and ambient light level were included in the model as explanatory
covariates.

Model fitting and parameter estimation was conducted using the MIXED

procedure of the statistical software SAS.
The data was log-transformed to meet the normality assumption.

Only age

(p=0.0006) and inside-light level of the legibility instrument (p<0.0001) were retained out
of all the possible covariates that were collected during the experiment, based on Type
III p-values (α=0.05).
Estimated least square means (LSM) and corresponding 95% confidence
intervals (LCL: Lower Confidence Limit and UCL: Upper Confidence Limit) are reported
in the original scale of a degree of rotation (Figure 23).

Relevant pairwise comparisons

were conducted, using Fisher’s LSD.
The analysis of variance identified a significant effect of information type on the
legibility readings as measured by the degrees of rotation of the polarizing filter
(p<0.0001).

Data suggest that brand name (LSM =10.3, LCL=9.5, UCL=11.2) and

product name (LSM =11.1, LCL=10.3, UCL=12.1) are significantly more legible than
other pieces of critical information: latex status, sterility status and expiration dating.
Expiration dating (LSM =13.7, LCL=12.7, UCL=14.9) is significantly more legible that
the other two pieces of critical information: latex status (LSM =16.0, LCL=14.7,
UCL=17.3) and sterility status (LSM =16.0, LCL=14.8, UCL=17.4). The analyzed data
suggested no evidence of a significant difference in legibility when latex status and
sterility status were compared (p=0.6544). More details on these comparisons are
presented in Figure 23.

75

Estimated degree of rotation

20.0

c

c

18.0

d

16.0
14.0

b

a

12.0
10.0
8.0
6.0
4.0
2.0
0.0
Brand name

Product name

Latex status

Sterility status

Expiration
dating

Figure 23. Estimated least square means (LSM) of degrees of rotation on legibility of
the five pieces of labeling information with estimated upper and lower limits. Differing
letters indicate statistical significance at α = 0.05.
3.4 Discussion and Conclusions
The purpose of this benchmarking study was to continue to explore the labeling
problems identified from Cai’s focus group study (Cai, 2012).
Out of 20 catheters purchased from commercial markets, not a single label
contained the four pieces of critical information (product name, latex status, sterility
status and expiration dating) in a single panel.

This scattered information placement

has the potential to be problematic when healthcare providers require critical
information in a busy and/or chaotic environment.

.

Results from the legibility experiment provide evidence that the three pieces of
critical information (latex status, sterility status and expiration dating) were significantly
less legible than brand name and product name (α=0.05).
76

This is not surprising given

the difference in font size that was found (see Figure 15).

This finding is consistent

with previous work that suggests that warning and safety-related information was not as
easy to read as brand name, claim statement, etc. (Bix et al., 2009).
The color contrasts on the three pieces of critical information were various and
not standardized: e.g. green/white, black/white, blue/white, white/blue, etc.
This benchmarking study helps to confirm data reported by Cai (2012).
Specifically, critical information was scattered, small font sizes hindered legibility, and
several color contrasts were problematic. It is not unreasonable to assume that these
contribute to difficulty in finding critical information on medical device labels that the
same groups reported.

77

CHAPTER 4
DESIGN EFFECTS (BOXING, GROUPING, SYMBOL AND COLOR
CODING) ON EARLY STAGES OF THE INFORMATION PROCESSING
MODEL USING CHANGE DETECTION
4.1 Objective & Hypothesis
•

Objective 1 – Investigate the efficacy of boxed information compared with
unboxed
o Evaluate the attentive behaviors of healthcare professionals regarding
three pieces of critical information presented in a “boxed format” vs.
“unboxed format” using a change detection methodology.

•

Objective 2 – Investigate the effect of a single location placement of three
pieces of critical information on the noticeability of said information using a
change detection methodology.

•

Objective 3 – Investigate the effect of symbols (presence vs. absence) of
three pieces of critical information on noticeability using a change detection
methodology.

•

Objective 4 – Investigate the effect of color coding (presence vs. absence) of
three pieces of critical information on noticeability using a change detection
methodology.

•

Objective 5 – Using the symbols identified in AAMI/ANSI/ISO 15223: 2007
A1:2008, evaluate comprehension using ISO 9186 – 1 2007: Graphical
symbols – Test methods – Part 1: Methods for testing comprehensibility.

78

•

Hypothesis - It is hypothesized that a standard location and format of
information deemed critical to care will attract attention more quickly in
early stages of information processing.

4.2 Methodology
4.2.1 Subjects
Healthcare professionals were recruited at the Association of Surgical
Technologists (AST) conferences in Savannah, GA and Denver, CO, and using a
targeted e-mail (see Appendix 12) of AST members within a 30 mile radius of Lansing,
MI.

The screening criteria in recruitment were:
•

Have no history of seizure

•

Be over 18 years of age

•

Not be legally blind

•

Be a healthcare professional, or a student in a healthcare field.

Eighty-six healthcare professionals (primarily surgical technologists and nurses)
participated in two experiments: change detection to evaluate the efficacy of varied
designs for critical information and symbol comprehension evaluation.
This study was conducted using procedures approved under IRB # 13-698.
Prior to testing, a research consent form was provided to subjects to acquire their
informed consent (see Appendix 9). Three tests were conducted to characterize
participants, namely: visual acuity, color blindness and health literacy. Further, a
research questionnaire was given to subjects to collect subject demographic and jobrelated information (see Appendix 7).

79

4.2.2 Equipment and Software: Change Detection
During the change detection trials (“flicker task”), subjects were comfortably
seated in front of a computer screen and asked to depress the computer’s space bar as
soon as they noticed the “flickering portion” of an image showing on the screen.
During each flicker trial, a control image (240ms) continuously alternated with the test
image (240ms) that had been slightly altered with a brief, gray screen image (80ms)
interleaving as shown in Figure 24.

The only difference between the control and test

images was the disappearance of a single piece of information on the label (for critical
trials, a piece of critical information). This sequence control-blank-test-blank looped
until the participant pressed the space bar, indicating that they had found the alternation,
or until they timed out at 1 minute.

Testing was conducted using E-Prime 2.0

(Psychology Software Tools, Inc.) and trials were randomized to mitigate any run order
effects.

Additionally, each trial image was divided into 4 sectors and the individual

section images for all trials were randomized to mitigate any location effects (Figure 25).

80

First screen: A control image appears for 240ms (Latex symbol and texts).

Section 1

Section 2

Section 3

Section 4

Second screen: A gray image appears for 80ms.

Third screen: A test image appears for 240ms (Latex symbol and texts are
disappearing).

Changed part
Section 1

Section 2

Section 3

Section 4

Fourth screen: A gray image for 80ms (Returning back to the control image if a
subject does not press the computer’s space bar).

Figure 24. Sequence of Change Detection images

81

Figure 24. (Cont’d)
First screen: A control image appears for 240ms (Sterile symbol and texts).

Section 1

Section 2

Section 3

Section 4

Second screen: A gray image appears for 80ms.

Third screen: A test image appears for 240ms (Sterile symbol and texts are
disappearing).
Changed part

Section 1

Section 2

Section 3

Section 4

Fourth screen: A gray image for 80ms (Returning back to the control image if a subject
does not press the computer’s space bar).

82

Figure 24. (Cont’d)
First screen: A control image appears for 240ms (Sterile symbol and texts)

Section 1

Section 2

Section 3

Section 4

Second screen: A gray image appears for 80ms.

Third screen: A test image appears for 240ms (Sterile symbol and texts are
disappearing).
Changed part

Section 1

Section 2

Section 3

Section 4

Fourth screen: A gray image for 80ms (Returning back to the control image if a
subject does not press the computer’s space bar).

83

A label was divided into 4 section images.

Section 1

Section 2

Section 3

Section 4

Locations of section images were randomized per subject.
Example 1

Example 2

Example 3

Figure 25. The same label appears with sections in different locations. Location was
randomized across subjects.

84

4.2.3 Material and Method: Efficacy of boxing, grouping, symbol and color
Trial labels were developed using Adobe Illustrator CS 3.0. Their size was
1,280 pixels wide by 768 pixels tall. Each trial label was divided into 4 sectors for
randomization as explained in the previous section. Each sector image in a label was
256 pixels wide by 192 pixels tall.
Each of the design factors (Boxing, Grouping, Symbol and Color) was evaluated
at two levels, present or absent. Conditions were crossed for a total of sixteen
treatments (2 x 2 x 2 x 2) of interest for each piece of critical information (Table 17):
•

(1-2) grouped information within a box with and without a symbol in a color-coded
format,

•

(3-4) grouped information unboxed with and without a symbol in a color-coded format,

•

(5-6) ungrouped information within a box with and without a symbol in a color-coded
format, and

•

(7-8) ungrouped information, unboxed with and without a symbol in a color-coded
format

•

(9-10) grouped information within a box with and without a symbol in a non-colorcoded format,

•

(11-12) grouped information unboxed with and without a symbol in a non-color-coded
format,

•

(13-14) ungrouped information within a box with and without a symbol in a non-colorcoded format, and

•

(15-16) ungrouped information, unboxed with and without a symbol in a non-colorcoded format (see Table 17).

85

Table 17. Experiment combinations of Change Detection
Critical information

sterility status, latex status and expiration dating

Boxing format

boxed vs. unboxed information: 2 levels

Grouping format

grouped vs. ungrouped information: 2 levels

Symbol and Text format
Color coding format

symbol with text (symbol present) vs. text only without symbol
(symbol absent) : 2 levels
color-coded vs. non-color-coded: 2 levels

Color coding consisted of color-coded and non-color-coded formats. In the
color-coded format, red (text and symbol) and white (background) colors were used
when latex information was present, green (text and symbol) and white (background)
colors, when sterile information was present, and black (text and symbol) and white
(background) colors, when expiration date was present (see Table 18). In the noncolor-coded format, black (text and symbol) and white (background) colors were used
for all three pieces of information (see Table 18).

86

Table 18. Color
C
codin
ng formats: Change De
etection
Critic
cal
informa
ation

Collor-coded

N
Non-color-c
coded

Sterility
statu
us

Latex sttatus

Expiration
datin
ng

*
*In color coded trials of expiration date, the latex and ste
erility status appeared ass color-code
ed
information; expiratio
on date (the changing element) appe
eared in blacck and white
e.

The
T sixteen treatments
s previously
y described were appliied for each
h of three
pieces of
o critical infformation (s
sterility stattus, latex sttatus and e
expiration da
ating), for a
total of 48
4 critical trrials (16 x 3)
3 for the ch
hange detecction testing
g (Table 19
9).

In addition

to the 48
8 trials (Figures 26 and
d 27) where
e changes occurred to
o critical infformation, 4
48
filler trials were created (Figurres 27 and 28).

As su
uch, there w
were 96 tria
als per

participa
ant; run order was rand
domized by
y subject; p
position of th
he quadran
nts within th
he
label wa
as also rand
domized as
s pilot testing suggeste
ed location effects.
the 96 trrials are sho
own in Table 19.

87

T
The details o
of

Table 19. Matrix Chart of Change Detection Trials
Label
image
Critical
information
trials
(Image
Type A)

Grouping
Ungrouped

Color
coding

Label
information

Colorcoded

Sterility
status
Latex status
Expiration
dating
Sterility
status
Latex status
Expiration
dating
Sterility
status
Latex status
Expiration
dating
Sterility
status
Latex status
Expiration
dating
Brand name
Product
name
Do not reuse
Do not
resterilize
Manufacturer
Do not use

Noncolorcoded

Critical
information
trials
(Image
Type B)

Grouped

Colorcoded

Noncolorcoded

Dummy
trials
(Image
Type B)

Dummy
trials
(Image
Type A)
Dummy
trials
(Image C)

Grouped

Noncolorcoded

Ungrouped

Batch code
Ungrouped

Noncolorcoded

Brand name
Product
name
Do not reuse
Do not
resterilize
Manufacturer
Do not use
Batch code
*Not all combinations of filler trials were tested.

88

Symbol with text

Text only
without symbol
Unboxe
Boxed
d

Boxed

Unboxed

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

NA*

NA*

49

50

NA*

NA*

51

52

53

54

55

56

57

58

59

60

61
65

62
66

63
67

64
68

69

70

71

72

NA*

NA*

73

74

NA*

NA*

75

76

77

78

79

80

81

82

83

84

85
89
93

86
90
94

87
91
95

88
92
96

Ungrouped, Unboxed, Symbol absent and Non-color-coded conditions

Ungrouped, Unboxed, Symbol absent and Color-coded conditions

Ungrouped, Unboxed, Symbol present and Non-color-coded conditions

Ungrouped, Unboxed, Symbol present and Color-coded conditions

Figure 26. Change Detection trials: Image Type A
89

Figure 26. (Cont’d)
Ungrouped, Boxed, Symbol absent and Non-colored conditions

Ungrouped, Boxed, Symbol absent and Color-coded conditions

Ungrouped, Boxed, Symbol present and Non-color-coded conditions

Ungrouped, Boxed, Symbol present and Color-coded conditions

90

Grouped, Unboxed, Symbol absent and Non-color-coded conditions

Grouped, Unboxed, Symbol absent and Color-coded conditions

Grouped, Unboxed, Symbol present and Non-color-coded conditions

Grouped, Unboxed, Symbol present and Color-coded conditions

Figure 27. Change Detection trials: Image Type B
91

Figure 27. (Cont’d)
Grouped, Boxed, Symbol absent and Non-color-coded conditions

Grouped, Boxed, Symbol absent and Color-coded conditions

Grouped, Boxed, Symbol present and Non-color-coded conditions

Grouped, Boxed, Symbol present and Color-coded conditions

92

Ungrouped, Unboxed, Symbol absent and non-color-coded conditions

Ungrouped, Unboxed, Symbol present and non-color-coded conditions

Ungrouped, Boxed, Symbol absent and Non-color-coded conditions

Ungrouped, Boxed, Symbol present and Non-color-coded conditions

Figure 28. Change Detection trials: Image Type C
93

4.2.4 Materials and Method: Comprehension of symbols
Symbols for warnings, cautions, etc. are commonly used for pharmaceutical and
medical device products to reduce or eliminate potential risks.

It is very important that

product users comprehend the correct meaning of symbols intended to convey
important information regarding many medical devices.
Comprehension testing quantifies the degree of understanding of symbols by
the target group and intends to answer the questions: “What do you think this means?”
or “What action would you take in response to this symbol?” (ISO 9186-1, 2007).
4.2.4.1 Stimulus Materials
A set of printed test sheets was prepared with 41 graphical symbols within a
square not less than 28 mm x 28 mm such that the graphical symbol filled the square
(ISO 9186-1 2007).

The 38 symbols, standardized and defined by AAMI/ANSI/ISO

15223: 2007 A1: 2008 were included in the comprehension test form. Along with those
38 symbols, three latex-free symbols that were identified from the previous
benchmarking study were tested (Experiment 1 – see Figure 17). The set of printed
test sheets was given to subjects with the following instruction (see Appendix 6):
“This study is intended to evaluate your comprehension level of medical device
symbols used for commercially available medical devices.”
4.2.4.2 Procedure
Participants were instructed to record their answer to the question: “What do you
think this symbol means?”

In addition, they were told to write the response “Don’t

know” if they were unable to assign a meaning to the symbol.
for them to fill out the symbol comprehension form.
94

There was no time limit

4.2.4.3 Categorization
According to ISO 9186-1:2007, subject responses were categorized as below:
•

1: Correct,

•

2a: Wrong, 2b: Wrong and the response given is the opposite of the
intended meaning

•

3: The response given is “Don’t know”

•

4: No response is given.

All the responses regarding symbol meaning were coded in an excel spreadsheet.
Three judges reviewed categorized codes, and in-depth discussion among three judges
was conducted to come up with consensus on unmatched response codes.
responses in category 1 were considered a correct answer.

The

Responses from

categories 2 to 4 were tallied as incorrect. Percentage by category code was
calculated for each symbol by dividing the number responses in a category by the total
number of responses for that symbol response. The criterion of 85% described in
ANSI Z535.3 was applied to evaluate participants’ comprehension level (ANSI Z535. 15 2011, & Liu et al., 2005). Symbols which generate more than an 85% response rate
in category 1 were considered as having an acceptable comprehension level;
conservatively, we defined this as having an Upper Confidence Limit (UCL) that
exceeded the 85% value.

95

4.3 Results
4.3.1 Subject demographics
Eighty-six healthcare professionals were recruited at the Association of Surgical
Technologists (AST) conferences in Savannah, GA and Denver, CO, and using a
targeted e-mail (see Appendix 12) of AST members within a 30 mile radius of Lansing,
MI.

The average age of participants was 44 years old (ranging from 18 to 66, median:

47). Figure 29 provides information regarding the age of the test population.
 
30.0%

26.7%

25.0%
20.0%

17.4%
14.0%

15.0%
10.5% 10.5%
10.0%

8.1%

7.0%
3.5%

5.0%
1.1%

0.0%

1.1%

0.0%
16‐20 21‐25 26‐30 31‐35 36‐40 41‐45 46‐50 51‐55 56‐60 61‐65 66‐70
years years years years years years years years years years years

 

 

Figure 29. Age of participants

Of the 86 participants, 17 were male; 69 female.
native speakers of English.

Eighty-four participants were

More details on the demographics are presented in Figure

30.

96

Ethnicity (%)

Gender (%)
100.0%
90.0%

80.2%

80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%

100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%

73.3%

4.7%

12.8%

8.1%

1.1%

19.8%

10.0%
0.0%
Male

Female

Native language
100.0%

97.7%

Education
100.0%
80.0%

80.0%

60.0%

60.0%

59.3%

40.0%

40.0%
20.0%

22.1%

15.1%

20.0%

3.5%

2.3%
0.0%

0.0%
English

Associate's Bachelor's Master's
Degree
Degree
Degree

Others

Others

Figure 30. Demographics information (%) of participants on gender, ethnicity, native
language and education (highest level achieved)
The three pre-tests regarding visual acuity, color blindness and health literacy
were conducted prior to the change detection and symbol comprehension experiments.
The details of the pre-test results are presented in Figure 31.

97

Visual acuity
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%

67.4%

16.3%

20/20

11.6%

20/30

4.7%

20/40

20/50

Color blindness
100.0%

100.0%

Health literacy

100.0%

80.0%

80.0%

60.0%

60.0%

40.0%

40.0%

20.0%

20.0%

0.0%

0.0%

98.8%

1.2%
Normal

At‐risk

Healthy

At‐risk

Figure 31. Subject characteristics (%) for visual acuity, color blindness and health
literacy
4.3.2 Descriptive statistics on questionnaire evaluation
4.3.2.1 Years of experience
On average, the subject population had 20 years of experience (ranging from 1
98

to 51, median: 22).

Figure 32 characterizes the subject population with regard to years

of experience.
25.0%
20.0%

19.8%
17.4%

16.3%

15.0%

11.6%
9.3%

10.0%

9.3%

9.3%
7.0%

5.0%
0.0%
0‐5
years

6‐10 11‐15 16‐20 21‐25 26‐30 31‐35 >35
years years years years years years years

Figure 32. Experience in years

4.3.2.2 Employment settings
Healthcare providers work in diverse care settings.

Figure 33 provides an

indication of the frequency of employment setting as self-reported by participants.

60.0%

52.3%

50.0%
40.0%
25.5%

30.0%
20.0%
10.0%

7.0%

1.2%

1.2%

4.7%

0.0%

Figure 33. Employment settings of participants (%)
99

8.1%

4.3.2.3 Position & role
Figure 34 depicts the frequencies with which participants reported their role in
healthcare.

60.0%
48.8%

50.0%
40.0%
30.0%

26.7%
16.3%

20.0%
10.0%

4.7%

3.5%

Student

Other

0.0%
Program Director,
Instructor,
Coordinator, or
Educator

Surgical
Technologist

Registered Nurse

Figure 34. Position & role of participants (%)

4.3.2.4 Critical pieces of labeling information
Participants were asked to report the information from medical device labeling
they deemed to be most important (see Survey - Appendix 7). Participants’ responses
were ranked from 1st (most important) to 4th (least important) items, and were
categorized into 14 response groups (Appendix 14). The total frequencies of the top
five responses, and their median and mode are presented in Table 20.

100

Table 20. Critical pieces of labeling information with top 5 out of 14 response groups
Expiration
dating

Latex
status

Sterility
status

Product
name

Use instructions

1 (Most important)

31

19

11

13

1

2

27

15

19

6

9

3

14

16

18

14

7

4 (Least important)

7

12

11

11

7

Total Frequency of
Participant
Responses

79

62

59

44

24

Median of Ranking

2

2

2

3

3

Mode of Ranking

1

1

2

3

2

4.3.2.5 Critical labeling problems
Participants were asked to write labeling problems on a response sheet (see
Survey - Appendix 7).

Responses were ranked from 1st (most problematic) to 4th (least

problematic) items in this category, and grouped into 10 response categories (Appendix
15). The total frequencies of the top five response groups, and their median and mode
are presented in Table 21.

101

Table 21. Critical labeling problems with top 5 out of 10 response groups
No
Labeling
standard
location for designs not
standardized
critical
information

No color
coding

No color
contrast

8

5

7

21

11

14

9

6

15

20

11

9

4 (Least problematic)

4

4

10

4

4

Total Frequency of
Participant
Responses

75

58

49

34

29

Median Ranking

1

2

3

2

2

Mode Ranking

1

2

3

2

2,3

Ranking

Small font
size

1 (Most problematic)

44

18

2

21

3

4.3.2.6 Medical errors due to labeling issues
Participants were asked to report medical errors that they experienced due to
labeling issues (see Survey - Appendix 7). Responses were categorized into 6 groups.
Their total frequencies are presented in Table 22.

Table 22. Medical errors participants experienced due to labeling issues
Response groups

Total frequencies

Wrong product/size opening or use
Expired product opening or use
Latex-containing product opening or use to latex-allergy patients
Incorrect dosage
Unsterile product opening or use
Other medical errors

37
30
21
11
9
12

4.3.2.7 Recommendations for labeling designs
Participants were asked to make suggestions regarding the resolution of
102

labeling problems (see Survey - Appendix 7).

Suggestions were then categorized into

8 response groups. Response frequencies are summarized in Table 23.
Table 23. Suggested recommendations to resolve labeling problems
Response groups

Total frequencies

Bigger or bolder font size
Color coding
Standard location for labeling information
Standardization for labeling designs
Highlighted critical information
Clear color contrast
Standardized symbols
Others

57
44
43
29
14
12
4
19

4.3.3 Statistical analysis on Change Detection
Two response variables were obtained for this experiment from the EPrime®
software for each change detection trial:
•

A binary variable: Successful detection of change (Yes/No) prior to
timing out at 60 seconds

•

A continuous variable: Time to successful change detection prior to
timing out at 60 seconds.

4.3.3.1 Binary Variable – Change Detected (Yes/No)
A generalized linear mixed model was fitted to this binary variable - change
detected (yes/no or timeout at 60 seconds) using a logit-link function to model the
probability of change detection (in %). Only critical trials were analyzed i.e. design
changes in three pieces of critical information.

Linear predictors in the model were four

design factors (grouped vs. ungrouped + boxed vs. unboxed + symbol presence vs.

103

symbol absence + color-coded vs. non-color-coded), and all possible 2-way, 3-way and
4-way interactions were analyzed.
None of the demographic covariates was retained in the final model since there
was no significant effect of those covariates, based on their Type III p values (α=0.05).
The model was fitted using the GLIMMIX procedure of SAS 9.3 (SAS Ins., Cary,
NC).

Relevant pair-wise comparisons were conducted using Fisher’s LSD.
A total of 4,128 trials (86 subjects x 48 trials) were analyzed as part of the

change detection experiment.

In 98.9% of the total trials, participants correctly

identified the location of change prior to timing out; 1.1% of trials resulted in incorrect
identification of location.
Although there was evidence of a main effect of Grouping (p=0.0294) and Color
(p=0.0499) on the probability of successful detection, these results were not practically
significant because of the high rates of successful detection, regardless of treatment
(e.g. grouped vs. ungrouped, and color-coded vs. non-color-coded) (Figure 35 and 36).
No 2-way, 3-way and 4-way interaction terms were significant statistically.

104

100.0%

a

b

98.9%

99.6%

Grouped

Ungrouped

98.0%
Estimated probability (%)

96.0%
94.0%
92.0%
90.0%
88.0%
86.0%
84.0%
82.0%
80.0%

Figure 35. The effect of ‘Grouping’ on the probability of successful change detection:
Estimated least square means (LSM) with estimated upper and lower limits. Differing
letters indicate statistical significance at α=0.05.

100.0%

a

b

99.0%

99.6%

Color‐coded

Non‐color‐coded

Estimated probability (%)

98.0%
96.0%
94.0%
92.0%
90.0%
88.0%
86.0%
84.0%
82.0%
80.0%

Figure 36. The effect of ‘Color’ on the probability of successful change detection:
Estimated least square means (LSM) with estimated upper and lower limits. Differing
letters indicate statistical significance at α=0.05.
105

4.3.3.2 Continuous Variable – Time to detect change (milliseconds)
For critical changes that were successfully detected prior to timing out at 60
seconds, a second variable, "time to detect change," was recorded in milliseconds.
Gender, age, ethnicity, education level, visual acuity, health literacy, color blindness,
native languages, and change location were included in the model as explanatory
covariates.

In order to meet necessary model assumptions, values were expressed on

a log scale.

Similar to the analysis for the previous variable, linear predictors in this

model were the four design factors (grouped vs. ungrouped + boxed vs. unboxed +
symbol presence vs. symbol absence + color-coded vs. non-color-coded), and all
possible 2-way, 3-way and 4-way interactions were analyzed.
Gender, age and ethnicity were retained in the final model, based on their Type
III p values (α=0.05). Change location, which was randomized in each trial by the
EPrime® Software, showed a significant effect on time to successful change detection,
and as such, was included as a random variable in the final model, based on its Type III
p values (α=0.05).
The model was fitted using the Mixed procedure of SAS 9.3 (SAS Ins., Cary, NC).
Estimated least square means (LSM) and corresponding 95% confidence intervals
(LCL=Lower Confidence Limit and UCL=Upper Confidence Limit) were reported in the
original millisecond scale.

Relevant pairwise comparisons were conducted using

Fisher’s LSD.
There was evidence for a main effect of three design factors on the ‘time to
detect change’ response: Boxing (p<0.0001), Symbol (p=0.0002) and Color (p<0.0001).
Several 2-way interaction terms showed a significant effect on the response time
106

variable: ‘Grouping by boxing’ (p<0.001), ‘Grouping by symbol presence’ (p=0.0253),
‘Grouping by color’ (p=0.0015), ‘Boxing by color’ (p=0.0003), and ‘Symbol by color’
(p=0.0028).

In addition, there was a significant effect of one 3-way interaction term,

‘Boxing by symbol by color’ (p=0.0323).

Significant two-way interaction terms which

are not included in the significant 3-way interaction term are reported below.
4.3.3.2.1 Significant 2-way interaction terms: Grouping x Boxing
This statistical analysis suggested that the time to detect changes depended on
both the Grouping and Boxing designs.

When designs were boxed changes took

significantly less time to detect when compared to their unboxed counterpart in the
grouped format (p=0.0086).

This positive effect of boxing also took place in the

ungrouped format (p<0.0001, Figure 37).

Specifically, changes were successfully

detected faster in the boxed, ungrouped condition (LSM=1740.2ms, LCL=1558.5ms,
UCL=1943.1ms), when compared with the boxed, grouped condition (LSM=1925.8ms,
LCL=1724.2ms, UCL=2150.3ms).

By contrast, the grouped, unboxed condition

(LSM=2051.6ms, LCL=1837.0ms , UCL=2290.9ms) outperformed the ungrouped,
unboxed condition (LSM=2268.8ms, LCL=2031.4ms, UCL=2533.4ms, relatively; Figure
37)
This is likely because the use of the box triggered bottom-up attention response.
The ungrouped, boxed condition likely triggered a search behavior in which participants
rapidly moved to a series of small boxed targets (in close proximity to the information
changing), quickly reaching the information that was changing.
condition, no such benefit was present.

107

In the unboxed

Estimated Time to detect change 
(milliseconds)

3000.0
2500.0

d
c

a
b

2000.0
1500.0
1000.0
500.0
0.0
Grouped + Boxed

Ungrouped +
Boxed

Grouped  +
Unboxed

Ungrouped +
Unboxed

Figure 37. The effect of ‘Grouping’ and ‘Boxing’ formats on Time to detect change:
Estimated least square means (LSM) with estimated upper and lower limits. Differing
letters indicate statistical significance at α=0.05.
4.3.3.2.2 Significant 2-way interaction terms: Color x Grouping
The time to detect changes also depended on both Color and Grouping.
Participants took less time to detect change when designs were color-coded as
compared to their non-color-coded counterparts in the grouped format (p<0.0001).
This positive effect of color also took place in the ungrouped format (p<0.0001, Figure
38).

When non-color-coded, the ungrouped condition outperformed the grouped

condition (LSM=2140.9ms, LCL=1917.3ms, UCL=2391.1ms vs. LSM=2260.5ms,
LCL=2024.0ms, UCL=2524.6ms, relatively).

The reverse was true of the colored

treatments, where those that were grouped were detected faster (LSM = 1747.4ms,
LCL=1564.9ms, UCL=1951.6ms vs. LSM=1843.7ms, LCL=1651.2ms, UCL=2059.2ms).
Again, it could be theorized that color is effective in triggering bottom-up processing and
the block of color in the grouped condition induces quick responses from participants.
108

Estimated Time to detect change 
(milliseconds)

3000.0

b

2500.0

d
c

a
2000.0
1500.0
1000.0
500.0
0.0
Color‐coded +
Grouped

Non‐color‐coded +
Grouped

Color‐coded +
Ungrouped

Non‐color‐coded +
Ungrouped

Figure 38. The effect of ‘Color’ and ‘Grouping’ and formats on Time to detect change:
Estimated least square means (LSM) with estimated upper and lower limits. Differing
letters indicate statistical significance at α=0.05.
4.3.3.2.3 Significant 2-way interaction terms: Grouping x Symbol
The time to detect changes also depended on both Grouping and Symbol.
Within grouped conditions, symbol presence (LSM=1888.9ms, LCL=1691.2ms,
UCL=2109.1ms) resulted in less time to detect changes than the symbol absent
condition (p<0.0001: LSM=2091.7ms, LCL=1872.8ms, UCL=2336.1ms, Figure 38).
However, there was no evidence of a significant difference in the time to detect change
between grouped and ungrouped designs in the symbol presence format (p=0.1159).
This was also true of the symbol absence format (p=0.1116, Figure 39).

109

2500.0
Estimated Time (milliseconds)

a

a,b

c

b,c

2000.0

1500.0

1000.0

500.0

0.0
Grouped +
Ungrouped +
Grouped +
Ungrouped +
Symobl presence Symbol presence Symbol absence Symbol absence

Figure 39. The effect of ‘Grouping’ and ‘Symbol’ formats on Time to detect change:
Estimated least square means (LSM) with estimated upper and lower limits. Differing
letters indicate statistical significance at α=0.05.

4.3.3.2.4 Significant 3-way interactions: Boxing x Symbol x Color
There was evidence of a significant interaction of Boxing x Symbol x Color on
time to successful change detection (p=0.0323).

As depicted in Figure 40, the design

combinations of ‘boxed/symbol present/color-coded’ and ‘boxed/symbol absent/colorcoded’ resulted in the fastest detection times (LSM = 1671.9ms, LCL=1489.7ms,
UCL=1875.9ms vs. 1740.2ms LCL=1550.2ms, UCL=1953.0m, respectively). There
was no evidence of a significant difference in the time to detect changes between these
two designs above (p=0.2418).
Although 3-way interactions, this one included, are challenging to interpret, there
are some interesting results provided that should be pointed out. The unboxed, noncolored, symbol absent treatment generated the largest response time (LSM=2669.9ms,
110

LCL=2378.5ms, UCL=2996.4ms), this was closely followed by treatments that were not
boxed, with symbol and no color coding (LSM=2273.0ms, LCL=2025.8ms,
UCL=2550.9ms).

These represent the current approach to labeling, yet performed

significantly worse than any other of the design combinations (α=0.05).

Generally

speaking, designs that had color-coding resulted in faster detection than those that did
not. The combination of boxing with color coding (both symbols present & absent)

Estimated Time to detect change (milliseconds)

generated the fastest responses.
3500.0

e

3000.0
2500.0

b
a

b

b

c

d

a

2000.0
1500.0
1000.0
500.0
0.0

Figure 40. The effect of ‘Boxing’, ‘Symbol’ and ‘Color’ formats on Time to detect
change: Estimated least square means (LSM) with estimated upper and lower limits.
Differing letters indicate statistical significance at α=0.05.

111

4.3.4 Percentage statistics on symbol comprehension evaluation
Responses of participants on the meaning of medical device symbols were
categorized according to the criteria described in the section of symbol evaluation
methodology. The percentage data of all 5 response categories for 41 symbols is
presented in Table 24.

112

Table 24. Percentage of each category response on medical device symbols
Symbol
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41

Symbol Description
Biological risks
Do not re-use
Consult instructions for use
Caution, consult accompanying
documents
Fragile, handle with care
Keep away from sunlight
Protect from heat and radioactive
sources
Keep away from rain
Lower limit of temperature
Upper limit of temperature
Temperature limitation
Use by date
Date of manufacture
Batch code
Catalog number
Serial number
Control
Negative control
Positive control
Sterile
Sterilized using ethylene oxide
Sterilized using aseptic processing
techniques
Sterilized using irradiation
Sterilized using steam or dry heat
Do not resterilize
Non-sterile
Do not use if package is damaged
In Vitro Diagnostic medical device
Patient number
Humidity limitation
Atmosphere pressure limitation
Sampling site
Fluid path
Non-pyrogenic
Contains or presence of natural rubber
latex
Drops per milliliter
Liquid filter with pore size
One-way valve
Latex-free
Latex-free
Latex-free

19.8% 
20.9% 
3.5% 

2b
(opposite
meaning)
0.0% 
1.2% 
0.0% 

3
(Don’t
know)
4.7% 
31.3% 
24.3% 

4
(No
response)
0.0% 
1.2% 
1.2% 

76.7% 

7.0% 

0.0% 

15.1% 

1.2% 

17.5% 
32.5% 

50.0% 
40.7% 

0.0% 
2.3% 

26.7% 
22.2% 

5.8% 
2.3% 

34.9% 

29.1% 

4.7% 

27.8% 

3.5% 

37.3% 
44.2% 
47.7% 
39.5% 
67.4% 
38.4% 
95.3% 
87.2% 
61.7% 
55.8% 
47.7% 
47.7% 
94.2% 
71.0% 

34.9% 
23.3% 
17.4% 
26.7% 
17.4% 
17.4% 
2.3% 
3.5% 
7.0% 
12.8% 
10.5% 
9.3% 
0.0% 
1.2% 

15.1% 
19.8% 
16.3% 
2.3% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 

8.0% 
11.5% 
15.1% 
28.0% 
15.2% 
41.9% 
1.2% 
8.1% 
29.0% 
22.1% 
32.5% 
33.7% 
2.3% 
23.1% 

4.7% 
1.2% 
3.5% 
3.5% 
0.0% 
2.3% 
1.2% 
1.2% 
2.3% 
9.3% 
9.3% 
9.3% 
3.5% 
4.7% 

38.4% 

2.3% 

1.2% 

48.8% 

9.3% 

43.0% 
72.1% 
70.9% 
97.7% 
39.6% 
2.3% 
18.6% 
14.0% 
2.4% 
0.0% 
0.0% 
24.4% 

5.8% 
4.7% 
23.3% 
0.0% 
27.9% 
27.9% 
32.6% 
24.4% 
22.1% 
50.0% 
7.0% 
15.1% 

2.3% 
0.0% 
0.0% 
1.2% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
1.2% 

40.8% 
17.4% 
5.8% 
1.1% 
27.8% 
61.7% 
43.0% 
54.6% 
67.4% 
43.0% 
82.5% 
51.2% 

8.1% 
5.8% 
0.0% 
0.0% 
4.7% 
8.1% 
5.8% 
7.0% 
8.1% 
7.0% 
10.5% 
8.1% 

96.5% 

1.2% 

0.0% 

1.1% 

1.2% 

5.8% 
0.0% 
0.0% 
98.8% 
100.0% 
98.8% 

68.6% 
29.1% 
11.6% 
1.2% 
0.0% 
0.0% 

3.5% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 

18.6% 
60.4% 
76.8% 
0.0% 
0.0% 
1.2% 

3.5% 
10.5% 
11.6% 
0.0% 
0.% 
0.0% 

1
(correct)

2a
(wrong)

75.5% 
45.4% 
71.0% 

113

The percentage data of the correct responses (category 1) was analyzed statistically,
using a “Proc Means” model of the statistical software SAS 9.3 (SAS Ins., Cary, NC).
Through the “Proc Means” data analysis, means and Lower Confidence Limits (LCL)
and Upper Confidence Limits (UCL) at the 95% confidence level were reported by
percentage for the 41 symbols tested (Table 25).

If the UCL of their percentage

exceeds 85%, it was considered as passing comprehension testing, based on the ANSI
85% criterion (Table 25).

The shaded rows in Table 26 depict “passing symbols”,

based on this criterion.

114

Table 25. Percentage of correct response category on medical device symbols:
Means and Upper & Lower Confidence Limits at 95% confidence level
Symbol
No.

Symbol Description

Caution, consult accompanying
documents
2
Batch code
3
Catalog number
4
Sterile
5
Non-sterile
Contains or presence of natural rubber
6
latex
7
Latex-free*
8
Latex-free*
9
Latex-free*
10
In Vitro Diagnostic medical device
11
Atmosphere pressure limitation
12
Sampling site
13
Fluid path
14
Liquid filter with pore size
15
One-way valve
16
Keep away from sunlight
17
Protect from heat and radioactive sources
18
Keep away from rain
19
Lower limit of temperature
20
Upper limit of temperature
21
Temperature limitation
22
Sterilized using radiation
23
Drops per milliliter
24
Biological risks
25
Do not re-use
26
Consult instructions for use
27
Fragile, handle with care
28
Use by date
29
Date of manufacture
30
Serial number
31
Control
32
Negative control
33
Positive control
34
Sterilized using ethylene oxide
35
Sterilized using aseptic processing techniques
36
Sterilized using steam or dry heat
37
Do not resterilize
38
Do not use if package is damaged
39
Patient number
40
Humidity limitation
41
Non-pyrogenic
*: non-standard symbols / shaded rows: passing symbols
1

115

Means

LCL

UCL

76.7%

67.6%

85.9%

95.3%
87.2%
94.2%
97.7%
96.5%

90.8%
80.0%
89.1%
94.4%

99.9%
94.4%
99.2%
100.0%

92.6%,

100.0%

98.8%
100.0%
98.8%
2.3%
2.4%
0.0%
0.0%
0.0%
0.0%
32.5%
34.9%
37.3%
44.2%
47.7%
39.5%
43.0%
5.8%
75.5%
45.4%
71.0%
17.5%
67.4%
38.4%
61.7%
55.8%
47.7%
47.7%
71.0%
38.4%
72.1%
70.9%
39.6%
18.6%
14.0%
24.4%

96.5%
100.0%
96.5%
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
0.0% 
22.5% 
24.6% 
26.8% 
33.5% 
36.9% 
29.0% 
32.3% 
0.0% 
66.3% 
34.6% 
61.1% 
9.3% 
57.3% 
27.9% 
51.1% 
45.1% 
36.9% 
36.9% 
61.1% 
27.9% 
62.4% 
61.1% 
29.0% 
10.2% 
6.5% 
15.2% 

100.0%
100.0%
100.0%
5.6%
5.6%
0.0%
0.0%
0.0%
0.0%
42.7%
45.2%
47.6%
54.9%
58.4%
50.1%
53.7%
10.9%
84.8%
56.1%
80.7%
25.6%
77.5%
48.9%
72.1%
66.5%
58.4%
58.4%
80.7%
48.9%
81.8%
80.7%
50.1%
27.0%
21.4%
33.7%

Alarmingly,
A
only
o
6 out of
o the 38 sta
andard testted symbols passed th
he
compreh
hension critterion specified by the
e ANSI stan
ndard at the
e 95% confidence leve
el
(see Tab
ble 26).
Table 26.
2 Passing
g symbols in comprehe
ension: Me
eans and Up
pper & Low
wer Confide
ence
Limits
L
at 95
5% confiden
nce level
Sym
mbol

Correct meaning
m

Means

LCL

UC
CL

Caution, co
onsult accom
mpanying doccuments

76.7%,

6
67.6%,

85.9%

Batch code
c

95.3%

9
90.8%,

99.8%

Catalog number
n

87.2%

8
80.0%

94.4%

Sterile

94.2%

8
89.1%

99.2%

Non-ste
erile

97.7%,

9
94.4%

100
0.0%

Contains or presence off natural rub
bber latex

96.5%

9
92.6%,

100
0.0%

 

In addition, all of th
he three non
n-standard symbols id
dentified in tthe benchm
marking stud
dy
t 85% criterion.
passed the

One common
n factor tha
at the passing symbolss had was tthe

incorporration of tex
xt within, ex
xcept for on
ne symbol o
of “Caution, consult acccompanyin
ng
documents”.

This
s calls to qu
uestion the efficacy
e
of recognized
d symbols th
hat do not

utilize te
ext, and is relevant given the currrent FDA prroposal whiich would a
allow for
incorporration of sta
and-alone symbols
s
(if accompanie
a
gend and re
ecognized b
by an
ed by a leg
international standa
ard).
The
T percen
ntage data of
o the corre
ect response
es (categorry 1) was also analyze
ed
116

for low comprehension. The symbols which had an LCL at or below 10% in category 1
were “In-vitro diagnostic medical device”, “Atmosphere pressure limitation”, “Sampling
site”, “Fluid path”, “Liquid filter with pore size”, “One-way valve”, “Drops per milliliter”,
“Humidity limitation”, and “Fragile, handle with care” (see Table 27).

Most of the

symbols with poor comprehension levels were pictorial symbols which did not
incorporate text in their symbol, except for “In-vitro diagnostic medical device”, “Liquid
filter with pore size” and “Drops per milliliter”. The symbols which had 0% in
comprehension were “Sampling site”, “Fluid path”, “Liquid filter with pore size”, and
“One-way valve”.

These symbols were newly added to the ANSI/AAMI/ISO 15223, in

2008, but are not included in ISO 15223.

117

Table 27.
2 Symbolls at less th
han 10% pe
ercentage in
n comprehe
ension: Mea
ans and Up
pper
& Lower Co
onfidence Limits
L
at 95%
% confiden
nce level
Sym
mbol

Correct meaning
m

Means

LCL

U
UCL

In-vitro diagnostic
c medical de
evice

2.3% 

0.0% 

5
5.6% 

Atmo
osphere pres
ssure limitattion

2.4% 

0.0% 

5
5.6% 

Samplin
ng site

0.0% 

0.0% 

0
0.0% 

Fluid path
p

0.0% 

0.0% 

0
0.0% 

Liquid filter with pore size
e

0.0% 

0.0% 

0
0.0% 

One-way
y valve

0.0% 

0.0% 

0
0.0% 

Drops perr milliliter

5.8%

0.0%

10
0.9%

Humidity limitation

14.0%

6.5%

21
1.4%

Fragile, handle with care

17.5%

9.3%

25
5.6%

 

 

 

 
Wrong
W
respo
onses that were
w
oppos
site to the intended me
eaning werre coded ass
“2b”. To
T be conse
ervative, we
e only listed
d symbols a
as “Criticallyy confusing
g symbols”
when their Lower Confidence
C
Limits exce
eed 5% in tthe 2b: the 5% cut off is defined b
by
ANSI Z5
535.3.
The
T percen
ntage data in the 2b ca
ategory (wro
ong & oppo
osite) for tessted symbo
ols
118

was ana
alyzed statis
stically, usin
ng a “Proc Means” mo
odel of the sstatistical ssoftware SA
AS
(Version
n 9.3, SAS Institute, Ca
ary, NC). Through th
he “Proc Me
eans” data analysis,
Lower Confidence
C
Limit (LCL)) and Upper Confidencce Limit (UC
CL) at the 9
95%
confiden
nce level arre reported in Table 29
9.

If the LC
CL of the 2b response
es exceeded

5%, it was considered as a “crritically con
nfusing sym
mbol” with re
egard to comprehensio
on
(Table 28).
2
Three
T
symb
bols were fo
ound to fall into the ca
ategory of “ccritically confusing
symbols
s”.

These symbols we
ere: “Keep away from sunlight”, ““Lower limitt of

tempera
ature”, and “Upper
“
limit of temperrature”.

Se
ee Table 28
8 for summation of failled

symbols
s.
Tablle 28. “Critically confusing symbo
ols”: Meanss and Upper & Lower C
Confidence
e
Limits
L
at 95%
% confiden ce level
Correct
C
meaning
m

Crittically confu
used
re
esponse (2 b)

Means

L
LCL

UCL
L

Keep
p away from
rain

proof, Imperm
meable,
Waterp
Imperv
vious to rain , Water
resistant,
r
etcc.

15.1%

7
7.3%

22.6%
%

wer limit of
Low
tem
mperature

Keep
p cool, Keep at low
tempera
ature, Store below x
temperrature, Ok to
o store in
co
old places, e
etc.

19.8%

11
1.0%

28.0%
%

Upper limit of
mperature
tem

Keep warm,
w
Above
e boiling,
Store above
a
x temp
p, Ok to
sto
ore in hot pla
aces

16.3%

8
8.2%

24.0%
%

Symbol

cussion
4.4 Disc
The
T sample
e population
n of this stu
udy was com
mprised of an experienced pool o
of
healthca
are providerrs (average
e experienc
ce years: 20
0, ranging ffrom 1 to 51
1). Further, by
sampling
g at confere
ences of a national, prrofessional organizatio
on, it is not unreasona
able
119

to assume that it included engaged providers from across the nation.
providers we recruited were generalists.

That said, the

It is likely that specifically targeting specialists

would yield different results.
The information focus groups in Cai’s study identified as critical were also
reported as critical in the survey results reported herein (see Table 21). In addition, the
recommendations for improvement to labeling design were also very similar to what Cai
suggested (see Table 23)
4.4.1 Change Detection
In order to evaluate the effect of four design factors (Grouping, Boxing, Symbol
and Color) on the three pieces of critical information as identified by Cai’s study (2012)
in early stages of information processing (i.e. attention), we employed a change
detection method. In doing so, we enumerated the effect of the varied designs on
attention.
The probability of successful change detection within the 60 second window was
found to be extremely high across all the four design factors. Even though there was
evidence of a main effect of two design factors (Grouping and Color), the difference in
successful detection was less than 1%.

This finding suggests that participants had

enough time to detect changes for all the treatments of evaluated labeling designs, that
is, participants were at ceiling.
We also employed time to successful change detection as a response variable.
Analysis revealed that multiple 2-way and one 3-way interactions were evident.
Participants responded more quickly to changes in the three pieces of critical
information when the format was boxed than when unboxed, in the grouped design.
120

(p=0.0086). This was also true in the ungrouped design (p<0.0001, Figure 37).
However, when boxed and ungrouped, rates of detection were significantly faster than
those boxed and grouped. The opposite was true in the unboxed condition.

That is,

when information was grouped, respondents found changes faster than when
ungrouped, which took the longest time overall.

However, the opposite effect of

grouping was indicated for non-colored designs, whereby those that were grouped took
significantly more time to successful detection (LSM=2260.5ms, LCL=2024.0ms,
UCL=2524.6ms) than those that not (LSM=2140.9ms, LCL=1917.3ms, UCL=2391.1ms).
And, there was no evidence of a significant effect of grouping to the time to detect
changes in the symbol present condition (p=0.1159).

This result was also true in the

symbol absent condition (p=0.1116).
This unexpected finding regarding effect of grouping might result from an
experimental design context of our change detection.

In bottom-up processing,

incoming data is a critical piece influencing attention and perception (Goldstein, 2007).
In our change detection experiment, the incoming data which participants needed to
detect involved a piece of flickering critical information in the varied designs.

In the

boxed, grouped design, only one out of three pieces of critical information in a large
rectangular box that encompassed all three pieces of information flickered; in the boxed,
ungrouped condition, the box surrounded nothing but the flickering information.
box is the salient feature of the scene, it could explain the result found herein.

If the
That is,

responses to the grouped, boxed condition took significantly longer than the ungrouped,
boxed condition (where the salient item was close by the flickering information).
Colored designs catalyzed significantly faster change detection in the grouped
121

treatment than non-colored designs (p<0.0001).
treatment (p<0.0001).

This was also true in the ungrouped

This result reflects findings of previous research on a significant

effect of color presence: color-coded nutrition information on a cereal Front-Of-Panel
(Sundar, 2013).
Three design factors (Boxing, Symbol and Color) significantly interacted when
the dependent variable was time to detect changes (p<0.0323).

The designs of

‘boxed/symbol present/color-coded’ and ‘boxed/symbol absent/color-coded’ enabled
participants to detect changes significantly faster than other mixed designs.

This

reflects that box and color were the important salient features, resulting in faster
detection times.
4.4.2 Symbol Comprehension
It is likely that many manufacturers will take advantage of the opportunity to gain
label space by utilizing symbols from recognized standards if the proposed rule is
enacted by the US FDA.
Our work supports the work of others (Liu et al., 2004 and Hermans et al., 2011)
that suggests that the comprehension level for internationally published symbols for
medical device packaging is quite poor.

This was despite the fact that we recruited

from an experienced pool of healthcare providers from throughout the nation (see
Figure 32).
25).

Only 6 out of 38 standard symbols passed the 85% criterion (see Table

This poor comprehension result echos those reported by Liu et al. (2004) and

Hermans et al. (2011), who tested comprehension levels of medical device symbols
with populations outside of the US.
122

A common characteristic of successful symbols was the inclusion of
supplementary text within the symbol (see Table 26). Most of the symbols that were
correctly defined by less than 10% of respondents did not incorporate text within the
symbols (see Table 27). Perhaps most concerning is the fact that 3 of the symbols
that we tested were categorized as “critically confusing” according to the ANSI Z535.3
criteria.

In other words, at least 5% of respondents (as defined with an LCL of more

than 5%) indicated a meaning opposite to the defined, intended meaning of the symbol
(see Table 28).
In light of a very limited body of work (Liu et al., 2004 and Hermans et al., 2011),
all of which suggests poor comprehension rates for standard symbols, policy changes
should be carefully considered.

123

CHAPTER 5
DESIGN FEATURES (BOXING, GROUPING, SYMBOL AND COLOR
CODING) INFLUENCE ON INFORMATION PROCESSING DURING A
FORCED CHOICE TASK
5.1 Objective & Hypothesis
o Objective – Investigate the efficacy of Boxing, Grouping, Symbol
presence and Color-coding to critical information, during most stages of
information processing
o Hypothesis - It is hypothesized that a standard location and format of
information deemed to critical to care will have a higher rate of correct
response and shorter time to correct response during most stages of
information processing.
5.2 Methodology
5.2.1 Subjects
Healthcare professionals were recruited at the Association of Surgical
Technologists (AST) conferences in Savannah, GA and Denver, CO, and using a
targeted e-mail (see Appendix 13) of AST members within a 30 mile radius of Lansing,
MI.

The screening criteria in recruitment were:
•

Be over 18 years of age

•

Not be legally blind

•

Be a healthcare professional, or a student in a healthcare field.

Eighty-nine perioperative personnel (primarily surgical technologists and nurses)
participated to evaluate the efficacy of varied designs for critical information common to
124

medical device packages.
under IRB # 13-698.

This study was conducted using procedures approved

A research consent form was provided to subjects to acquire

their informed consent (see Appendix 10). Prior to the forced choice task, participants
were characterized in numerous ways, including: visual acuity, color blindness, health
literacy and demographics.

A questionnaire was given to subjects to collect subject

demographic and job-related information (see Appendix 7).
5.2.2 Materials and methods
Labels were developed using Adobe Illustrator CS 3.0.

Their size was 1280

pixels wide by 768 pixels tall. Each label was divided into 4 sectors for randomization
as explained in the methodology section of the change detection experiment.

A size of

one sector image in a label was 256 pixels wide by 192 pixels tall.
Test labels were created in combinations of boxing, grouping, symbols (absence
and presence) and color. Two color coding treatments (color-coded vs. non-colorcoded) were developed using green/white, red/white and black/white colors as
presented in Table 29.

125

Table
T
29. Color
C
coding
g formats: F
Forced Cho
oice Task
Critic
cal
informa
ation
Late
ex

Colo
or-coded

N
Non-color-c
coded

Latex-ffree

Sterile

Non-ste
erile

Expirred
*
Unexpired
*
*In color coded trials of expiration date, the latex and ste
erility status appeared ass color-code
ed
information; expiratio
on date appe
eared in blac
ck and white
e.

For
F the sake
e of comparrison, two commercial
c
labels werre broken in
nto 4 sectorrs
(see Figures 43 and 44).

The
ey comprised six force
ed choice trrials (2 bran
nds x 3 piecces

al informatio
on, see Tab
ble 31).
of critica
Labels that we
w designe
ed were also
o created.

For these
e labels, we
e evaluated four

f
eac
ch at two lev
vels: Boxing
g (boxed an
nd unboxed
d), Groupin
ng (grouped
d and
design factors,

126

ungrouped), Symbol (absent or present) and Color coding (absent or present).
Conditions were crossed for a total of sixteen treatments of interest (2 x 2 x 2 x 2).
During each forced choice task, two labels appeared on the screen (Figure 41
and 42). These labels only differed in one aspect, a single piece of critical information
was changed (e.g. one was sterile, the other not).

Trials were conducted with mock

brands (16 treatments x 3 pieces of information) (Table 30) and six trials with two labels
which emulated commercial labels (Figures 43 and 44) based on our benchmarking
results.

As such, there were 54 trials in total (48 trials for newly developed labels + 6

commercial trials (2 brands x 3 critical information) for this forced choice task (Table 30
and 31).
Testing was conducted using E-Prime 2.0 (Psychology Software Tools, Inc.),
and trial order was randomized to mitigate any run order effects. As with the change
detection trials, images were divided into 4 sections, with individual sections
randomized to mitigate any location effects (Figure 41 and 42).

However, these

randomizations were “yoked” such that the sections, and therefore the two comparative
labels were the same in all aspects other than the difference involving the selection
question for each pair in the choice.

127

Table 30. Matrix chart of Forced Choice Tasks: Newly developed labels
Critical
information

Grouping

Ungrouped
Sterile vs. Nonsterile

Grouped

Ungrouped
Latex vs. Latexfree

Grouped

Ungrouped
Expired vs.
Unexpired

Grouped

Symbol with text

Color
coding

Text only without
symbol
Boxed
Unboxed

Boxed

Unboxed

Color-coded

1

2

3

4

Non colorcoded

5

6

7

8

Color-coded

9

10

11

12

Non-color
coded

13

14

15

16

Color-coded

17

18

19

20

Non colorcoded

21

22

23

24

Color-coded

25

26

27

28

Non colorcoded

29

30

31

32

Color-coded

33

34

35

36

Non colorcoded

37

38

39

40

Color-coded

41

42

43

44

Non colorcoded

45

46

47

48

Table 31. Matrix chart of Forced Choice Tasks: Commercial labels
Image Type
Commercial
label A
Commercial
label B

Critical Information
‘Sterile’ vs. ‘Non-sterile’ label
‘Latex’ vs. ‘Latex-free’ label
‘Expired’ vs. ‘Unexpired’ label
‘Sterile’ vs. ‘Non-sterile’ label
‘Latex’ vs. ‘Latex-free’ label
‘Expired’ vs. ‘Unexpired’ label

Cell #
1
2
3
4
5
6

Just prior to each trial, participants were provided with instructions to select a
specific product (e.g. select the sterile device; select the latex containing device; select
the expired device) as quickly as possible (see Table 32 and Figure 41 & 42).
Selection was made by depressing either “ ↑ ”

128

(UP ARROW) or “ ↓ ” (DOWN

ARROW) on a keyboard entry system (corresponding with the label in the upper or
lower position, respectively) within 1 minute.
Table 32. Questions of Forced Choice Tasks
For the next pair, please select the device that IS STERILE.
‘Sterile’ vs. ‘Non-sterile’
label

Press “

↑ ” (UP ARROW) for the top device or

Press “ ↓ ” (DOWN ARROW) for the bottom device.
For the next pair, please select the device that HAS
LATEX.
‘Latex’ vs. ‘Latex-free’ label

Press “

↑ ” (UP ARROW) for the top device or

Press “ ↓ ” (DOWN ARROW) for the bottom device.
For the next pair, please select the device that IS EXPIRED.
‘Expired’ vs. ‘Unexpired’
label

Press “

↑ ” (UP ARROW) for the top device or

Press “

↓ ” (DOWN ARROW) for the bottom device.

The position of the correct choice was counter-balanced between subjects for
each combination of treatments. For instance, if the latex containing product for a label
that had color, grouping, symbol and boxed information appeared on top for subject one,
it would appear on the bottom for subject two.

For each subject, a correct choice for

27 trials took place at the top location and the remaining trials took place at the bottom
location.

This was accomplished with a set of stimulus and A & B sets, which were

rotated between subjects.

In this experiment, 44 subjects participated in the type-A

forced choice task; 45 subjects, in the type-B forced choice task. Order of presentation
of the complete set of 54 choices and the position of the section containing the
information critical to the choice, was randomized.

129

Top image.

Section 1

Section 2

Section 3

Section 4

Bottom image

Section 1

Section 2

Section 3

Section 4

Figure 41. Test cell # 26 from Table 30 (Latex vs. Latex free information)
130

Top image.

Section 1

Section 2

Section 3

Section 4

Bottom image

Section 1

Section 2

Section 3

Section 4

Figure 42. Test cell # 25 from Table 30 (Latex vs. Latex free information)
131

Top image.

Sec
ction 1

Section 2

Section 3

Section 4

Bottom imag
ge

Sec
ction 1

Section 2

Section 3

Figure 43
3. Commerciall label A (Latex
x vs. Latex free
e)

132

Section 4

Top image.

Sec
ction 1

Section 2

Section 3

Section 4

Bottom imag
ge

Sec
ction 1

Section 2

Section 3

Figure 44
4. Commerciall label B (Latex
x vs. Latex free
e)

133

Section 4

5.3 Results
5.3.1 Subject demographics
Eighty-nine healthcare professional were recruited at the Association of Surgical
Technologists (AST) conferences in Savannah, GA and Denver, CO, and using a
targeted e-mail (see Appendix 12) of AST members within a 30 mile radius of Lansing,
MI.

The average age of participants was 45 years old (ranging from 18 to 66, median:

47). Figure 45 provides information regarding the age of the test population.

23.6%

25.0%
20.0%
15.7%

16.9%

15.0%

12.4%

10.0%
5.0%

7.9%

2.2%

9.0%
6.7%

3.4%
1.1% 1.1%

0.0%
16‐20 21‐25 26‐30 31‐35 36‐40 41‐45 46‐50 51‐55 56‐60 61‐65 66‐70
years years years years years years years years years years years

Figure 45. Age of participants

Of the 89 participants, 16 were male; 73, female.

Eighty-seven participants

used English as a native language; two subjects reported English as a secondary
language.

More details on the demographics information are presented in Figure 46.

134

Ethnicity (%)

Gender (%)

100.0%

100.0%
80.0%
60.0%
40.0%
20.0%
0.0%

82.0%

80.0%
60.0%
40.0%
20.0%

18.0%

70.8%

3.4%

15.7%

9.0%

1.1%

0.0%
Male

Female

Education

Native language
100.0%
100.0%

97.8%

80.0%
60.0%

80.0%

40.0%

60.0%

20.0%

40.0%

0.0%

20.0%

50.6%
20.2%

9.0%

19.1%
1.1%

2.2%

0.0%
English

Others

Figure 46. Demographics information (%) of participants on gender, ethnicity, native
language and education (highest level achieved)

Three pre-tests regarding visual acuity, color blindness and health literacy were
conducted prior to the forced choice task. The details of the pre-test results are
presented in Figure 47.

135

Visual acuity
100.0%
90.0%
80.0%
70.0%

59.6%

60.0%
50.0%
40.0%
30.0%
20.0%

15.7%

14.6%

7.9%

10.0%

2.2%

0.0%
20/20

20/30

20/40

20/50

Color blindness
100.0%

96.7%

Health literacy
100.0%

80.0%

80.0%

60.0%

60.0%

40.0%

40.0%

20.0%

Unknown

97.8%

20.0%
1.1%

2.2%

0.0%

0.0%

2.2%

At‐risk

Unknown

0.0%
Normal

At‐risk

Unknown

Healthy

Figure 47. Subject characteristics (%) on visual acuity, color blindness and
health literacy

136

5.3.2 Descriptive statistics on questionnaire evaluation
5.3.2.1 Years of experience
On average, the subject population had 21 years of experience (ranging from 0
to 51, median: 22).

Figure 48 characterizes the subject population with regard to years

of experience.
20.0%
18.0%
16.0%

18.0%
15.7%

14.0%
12.0%

10.1%

10.1%

6‐10
years

11‐15
years

12.4%

11.2%

12.4%
10.1%

10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
0‐5 years

16‐20
years

21‐25
years

26‐30
years

31‐35 >35 years
years

Figure 48. Experience in years

5.3.2.2 Employment settings
Healthcare providers work in diverse care settings.

Figure 49 provides an

indication of the frequency of employment setting as self-reported by participants.

137

45.0%
40.0%
35.0%
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%

41.6%
37.1%

7.9%

6.7%
1.1%

3.4%

2.2%

Figure 49. Employment settings of participants (%)

5.3.2.3 Position & role
Figure 50 depicts the roles of participants in healthcare.

45.0%
40.0%
35.0%
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%

39.3%
33.7%

18.0%
5.6%

Program
Director,
Instructor,
Coordinator,
or Educator

Surgical
Technologist

Registered
Nurse

Student

Figure 50. Position & role of participants (%)

138

3.4%
Others

5.3.2.4 Critical pieces of labeling information
Participants were asked to record the information from medical device labeling
from most important (1) to least important (4) (see Survey - Appendix 7).

Researchers

categorized participant responses (post-hoc) into 15 response groups (see Appendix
16).

The total frequencies of the top five responses, and their median and mode are

presented in Table 33.

Table 33. Critical pieces of labeling information with top 5 out of 15 response groups
Expiration
dating

Sterility
status

Latex
status

Product
Name

Use
Instructions

1 (Most important)

22

20

10

18

4

2

25

20

17

9

7

3

20

15

15

10

7

4 (Least important)

11

4

17

5

8

Total Frequency of
Participant Responses

78

59

59

42

26

Median Ranking

2

2

3

2

3

Mode Ranking

2

1,2

2,4

1

4

This result closely parallels results collected from participants of the change
detection experiment (see Table 20).
5.3.2.5 Critical labeling problems
Participants were asked to record problems that resulted from labeling on a
response sheet (see Survey - Appendix 7).

Responses were ranked from 1st (most

problematic) to 4th (least problematic) items in this category, and grouped into 10
139

responses (see Appendix 17). The total frequencies of the top five response groups,
and their median and mode are presented in Table 34.

Table 34. Critical labeling problems with top 5 out of 10 response groups

Ranking

Small font
size

No
standard
location for
critical
information

1 (Most problematic)

47

16

9

5

1

2

20

14

17

15

6

3

10

19

12

9

6

4 (Least problematic)

3

5

3

4

5

Total Frequency of
Participant
Responses

80

54

41

33

18

Median Ranking

1

2

2

2

3

Mode Ranking

1

3

2

2

2,3

No color
contrast

No color
coding

Labeling
designs not
standardized

These top 5 response groups were also reported in the change detection
experiment survey. But, the ranking of frequencies of ‘no color contrast’, ‘no color
coding’ and ‘labeling designs not standardized’ was identical each other between the
surveys (see Tables 21and 34).
5.3.2.6 Medical errors due to labeling issues
Participants were asked to record medical errors involving labeling that they had
been involved with (see Survey - Appendix 7).
categorized into 6 groups.

Participants’ responses were

Their total frequencies are presented in Table 35.

140

Table 35. Medical errors participants experienced due to labeling issues
Response groups

Total frequencies

Wrong product/size opening or use
Expired product opening or use
Unsterile product opening or use
Incorrect dosage
Latex-containing product opening or use to latex-allergy patients
Other medical errors

31
27
17
15
13
20

The five shaded responses on medical errors were those that were reported in
the change detection experiment survey.

The ranking of frequencies of ‘unsterile

product opening or use’, ‘incorrect dosage’ and ‘latex-containing product opening or use
to latex-allergy patients’ was not identical each other between the surveys (see Tables
22 and 35).
5.3.2.7 Recommendations on labeling designs
Suggestions regarding the resolution of labeling problems were also collected
from research participants.

Suggestions were then categorized into 8 response groups.

Response frequencies are summed in Table 36
Table 36. Suggested recommendations to resolve labeling problems
Response groups

Total frequencies

Bigger or bolder font size
Color coding
Standard location for labeling information
Clear color contrast
Standardization for labeling designs
Highlighted critical information
Standardized symbols
Others

60
46
40
23
14
11
3
16

The seven shaded recommendations to improve labeling related problems were
the same as those that were reported in the change detection experiment survey.
141

The

ranking of frequencies of ‘clear color contrast’, ‘standardization for labeling designs’ and
‘highlighted critical information’ was identical each other between the surveys (see
Tables 23 and 36).
5.3.3 Statistical analysis on Forced Choice Tasks
Two response variables were collected for analysis from E Prime® software for
each forced choice task trial:
•

A binary variable: Correct choice (Yes/No) prior to timing out at 60
seconds

•

A continuous variable: Time taken to make a correct choice (milliseconds)
prior to timing out at 60 seconds.

5.3.3.1 Binary Variable – correct choice (Yes/No)
A generalized linear mixed model was fitted to this binary variable – correct
choice (yes/no or timeout at 60 seconds) using a logit-link function to model the
probability of correct choice (in %). Commercial label trials were not included in this
data analysis.

Linear predictors in this model were four design factors (grouped vs.

ungrouped + boxed vs. unboxed + symbol presence vs. symbol absence + color-coded
vs. non-color-coded), and all possible 2-way, 3-way and 4-way interactions were
analyzed.
From the demographic information collected, ethnicity (p=0.0230) was retained in
the final model, based on their Type III p values (α=0.05).
The model was fitted using the GLIMMIX procedure of SAS 9.3 (SAS Ins., Cary,
NC).

Estimated least square means (LSM) and corresponding 95% confidence
142

intervals (LCL=Lower Confidence Limit and UCL=Upper Confidence Limit) were
reported in the percentage of probability of correct choices.
4,053 (94.9%) out of the 4,272 trials (48 trials x 89 subjects) resulted in correct
choices; 219 trials (5.1%) had the incorrect product chosen.
There was evidence of a main effect of two factors on the probability of correct
choice: Boxing (p=0.0101) and Symbol (p<0.0001). The LSM difference between the
boxed and unboxed treatments was 1.4%. The unboxed treatment resulted in a higher
rate of correct choices that the boxed treatment (Figure 51).

The LSM difference

between the symbol presence and symbol absence treatments was 3.3%. The symbol
presence treatment resulted in a higher rate of correct choices than the symbol absence
treatment (Figure 52).

Estimated Probability  of correct 
choices (%)

100.0%

a

b

95.7%

97.1%

Boxed

Unboxed

95.0%
90.0%
85.0%
80.0%
75.0%
70.0%
65.0%
60.0%
55.0%
50.0%

Figure 51. The effect of ‘Boxing’ on Probability of correct choice: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters indicate
statistical significance at α=0.05.

143

Estimated Probability of correct choices 
(%)

100.0%

a

b

97.7%

94.4%

Symbol presence

Symbol absence

95.0%
90.0%
85.0%
80.0%
75.0%
70.0%
65.0%
60.0%
55.0%
50.0%

Figure 52. The effect of ‘Symbol’ on Probability of correct choice: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters indicate
statistical significance at α=0.05.

5.3.3.2 Continuous Variable: Time taken to make a correct choice
(milliseconds)
For critical trials that were correctly chosen prior to timing out at 60 seconds, a
second variable, "time to make a correct choice," was recorded in milliseconds.
Commercial label trials were not included in this data analysis. Gender, age, ethnicity,
education level, visual acuity, health literacy, color blindness, ethnicity, native languages,
and choice location were included in the model as explanatory covariates.
was log-transformed to meet normality assumptions.

The data

Similar to the analysis for the

previous variable, linear predictors in this model were the four design factors (grouped
vs. ungrouped + boxed vs. unboxed + symbol presence vs. symbol absence + colorcoded vs. non-color-coded, and all possible 2-way, 3-way and 4-way interactions were
analyzed.
144

Age (p<0.0001) and ethnicity (p=0.0084) were retained in the final model, based
on their Type III p values (α=0.05).

Correct choice location, which was randomized in

each trial by EPrime® software, had a significant effect (p=0.0005) on the dependent
variable (time to make a correct choice), and, as such, was included as a random
variable in the final model, based on its Type III p values (α=0.05).
The model was fitted using the Mixed procedure of SAS (Version 9.3, SAS
Institute, Cary, NC). Estimated least square means (LSM) and corresponding 95%
confidence intervals (LCL=Lower Confidence Limit and UCL=Upper Confidence Limit)
were reported in the original millisecond scale. Relevant pairwise comparisons were
conducted, using Fisher’s LSD.
There was significant evidence for a main effect of three design factors: Grouping
(p=0.0104), Symbol (p<0.0001) and Color (p<0.0001).

No interaction terms yielded

evidence of significant differences.
5.3.3.2.1 Significant main terms: Grouping, Symbol and Color
Participants took significantly less time to make a correct choice when the pieces
of critical information were grouped (LSM=4202.4ms, LCL=3637.5ms, UCL=4856.2ms,
when compared with those that were ungrouped (LSM=4407.6ms, LCL=3814.2,
UCL=5093.3; p<0.0104, see Figure 53).

145

Estimated Time to make a 
correct choice (milliseconds)

6000.0
5000.0

b

a

4000.0
3000.0

4202.4

4407.6

Grouped

Ungrouped

2000.0
1000.0
0.0

Figure 53. The effect of ‘Grouping’ on Time to make a correct choice: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters indicate
statistical significance at α=0.05.
Symbol use also positively impacted the time to correctly select a product (LSM=3820.3,
LCL=3306.7, UCL=4413.7; p<0.0001, see Figure 54).

146

b

Estimated Time to make a correct 
choice (milliseconds)

6000.0
5000.0

a

4000.0
3000.0
2000.0

4848.4
3820.3

1000.0
0.0
Symbol Presence

Symbol Absence

Figure 54. The effect of ‘Symbol’ on Time to make a correct choice: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters indicate
statistical significance at α=0.05.
Color also decreased time to correct selection of products (LSM=3922.8, LCL=3394.7,
UCL=4532.1; p<0.0001, see Figure 55).

Estimated Time to make a correct 
choice (milliseconds)

6000.0
5000.0

b
a

4000.0
4722.8

3000.0
3922.8
2000.0
1000.0
0.0
Color‐coded

Non‐color‐coded

Figure 55. The effect of ‘Color’ on Time to make a correct choice: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters indicate
statistical significance at α=0.05.
147

5.3.3.3 Pairwise comparisons between optimal (grouped + symbol present
+ color-coded) label and commercial labels
Since grouping, symbol presence and color-coding design indicated evidence of
a significant benefit on the time to make a correct choice, the combined design label of
these three factors (referred to as “optimal label”, see Figures 56 and 57) on the
probability of making a correct choice and the time to make a correct choice was
compared with two commercial labels (see Figures 43 and 44).
5.3.3.3.1 Binary Variable: Probability of correct choice (Yes/No)
The optimal label (LSM=97.3%, LCL=95.5%, UCL=98.4%) resulted in a
significant positive benefit regarding the probability of correct choice during the forced
choice task, as compared to the two commercial labels we tested (LSM=92.0%,
LCL=87.9%, UCL=94.7%, and LSM=89.8%, LCL=85.3%, UCL=93.0%; p’s<0.0001, see
Figure 56).
a
Estimated Probability of correct 
choice (%)

100.0%

b

97.3%

b

92.0%
89.8%

95.0%
90.0%
85.0%
80.0%
75.0%
Optimal label

Commerical A

Commerical B

Figure 56. The effect of ‘Grouped + Symbol presence + Color-coded’ design on
Probability of correct choice, compared to commercial label designs: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters indicate
statistical significance at α=0.05.
148

5.3.3.3.2 Continuous Variable: Time taken to make a correct choice
(milliseconds)
Time to correct selection was also positively impacted in our optimal design
(LSM=3525.3ms, LCL=3260.6ms, UCL=3811.5ms), as compared to the two commercial
labels we tested (LSM=8922.8ms, LCL=8105.9ms, UCL=9824.3ms and LSM=8260.4ms,
LCL=7497.2ms, UCL=9101.2ms: p’s<0.0001, see Figure 57).

Esimated Time to make a correct 
choice (milliseconds)

12000.0

b

10000.0

b

8000.0
8922.8
6000.0

8260.4

a
4000.0
2000.0

3525.3

0.0
Optimal label

Commerical A

Commerical B

Figure 57. The effect of ‘Grouped + Symbol presence + Color-coded’ design on Time
to make a correct choice, compared to commercial label designs: Estimated least
square means (LSM) with estimated upper and lower limits. Differing letters indicate
statistical significance at α=0.05.

5.4 Discussion
Like the change detection study, the sample population of this study was
comprised of an experienced pool of healthcare providers (average experience years:
21, ranging from 0 to 51).
Results from both experiments bolster findings reported by Cai’s study with
149

regard to what constitutes critical information (Table 20 and 33).

Both sets of

participants (n=86, change detection, n=89, forced choice task) reported expiration
dating, sterility status, latex status and product name to be within the top 5 most critical
pieces of information they use.

Further, by sampling at conferences of a national,

professional organization, it is not unreasonable to assume we were getting engaged
providers from across the nation.

In addition, the recommendations for improvement to

labeling design were also very similar to what Cai suggested: i.e. noticeable texts
employing bigger or bolder font, highlighting critical information in a single standard
location (Table 23 and 36).
Our objective was to evaluate the effect of four design factors (Grouping, Boxing,
Symbol and Color) on the three pieces of critical information during most stages of
information processing (i.e. attention, perception, comprehension).

It was expected

that these design factors would enhance the information processing tasks of
participants as indicated by a higher probability of correct choice and faster rates of
correct choices.
Two design factors suggested a main effect on the probability of correct choices:
Symbol (p<0.0001) and Boxing (p=0.0101).

In analyzing the dependent variable,

probability of correct choices, it was evident that symbol presence (LSM=97.7%) helped
participants to make a higher rate of correct choices than symbol absence (LSM=94.4%,
p<0.0001).

This result suggests that symbols accompanying their supplementary text

were helpful for participants to select correct devices. However, comprehension levels
of stand-alone symbols were quite poor at the symbol-alone comprehension testing.
By contrast, the boxed design (LSM=95.7%) resulted in a reduced probability of correct
150

choice, compared to the unboxed design (LSM=97.1%, p=0.0101).
When the time to make correct choices was analyzed, it clearly suggested that
the three design treatments: grouping, symbol presence and color-coding, enabled
participants to attend, perceive and comprehend the three pieces of critical information
significantly faster, compared to other design treatments (α=0.05, Figures 53, 54 and
55).

However, there was no evidence of a main effect of boxing on the time to make

correct choices (p=0.4450).
The findings regarding boxing are interesting. Specifically, the presence of a
box resulted in significantly reduced proportion of products chosen correctly, compared
to those that were not boxed (LSM=95.7%, LCL=92.5%, UCL=97.5% vs. LSM=97.1%,
LCL=94.9%, UCL=98.4%; p=0.0101).
the selection time was evident.

However, among correct choices, no benefit to

By contrast, in the change detection experiment,

changes were found faster in the boxed conditions (Figure 37), with designs that were
ungrouped and boxed generating faster times than any of the other 4 combinations of
boxing and grouping.
Perhaps most striking is the comparison of our theorized “optimal label”
(grouping, symbol presence and color-coding), compared to two commercial labels
(α=0.05, Figures 56 and 57).

The commercial labels were created based on a

synthesis of benchmarking studies comprised of 20 labels from manufacturers.

Our

optimal label increased the probability rate of correct choices and reduced the time to
make a correct choice, when compared to the two commercial labels.

Specifically,

participants responded correctly in trials testing the optimal labels in approximately half
the time, compared to the two commercial labels.
151

From a regulatory standpoint, required labeling information should be
prominently placed with appropriate conspicuousness (as compared with other words,
statements, designs, or devices, in the labeling) and should be likely to be read and
understood by the ordinary individual under customary conditions of purchase and use
(Federal Food, Drug and Cosmetic Act, Section 502).

Our study provides evidence

that improvements can be made with regard to correct selection and time to selection by
leveraging design insights that have been recommended by healthcare providers tested
here.

152

CHAPTER 6
CONCLUSIONS AND LIMITATIONS & FUTURE WORK
6.1 Conclusions
The efficacy of four design elements (Boxing, Grouping, Symbol and Color) was
assessed on the three pieces of critical information for enhanced attention and
comprehension.

Two experiments were conducted to evaluate the effectiveness of the

16 mixed treatments (2 x 2 x 2 x 2) of these four design factors using the change
detection and forced choice task methodologies.
During the change detection analysis, two elements emerged as having a
consistently positive impact on time to detect changes: boxing (Figure 37) and color
(Figure 38).

That is, participants responded significantly faster to changes when color

was present and when critical information was boxed.

However, grouping of the

information significantly affected response times in the presence of these other
conditions.

Responses to changes were significantly faster in the grouped condition

(LSM=1747.4ms, LCL=1564.9ms, UCL=1951.6ms) than those not grouped
(LSM=1843.7ms, LCL=1651.2ms, UCL=2059.2ms) when treatments were colored. By
contrast, ungrouped treatments resulted in faster response times (LSM=1740.2ms,
LCL=1558.5ms, UCL=1943.1ms) than grouped when boxes were present
(LSM=1925.8ms, LCL=1724.2ms, UCL=2150.3ms).
These findings become more intriguing when coupled with those obtained from
the forced choice experiment. Boxing proved to be a significant factor in making a
correct choice, resulting in a reduced probability to correct selection (95.7% for boxed
designs vs 97.1% for unboxed designs).

Taken in total, this may suggest that visually
153

salient factors that are not imbedded within the message itself may actually subvert
information processing.
By contrast, when design elements were imbedded within the message (i.e.
color and symbol), the effect was clear and positive.

When examining the four

possible combinations of color and grouping, colored/grouped information
(LSM=1747.4ms, LCL=1564.9ms, UCL=1951.6ms) outperformed all other treatments.
Colored, ungrouped treatments (LSM=1843.7ms, LCL=1651.2ms, UCL=2059.2ms)
resulted in significantly less time to detect changes than both the grouped
(LSM=2260.5ms, LCL=2024.0ms, UCL=2524.6ms) and ungrouped treatments
(LSM=2140.9ms, LCL=1917.3ms, UCL= 2391.1ms) with no color (see Figure 38).
Symbol behaved similarly to color, and the symbol present, grouped treatment
(LSM=1888.9ms, LCL=1691.2ms, UCL=2109.1ms) outperformed other symbol absent
combinations: both ungrouped (LSM=2012.8ms, LCL=1802.2ms, UCL=2248.0ms) and
grouped treatments (LSM=2091.7ms, LCL=1872.8ms, UCL=2336.1ms) without symbol
presence (see Figure 39).

Forced choice data also suggested clear benefits of these

elements. Symbol presence positively influenced correct selection (97.7% when
present versus 94.4% when absent; see Figure 52) and time to correct selection
(3820.3ms when present vs 4848.4ms when absent; see Figure 54).

A time to correct

selection advantage was evident in colored treatments (3922.8ms for colored
treatments vs 4722.8ms for non-colored treatments; see Figure 55).
In total, these findings lead us to theorize that visually salient design elements
that are imbedded within the message (i.e. symbol and color) excel at attracting
attention to the message being conveyed.

Although visually salient features that are
154

not imbedded, but in close proximity (i.e. the box) provide advantage in early stages of
information processing (i.e. attention), they have the potential to distract during the late
stages (as evidenced by the reduced probability of correct choice).
Thus, the grouped, symbol-present and color-coded format enabled participants
to choose a product that had latex, was sterile or was expired, at a significantly higher
rate of correct selection and faster than the two commercial labels tested in this study.
In addition, the comprehension level of recognized symbols in the
AAMI/ANSI/ISO 15223 was quite poor.

Three out of 38 symbols tested in this study

were critically confused and participants’ responses were opposite of the intended,
defined meaning of those symbols in the standard. The FDA rule currently proposed
regarding the stand-alone graphical representation should be carefully taken into
consideration before it is enacted.
6.2 Limitations & Future Study
Though boxing had a negative effect on correct choice (LSM=95.7%,
LCL=92.5%, UCL=97.5% vs. LSM=97.1%, LCL=94.9%, UCL=98.4%; p=0.0101) in the
forced choice experiment, it had a positive effect on time to detect change during
change detection testing (see Figure 37). Along with the visual salient design not
embedded within message, another assumed reason why this curious result occurred is
that the multiple box design (see Figure 41) highlighting all pieces of critical information
might not work as expected, if top-down processing predominated.

As such, future

work is recommended to test both a single box system to highlight targeted piece (s) of
critical information and the multiple box system for comparison purposes.
In the change detection and forced choice task experiments, locations of change
155

detection and forced choice were randomized to mitigate their effects.

Both

experiments suggested that there was evidence of a significant effect of location on the
dependent variables: time to detect change and time to correct choice.

Previous

research (Bix et al., 2010) reported the effect of locations on detection time. Thus, it is
recommended to analyze this research data so as to see behaviors of participants to
detect change at different image locations.
The change detection and forced choice task experiments were conducted on a
computer monitor to simulate a context of a medical device being used.

There could

be some difference of attentional behaviors of healthcare professionals between this
simulated environment and a real context in the use of medical devices.

In more

realistic environments, real context research is recommended to assess the effect of
Grouping, Symbol and Color coding.
In our symbol comprehension study, an open-ended test was used to assess
comprehension levels of medical device symbols.

Participants were asked to write the

meaning of medical device symbols on a response form. Some literature (Vukelich &
Whitaker, 1993 and Wolff & Wogalter, 1998) suggests that the presence of context
(depicting the probable environment where a symbol would be seen) enhances the level
of symbol comprehension. It would be worthwhile to evaluate symbol comprehension
in a context-based test form (e.g. symbols being embedded on a real medical device
label) in future research.

In addition, participants who were recruited for our symbol

comprehension study were generalists (e.g. surgical technologists, registered nurses,
etc.) in the healthcare industry.

Twenty-two out of the tested 38 symbols were

included in the FDA guidance: Use of symbols on labels and in labeling In-Vitro
156

Diagnostic devices intended for professional use (2004).

Some specialized symbols

(e.g. In Vitro Diagnostic medical device, Lower limit of temperature, Upper limit
temperature, Temperature limitation, etc.) used for In-Vitro Diagnostic devices may not
be commonly used in the environments where our participants work.

That said, results

were similar to those reported by Hermans, et al. (2011) who did tailor their test
population to a specific environment and also utilized context testing.

157

APPENDICS

158

APPENDIX 1. Proposed Rules of FDA on medical device labeling
1. “Unique Device Identification (UDI) System”
In July 2012, the US Food and Drug Administration (FDA) proposed a rule
requiring a Unique Device Identifier (UDI) for the vast majority of medical devices (FDA,
UDI Proposed Rule, 2012), and a final rule of the UDI was published in the Federal
Register, in September, 2013 (FDA, UDI Final Rule, 2013).

Its final rule requires a

device identifier and a production identifier to be provided in an easily readable, plaintext version and in a form that uses Advanced Identification Data Capture (AIDC)
technology on the label and package of a medical device (FDA, UDI Final rule, 2013).

• Device identifier: The specific version or model of a device and the labeler of
that device

• Production identifier: One or more of the followings should be present on the
label of a device:
9 The lot or batch within which a device was manufactured
9 The serial number of a specific device
9 The expiration date of a specific device
o Date format: YYYY-MM-DD (e.g., 2013-09-30)
9 The date a specific devices was manufactured
9 The distinct identification code required by 21 CFR 1271. 290 (c) for a
human cell tissue, or cellular and tissue-based products (HCT/P)
regulated as a device.

159

For certain categories of medical devices which may be used for an extended
periods of time or may become separated from their labeling (examples: implantable
devices, multiple use devices and devices sterilized before each use or stand-alone
software), direct marking of the UDI on the device itself is required (FDA, UDI Final rule,
2013).
FDA expects that several important public health benefits will be generated
through adequate identification of medical devices at the time of distribution and use
(FDA, UDI Proposed rule, 2012).

The key benefits anticipated to be achieved with the

implementation of a UDI system are listed in the following table (FDA, UDI, Proposed
rule, 2012).
Table 37. Key benefits through UDI implementation
Benefits
Reduce Medical Errors

Simplify the Integration of
Device Use Information
Into Data Systems
Provide for More Rapid
Identification of Medical
Devices With Adverse
Events
Provide for More Rapid
Development of Solutions
to Reported Problems
Provide for More Rapid,
More Efficient Resolution
of Device Recalls

Details
The presence of a UDI that is linked to device information in the
Global Unique Device Identification Database (GUDID)
database will facilitate rapid and accurate identification of a
device, thereby removing a cause of confusion that can lead to
inappropriate use of a device (e.g., confusion as to whether a
device is packaged as sterile, or failure to recognize that a
device is the subject of a recall or enforcement action).
UDIs, particularly when provided through AIDC technology,
would allow rapid and accurate data acquisition, recording, and
retrieval.
The inclusion of UDIs in adverse event reports would lead to
greater accuracy in reporting, by eliminating uncertainty
concerning the identity of the device that is the subject of a
report.
The inclusion of UDIs in adverse event reports would allow
manufacturers and FDA to more rapidly review, aggregate, and
analyze related reports regarding a particular device.
Delays in identifying recalled devices can result in the continued
use of those devices on patients and involves an increased risk
for patient harm. A device labeled with a UDI can be identified
rapidly and with great precision and the UDI, particularly when
combined with AIDC technology, will hasten the identification of
devices that are the subject of a recall.

160

Table 37. (Cont’d)
Better-Focused and More
Effective FDA Safety
Communication
Provide an EasilyAccessible Source of
Definitive Device
Identification Information
Standard Format for Dates
Provided on a Device
Label or Package

By citing UDIs, FDA would be able to more precisely focus
safety alerts, public health notifications, or other
communications, eliminating confusion with similar devices and
allowing more rapid responsive action.
The inclusion of device identifiers could allow the document to
focus on its important core messages without the distraction of
greater complexity, while a reader who wants those additional
details could use the UDI to obtain information from the GUDID.
The rule would also contribute to improved identification of
medical devices, and at the same time, better ensure the safe
use of devices, by requiring dates on medical device labels to
conform to a standard format to ensure dates are unambiguous
and clearly understood by device users.

161

2. “Use of Certain Symbols in Labeling” to allow for the inclusion of stand-alone
graphical representations of information, or symbols
The “Use of Certain Symbols in Labeling”, another proposed rule, would allow
the inclusion of stand-alone graphical representations of information, accompanied by a
symbols glossary, provided they are recognized standards, and the use of “Rx” only
without accompanying explanation text (FDA, Use of Certain Symbols, Proposed rule,
2013).

Based on current, general labeling requirements (CFR Title 21, Part 801),

graphics, pictures or symbols have to be accompanied by explanatory English text
adjacent to those graphical representations except for In Vitro Diagnostic Devices (IVD)
intended for professional use. The intention of the proposed rule is to make labeling
more user-friendly by replacing small, difficult-to-read text with pictorial information and
to harmonize the labeling requirements of U.S. and other regulatory bodies (FDA, Use
of Certain Symbols, Proposed rule, 2013). Incongruence in European and U.S.
requirements regarding symbol representation on medical device labeling is a
recognized issue.

In Europe, stand-alone graphical representations are used for

medical sold in multiple countries in order to avoid multiple languages on their label.

If

implemented, this regulatory difference would be harmonized to avoid the development
of different labels of a medical device which may be sold in US and Europe (FDA, Use
of Certain Symbols, Proposed rule, 2013).

162

APPENDIX 2. Misbranding (specified by section 502 of the Federal, Food, Drug
and Cosmetic Act)
Table 38. Misbranding items
No.
1
2
3
4
5
6
7
8
9
10
11

12
13
14
15

16
17

18
19
20
21
22
23

Contents
Remarks
False or misleading label
Labeling-related
Package form; contents of label
Labeling-related
Prominence of information on label
Labeling-related
Designation of drugs or devices by established Names
Labeling-related
Directions for use and warnings on label
Labeling-related
Representations as recognized drug; packing and
drug-related
labeling; inconsistent requirements for designation of drug
Deteriorative drugs; packing and labeling
drug-related
Drug; misleading container; imitation; offer for sale under
drug-related
another name
Health-endangering when used as prescribed
Labeling-related
Color additives; packing and labeling
Labeling-related
Prescription drug advertisements: established name;
drug-related
quantitative formula; side effects, contraindications, and
effectiveness; prior approval; false advertising; labeling;
construction of the Convention on Psychotropic
Substances
Drugs or devices from nonregistered establishments
Labeling-related
Packaging or labeling of drugs in violation of Regulations
drug-related
Restricted devices using false or misleading advertising
or used in violation of regulations
Restricted devices not carrying requisite accompanying
statements in advertisements and other descriptive
printed matter
Devices subject to performance standards not bearing
requisite labeling
Devices for which there has been a failure or refusal to
give required notification or to furnish required material or
information
Identification of manufacturer
Reprocessed single-use devices
New animal drugs
Nonprescription drugs
Drugs subject to approved risk evaluation and mitigation
strategy
Post-market studies and clinical trials; new safety
information in labeling
163

 

Labeling-related 
Labeling-related 

Labeling-related
drug-related
drug-related
drug-related
drug-related

APPENDIX 3. Latex glove manufacturing process
Table 39. Latex glove manufacturing process
(Zalglaniczny, 2001; Yunginger, 1998)
Process
Description
Natural rubber latex
Natural-latex containing protein is harvested from
harvesting
Hevea brasilienis rubber tree.
Collection with ammonia
Autocoagulation of natural latex is prevented by the
addition of ammonia.
Concentration from 30% Natural latex is centrifuged and concentrated from 30%
60% solid
to 60% solids. Removal of serum phase reduces the
concentration of water soluble proteins.
Compounding
Processing and attributes of the finished device depend
on the addition of many chemicals to the natural latex
(compounding). Significant Type IV allergens include
the accelerators and antioxidants.
Brush or ultrasonic former Porcelain formers attached to a continuous chain are
cleaning
cleaned to remove debris to a previous cycle.
Coagulant dipping
Formers are dipped in an emulsion to apply corn starch
as a releasing agent and a compound that coagulates
liquid natural latex on contact.
Agent drying
Releasing agent and coagulant are oven-dried.
Latex dipping
Formers dip into natural latex and a uniform film is
deposited.
Oven heating
The coagulant and heat convert the natural latex from
liquid to solid.
Bead rolling
Rotating brushes contact the rotating formers and a
cuff is rolled onto the glove.
Leaching in water tank
Formers pass through water baths to remove watersoluble protein and excess additives.
Vulcanization in oven
Cross-linking of the polyisoprene polymers is catalyzed
by heat and requires an accelerator.
Application of corn starch
Corn starch is applied as slurry to the outer surface of
power
the natural rubber latex as a detacking agent.
Residual rubber proteins may elute from the gloves at
this point and bind to the corn starch particles.
Stripping
The gloves are stripped from the porcelain formers.

164

APPENDIX 4. Stimulus materials for Legibility test
Brand A
Brand name/Product name/Latex (Location 1)

Sterile (Location 2)

Sub-product name/Expiration dating (Location 4)

Figure 58. Legibility stimulus materials for Brand A
165

Brand B
Brand name/Sterile(text) (Location 1)

Product name/Sub-product name/Latex/Sterile(symbol)/Expiration dating
(Location 4)

Figure 59. Legibility stimulus materials for Brand B

166

Brand C
Brand name/Product
name/Sterile/Latex
(Location 1)

Sub-product name/Expiration dating
(Location 4)

Brand D
Sub-product name/Expiration dating
(Location 4)

Brand name/Product
name/Sterile/Latex
(Location 1)

Figure 60. Legibility stimulus materials for Brand C and D

167

Brand E
Brand name/Product name (Location 1)

Sterile/Latex (Location 3)

Sub-product name/Expiration dating (Location 4)

Figure 61. Legibility stimulus materials for Brand E
168

Brand F
Brand name/Product name (Location 1)

Sterile (Symbol)/Latex (Location 4)

Sub-product name/Expiration dating (Front side)

Figure 62. Legibility stimulus materials for Brand F
169

APPENDIX 5. Data collection sheet: Legibility test

Research Questionnaire/Data collection Form 
Legibility of Medical Device Labels 
 

 

 

 

 

 

 

 

Subject #:    ________     

A. Data collection sheet 
Subject #:                                          Gender:                                            Age:                                                       
Health Literacy:                                                    Color differentiation ability:                         
Inside Light Level:                      Ambient Light Level:                        Visual Acuity:   

B. Questionnaire 
1. What is your gender? 
Female                                        Male    ________             
 
2. What is your age? 
        ______________
3. What is your ethnicity? 
 
American Indian/Alaskan Native_______, Asian or Pacific Islander_______ 
Black, non‐Hispanic _______,   
Hispanic                            _______ 
White, non‐Hispanic_______,       
Other                                  _______ 
 
4. What is the highest level of education you achieved? 
Associate’s Degree      _______,   
Bachelor’s Degree _______ 
 
Master’s Degree        _______,                Doctor Degree      ________     
 
Other            ______________________________                                                                                     
 
5. What is your native language?     
________                  _____

170

Thanks for your effort to fill out this questionnaire form above.
page will be filled out by the researcher of this study.

The next

Table 40. Legibility data sheet
 
Run order 

Test Stimulus 

1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
29 
30 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Labeling 
Information 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 

171

Required 
Degrees of 
Rotation 

Distance (Wall to 
Easel) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

APPEND
DIX 6. Data
a collection sheet: Comprehen
nsion test

Data Collection Sheet 
Comprehensio
on level on medical deviice symb
bols                 
Subject #
#:                                            Genderr: 

            Age::                                                   

Health Literacy:                                                            Col or differentiation abilityy:                         
Visual Accuity:   
Instructtions: 
This study is intended to evaluaate your com
mprehensionn level on meedical devicee symbols ussed 
for com
mmercially avvailable med
dical devices.    Please vieew the medical device ssymbols 
presentted in the folllowing table
e, and fill in the box of ““Answer” witth a meaning of each 
symbol which you think.    If you are not ab
ble to assign a meaning o
of a symbol (s), you can 
simply w
write “Don’t know” in th
he answer bo
ox.    But, it is highly recommended to write 
whateve
er you think each symbo
ol means. 

Symbo
ol 1: 

Symbo
ol 2:

 

 
Answer:       

 

 
Answeer: 

 

 

Symbo
ol 3: 

Symbo
ol 4:

 
Answer:       

 

 

 
Answeer: 
 

 

Figu
ure 63. Me
edical devicce symbolss
172

Figure
e 63. (Contt’d)
ol 6:
Symbo

Symbo
ol 5: 

 
Answer:       

 

 

 
Answeer: 

 

 

Symbo
ol 7: 

Symbo
ol 8:

 

 
Answer:       

 

 
Answeer: 

 

 

Symbo
ol 9: 

Symbo
ol 10:

 
Answer:       

 

 

 
Answeer: 
 

 

 
 
173

Figure
e 63. (Contt’d)
 
 
 
ol 11: 
Symbo

 
 
ol 12: 
Symbo

 
Answer:       

 

 

 
Answeer: 

 

 

Symbo
ol 13: 

Symbo
ol 14:
 

 
 
 
Answer:     
 

 

 

 
 

 
Answeer: 
 

Symbo
ol 15: 
 

Symbo
ol 16:
 

 
Answer:       

 

 

 
Answeer: 
 

 

 
174

Figure
e 63. (Contt’d)
ol 18:
Symbo
 

Symbo
ol 17: 
 
 

 
Answer:       

 

 
Answeer: 

 

 

Symbo
ol 19: 
 

Symbo
ol 20:
 

 
Answer:       

 

 

 
Answeer: 
 

 

Symbo
ol 21: 
 

Symbo
ol 22:
 

 
Answer:       

 

 

 
Answeer: 

 

 

Symbo
ol 23: 
 

Symbo
ol 24:
 

 
Answer:       

 

 

 
Answeer: 

 

 

 
175

Figure
e 63. (Contt’d)
ol 26:
Symbo

Symbo
ol 25: 

 

 
Answer:       

 

 

 
Answeer: 
 

 

Symbo
ol 27: 

Symbo
ol 28:
 

 

 
Answer:       

 

 
 

 
Answeer: 
 

 

Symbo
ol 29: 

Symbo
ol 30:

 
Answer:       

 

 

 
Answeer: 

 

 

 
 
176

Figure
e 63. (Contt’d)
ol 32:
Symbo

Symbo
ol 31: 

 
 
Answer:       

 

 

 

 
Answeer: 
 

 

Symbo
ol 33: 

Symbo
ol 34:

 
 
Answer:       

 
 

 
Answeer: 
 

 

Symbo
ol 35: 

Symbo
ol 36:

 
Answer:       

 

 

 
Answeer 

 

 

 
177

Figure
e 63. (Contt’d) 
Symbo
ol 38:

Symbo
ol 37: 

 

 
 
Answer 

 

 
 
Answeer 

 

 
 
 
 
Symbo
ol 39: 
 

 

Symbo
ol 40:
 
 

 

 

 
 

 

 

 
Answer 

Answeer: 

 

 

 

Symbo
ol 41: 
 

 
Answer 

 
 
 

178

 

APPENDIX 7. Research questionnaire form: Change Detection/Forced
Choice Task tests

Research Questionnaire Form 
 

 

 

 

 

 

 

 

Subject #:    ________     

A. Demographic Survey 
1. What is your gender? 
Female                                        Male    ________             
 
2. What is your age? 
        ______________
3. What is your ethnicity? 
American Indian/Alaskan Native_______, Asian or Pacific Islander_______ 
Black, non‐Hispanic_______,   
  Hispanic                          _______ 
White, non‐Hispanic _______,       
  Other                              _______ 
 
4. What is the highest level of education you achieved? 
  Associate’s Degree    _______,   
Bachelor’s Degree _______ 
 
Advanced Practice Nurse certificate    _______, Master’s Degree        _______
 
Doctor Degree_______,       
                            Other                        _______ 
 
5. What is your native language?     
________                  _____

B. Job‐related Survey 
6. What is your current employment setting? 
 
Acute Care Hospital _______,   
Physician’s office _______ 
              
Ambulatory/Day Surgery Center_______  
 
 
 
              
Nursing home/long‐term care facility _______   
 
        
Public/community health   _______                                                                                                       
Student ________ _,          Other ___________________________________         

179

7. What is your position and role in your employment? 
________                  _________________________________________
 
8. How many years have you been working in the healthcare industry? 
        ______________
 
9. What labeling information do you care about on packages of medical devices? (Mark 
at least 5 items) 
1) Brand name    _________            2) Product name      ___________ 
3) Latex              __________            4) Sterile                    __________ 
5) Use instruction__________          6) Expiration dating _________ 
7) Serial # __________                      8) Batch code                        __________ 
9) Manufacturer name ________                      10) Do not reuse            __________ 
11) Do not resterilize        ________                  12) Date of manufacture ______ 
13) Do not use if package is opened or damaged _________                                                     
10.   What is the top 4 labeling information you really care about before using medical 
devices among the marked items at question 9? 
1) 
______________________________________________________ 
2)   
______________________________________________________ 
3)   
______________________________________________________ 
4)                    ______________________________________________________ 
 
11. What problems do you have to capture the labeling information of medical device 
packages easily? (e.g. small font size, no color contrast, no color coding, no standard 
location for critical information, etc.) 
1) 
______________________________________________________ 
2)   
______________________________________________________ 
3)   
______________________________________________________ 
4)            ______________________________________________________ 
5)            ______________________________________________________ 
 
180

12. What medical errors have you experienced because of a labeling issue(s) of medical 
devices? 
1) 
______________________________________________________ 
2)   
______________________________________________________ 
3)   
______________________________________________________ 
4)            ______________________________________________________ 
5)            ______________________________________________________ 
 
13. What would be your recommendations to resolve a problem(s) you have in capturing 
the labeling information of medical device packages? (e.g. Bigger font size, color 
contrast, color coding, standard location for critical information, etc.) 
______________________________________________________ 
______________________________________________________ 
______________________________________________________ 
______________________________________________________ 
______________________________________________________ 
 
 
 

181

APPEND
DIX 8. Con
nsent form: Legibility
y test

182

183

184

APPEND
DIX 9. Con
nsent form: Change Detection/C
D
Comprehension tests
s

185

186

187

APPEND
DIX 10. Co
onsent form
m: Forced Choice
C
Tas
sk test

188

189

190

APPEND
DIX 11. Recruitment flyer: Legiibility test

Figure 64. Recruiitment flyerr: Legibility test
191

APPEND
DIX 12. Recruitment flyer: Change Detecttion/Comp
prehension
n tests

Figure 65
5. Recruitment flyer: Change
C
Dete
ection/Com
mprehension
n tests
192

APPEND
DIX 13. Recruitment flyer: Forc
ced Choice
e Task testt

Fiigure 66. Recruitment
R
t flyer: Forcced Choice Task test
193

APPENDIX 14. Critical pieces of labeling information (Change Detection)

Table 41. Critical pieces of labeling information (Change Detection)

1 (Most important)
2
3
4 (Least important)
Total Frequency of
Participant Responses
Median of Ranking
Mode of Ranking

1 (Most important)
2
3
4 (Least important)
Total Frequency of
participant responses
Median of Ranking
Mode of Ranking

Expiration
dating
31
27
14
7

Sterility
status
19
15
16
12

79

Latex status

Product Name

Use Instruction

Brand Name

Serial #

11
19
18
11

13
6
14
11

1
9
7
7

0
2
2
6

1
2
1
4

62

59

44

24

10

8

2
1

2
1

2
2

3
3

3
2

4
4

3.5
4

Date of
Manufacture

Do not
resterilize

Name

Manufacturer
Name

Do not use if
package is
opened or
damaged

Batch Code

Other

0
1
3
4

0
0
1
5

3
0
1
2

0
0
3
0

0
0
1
1

0
0
1
0

2
1
0
3

8

6

6

3

2

1

6

3.5
4

4
4

2
1

3
3

3.5
3,4

3
3

3
4

194

APPENDIX 15. Critical labeling problems (Change Detection)
Table 42. Critical labeling problems (Change Detection)

44
21
6
4

No standard location for
critical information
18
21
15
4

Labeling designs not
standardized
8
11
20
10

75

58

1
1

Small font size
1 (Most important)
2
3
4 (Least important)
Total Frequency of
Participant Responses
Median of Ranking
Mode of Ranking

1 (Most important)
2
3
4 (Least important)
Total Frequency of
participant responses
Median of Ranking
Mode of Ranking

No color coding

No color contrast

5
14
11
4

7
9
9
4

49

34

29

2
2

3
3

2
2

2
2,3

Hard to get
expiration dating
information easily

Hard to get latex status
information easily

Hard to get sterile
statue information
easily

Hard to
comprehend
symbol meaning
easily

Other

2
2
3
3

0
3
0
0

0
0
2
1

0
1
0
0

0
1
2
4

10

3

3

1

7

3
3,4

2
2

3
3

2
2

4
4

195

APPENDIX 16. Critical pieces of labeling information (Forced Choice Task)
Table 43. Critical pieces of labeling information (Forced Choice Task)

1 (Most important)
2
3
4 (Least important)
Total Frequency of
Participant Responses
Median of Ranking
Mode of Ranking

1 (Most important)
2
3
4 (Least important)
Total Frequency of
participant responses
Median of Ranking
Mode of Ranking

Expiration
dating
22
25
20
11

10
17
15
17

Product
Name
18
9
10
5

Use
Instruction
4
7
7
8

Do not reuse

Name

1
2
1
5

4
0
1
4

Do not
resterilize
0
0
3
5

59

59

42

26

9

9

8

2
2

2
1,2

3
2,4

2
1

3
4

4
4

3
1,4

4
4

Serial #

Brand
Name

Date of
Manufacture

Batch Code

Manufacturer
Name

Do not use if
package is
opened or
damaged

Other

0
0
2
5

3
0
1
2

0
0
2
4

1
0
1
3

0
0
3
2

0
0
0
4

3
6
3
3

7

6

6

5

5

4

15

4
4

2
1

4
4

4
4

3
3

4
4

2
2

Sterile

Latex

20
20
15
4

78

196

APPENDIX 17. Critical labeling problems (Forced Choice Task)
Table 44. Critical labeling problems (Forced Choice Task)

47
20
10
3

No standard location for
critical information
16
14
19
5

80

54

41

33

18

1
1

2
3

2
2

2
2

3
2,3

Hard to get
expiration dating
information easily

Hard to get latex status
information easily

Hard to comprehend
symbol meaning easily

Hard to get sterile
statue information
easily

Other

1
6
4
3

1
0
0
3

0
1
2
0

0
0
0
2

8
4
6
8

14

4

3

2

26

3
2

4
4

3
3

4
4

3
1,4

Small font size
1 (Most important)
2
3
4 (Least important)
Total Frequency of
Participant Responses
Median of Ranking
Mode of Ranking

1 (Most important)
2
3
4 (Least important)
Total Frequency of
participant responses
Median of Ranking
Mode of Ranking

197

No color contrast

No color coding

9
17
12
3

5
15
9
4

Labeling designs
not standardized
1
6
6
5

A
APPENDIX 18
8. Benchmarking study labels
Sample 1

Sample 2

Sample 3

Figure 67
7. Benchmarking study labels
198

Figure 67. (Cont’d))
Sample 4

Sample 5

Sample 6

199

Figure 67. (Cont’d)
Sample 7

Sample 8

Sample 9

200

Figure 67. (Cont’d)
Sample 10

Sample 11

Sample 12

201

Figure 67. (Cont’d)
Sample 13

Sample 14

Sample 15

202

Figure 67. (Cont’d)
Sample 16

Sample 17

Sample 18

203

Figure 67. (Cont’d)
Sample 19

Sample 20

204

APPENDIX 19. Placement of critical information
Table 45. Placement of critical information
No.

Product
name

Sub-product
name

Latex

Sterile

Expiration
dating

1

Location 1

Location 4

NA*

Location 2

Location 4

2

Location 1 

Location 4 

Location 1 

Location 2 

Location 4 

3

Location 1 

Location 4 

Location 1 

Location 2 

Location 4 

4

Location 1 

Location 4 

Location 2 

Location 2 

Location 4 

5

Location 1 

Location 4 

Location 1 

Location 2 

Location 4 

6

Location 1 

Location 4 

Location 1 

Location 2 

Location 4 

7

Location 1 

Location 4 

Location 1 

Location 2 

Location 4 

8

Location 1 

Location 4 

Location 1 

Location 2 

Location 4 

9

Location 1 

Location 4 

NA* 

Location 2 

NA*

10

Location 4

Location 4

Location 4

Location 1** 

Location 4 

11

Location 4 

Location 4 

Location 4 

Location 1** 

Location 4 

12

Location 4 

Location 4 

Location 4 

Location 1** 

Location 4 

13

Location 4 

Location 4 

Location 4 

Location 1** 

Location 4 

14

Location 4 

Location 4 

Location 4 

Location 1** 

Location 4 

15

Location 4 

Location 4 

Location 4 

Location 1** 

Location 4 

16

Location 1

Location 4

Location 1 

Location 1 

Location 4

17

Location 1

Location 4

Location 2

Location 2

Location 4

18

Location 1

Location 4

Location 3

Location 3

Location 4

19

Location 1

Location 4

Location 1

Location 1

Location 4

20

Location 1

Front side

Location 3

Location 3

Front side

* Related information was not available.
** Different locations for texts and symbol: Texts were placed in location 1 and symbol, in
location 4.

205

APPENDIX 20. Evaluation of text size of critical information
Unit: mm
Table 46. Text size of critical information
No.

Product name

Latex

Sterile

Expiration dating

1

5.162

NA*

1.260

2.248

2

2.182

1.222

1.264

2.234

3

5.134

1.268

1.242

2.202

4

5.032

1.196

1.272

2.222

5

5.182

1.324

1.176

2.170

6

5.130

Symbol only

1.308

2.606

7

5.114

1.174

1.196

2.168

8

5.212

1.420

1.272

2.170

9

5.162

NA*

1.266

NA*

10

2.278

1.474

1.642

2.770

11

2.326

1.870

1.602

2.628

12

2.148

1.572

1.58

2.744

13

2.394

1.516

1.546

2.872

14

2.334

Symbol only

1.642

2.218

15

1.370

Symbol only

1.608

2.022

16

1.420

1.506

2.180

17

3.628
4.362

2.000

Symbol only

1.644

18

5.490

2.010

1.740

2.262

19

6.072

1.650

1.656

1.736

20

3.098

1.136

Symbol only

1.142

3.941

1.483

1.432

2.223

6.072

2.010

1.740

2.872

1.370

1.136

1.176

1.142

Avg
.
Max
.
Min.

* Related information was not available.

206

APPENDIX 21. Evaluation of text leading of critical information
Unit: mm
Table 47. Text leading of critical information
No.

Product name

Latex

Sterile

Expiration dating

1

11.168

NA

NA

NA

2

NA

2.470

2.540

NA

3

11.170

NA

NA

NA

4

11.228

2.552

NA

NA

5

11.220

NA

NA

NA

6

11.306

NA

NA

NA

7

11.248

NA

NA

NA

8

NA

NA

NA

NA

9

7.898

NA

NA

NA

10

3.004

2.554

2.496

NA

11

3.968

2.618

2.454

NA

12

2.738

3.276

2.356

NA

13

3.228

2.49

2.392

NA

14

NA

NA

2.452

NA

15

2.858

NA

2.422

NA

16

5.958

NA

NA

NA

17

NA

NA

NA

NA

18

NA

3.330

2.914

NA

19

NA

3.638

NA

NA

20

5.704

2.018

NA

NA

Avg.

7.335

2.772

2.503

NA

Max.

11.306

3.638

2.914

NA

Min.

2.738

2.018

2.356

NA

NA: Leading for single line statement or symbol only was not available, or related information
did exist.

207

APPENDIX 22. Evaluation of text color contrast of critical information

Table 48. Text color contrast of critical information
No.

Product name

Latex

Sterile

Expiration dating

1

Green/White

NA*

Green/White

Black/White

2

Black/White

Black/White

Black/White

Black/White

3

Green/White

Green/White

Green/White

Black/White

4

Green/White

Green/White

Green/White

Black/White

5

Green/White

Green/White

Green/White

Black/White

6

Green/White

Symbol only

Green/White

Black/White

7

Green/White

Green/White

Green/White

Black/White

8

Green/White

Green/White

Green/White

Black/White

9

Blue/White

NA*

Blue/White

NA*

10

Black/White

Black/White

Blue/White

Black/White

11

Black/White

Black/White

Blue/White

Black/White

12

Black/White

Black/White

Blue/White

Black/White

13

Black/White

Black/White

Blue/White

Black/White

14

Black/White

Symbol only*

Blue/White

Black/White

15

Black/White

Symbol only*

Blue/White

Black/White

16

Blue/White

Blue/White

Blue/White

Blue/White

17

White/Blue

Blue/White

Symbol only*

Blue/White

18

White/Blue

Blue/White

Blue/White

Black/White

19

Green/White

Blue/White

Blue/White

Blue/White

20

Blue/White

Blue/White

Symbol only*

Black/White

* Related information was not available.

208

APPENDIX 23. Evaluation of symbol size of critical information
Unit: mm
Table 49. Symbol size of critical information
No.

Latex

Sterile

Expiration dating

Width

Height

Width

Height

Width

Height

1

NA

NA

NA

NA

3.274

4.432

2

9.154

8.280

19.310

4.456

3.344

4.600

3

NA

NA

NA

NA

3.282

4.450

4

NA

NA

17.316

3.622

3.328

4.430

5

NA

NA

NA

NA

3.252

4.338

14.630

3.428

2.954

2.954

6

Non-standard symbol

7

NA

NA

NA

NA

3.266

4.478

8

NA

NA

NA

NA

3.350

4.374

9

NA

NA

NA

NA

NA

NA

10

NA

NA

17.656

4.068

4.048

4.592

11

NA

NA

17.634

5.114

3.636

5.668

12

NA

NA

16.618

4.46

4.278

5.850

13

NA

NA

16.916

5.548

4.038

5.506

14

Non-standard symbol

16.710

4.244

2.110

3.304

15

Non-standard symbol

14.860

3.150

4.496

NA

NA

3.020
NA

30.398

6.58

16
17

NA

NA

Non-standard symbol

NA

5.224
5.038
Non-standard symbol

18

NA

NA

28.856

6.77

19

NA

NA

NA

NA

NA

NA

20

8.446

7.158

16.130

4.534

3.036

5.158

Avg.

8.800

7.719

18.920

4.665

3.453

4.697

Max.

9.154

8.280

30.398

6.770

5.038

5.850

Min.

8.446

7.158

14.630

3.150

2.110

3.304

NA: Related symbol was not available.
A bolded dimension is a dimension of the square lines to contain each symbol.

209

APPENDIX 24. Evaluation of symbol color contrast of critical information
Table 50. Symbol color contrast of critical information
No.

Latex

Sterile

Expiration dating

1

NA

NA

Black/White

2

Black/White

Black/White

Black/White

3

NA

NA

Black/White

4

NA

Green/White

Black/White

5

NA

NA

Black/White

6

Green/White

Green/White

Black/White

7

NA

NA

Black/White

8

NA

NA

Black/White

9

NA

NA

NA

10

NA

Black/White

Black/White

11

NA

Black/White

Black/White

12

NA

Black/White

Black/White

13

NA

Black/White

Black/White

14

Black/White

Black/White

Black/White

15

Black/White

Black/White

Black/White

16

NA

NA

NA

17

Blue/White

Blue/White

Blue/White

18

NA

Blue/White

Black/White

19

NA

NA

NA

20

Blue/White

Blue/White

Black/White

NA: Related symbol was not available

210

APPENDIX 25. Evaluation of originating symbol standard of critical information

Table 51. Originating symbols standard of critical information
No.

Latex

Sterile

Expiration dating

1

NA

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

3

NA
ANSI/AAMI/ISO15223-1
NA

NA

ANSI/AAMI/ISO-15223-1

4

NA

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

5

NA

NA

ANSI/AAMI/ISO-15223-1

6

Non-standard

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

7

NA

NA

ANSI/AAMI/ISO-15223-1

8

NA

NA

ANSI/AAMI/ISO-15223-1

9

NA

NA

NA

10

NA

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

11

NA

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

12

NA

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

13

NA

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

14

Non-standard

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

15

Non-standard

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

16

NA

NA

NA

17

Non-standard

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

18

NA

ANSI/AAMI/ISO-15223-1

Non-standard

19

NA
ANSI/AAMI/ISO15223-1

NA

NA

ANSI/AAMI/ISO-15223-1

ANSI/AAMI/ISO-15223-1

2

20

NA: Related symbol was not available.

211

APPENDIX 26. Evaluation of presence/absence of symbols for critical information
Table 52. Presence/absence of symbols for critical information
No.

Latex

Sterile

Expiration dating

1

No symbol/No text

Text only

2

Symbol/Text

Symbol/Text

Symbol/Text
Symbol/Text 

3

Text only

Text only

Symbol/Text 

4

Text only

Symbol/Text

Symbol/Text 

5

Text only

Text only

Symbol/Text 

6

Symbol only

Symbol/Text

Symbol/Text 

7

Text only

Text only

Symbol/Text 

8

Text only

Text only

Symbol/Text 

9

Text only

10

No symbol/No text
Text only

11

Text only

Symbol/Text
Symbol/Text 

No symbol/No text
Symbol/Text 

12

Text only

Symbol/Text 

Symbol/Text 

13

Text only

Symbol/Text 

Symbol/Text 

14

Symbol only

Symbol/Text 

Symbol/Text 

15

Symbol only

Symbol/Text 

Symbol/Text 

16

Text only

Text only 

Text only

17

Symbol/Text

Symbol only

Symbol/Text

18

Text only

Symbol/Text

Symbol/Text

19

Text only

Text only

Text only

20

Symbol/Text

Symbol only

Symbol/Text

212

Symbol/Text 

BIBLIOGRAPHY

213

BIBLIOGRAPHY

ANSI/AAMI/ISO 15223-1:2007/A1: 2008: Medical devices – Symbols to be used with
medical device labels, labeling, and information to be supplied – Part 1: General
requirements ISO 9186 - 1 2007: Graphical Symbols – Test methods – Part 1:
Method for testing comprehensibility.
ANSI Z535. 1-5 2011: Criteria for Safety Symbols.
Association of periOperative Registered Nurses (AORN Inc.) (2010a). Recommended
practices for hand hygiene in the perioperative setting. Perioperative
Standards and Recommended Practices. Denver, CO: AORN, Inc.: 75-89.
Association of periOperative Registered Nurses (AORN Inc.) (2010b). Recommended
practices for maintaining a sterile field. Perioperative Standards and
Recommended Practices. Denver, CO: AORN, Inc.: 91-99.
Association of Surgical Technologists (AST) (2011). Standards for Practices for
Creating the Sterile Field. [Cited August 4th 2014]. Available from:
http://www.ast.org/uploadedFiles/Main_Site/Content/About_Us/Standard_Creati
ng_Sterile_Field.pdf
ASTM 7298-06: 2011: Standard Test Method of Comparative Legibility by Means of
a Polarizing Filter.
Alwilda, Scholler-Jaquish, Cynthia, J. Weiss-Kaffie & Alyce, Ashcraft. (2003). The Latex
Allergic, Psychiatric Patient. Journal of Psychosocial Nursing, Vol 41:
No. 2: 31 – 36.
Base III, P. F., Wilson, J. F., & Griffin, C. H. (2003). A shortened instrument for literacy
screening. J Gen Intern Med, 181: 1036-1038.
Becker, D., Heinrich, J., von Sichowsky, J. & Wendt, D. (1970). Reader Preferences for
Typeface and Leading. The Journal of Typographic Research, Vol. 4. (1): 61-66.
Bix, L. (1998). The effect of subject age on legibility. Master thesis. Michigan State
University.
Bix, L. (2001). Toward a performance standard for typeface legibility: the Lockhart
Legibility Instrument. Ph. D dissertation. Michigan State University.

214

Bix, L., Lockhart, H., Selke, S., Cardoso, F., & Olejnik, M. (2003). Is x-height a better
indicator of legibility than type size for drug labels? Packaging Technology and
Science, 16(5): 199–207.
Bix, L., Bello, N. M., Auras, R., Ranger, J., & Lapinski, M. K. (2009). Examining the
Conspicuousness and Prominence of Two Required Warnings on OTC Pain
Relievers. Proceedings of the National Academy of Sciences of the United States
of America, 106(16): 6550–6555.
Bix, L., Kosugi, W., Bello, N., Sundar, R., & Becker, M. (2010). The use of change
detection as a method of objectively evaluating labels. Packaging Technology
and Science, 23(7): 393–401.
 

Bix, L., Seo, W., & Sundar, R. P. (2013). The Effect of Colour Contrast on Consumers’
Attentive Behaviors and Perception of Fresh Produce. Packaging Technology
and Science, 26(2): 96–104.
Borgmeier, I., & Westenhoefer, J. (2009). Impact of different food label formats on
healthiness evaluation and food choice of consumers: a randomized-controlled
study. BMC Public Health, 9 (184).
Bowyer, Rebecca V. ST. L. (1998). Latex allergy: how to identify it and the people at risk.
Journal of Clinical Nursing, 8: 144-149.
Brantley, K. A., & Wogalter, M. S. (1999). Oral and Written Symbol Comprehension
Testing: The Benefits of Cognitive Interview Probing. Proceedings of the Human
Factors and Ergonomics Society 43rd Annual Meeting: 1060-1064.
Buswell, G.T. (1935). How People Look at Pictures: A study of the psychology of
perception in Art. Chicago: University of Chicago Press.
Butschli, J. (2008). Surgical nurses survey medical device packaging.
[Cited August 16th 2013]. Available from:
http://www.healthcarepackaging.com/applications/healthcare/
surgical-nurses-survey-medical-device-packaging
Cai, J. (2012). Perceptions of medical device packaging used by operating room
personnel. Master thesis. Michigan State University
Centers for Disease Control and Prevention (CDC) (2009). The direct medical costs of
healthcare-associated infections in U.S. Hospitals and the benefits of prevention.
[Cited September 17th 2013] Available from: http://www.cdc.gov/hai/burden.html
Centers for Disease Control and Prevention (CDC) (2012a). 2011 National and State
healthcare-associated infections standardized infection ratio report. [Cited
215

September 17th 2013] Available from: http://www.cdc.gov/hai/national-annualsir/index.html
Centers for Disease Control and Prevention (CDC) (2012b). Health-care associated
infections (HAIs). [Cited November 30th 2012]; Available from:
http://www.cdc.gov/hai/
 

Cheng, L., & Lee, D. (1999). Review of latex allergy. J Am Board Fam Pract, 1999;
12(4): 285-292.
 

CFR Title 21, Part 801. General Labeling Provisions [Cited February 23rd 2013];
Available from
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?CFRPar
t=801
CFR Title 21, Part 801.1. Medical devices; name and place of business of
manufacturers, packer or distributor. [Cited February 23rd 2013]; Available from 
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/cfrsearch.cfm?cfrpart=80
1&showfr=1
CFR Title 21, Part 801.4. Meaning of intended uses. [Cited February 23rd 2013];
Available from
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=801.4
CFR Title 21, Part 801.5. Medical devices; adequate directions for use. [Cited February
23rd]; Available from
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=801.5
CFR Title 21, Part 801.6. Medical devices; misleading statements. [Cited February 23rd
2013]; Available from
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=801.6
CFR Title 21, Part 801.15. Medical devices; prominence of required label statements.
[Cited February 23rd 2013]; Available from
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/
CFRSearch.cfm?fr=801.15
CFR Title 21, Part 801.16. Medical devices; Spanish-language version of certain
required statements. [Cited February 23rd 3013]; Available from
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/
CFRSearch.cfm?fr=801.16
CFR Title 21, Part 801.437. User labeling for devices that contain natural rubber.
216

[Cited February 23rd 2013]; Available from
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=801.4
37#.T8zlRahNAgc.email
CFR Title 21, Part 809. In vitro diagnostic products for human use. [Cited February 23rd
2013]; Available from
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/
CFRSearch.cfm?CFRPart=809&showFR=1
CFR Title 21, Part 820. Quality system regulations. [Cited February 23rd 2013];
Available from
http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/Overview/
DeviceLabeling/QualitySystemRegulationLabelingRequirements/
default.htm#sterile
CFR Title 21, Part 809 Invitro diagnostic products for human use. [Cited February 23rd
2013]; Available from
http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?CFRPar
t=809
Craig, J. (1980). Designing with Type: A basic course in typography. New York:
Watson-Guptill Publications.
Davies, S., Haines, H., & Wilson, J. (1998). Safety pictograms: Are they getting the
message across? Applied ergonomics, 29(1): 15-23.
DeHenau, C. J. (2010). Applying change detection to test the noticeability of
components of medical labels. Master thesis. Michigan State University.
Dejoy, D. M. (1991). A revised model of the warnings process derived from valueexpectancy theory. In Proceedings of the Human Factors Society 35th Annual
Meeting: 1043-1047.
de la Fuente, C. J. (2013). Usability of tabs in semi-rigid packaging. Ph. D dissertation.
Michigan State University.
 

Dillard, S. F. & MacCollum, M. A. (1992). Reports to FDA: allergic reactions to latex
containing medical devices. International Latex Conference: Sensitivity to Latex
in Medical devices. Baltimore, MD, 23, cited in Steelman, V.M. (1995) Latex
allergy precautions: a research based protocol. Nursing Clinics of North America,
30(3): 479.
EN 980: 2008: Symbols for use in the labeling of medical devices.
217

EU Council Directive 93/42/EEC: 1993: Concerning medical devices.
[Cited August 3rd 2014]; Available from http://eur-lex.europa.eu/legalcontent/EN/TXT/PDF/?uri=CELEX:31993L0042&from=EN
Farnham, John J., Tomazic-Jezic, Vesna J., & Stratmeyer, Mel. E. (2002). Regulatory
initiatives for natural latex allergy: US perspectives. Academic Press. Method 27:
87-92.
Federal Food, Drug, and Cosmetics Act (FFDCA), Section 502. Misbranded drugs and
devices. [Cited August 16TH 2013]; Available from
http://www.fda.gov/RegulatoryInformation/Legislation
/FederalFoodDrugandCosmeticActFDCAct/FDCActChapterVDrugsandDevices
/default.htm
Filik, R., Purdy, K., Gale, A. & Gerrett, D. (2006). Labeling of Medicines and Patient
Safety: Evaluating Methods of Reducing Drug Name Confusion. Human Factors,
48(1): 39-47.
Food and Drug Administration (FDA). Use of symbols on labels and in labeling of In
Vitro Diagnostic devices intended for professional use; Guidance for Industry and
FDA staff: November 2004; [Cited February 7th 2013]; Available from: 
http://www.fda.gov/medicaldevices/deviceregulationandguidance/guidancedocum
ents/ucm085404.htm
Food and Drug Administration (FDA). Unique Device Identification System (UDI);
Proposed rule: July 2012. [Cited February 7th 2013]; Available from:
http://www.fda.gov/medicaldevices/deviceregulationandguidance/uniquedeviceid
entification/default.htm
Food and Drug Administration (FDA). Unique Device Identification System (UDI); Final
rule: September 2013. [Cited October 21st 2013]; Available from:
http://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/Reports/Econo
micAnalyses/UCM368961.pdf
Food and Drug Administration (FDA). Use of certain symbols in labeling; Proposed rule:
April 2013. [Cited April 27th 2013]; Available from:
http://www.gpo.gov/fdsys/pkg/FR-2013-04-19/html/2013-09175.htm
Food and Drug Administration (FDA). Labeling Requirements-Misbranding. [Cited
November 20th 2013]; Available from
http://www.fda.gov/medicaldevices/deviceregulationandguidance/overview/devic
elabeling/generaldevicelabelingrequirements/ucm052190.htm
Food and Drug Administration (FDA). Regulations of Natural Rubber Latex Gloves.
218

[Cited February 23rd 2013]; Available from
http://www.fda.gov/NewsEvents/Testimony/ucm115083.htm
Gaschler, R., Mata, J., Störmer, V. S., Kühnel, A., & Bilalić, M. (2009). Change
detection for new food labels. Food Quality and Preference, 21(1): 140–147.
Garcia, M. (1981). Newspaper Design. Englewood Cliffs: Prentice-Hall, Inc.
Goldstein, E. B. (2007) Sensation & Perception (7th edition). Thomson Wadsworth.
Gomes, T., Hurley, R. A., Duchowski, A., Darby, D., Outzts, A. (2014). The Effect of Full
Body Versus Partial Body Graphic Labeling on Beverage Packaging. Packaging
Technology and Science, DOI: 10.1002/pts 2070.
Graham, D. J, Robert, W., & Jeffery W. (2011). Eye-Tracking evidence that consumers
preferentially view prominently positioned nutrition information. Journal of the
American Diebetic Association, 111: 1704-1711.
Hermans, V., Monzote, L., et al. (2011). Assessment of the knowledge of graphical
symbols labeled on malaria rapid diagnostic tests in four international settings.
Malaria Journal, 10: 331.
Herpen, E. V., & Trijp, H. C. M. (2011). Front-of-pack nutrition labels. Their effect on
attention and choice when consumers have varying goals and time constraints.
Appetite, 57: 148-160.
Hopper, William R., & Moss, Rose. (2010). Common Breaks in Sterile Technique:
Clinical Perspectives and Perioperative Implications. AORN J, 91: 350-364.
Hurley, R. A., Galvarino, J., Thackston, E., Ouzts, A., & Pham, A. (2012). The Effect of
Modifying Structure to Display Product Versus Graphical Representation on
Packaging. Packaging Technology and Science. DOI: 10.1002/pts 1996.
 

Hurley, R. A., Quzts A., Fischer J., & Gomes T. (2013) Effects of private and public label
packaging on consumer purchase patterns. Packaging Technology and Science.
DOI:10.1002/pts 2012.
 

International Paper. (1997). Pocket Pal: A graphic arts production handbook (17th ed.).
ISO 15223-1. 2007: Medical devices - Symbols to be used with Medical Device Labels,
Labeling and Information to be supplied.
ISO 3864, 2004: International Standard for Safety Colours and Safety Signs.
219

ISO 9186-1 2007: Graphical Symbols – Test methods – Part 1: Method for testing
comprehensibility.
Kam, P., Lee, M., & Thompson, J. (1997). Latex allergy: an emerging
clinical and occupational health problem. Anesthesia, 52: 570-575.
Klevens, R. Monina, Edwards, Jonathan R., et al. (2007). Estimating Health CareAssociated Infections and Deaths in US Hospitals, 2002. Public Health Reports;
March-April 2007; 122: 160-166.
Kellett, P. B. (1997). Latex allergy: a review. Journal of Emergency Nursing, 23(1): 31.
Liu, L., Hoelscher, U., & Gruchmann, T. (2005). Symbol comprehension in different
countries: Experience gained from medical device area. WorkshopsProceedings der 5 fachübergreifenden Konferenz Menschund Computer. Wien:
Oesterreichische 2005, 81:7.
McKenzie, Craig R. M., Wixed, John T., Noelle, David C. (2001). Relation between
confidence in yes-no and Forced-Choice Tasks. Journal of Experimental
Psychology, Vol 130, No. 1: 140-155.
Michelson, J. D., Lotke, P. A. & Steinberg, M. E. (2004). Urinary-bladder management
after total joint-replacement surgery. N Engl J Med., 319: 321-326.
Neuendorf, K. A. (2002). The content analysis guidebook. Thousand oaks, CA: Sage
Publications.
Noton, D., & Stark, L. (1971). “Eye Movements ad Visual Perception.” Scientific
American 224: 34 – 43.
Oh, C. H. (2010). Measuring the relative prominence of graphic symbols vs. text for
nutrition labels using eye tracking. Master thesis. Michigan State University.
 

Perles, P. (1977). Readability, Has it gone? Or, Pity the poor reader. Direct marketing,
39(10): 32-40.
Perry, J. (2003). Graphical symbols to address consumer needs. ISO Bulletin, March:
8-11.
Pettit, G. (2000). PKG 330 (Packaging printing) Lecture. Michigan State University.
Petty, R. E., Cacioppo, J. & Schumann, D. (1983) Central and Peripheral Routes to
Advertising Effectiveness: The Moderating Role of Involvement. Journal of
Consumer Research, 10 (September): 135-146.
220

Pilchik, Ronald. (2003). Validating Medical Packaging, CRC Press.
Queiroz, M. D., et al. (2009). Latex allergy in children: modalities and prevention.
Pediatric Anesthesia, 19: 313-319.
Rayner, Keith. (1998). “Eye movements in reading and information processing: 20 years
of research.” Psychological Bulletin, 124 (3): 372 – 422.
Rensink, R. A., O’Regan, J. K., & Clark, J. J. (1997). To See or not to See: The Need
for Attention to Perceive Changes in Scenes. Psychological Science, 8(5), 368–
373.
Ross, V., & Kalokerinos, S. M. (1985). Packaging and medical device recalls.
Polymers, Lamination and Coatings Conference.
Rousseau, G. K., Lamson, N., & Rogers, W. A. (1998). Designing warnings to
compensate for age-related changes in perceptual and cognitive abilities.
Psychology and Marketing, 15(7): 643–662.
 

Sherman, Max. (1998). Medical device packaging handbook. Marcel Dekker, INC.
Skelly, J. M., Guyatt, G. H., Kalbfleisch, R., et al. (1992) Management of urinary
retention after surgical repair of hip fracture. Can Med Assoc J.,146: 1185-1188.
Sorg, J. (1985). Designing warnings to compensate for age related changes in
perceptual and cognitive abilities. Psychology & Marketing, 15(7): 643.
Stone, P. W, Braccia, D, & Larson, E. (2005). Systematic review of economic analyses
of health care-associated infections. AM J Infect Control, 33: 501-509.
Sundar, R. P. (2009). The effect of color contrast of text on the legibility and noticeability
of prescription drug labels. Master thesis. Michigan State University.
Sundar, R. P. (2013). Investigating the effect of color and icon on information
processing behaviors related to Front-Of-Package nutrition labels. Ph. D
dissertation. Michigan State University.
Virant, F. S. (1996). Radiocontrast, local anesthetic and latex reactions. In: Bierman CW,
Pearlman DS, Shapiro GG, et al., eds. Allergy; Asthma and Immunology from
Infancy to Adulthood. Philadelphia, Pa: W.B. Saunders Co: 361-365.
Vukelich, M., & Whitaker, L. A. (1993). The effects of context on the comprehension of
graphic symbols. In Proceedings of the Human Factors and Ergonomics Society
37th Annual Meeting: 511-515.
221

Wedel, M., & Peiters, R. (2008). A review of eye-tracking research in marketing.
In Review of marketing research. Emerald Group: Bingley, UK.
Wells, D. L., & Saltmarche, A. (2004). Voiding dysfunction in geriatric patients with hip
fracture; prevalence rate and tentative nursing interventions. Orthopedic Nursing,
5: 25-28.
White, I. R. (1996). Occupational dermatitis. ABC of work related disorder. British
Medical Journal, 13: 487-489.
Woff, J. S., Wogalter, M. S. (1998). Comprehension of pictorial symbols: Effects of
context and test method. Human factors, June: 173-186.
Young, L. R., & Sheena, D. (1975). Survey of eye movement recording methods:
Behavioral research methods and instrumentation, 7: 397-429.
Yunginger, J. W. (1998). Natural rubber latex allergy. Ln: Middleton E, Reed CE, Ellis
EF, et al. Allergy Principles & Practice, Vol. II St. Louis, Mo: Mosby:
1073-1078.
Zaglaniczny, Karen (2001). Latex allergy: Are you at risk?. AANA Journal, 69(5):
413-424.

222