EFFECTS OF HOLE SIZE, PRESSURE DIFFERENTIAL, AND SECONDARY PACKAGING ON
MICROBIAL INGRESS OF STERILE MEDICAL DEVICE TRAYS
By
Ondrea Kassarjian

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Packaging
2011

ABSTRACT
EFFECTS OF HOLE SIZE, PRESSURE DIFFERENTIAL, AND SECONDARY PACKAGING ON
MICROBIAL INGRESS OF STERILE MEDICAL DEVICE TRAYS
By
Ondrea Kassarjian
Nosocomial infections, also referred to as healthcare-associated or hospitalacquired infections (HAIs), are those which hospitalized patients acquire during medical
treatment for another condition that were not present or incubating at the time of
admission, unless related to a previous admission. These infections are a significant
cause of morbidity and mortality that drive up the cost of healthcare in the United
States. Despite all of the efforts to prevent nosocomial infections from occurring, they
have not been eliminated and are becoming increasingly difficult to treat. It is also
difficult determine factors of significance in the causal pathway, such as tracing an
infection back to its source. Devices have been cited as sources of infection, both as
modes of transmission and reservoirs, but the origin of the pathogens and their transfer
to the device are unknown and not widely studied. This raises questions as to whether
the devices become contaminated after they are opened, or if their sterile barriers are
breached prior to use.
The objective of this research was to examine the effects of hole size, pressure
differential, and secondary packaging on microbial ingress of sterile medical device
trays.
The methods involved aseptically filling sealed, sterile device packages with a known
volume of an appropriate growth medium, exposing the packages to an aerosolized

microbial challenge, incubating the packages, and inspecting for growth. After creating
and refining the test techniques, the research explored the impacts of hole size (10 and
100 µm), pressure differential (0 and -3.78 psi), and secondary packaging (pouches and
cartons) on microbial penetration. The specific pressure differential examined simulates
an aircraft descending from 8,000 feet or a ground shipment descending from the same
elevation.
Hole size, pressure differential, and secondary package type all had a statistically
significant effect on microbial penetration of the sterile medical device test trays used in
the study.

ACKNOWLEDGEMENTS

There are several contributors who have made this research effort possible.
First, I would like to thank my fabulous guidance committee for their collaborative
support and express my sincere gratitude for each of their unique contributions to this
research. Dr. Melinda Frame, thank you for sharing your (pleasant) expertise so that we
could pioneer into the beautiful packaging microcosm. Dr. John Linz, thank you for the
extensive microbiological training and sharing your laboratory. You have made me a
real scientist and I will miss working in your lab. Dr. Gary Burgess, thank you for your
brilliance, candor, and wit. Dr. Susan Selke, thank you for keeping wind in my sails when
the air was still. You truly made the ability for me to remain at the School of Packaging a
reality and I am so very grateful. And last but certainly not least, Dr. Laura Bix, my major
professor and friend, thank you for so many things, but most of all for the lesson of
small bites.
I would also like to give special acknowledgement to Dr. Nora Bello. Thank you
for your knowledge and time in providing statistical expertise for the experimental
design and data analysis of this research. I appreciate both your enthusiasm and
mentorship.
I would also like to acknowledge the contributions from the following individuals,
associations, and companies: Abbott Laboratories, Amcor Flexibles, ATC, Becton
Dickinson, Boston Scientific, Cardinal Health, Cory Wilson, Curt Larsen, DDL, Dennis
Young, DePuy Orthopaedics, Edwards Lifesciences, Ethicon Endosurgery, the Flexible
iv

Packaging Association, Guidant, Hal Miller, John Spitzley, Ken Gruhl, Lansmont, Mangar,
Medtronic, Perfecseal, Smith & Nephew, and the Sterile Barrier Association.
There are also a handful of people who have made my life pivotal, and I would
like to acknowledge Eric Joneson for the pivot point into this endeavor. Thank you for
your insight and value.
Thank you to my family of mentors, peers, and colleagues at the School of
Packaging. You have brought joy to my life.
Lastly, I would like to give an enormous thank you to my friends and family for
your unconditional support and patience in the good times and not-so-good times. The
unwritten pieces of this belong to you.

v

TABLE OF CONTENTS

LIST OF TABLES……………………………………………………………………………………………………………v
LIST OF FIGURES…………………………………………………………………………………………………………vi
1

INTRODUCTION………………………………………………………………..………………………………….1

2

LITERATURE REVIEW…………………………………………………………………………………………….2
2.1 Nosocomial Infections…………………………………………………………………………..……..2
2.2 Nosocomial Infection Rates………………………………………………………………………….3
2.2.1 Centers for Disease Control and Prevention Rates……………….………..4
2.2.2 United States Department of Health and Human Services Rates…...5
2.2.3 National Nosocomial Infections Surveillance System Rates…………...5
2.2.3.1 Device-associated Nosocomial Infection Rates in Adult
and Pediatric Intensive Care Units………………………………………….……6
2.2.3.2 Device-associated Nosocomial Infection Rates in High-risk
Nurseries…………………………………………………………………………..…………6
2.2.3.3 Surgical Site Infection Rates…………………………………………….7
2.3 Causes of Nosocomial Infections……………………………………………………………..…..8
2.3.1 Hospital Environment……………………………………………………………………9
2.3.2 Adherence to Infection Control Procedures………………………….……….9
2.3.3 Patient Immune Status………………………………………………………………..10
2.3.4 Medical Devices…………………………………………………………………………..11
2.4 Consequences of Nosocomial Infections…………………………………………………….12
2.4.1 Length of Stay………………………………………………………………………………12
2.4.2 Patient Costs…………………………………………………………….…………………14
2.4.3 Hospital Costs……………………………………………………………………………..17
2.4.4 Socioeconomic Impact………………………………………………….……………..18
2.4.5 Mortality………………………………………………………………………..……………18
2.4.6 Staph Infections…………………………………………………………………………..20
2.4.6.1 MRSA…………………………………………………….………………………21
2.4.6.1.1 Impact of MRSA………………………………………………24
2.4.6.2 Staphylococcus epidermidis………………………………..…………25
2.5 Prevention of Nosocomial Infections……………………………………………………..…..25
2.5.1 Infection Control Programs………………………………………………………….26
2.5.1.1 Antimicrobial Use and Resistance…………………….…………..27
2.5.1.2 Other Prevention Measures………………………….………………32
2.5.2 Device and Packaging Design…………………………………….…………………33
2.5.3 Successful Prevention Initiatives………………………….………………………34
2.6 Packaging…………………………………………………………………………………………………..35
2.6.1 Package Testing………………………………………………………………….……….36
vi

2.6.2 Factors that Can Affect Microbial Ingress…………………………………....39
2.6.2.1 Packaging Characteristics…………………..………………………….40
2.6.2.2 Packaging Defects……………………………………………………..….41
2.6.2.2.1 Creating and Measuring Hole Defects in
Packaging……………………………………………………………………….42
2.6.2.2.2 Microscopy Methods for Measuring Holes……..45
2.6.2.3 Torturous Path and Layers of Packaging………………………..49
2.6.2.4 Microorganisms…………………………………………………………….53
2.6.2.4.1 Nosocomial Microorganisms and
Pathogenicity…………………………………………………………………53
2.6.2.4.2 Test Microorganisms and Microbial
Concentration…………………………………………………………..……55
2.6.2.4.3 Viability on Packaging…………………………..………..56
2.6.2.5 Methods of Microbial Challenge……………………………………56
2.6.2.6 Transport, Storage, and Handling………………………………….57
2.6.2.6.1 Pressure Differential……………………………………...59
2.6.2.6.1.1 Pressure Differential Research…….…..60
2.6.2.6.1.2 Number of Pressure Changes………..…61
2.6.2.6.1.3 Pressure Differential and Microbial
Concentration……………………………………….……………62
2.6.2.6.1.4 Pressure Differentials Simulating
Transport and Storage…………………………..……………62
2.6.3 Investigation of Microbial Ingress through the Sterile Barrier………64
3

PART 1: EFFECT OF PRESSURE DIFFERENTIAL ON MICROBIAL INGRESS OF STERILE
MEDICAL DEVICE TRAYS WITH 100 µm PINHOLES…………………………………………….…66
3.1 Objectives……………………………………………………………………………………………….…66
3.2 Materials and Methods……………………………………………………………………………...66
3.2.1 Holes………………………………………………………………………………………..….68
3.2.2 Tray Sealing…………………………………………………………………………….…..68
3.2.3 Sterilization…………………………………………………………………………….…..69
3.2.4 Non-destructive Leak Testing…………………………………………….………..70
3.2.5 Agar Injection of Lidded Trays………………………………………………….….71
3.2.6 Preparation of Dilution Series Agar Plates………………………………..….76
3.2.7 Microbial Solution Preparation………………………………………….………..77
3.2.8 Dilution Series……………………………………………………………..………………79
3.2.9 Inoculation……………………………………………………………………………….….83
3.2.10 Microbial Challenge and Induction of Pressure Differential…..…..84
3.2.11 Incubation……………………………………………………………………..………….89
3.2.12 Colony Counting of Test Samples……………………………………………….89
3.2.13 Colony Counting of Agar Plates………………………………………………….90
3.2.14 Enterotubes…………………………………………………………………….………..90
3.3 Results and Discussion……………………………………………………………………………….91
vii

4

PART 2: EFFECTS OF PRESSURE DIFFERENTIAL AND SECONDARY PACKAGING ON
MICROBIAL INGRESS OF STERILE MEDICAL DEVICE TRAYS………………………..…………96
4.1 Objective………………………………………………………………………………………………..….96
4.2 Materials and Methods…………………………………………………………………………..….96
4.2.1 Sample Preparation for Unlidded Trays in Pouches………………….….98
4.2.2 Sample Preparation for Unlidded Trays in Cartons…………………….101
4.2.3 Sample Preparation for Lidded Trays, with 100 µm Pinholes, in
Cartons……………………………………………………………………………………………….103
4.2.4 Agar Injection for Unlidded Trays in Pouches and Unlidded Trays
in Cartons…………………………………………………………………………………………...105
4.2.5 Agar Injection for Lidded Trays, with 100 µm Pinholes, in
Cartons……………………………………………………………………………………………….106
4.2.6 Microbial Challenge and Induction of Pressure Differential……….107
4.2.7 Colony Counting of Test Samples…………………………………….…………107
4.3 Results and Discussion……………………………………………………………………………..108

5

PART 3: EFFECTS OF HOLE SIZE, PRESSURE DIFFERENTIAL, AND SECONDARY
PACKAGING ON MICROBIAL INGRESS OF STERILE MEDICAL DEVICE TRAYS……….115
5.1 Objective………………………………………………………………………………………………….115
5.2 Materials and Methods………………………………………………………………………….…115
5.3 Results and Discussion……………………………………………………………………………..120
5.4 Conclusions………………………………………………………………………………………………126
5.5 Limitations……………………………………………………………………………………………….129
5.6 Future Research……………………………………………………………………………………….129

6

PINHOLE CHARACTERIZATION………………………………………………………………………..…130
6.1 Objectives………………………………………………………………………………………….…….130
6.2 Materials and Methods…………………………………………………………………………….130
6.3 Results and Discussion………………………………………………………………………..……137

APPENDICES………………………………………………………………………………………….………………..141
Appendix A: Materials List……………………………………………………………………………….……142
Appendix B: Non-destructive Leak Test Results………………………………………..…………..146
Appendix C: Part 1 Test Sample Results…………………………………………………………………152
Appendix D: Part 1 Agar Plate Colony Counts………………………………………………………..153
Appendix E: Part 2 Agar Plate Colony Counts…………………………………………………………154
Appendix F: Part 2 Test Sample Results…………………………………………………………………156
Appendix G: Part 3 Agar Plate Colony Counts………………………………………………………..160
Appendix H: Part 3 Test Sample Results………………………………………………………………..161
BIBLIOGRAPHY……………………………………………………………………………………………………….170

viii

LIST OF TABLES

Table 1. Rates of infection by infection type………………………………………………………………….4
Table 2. Excess lengths of stay due to nosocomial infection by infection type…………….14
Table 3. Attributable costs due to nosocomial infections by infection type………………...16
Table 4. Attributable mortality rates due to nosocomial infections by infection type.…19
Table 5. Percent of total trays with CFU………………………………………………………………………92
Table 6. Percent of total trays with CFU…………………………………………………………………….108
Table 7. Percent of total samples with CFU……………………………………………………………….121
Table 8. Entry side diameters for thermal and excimer laser holes…………………………….139
Table 9. Non-destructive leak test results for trays without pinholes………………………..146
Table 10. Non-destructive leak test results for trays with 10 µm pinholes…….…………..147
Table 11. Non-destructive leak test results for trays with 100 µm pinholes…………….…149
Table 12. Part 1 test sample colony counts (CFU)…………………….………………………………..152
Table 13. Part 1 microbial chamber agar plate colony counts (CFU per plate)……….…..153
Table 14. Microbial chamber agar plate colony counts (CFU per plate) for runs testing
unlidded trays in pouches…………………………………………………………………………………………..154
Table 15. Microbial chamber agar plate colony counts (CFU per plate) for runs testing
unlidded trays in cartons…………………………………………………….……………………………………..154
Table 16. Microbial chamber agar plate colony counts (CFU per plate) for runs testing
lidded trays, with 100 µm pinholes, in cartons……………………………………………………………155
Table 17. Colony counts (CFU) of unlidded trays in pouches……………………………………..156
Table 18. Colony counts (CFU) of unlidded trays in cartons……………………………………….157
Table 19. Colony counts (CFU) of lidded trays, with 100 µm pinholes, in cartons……...158
ix

Table 20. Part 3 microbial chamber agar plate colony counts (CFU per plate)…………...160
Table 21. Colony counts (CFU) of split plot test samples…………………………………………...161

x

LIST OF FIGURES

Figure 1. Average of pooled means of NNIS System infection rates for device-associated
urinary tract infections (UTI), bloodstream infections (BSI), and pneumonia in ICUs (NNIS
System 2001; NNIS System 2004)…………………………………………………………………………………..6
Figure 2. Average of pooled means of NNIS System device-associated infection rates for
bloodstream infections (BSI) and pneumonia in HRNs (NNIS System 2001; NNIS System
2004)……………………………………………………………………………………………………………………………..7
Figure 3. Average of pooled means of NNIS System surgical site infection rates (NNIS
2001 and 2004)………………………………………………………………………………………………………………8
Figure 4. Percent of methicillin-resistant staph infections (Division of Healthcare Quality
Promotion 2007)…………………………………………………………………………………………….……………23
Figure 5. Comparison of NNIS System antimicrobial resistance rates in 2001 and 2004
(NNIS System 2001; NNIS System 2004)…………………………………………………..…………………..30
Figure 6. Comparison of NNIS System increase in antimicrobial resistance in 2001 and
2004 (NNIS System 2001; NNIS System 2004)…………………………..………………………………….31
Figure 7. Comparison of antimicrobial-resistance rates between 1990-1994 and 20002004 (Klevens, Edwards et al. 2008)……………………………………………………………………..………32
Figure 8. Effect of the number of pressure changes on the percentage of contaminated
pouches (Dunkelberg and Rohmann 2006)…………………………………………………….…………….62
Figure 9. Effects of pressure differentials for paper and nonwoven material rates of
contamination (Dunkelberg and Schmelz 2009)……………………….…………………………………..63
Figure 10. Test trays…………………………………………….………………………………………………………68
Figure 11. Tray lids……………………………………………………………………………………..……………….69
Figure 12. Sealing positions…………………………………………………………………………………………69
Figure 13. ATC, Inc. Model VE2 non-destructive leak tester………………….……………………..70
Figure 14. Glass bottles containing sterile nutrient growth agar…………………………..……..71
Figure 15. Microbial challenge methodology………………………………………………………………..73
xi

Figure 16. Injection frame with test samples: top view……………………………………………..74
Figure 17. Injection frame with test samples: bottom view…………………………………………74
Figure 18. Escherichia coli K12 ATCC Number 29181 (Severin 2006)…………………..……….77
Figure 19. Vials of E. coli stock……………………………………………………………………….…………….79
Figure 20. Dilution series……………………………………………………………………………….…………….81
Figure 21. Quebec Colony Counter illuminating a dilution series agar plate…………………82
Figure 22. Microbial challenge chamber………………………………………………………………………85
Figure 23. Interior view of microbial challenge chamber…………………………………..…………85
Figure 24. Pressure differential syringe placement……………………………………………………...87
Figure 25. Retraction bar and agar plate placement…………………………………………………….88
Figure 26. Colony viewing window cut from tray…………………………………………………………89
Figure 27. Quebec Colony Counter illuminating a microbial challenge agar plate……..…90
Figure 28. Example of enterotube and color key…………………………………….…………………..91
Figure 29. Example of tray with colony growth………………………………………………………..….92
Figure 30. Comparison of predicted probabilities of microbial penetration when
pressure differential is the independent variable…………………………………………………..…….94
Figure 31. Comparison of estimated number of CFU per tray when pressure differential
is the independent variable………………………………………………………………………………………….95
Figure 32. Unlidded tray with pressure differential septum…………………………………..……..99
Figure 33. Example of test pouch………………………………………………………………………………..99
Figure 34. Unlidded tray in pouch test configuration with septum placement…….…….100
Figure 35. Unlidded tray in carton test configuration with septum placement: open
view……………………………………………………………………………………………..…………………………….102

xii

Figure 36. Unlidded tray in carton test configuration with septum placement: closed
view……………………………………………………………………………………………………………………………102
Figure 37. Injection points on pressure differential septums……………………………….…….102
Figure 38. Septum placement for lidded tray in carton test configuration…………………103
Figure 39. Lidded tray in carton test configuration…………………………………………….………104
Figure 40. Lidded tray in carton septum placement: bottom view…………………..……….104
Figure 41. Lidded tray in carton septum placement: top view…………………………………..104
Figure 42. Example of agar injection into unlidded tray in pouch configuration…….….105
Figure 43. Test sample placement on agar injection frame (bottom view)………………...106
Figure 44. Comparison of predicted probability of microbial penetration of pouches and
cartons……………………………………………………………………………………………………………………….110
Figure 45. Comparison of estimated number of CFU per tray in pouches and cartons.111
Figure 46. Comparison of predicted probability of microbial penetration of unlidded and
lidded trays in cartons…………………………………………………………………….………………………….113
Figure 47. Comparison of estimated number of CFU per tray of unlidded and lidded
trays in cartons…………………………………………………..………………………………………………………114
Figure 48. Eight test sample configurations combining levels of hole size, presence or
absence of a carton, and presence or absence of a pressure differential……………………117
Figure 49. Microbial challenge chamber (assigned to either 0 or -3.78 psi) containing
four test samples (assigned to combinations of hole size and presence or absence of a
carton)……………………………………………………………………………………………………………..………..118
Figure 50. Microbial challenge chamber (assigned to -3.78 psi) containing four test
samples (assigned to combinations of hole size and presence or absence of a carton)118
Figure 51. Power analysis for predicted probability of microbial ingress…………..……….119
Figure 52. Power analysis for number of CFU………………………………………………………….…120
Figure 53. Percent growth for the eight test sample configurations………………..…………121

xiii

Figure 54. Comparison of probabilities of microbial penetration when pressure
differential is the main effect…………………………………………………………………………….……….123
Figure 55. Comparison of probabilities of microbial penetration when hole size is the
main effect………………………………………………………………………………………………………………...124
Figure 56. Comparison of probabilities of microbial penetration when secondary
packaging is the main effect…………………………………………………………………………………….…125
Figure 57. Confocal light pathways (Olympus 2011)…………………..………………………………131
Figure 58. Osmium coating of an excimer laser drilled hole……………………………………...132
Figure 59. Sample inside a Lab-Tek II #1.5 Coverglass Chamber……………….…………………133
Figure 60. Screen capture of imaging parameters………………………………………………………134
Figure 61. Thermal entry side laser hole images: fluorescent (top left), Brightfield (top
right), and overlay of both (bottom)…………………………….…………………………………………….135
Figure 62. Excimer entry side laser hole images: fluorescent (top left), Brightfield (top
right), and overlay of both (bottom)…………………………………………………………………………..135
Figure 63. Thermal laser hole z-series images shown in stepwise progression…………..136
Figure 64. Excimer laser hole z-series images shown in stepwise progression……………137
Figure 65. Side views of thermal laser hole 3D renderings…………………………………………138
Figure 66. Side views of excimer laser hole 3D renderings……………………………….………..138
Figure 67. Entry side images and diameters of thermal laser drilled holes………….……..139
Figure 68. Entry side images and diameters of excimer laser drilled holes…………………139

xiv

INTRODUCTION

Healthcare-associated infections (HAIs) occur in patients who are undergoing
treatment for another condition. Consequences of these include increased morbidity,
mortality, and costs. HAIs are difficult to treat and, despite the efforts of infection
control programs, are not completely preventable. It is also difficult to trace HAIs back
to a source.
Medical devices have been identified as sources of infection; however, even
when this has been possible, it is still unknown how the device became contaminated.
As a result, the creation and maintenance of sterile barrier systems (SBS) and aseptic
technique are becoming increasingly important.
The investigation of packaging as a source of infection is limited. Varying results
exist as to what types and sizes of defects can be penetrated by pathogenic
microorganisms. Additionally, there is a limited amount of research available examining
the effects of other variables within the in-use environments, such as pressure
differentials and secondary packaging.
This research utilizes microbial challenge methodologies to investigate the
effects of hole size, pressure differential, and secondary packaging on the microbial
ingress of sterile medical device trays. Desired outcomes include increasing the
knowledge base of what facilitates the ability of microorganisms to traverse the sterile
barrier. This knowledge can then be applied toward tracing sources of infection and
improving package/device designs to maintain the sterile barrier and aid in aseptic
technique in an effort to reduce healthcare-associated infections.
1

LITERATURE REVIEW

2.1 Nosocomial Infections
Nosocomial infections, also referred to as healthcare-associated or hospitalacquired infections (HAIs), are those which hospitalized patients acquire during medical
treatment for another condition (Bascetta, Edwards et al. 2008) that were not present
or incubating at the time of admission, unless related to a previous admission (Murphy,
Whiting et al. 2007). In addition to being extremely costly and a cause of unnecessary
suffering (Bascetta, Edwards et al. 2008) nosocomial infections are a significant cause of
morbidity and mortality in the United States (Klevens, Edwards et al. 2007). Prolonged
hospitalization, the need for ICU care, surgery, or other procedures, reduced activity
levels at discharge, and the loss of work (Cosgrove 2006) are among the consequences
of HAIs. In fact, for an average 250-bed hospital, these infections result in 2,000 extra
days of hospitalization, 20 deaths, and $1 million in excess costs annually (Jarvis 1996).
Nosocomial infections are becoming more difficult to treat and complete
prevention of them has not yet become possible. As such, policy makers and healthcare
providers have spent significant attention and resources on programs intended to
reduce the prevalence and impact of these infections. Though not emphasized to the
same degree as those that concentrate on things like hygienic technique or the use and
development of antimicrobials, the creation and maintenance of the sterile barrier
1

system (SBS) , as well as its ability to be presented aseptically, are increasingly
important.
1

Sterile Barrier System (SBS): the “minimum package that prevents ingress of
microorganisms and allows aseptic presentation of the product at the point of use” (ISO
2
11607).

2.2 Nosocomial Infection Rates
Nosocomial infections rates have been examined and reported by both
independent researchers and departments within the US government. The reports vary
widely, with some authors suggesting that nosocomial infections rates appear to be
improving (Murphy, Whiting et al. 2007), while others suggest that they have increased
by 36% in the last 20 years (Schwegman 2008). General rates that have been published
range from 1.2% to 51.2% (Table 1).
Not only do the published general infection rates vary, the rates that are specific
to type of infection range significantly as well. The Centers for Disease Control and
Prevention (CDC) identifies four types of infections as being the most prevalent: urinary
tract infections (UTIs), surgical site infections (SSIs), pneumonia or other lung infections,
and bloodstream infections (BSIs). Of the total number of nosocomial infections that
occur annually in the United States, UTIs are estimated to account for 32%, SSIs for 22%,
pneumonia and other lung infections for 15%, and BSIs for 14% (Klevens, Edwards et al.
2007). Ranges of rates published by independent researchers for these types of
infections are: 0.6% to 15.5% for pneumonia, 0.27% to 6% BSIs, 2.7% to 7% for SSIs, and
2.39% to 13% for UTIs (Table 1).

3

Table 1. Rates of infection by infection type.
Infection Type
Authors and Year
(Craig and Connelly 1984)
(Haley, Culver et al. 1985)
Pneumonia
(Craven, Kunches et al. 1988)
(Leu, Kaiser et al. 1989)
(Kollef 1993)
(Townsend and Wenzel 1981)
(Craven, Kunches et al. 1988)
Bloodstream
(Haley, Culver et al. 1985)
Infections (BSIs)
(Pittet, Tarara et al. 1994)
(Wisplinghoff, Bischoff et al. 2004)
(Haley, Culver et al. 1985)
(Craven, Kunches et al. 1988)
Surgical Site
(Kirkland, Briggs et al. 1999)
Infections (SSIs)
(Whitehouse, Friedman et al.
2002)
(Haley, Culver et al. 1985)
Urinary Tract
Infections (UTIs) (Craven, Kunches et al. 1988)
(Craven, Kunches et al. 1988)
(Emori and Gaynes 1993)
(Fagon, Novara et al. 1994)
Overall Infections
(Malone and Larson 1996)
Due to Medical
(Richards, Edwards et al. 2000)
Care
(Christensen and Jepsen 2001)
(Graves 2004)
(Murphy, Whiting et al. 2007)

Rate of Infection
8.8%
0.6%
9%
8.6%
15.5%
4%
6%
0.27%
2.7%
0.6%
2.79%
7%
2.7%
2.8%
2.39%
13%
51.2%
5.7%
5%
3.9%
6.1%
17%
10%
1.2%

2.2.1 Centers for Disease Control and Prevention Rates
The CDC reported that in 2002 there were 1.7 million HAIs which resulted in
176.4 million patient-days. This equates to a rate of 9.3 infections for every 1000
patient-days, or 4.5 infections for every 100 admissions. Of the patients who acquired
an infection during this period, adults and children were 93.1% of the population and
newborns were 6.9% (Klevens, Edwards et al. 2007).

4

2.2.2 United States Department of Health and Human Services Rates
In 2010, the US Department of Health and Human Services (DHHS) reported that
little progress has been made on eliminating HAIs and that some types have actually
increased. According to the DHHS Annual Quality and Disparities Report, the increases
in infection rates are as follows: postoperative BSIs by 8%, postoperative catheterassociated UTIs by 3.6%, and selected infections due to medical care by 1.6%. The same
data suggests there was no change in the number of central venous catheter BSIs and
there was improvement with rates of postoperative pneumonia, which decreased by
12% (U.S. Department of Health and Human Services Agency for Healthcare Research
and Quality 2010).
2.2.3 National Nosocomial Infections Surveillance System Rates
The National Nosocomial Infections Surveillance (NNIS) System has also
published data which suggests that some types of infections have rates that are
decreasing while other types are increasing. The NNIS System was established in 1970
and is composed of the Division of Healthcare Quality Promotion, National Center for
Infectious Diseases, Centers for Disease Control and Prevention, Public Health Service,
and US Department of Health and Human Services, Atlanta, Georgia. The system
collects data from more than 300 hospitals that provide general medical-surgical
inpatient services to adults or children within the United States. The surveillance
components reported on are: adult and pediatric intensive care units (ICUs), high-risk
nurseries (HRNs), and surgical patients (NNIS System 2001; NNIS System 2004).

5

2.2.3.1 Device-associated Nosocomial Infection Rates in Adult and Pediatric Intensive
Care Units
The NNIS has most recently published reports on device-associated nosocomial
infections in 2001 and 2004. When comparing these reports, the rates for deviceassociated UTIs, BSIs, and pneumonia in ICUs have decreased. The average of the
pooled means for each infection category suggest that rates decreased for UTIs from 5.9
to 4.9, BSIs from 5.5 to 4.9, and pneumonia from 10.4 to 7.5 (Figure 1) (NNIS System

Infection Rate per
1000 Device-days

2001; NNIS System 2004).

11
10
9
8
7
6
5
4
3
2
1
0

10.4

7.5
5.9
4.9

5.5

4.9

1995-2001
2002-2004

UTI

BSI

Pneumonia

Figure 1. Average of pooled means of NNIS System infection rates for deviceassociated urinary tract infections (UTI), bloodstream infections (BSI), and pneumonia
in ICUs (NNIS System 2001; NNIS System 2004).
2.2.3.2 Device-associated Nosocomial Infection Rates in High-risk Nurseries
When comparing reports published in 2001 and 2004, rates for deviceassociated BSIs and pneumonia in high-risk nurseries (HRNs) have also decreased. The
average of the pooled means for each infection category suggest that rates decreased
6

for BSIs from 6.5 to 5.5 and pneumonia from 3.5 to 2.3 (Figure 2) (NNIS System 2001;
NNIS System 2004).

7

6.5

Infection Rate per
1000 Device-days

6

5.5

5
4

3.5

3

1995-2001
2.3

2002-2004

2
1
0
BSI

Pneumonia

Figure 2. Average of pooled means of NNIS System device-associated infection rates
for bloodstream infections (BSI) and pneumonia in HRNs (NNIS System 2001; NNIS
System 2004).
2.2.3.3 Surgical Site Infection Rates
While device-associated UTIs, BSIs, and pneumonia rates have all decreased,
surgical site infection rates have not had the same improvement according to NNIS
System data. Comparisons of the averaged pooled means from 2001 and 2004 NNIS
System reports suggest that the rate for surgical site infections (SSIs) has increased from
2.53 to 2.73 (Figure 3) (NNIS System 2001; NNIS System 2004).

7

3

2.73

2.53
Infection Rate per
100 Surgeries

2.5
2
1.5
1
0.5
0
1992-2001

1992-2004

Figure 3. Average of pooled means of NNIS System surgical site infection rates (NNIS
2001 and 2004).
The majority of nosocomial infections occurs within only four categories
(infection types) and affects all age populations, including newborns. Despite the
variation of infection rates, and whether the reports indicate they are increasing or
decreasing, they are still occurring. Knowledge exists on sterile environments and
infection control programs and procedures are in place, but again, the infections
continue to happen. This raises questions as to what the unidentified causes of these
infections are and why they have not been eliminated.
2.3 Causes of Nosocomial Infections
The word “nosocomial” is derived from the Greek and Latin words for hospital.
In Roman times, hospital orderlies were referred to as nosocomi and hospital infections
were believed to be caused by miasmi, which is a vaporous exhalation of bad air (Wilcox
2003). Now, obviously, there is a much greater understanding of the real causes of
nosocomial infections; they are most frequently associated with high-risk medical
8

interventions such as surgical procedures and the use of invasive devices (Emori and
Gaynes 1993). The sources of infection and modes and rates of transmission can vary in
relation to the hospital environment, how well hospital employees adhere to infection
control procedures, and the immune statuses of patients (Schwegman 2008).
2.3.1 Hospital Environment
Hospitals are “more likely to spread infections than to stop them” and patients
are now more scared of catching an infection than going under the knife” (Cole 2008).
This is because bacteria are infecting patients despite the efforts of procedural
techniques and the maintenance of a sterile environment (Schierholz and Beuth 2001).
Not only are microorganisms transmitted through direct contact between people, but
contaminated intermediate objects serve as a means of indirect contact. Non-sterile
medical devices including blood pressure meters, thermometers, and stethoscopes have
all served in the spread of nosocomial infections (Schwegman 2008). Nosocomial
pathogens are also able to be transmitted through both wet media and dry surfaces.
Staphylococci are able to survive under dry conditions for long periods of time;
therefore, common items within a hospital environment such as furniture, equipment,
instruments, and even sterile packaging can become reservoirs for microorganisms
(Dietze, Rath et al. 2001).
2.3.2 Adherence to Infection Control Procedures
Extrinsic factors that contribute to the spread of HAIs include the practices of
individuals and hospitals (Emori and Gaynes 1993). Hospital personnel can fail to
adhere to infection control procedures for hand hygiene and surgical hand antisepsis,
9

intravenous site preparation and selection, dressing changes, and barrier protection
(Gilmore 2003); barrier protection is defined to include proper use of gloves, protective
eyewear, gowns, and masks (Emori and Gaynes 1993). Other factors outside the point
of care include isolation facilities, bed occupancy, staffing levels, education and training
(Cole 2008), antimicrobial use (Weinstein 1998), and the care of devices (Stone, Larson
et al. 2002).
Most of these extrinsic factors are not directly related to packaging; however,
the care (and use) of devices has the potential to be impacted. Adherence to infection
control procedures is dependent on the users of sterile devices and the environment
they are working in. Understanding how sterile barriers fail can be applied to improve
sterile barrier maintenance within these environments. Additionally, improving
package/device designs which aid in aseptic technique can also facilitate better
adherence to infection control procedures.
2.3.3 Patient Immune Status
Risk of infection can be decreased with shorter lengths of stay in the hospital and
by reducing inpatient surgeries altogether; risk increases with longer lengths of stay and
in patients who are elderly or critically ill (Malone and Larson 1996).
Immunocompromised patients are at a greater risk of infection due to their age,
underlying diseases, and the medical or surgical treatments they receive. Even more
alarming is that the patient population in the hospital is becoming more
immunocompromised because the sickest patients remain in the hospital while the
others receive outpatient care (Weinstein 1998). Not only are these patients the most
10

susceptible, but they contribute to the spread of infection as well; however, HAIs should
not be an acceptable and expected outcome of treating an older, sicker patient
population (Murphy, Whiting et al. 2007). Patients who are immunocompromised
cannot be controlled or changed, but the environment around them, and the devices
used to treat them, can be.
2.3.4 Medical Devices
Invasive medical devices have been examined for their role in nosocomial
infections. From 1992 to 1998, 54% of nosocomial infections occurred in surgical
patients with most of them linked to a particular invasive medical device. Central
intravenous lines were associated with 87% of BSIs, mechanical ventilators were
associated with 83% of nosocomial pneumonias, and catheters were associated with
97% of UTIs (Richards, Edwards et al. 2000). These types of devices allow pathogens
easy entry into the body (Wenzel 2000) by serving as a pathway from the environment
or one part of the patient’s body to another. They also serve as reservoirs where
microbes colonize, protected from the patient’s immune defenses; for instance, bacteria
colonize on catheters (Hu, Veenstra et al. 2004). As a result, the higher rates of
pneumonia, UTIs, and BSIs have been linked to the use of these devices both during and
following surgeries (Emori and Gaynes 1993).
Implanted devices have also been implicated as source of 45% of nosocomial
infections (Schierholz and Beuth 2001). Here, the use of implant devices either damages
or invades epithelial or mucous barriers in the patient. Again, they serve as a reservoir
for microbial colonization and are protected from or impair the patient’s immune
11

defenses which results in chronic infection or tissue necrosis. Implant infections are also
influenced by the patient’s immune and cellular defenses. Interestingly, the ability of
bacteria to adhere to a device is dependent on the device material’s topography, trace
chemicals, and ionic and glycoprotein sequences and, despite adherence to sterile
procedures for implants, device-associated infections are still occurring (Schierholz and
Beuth 2001).
2.4 Consequences of Nosocomial Infections
There are a multitude of consequences that result from nosocomial infections.
These include increased lengths of stay in the hospital for the infected patients, which
drive up costs for both the patients and hospitals. There is also a series of
socioeconomic costs, which are difficult to measure and affect both the patients and the
people associated with them, such as family and employers. The most severe
consequence, however, is mortality that is attributable to these infections.
2.4.1 Length of Stay
Nosocomial infections affect both the intensity and the duration of care. Excess
lengths of stay (LOS) are additional, beyond the usual LOS (and costs) for the original
admissions, and would be avoided if the infections did not occur (Murphy, Whiting et al.
2007). The risk of death has also been indicated to increase with duration of stay due to
the increased risk of acquiring a fatal infection during hospitalization (Fagon, Novara et
al. 1994). Patients with HAIs tend to have longer LOS while sicker patients who require
longer LOS are at an increased risk of acquiring an infection. This has led some authors
to questions whether nosocomial infections are the cause of significant extra LOS and
12

costs, or if they are relatively inexpensive and inevitable consequences of long and
expensive hospitalizations (Kilgore, Ghosh et al. 2008).
Nosocomial infections can create a significant amount of excess patient days in
the hospital that can range from days to weeks per infection. In 1985, it was estimated
that HAIs created an excess 7.5 million patient days nationwide (Murphy, Whiting et al.
2007). This equates to 20,548 years of unintended hospitalization for the American
public! In 2005, the Pennsylvania Healthcare Cost Containment Council (PHC4) reported
that they had approximately 1.9 million hospital admissions without HAIs and 24,000
with HAIs. The average lengths of stay were less than 5 days and 23 days, respectively.
On average, the patients with infections had to spend an additional 18 days
hospitalized!
Published studies indicate ranges of excess LOS attributable to nosocomial
infections for the four main types of infections, infections overall, and for patients who
acquire multiple infections (Table 2). Excess LOS for pneumonia, BSIs, SSIs, and UTIs
range from 4.7 to 30 days, 7 to 25.8 days, 7 to 14 days, and 1 to 12.5 days, respectively.
Overall excess LOS for infections due to medical care ranges from 4 to 18 days.
Additionally, those patients who have multiple infections have excess LOS of
approximately 35.9 days.

13

Table 2. Excess lengths of stay due to nosocomial infection by infection type.
Infection Type
Authors and Year
Excess LOS (days)
(Craig and Connelly 1984)
7.7
(Haley, Culver et al. 1985)
6
(Leu, Kaiser et al. 1989)
7
Pneumonia
(Jarvis 1996)
6.8 - 30
(Heyland, Cook et al. 1999)
4.7
(Warren, Shukla et al. 2003)
25
(Murphy, Whiting et al. 2007)
25.7
(Townsend and Wenzel 1981)
20
(Haley, Culver et al. 1985)
7
(Pittet, Tarara et al. 1994)
24
Bloodstream
(Jarvis 1996)
7 - 21
Infections (BSIs)
(DiGiovine, Chenoweth et al. 1999)
10
(Zhan and Miller 2003)
10.89
(Wisplinghoff, Bischoff et al. 2004)
7.5 - 25
(Murphy, Whiting et al. 2007)
25.8
(Haley, Culver et al. 1985)
7
(Jarvis 1996)
7 – 8.2
Surgical Site
(Kirkland, Briggs et al. 1999)
12
Infections (SSI)
(Whitehouse, Friedman et al. 2002)
14
(Murphy, Whiting et al. 2007)
7.5
(Givens and Wenzel 1980)
2.4
Urinary Tract
(Haley, Culver et al. 1985)
1
Infections (UTIs)
(Jarvis 1996)
1-4
(Murphy, Whiting et al. 2007)
12.5
(Haley, Culver et al. 1985)
4
(Malone and Larson 1996)
6.5 - 7
(Richards, Edwards et al. 2000)
2.3
Overall Infections
Due to Medical
(Zhan and Miller 2003)
9.58
Care
(Murphy, Whiting et al. 2007)
18
(Kilgore, Ghosh et al. 2008)
5.4
(Schwegman 2008)
7.4 - 9.4
Multiple Infections (Murphy, Whiting et al. 2007)
35.9

2.4.2 Patient Costs
Nosocomial infections drive up patient costs due the additional treatments and
duration of care, which also contributes to the high cost of healthcare in the United
States. Total healthcare costs and their proportion of the gross domestic product (GDP)
14

have been increasing in the United States and healthcare spending is expected to
outpace economic growth by 1.9 percentage points annually (Keehan, Sisko et al. 2008).
In 2002, total US healthcare costs accounted for approximately 13.5% of the GDP, or
$1.1 trillion (Stone, Larson et al. 2002). In 2007, total healthcare spending in the US was
$2.4 trillion, or $7,900 per person, and 17% of the GDP. In 2009, it increased to $2.5
trillion, or 17.6% GDP (National Coalition on Health Care 2009). By 2017, it is expected
to increase to $4.3 trillion, or 20% of the GDP (National Coalition on Health Care 2009).
Estimates of excess healthcare costs due to nosocomial infections are reported
to be between $4.5 and $5.7 billion annually (Soule 2008); however, estimates are
difficult to obtain, as the underlying condition of each infected patient affects the
duration, type, and costs of their treatment (Anderson, Kirkland et al. 2007). It is also
believed that many figures are underestimated, since they do not always include the
additional costs of outpatient care, in-home care, readmission (Kirkland, Briggs et al.
1999), additional procedures, physical therapy, and intravenous antibiotic therapy
(Whitehouse, Friedman et al. 2002).
Scott (2009) estimates total annual costs (in 2007 US $) for inpatient services in
US hospitals associated with specific sites of HAIs as:
Ventilator-associated pneumonia (VAP): $780 million to $1.5 billion
Central line-associated bloodstream infection (CLABSI): $590 million to $2.68
billion
Surgical site infection (SSI): $3.2 to $10 billion
Catheter-associated urinary tract infection (CAUTI): $340 million to $450 million
15

Clostridium difficile-associated disease (CID): $1.01 to $1.62 billion (Scott 2009)
Published studies indicate ranges of attributable costs (services provided and
billed to a patient that were caused by an HAI (Murphy, Whiting et al. 2007) by type of
infection and overall (Table 3). Estimates of attributable costs per patient range from:
$4,947 to $25,072 for pneumonia, $3,061 to $52,727 for BSIs, $2,734 to $27,969 for
SSIs, and $558 to $1,006 for UTIs. Estimates of overall attributable costs per patient due
to medical care range from $1,833 to $38,656.
Table 3. Attributable costs due to nosocomial infections by infection type.
Infection Type
Authors and Year
Attributable Costs
(Haley, Culver et al. 1985)
$4,947
(Stone, Larson et al. 2002)
$17,677
Pneumonia
(Warren, Shukla et al. 2003)
$11,897
(Stone, Braccia et al. 2005)
$9,969
(Anderson, Kirkland et al. 2007)
$25,072
(Haley, Culver et al. 1985)
$3,061
(Pittet, Tarara et al. 1994)
$40,000
(DiGiovine, Chenoweth et al. 1999)
$34,508
(Stone, Larson et al. 2002)
$38,703
Bloodstream
Infections (BSIs)
(Zhan and Miller 2003)
$52,727
(Hu, Veenstra et al. 2004)
$5,374 - $22,939
(Stone, Braccia et al. 2005)
$36,441
(Anderson, Kirkland et al. 2007)
$23,242
(Haley, Culver et al. 1985)
$2,734
(Kirkland, Briggs et al. 1999)
$5,038
(Hollenbeak, Murphy et al. 2002)
$14,211 - $20,103
Surgical Site
(Stone, Larson et al. 2002)
$15,646
Infections (SSIs)
(Whitehouse, Friedman et al. 2002)
$27,969
(Stone, Braccia et al. 2005)
$25,546
(Anderson, Kirkland et al. 2007)
$10,443
(Givens and Wenzel 1980)
$558
(Haley, Culver et al. 1985)
$593
Urinary Tract
Infections (UTIs)
(Stone, Braccia et al. 2005)
$1,006
(Anderson, Kirkland et al. 2007)
$758

16

Table 3 (cont’d).
Overall Infections
Due to Medical Care

(Haley, Culver et al. 1985)
(Stone, Larson et al. 2002)
(Zhan and Miller 2003)
(Murphy, Whiting et al. 2007)
(Kilgore, Ghosh et al. 2008)

$1,833
$13,973
$38,656
$8,832
$12,197

2.4.3 Hospital Costs
Nosocomial infections also consume hospital resources as a result of the
additional diagnostic and therapeutic interventions they require (Graves 2004). The
overall direct cost of HAIs for US hospitals has been estimated between $28.4 and $45
billion annually (Scott 2009). HAIs have also been estimated to reduce overall net
inpatient margins by $1,779 (Jarvis 1996) to $5,018 per infected patient (Murphy,
Whiting et al. 2007).
Malpractice lawsuits add to the mounting costs. The average 250-bed hospital
(or its malpractice carrier) spends between $300,000 and $1 million annually defending
malpractice lawsuits, not including the settlement and judgment amounts. Nosocomial
infections are listed as the eighth leading malpractice claim. Glabman indicates the
following:
1. Medication Errors
2. Diagnosis Failures
3. Negligent Supervision
4. Delayed Treatment
5. Failure to Obtain Consent
6. Lack of Proper Credentialing or Technical Skill
17

7. Unexpected Death
8. Iatrogenic Injury, Nosocomial and Wound Infections, Fractures
9. Pain and Suffering, Emotional Distress
10. Lack of Teamwork, Communication (Glabman 2004)
2.4.4 Socioeconomic Impact
Socioeconomic costs are composed of not only direct medical costs for
treatment, but also indirect costs related to loss of productivity for the patient and nonmedical costs and intangible costs related to diminished quality of life. While the direct
medical costs are more easily quantified, the indirect and intangible costs consist of a
series of rippling consequences that are often overlooked and difficult to calculate.
Indirect costs include not only lost time, wages, and productivity for the patient, but
also for their family members who take time off from work to travel to visit the patient
and become caregivers. Additionally, patient morbidity also reduces leisure time and
can lead to death. Intangible costs are those that have no monetary value but can have
a permanent impact on the patient and their loved ones. These include psychological
pain and suffering and changes in social functioning and the ability to perform daily
routines (Scott 2009).
2.4.5 Mortality
Nosocomial infections are a significant cause of mortality in the United States
(Klevens, Edwards et al. 2007). They are fourth leading cause of death in the US
following heart disease, cancer, and stroke (Jarvis 1996). In 2002, there were 1.7 million
HAIs resulting in 98,987 deaths (Klevens, Edwards et al. 2007). It is estimated that
18

100,000 deaths occur annually in the US as a result of nosocomial infections (Murphy,
Whiting et al. 2007).
As with the other estimates reviewed, reported attributable mortality rates vary
widely and depend largely on the type of infection (Table 4). In 1994, the literature
stated that mortality was highly probable for pneumonia, doubtful for bacteremia
(bloodstream infections), and uncertain for urinary tract infections (Fagon, Novara et al.
1994); however, further review of the literature indicates that pneumonia and
bloodstream infections have high rates of attributable mortality while UTIs have not yet
had attributable mortality rates linked to them in a publication. The ranges of
attributable mortality rates for patients with infections are: 5.8% to 60% for
pneumonia, 4.4% to 35% for BSIs, and 3.4% to 4.3% for SSIs. Attributable mortality
rates due to infections in surgical and medical ICUs are estimated at 7% while overall
nosocomial infection mortality rates range from 4.31 to 5%.
Table 4. Attributable mortality rates due to nosocomial infections by infection type.
Infection Type
Authors
Attributable Mortality
(Craig and Connelly 1984)
14.7%
(Leu, Kaiser et al. 1989)
6.8%
(Kollef 1993)
28.7%
Pneumonia
(Jarvis 1996)
6.8 - 30%
(Heyland, Cook et al. 1999)
5.8%
(Warren, Shukla et al. 2003)
16%
(Murphy, Whiting et al. 2007)
20-60%

19

Table 4 (cont’d).

Bloodstream
Infections (BSIs)

Surgical Site
Infections (SSIs)
Infections in Surgical
and Medical ICUs
Overall Infections
Due to Medical Care

(Townsend and Wenzel 1981)
(Smith, Meixler et al. 1991)
(Pittet, Tarara et al. 1994)
(Jarvis 1996)
(DiGiovine, Chenoweth et al. 1999)
(Zhan and Miller 2003)
(Wisplinghoff, Bischoff et al. 2004)
(Murphy, Whiting et al. 2007)
(Kirkland, Briggs et al. 1999)
(Whitehouse, Friedman et al. 2002)
(Craven, Kunches et al. 1988)
(Zhan and Miller 2003)
(Schwegman 2008)

21%
29.5%
35%
16.3 - 35%
4.4%
21.92%
27%
20 - 60%
4.3%
3.4%
7%
4.31%
5%

2.4.6 Staph Infections
Staphylococcus aureus is reported as the most common cause of nosocomial
infection in the United States (Cosgrove, Qi et al. 2005); as a result, researchers often
study its impact as related to the previously discussed consequences. Staph infections,
including methicillin-resistant staphylococcus aureus (MRSA), occur most frequently in
hospitals and healthcare settings in patients who have already weakened immune
systems (Division of Healthcare Quality Promotion 2010). They are the primary cause of
nosocomial pneumonia and surgical site infections and the second leading cause of
bloodstream infections (Cosgrove, Qi et al. 2005). Wisplinghoff et al. estimate that 20%
of bloodstream infections in hospital settings are caused by S. aureus (Wisplinghoff,
Bischoff et al. 2004).
Some researchers have studied how nosocomial S. aureus infections impact
economic burden, length of stay, and mortality. Rubin et al. estimated that S. aureus

20

infections double the length of stay, number of deaths, and costs of typical
hospitalizations. Researchers estimated the direct costs to be $28,800 per patient with
a mortality rate of 10.1%. When S. aureus infections were methicillin-resistant,
treatment costs increased to $31,400 per patient. Comparatively, a typical hospital stay
(without infection) cost $13,263, had a length of stay of 9 days, and had a mortality rate
of 4.1% (Rubin, Harrington et al. 1999).
Noskin et al. also evaluated S. aureus economic burden and mortality rates and
compared inpatient stays with surgical stays. Researchers estimated that the cost per
infected patient was $37,352 for all inpatient stays and $40,637 for all surgical stays.
Mortality rates were approximately equal, with a rate of 5.6% for all inpatient stays and
5.5% for surgical stays (Noskin, Rubin et al. 2007).
2.4.6.1 MRSA
Multidrug-resistant organisms (MDROs) are becoming a greater concern for HAIs
because they are becoming more prevalent. They can cause any type of HAI including
skin infections, BSIs, pneumonia, SSIs, and UTIs. They also exacerbate the consequences
of HAIs leading to even longer hospital stays, higher mortality rates, and higher
treatment costs because they are more difficult to treat (Bascetta, Edwards et al. 2008).
In fact, antimicrobial resistant infections are estimated to cost approximately $6,000 to
$30,000 more to treat than antimicrobial susceptible organisms (Cosgrove 2006).
Methicillin-resistant staphylococcus aureus (MRSA) is considered an MDRO and
is resistant to methicillin, oxacillin, penicillin, and amoxicillin (Division of Healthcare
Quality Promotion 2010). MRSA is one of the most common antimicrobial resistant
21

pathogens and occurs most frequently in patients who have undergone invasive
procedures or have weakened immune systems and are treated in hospitals and
healthcare facilities (Williams 2008). It reportedly causes three of the top four
categories of nosocomial infections: bloodstream, surgical site, and pneumonia
(Division of Healthcare Quality Promotion 2007). The main mode of transmission to
patients is through healthcare workers’ hands which can become contaminated and
transport microorganisms between patients, personnel, devices, other items, and
surfaces (Division of Healthcare Quality Promotion 2007).
One intermediate object that can serve as a pathway for microorganisms is cell
phones. In 2009, Ulger et al. published a study on health care workers’ cell phone
contamination. Two hundred participants and their cell phones were cultured and
nosocomial pathogens were found on several of the phones. Researchers found that
94.5% of healthcare workers’ cell phones had bacterial contamination and that rates
were similar between hands and phones. Fifty-two percent of the S. aureus strains
isolated from the phones and 37.7% from the healthcare workers’ hands were MRSA.
Researchers concluded that cell phones may be a source of nosocomial infections in
hospitals (Ulger, Esen et al. 2009).
MRSA infections in US intensive care units increased from 35.5% to 64.4%
between 1992 and 2003 (Klevens, Edwards et al. 2006). The Division of Healthcare
Quality Promotion reported that in 1974, MRSA represented 2% of all staph infections
and increased to 22% by 1995, and 63% by 2004 (Figure 4) (Division of Healthcare

22

Quality Promotion 2007). Clearly, the resistance of this organism is a growing cause for
concern.
70%

63%

Percent MRSA, %

60%
50%
40%
30%

MRSA

22%

20%
10%

2%

0%
1974

1995

2004

Figure 4. Percent of methicillin-resistant staph infections (Division of Healthcare
Quality Promotion 2007).
The national impact that MRSA will have is unknown (Kuehnert, Hill et al. 2005).
Evidence does suggest, however, that resistant infections result in higher mortality rates
(Rubin, Harrington et al. 1999). Stone et al. reported that the attributable cost of a
MRSA infection is $35,367 (Stone, Larson et al. 2002). MRSA has also been estimated to
add 2.7 million extra days in the hospital annually (Schwegman 2008).
Klevens et al. examined healthcare-associated infections in 16.5 million patients,
or approximately 5.6% of the US population. There were 8,987 cases of MRSA reported
during the surveillance. Of these, 92.4% of the patients were hospitalized, 17.8% died,
and 12.9% developed recurrent invasive infections. After adjusting for age, race, and
sex to the US population, researchers estimated that there were 94,360 invasive MRSA
infections in the United States in 2005 that resulted in 18,650 deaths (Klevens, Morrison
23

et al. 2007). The CDC estimates that 86% of MRSA infections were healthcare
associated and uses these figures as an annual estimate of the prevalence and impact of
MRSA (Division of Healthcare Quality Promotion 2007).
2.4.6.1.1 Impact of MRSA
Researchers have published reports comparing the impacts of drug resistant and
drug susceptible S. aureus infections, MRSA and MSSA, respectively. Cosgrove et al.
compared the impacts of methicillin-susceptible staphylococcus aureus (MSSA) and
MRSA on length of stay, mortality, and costs. Researchers examined data collected from
348 patients with S. aureus bloodstream infections; 73% contracted MSSA while 28%
contracted MRSA. Cosgrove et al. concluded that the median attributable impact of
MRSA, compared to MSSA, included an additional 2 days LOS (for patients who survived)
(p = .045), hospital charges of $6,916 per patient (p = .008), and hospital costs of $3,836
per patient. Mortality rates were similar (p = .53) for both MSSA (19.8%) and MRSA
(22.9%) (Cosgrove, Qi et al. 2005).
Engemann et al. compared the impacts of MSSA and MRSA on length of stay,
mortality, and costs for patients who underwent a surgical procedure and subsequently
tested positive for S. aureus. There were 479 usable patients: 40% were control
subjects (uninfected patients who underwent similar procedures), 35% had a MSSA SSI,
and 25% had a MRSA SSI. Researchers reported that the median LOS after surgery was
5, 14, and 23 days, respectively. The median LOS after infection was 10 days for MSSA
and 15 days for MRSA. The mortality rates were 2.1% for the control group, 6.7% for
MSSA patients, and 20.7% for MRSA patients. Researchers concluded that the adjusted
24

mean attributable excess hospital charges per patient were $13,901 for MRSA
compared with MSSA and $41,274 for MRSA compared with the control group
(Engemann, Carmeli et al. 2003).
2.4.6.2 Staphylococcus epidermidis
In 2002, Vuong and Otto identified Staphylococcus epidermidis as becoming the
most important cause of nosocomial infections. It is linked with bloodstream,
cardiovascular, eye, ear, nose, and throat infections, typically in patients who are
already compromised. While it is a normal bacterial flora that colonizes on the skin and
mucous membranes of humans, it can change from being innocuous to infectious
through the formation of biofilm (the main virulence factor) on indwelling intravascular
catheters and medical devices. Treatment is difficult because the biofilm protects
against attacks from the immune system and creates an impermeable barrier to many
antibiotics. Approximately 80% of nosocomial S. epidermidis infections are resistant to
methicillin and other antibiotics. When these infections occur, the device usually has to
be removed and replaced (Vuong and Otto 2002).
2.5 Prevention of Nosocomial Infections
Opposing philosophies exist regarding HAIs. Some believe that “many infections
are inevitable, although some can be prevented” while others believe “each infection is
potentially preventable until proven otherwise” (Gerberding 2002). Either way,
however, “prevention is better than cure” because it saves money and prevents pain,
suffering, and death (Harbath, Sax et al. 2003).

25

Nosocomial infections are a preventable cause of morbidity and mortality.
Harbath estimates that the range of preventable nosocomial infections is between 1070% and at least 20% of are “probably preventable” (Harbath, Sax et al. 2003). Jarvis
estimates that full implementation of current infection control recommendations would
reduce approximately one third of preventable nosocomial infections (Jarvis 1996).
In addition to the noble goals of reducing morbidity and mortality, most
prevention strategies are likely cost-attractive as well (Harbath, Sax et al. 2003). In the
US, it has been estimated that a 20% prevention rate of HAIs would save between $5.7
to $6.8 billion annually, and $25 to $31.5 billion would be saved if 70% of HAIs were
prevented (Scott 2009).
2.5.1 Infection Control Programs
In 1976, the Joint Commission on Accreditation of Healthcare Organizations
published infection control standards. These standards demanded the creation of
administrative and financial support for infection control programs. Early data,
published by the CDC in 1985, examined the efficacy of nosocomial infection control.
The report indicated that rates were reduced by one third when hospitals had an
effective epidemiologist, one infection control practitioner for every 250 beds, active
surveillance methods, and ongoing control efforts (Weinstein 1998).
CDC recommendations have evolved to now encompass 13 guidelines (based on
scientific evidence) for infection control and prevention for hospitals. Examples include:
Guideline for Prevention of Catheter-associated Urinary Tract Infections (1981)
Guideline for Prevention of Surgical Site Infection (1999)
26

Guidelines for Environmental Infection Control in Health-Care Facilities (2003)
Guidelines for Preventing Health-Care-Associated Pneumonia (2003)
Management of Multidrug-Resistant Organisms in Healthcare Settings (2006)
Guideline for Isolation Precautions: Preventing Transmission of Infectious Agents
in Healthcare Settings (2007)
These guidelines include nearly 1,200 implementation practices to prevent HAIs,
such as using alcohol-based hand sanitizers, isolating infected patients, properly
sterilizing equipment, treating patients with antibiotics prior to surgeries, and annually
vaccinating healthcare workers for the flu (Bascetta, Kohn et al. 2008). Notably, there is
no reference to packaging or sterile barriers and the transmission of HAIs. Healthcare
workers seem to trust that the devices are sterile at the point of use and are not
suspicious otherwise.
2.5.1.1 Antimicrobial Use and Resistance
A significant area of emphasis in infection control programs is the appropriate
use of antimicrobials. This emphasis is largely the result of reports that suggest that a
severe threat exists for a post-antibiotic era and aggressive antibiotic control programs
are needed to prevent this (Weinstein 1998). The World Health Organization (WHO)
reported in 2002 that some diseases will have no effective therapies within the next ten
years and issued the following statement: “Most alarming of all are diseases where
resistance is developing for virtually all currently available drugs, thus raising the spectre
of a post anti-biotic era” (World Health Organization 2002).

27

One of the causes of the rapid increase in resistance is from widespread overuse
of antibiotics by medical professionals. Hospitals are identified as critical contributors to
antimicrobial resistance because they practice routine intensive antimicrobial use,
contain already susceptible patients, and have occurrences of cross-infection. These
combined factors have resulted in nosocomial infections which are highly resistant to
treatment and are not only expensive to control, but extremely difficult to eradicate as
well (World Health Organization 2002). Additionally, one half of the antibiotics used in
the US to treat humans are also used to treat domestic animals. This concurrent use
enhances selection for drug-resistant microbes, further exacerbating the problem of
resistance (Pimentel, Tort et al. 1998).
Due to the development of antimicrobial resistance, antibiotic stewardship is
becoming increasingly important. Antibiotic control methods include prescription
tracking and restricting the use of certain antibiotics (Bascetta, Edwards et al. 2008);
however, prompt and appropriate antimicrobial therapy is also required when
infections do develop (Gerberding 2002).
To treat infections, inexpensive and effective first-line drugs are typically used;
however, these drugs tend to be the treatments that microorganisms are becoming
resistant to. When first-line drugs fail, second- or third-line drugs are implemented, but
these are almost always more expensive (sometimes by 100 times) and can also be
more toxic (World Health Organization 2002).
When antibiotic therapies fail, the patients and American society both
experience additional burdens. Healthcare system expenses increase (Division of
28

Healthcare Quality Promotion 2007) and infected patients are more likely to experience
prolonged illnesses that result in increased lengths of stay, costs, and risk of death
(World Health Organization 2002; Division of Healthcare Quality Promotion 2007).
Additional factors that may contribute to these consequences include increased toxicity,
improper dosing, and delay in treatment (Cosgrove 2006).
Antimicrobial resistance develops when microbes adapt as a result of the drug
treatment used; the microorganisms are forced to either adapt or die. Those that adapt
then carry genes for resistance and can no longer be killed with standard antimicrobial
treatments. Bacteria are efficient in creating resistance because they can multiply
quickly and pass on their resistance genes during replication. Additionally, resistant
bacteria can pass on their resistance genes through a process called conjugation; here,
plasmids carrying the genes jump from one bacterium to another (World Health
Organization 2002).
The NNIS System, referenced previously, has reported on antimicrobial
resistance of pathogens associated with nosocomial infections in ICUs. Researchers
suggest that antimicrobial resistance is continuing to increase for 8 of the 9
drug/pathogen combinations reported (Figure 5) (NNIS System 2001; NNIS System
2004).

29

87.5
89.1

100%
90%

60%

10%

3.4
5.8

20%

26.4
31.9

27.3
29.5

30%

17.7
21.1

40%

11.2
20.6

50%
26.3
28.5

% Resistance

70%

34.9
31.1

55.3
59.5

80%

2001

2004

0%

Figure 5. Comparison of NNIS System antimicrobial resistance rates in 2001 and 2004
(NNIS System 2001; NNIS System 2004).
While antimicrobial resistance has been increasing, the rate at which it has been
occurring appears to have slowed. When comparing reports, the rate at which
resistance progressed has decreased in 8 of the 9 drug/pathogen combinations (Figure
6) (NNIS System 2001; NNIS System 2004).

30

53
47
23
15

15

2001

9

20%

24
20

29

30%

11

40%

31

50%

12

5

10%

2004

0

1
1

% Increase in Resistance

60%

-1

0%
-6

-10%

Figure 6. Comparison of NNIS System increase in antimicrobial resistance in 2001 and
2004 (NNIS System 2001; NNIS System 2004).
In 2008, Klevens et al. used the NNIS System data to compare antimicrobialresistant infection levels for three specific types of infection between the periods of
1990-1994 (207 hospitals) and 2000-2004 (292 hospitals). A significant increase
(p<.001) in the proportion of antimicrobial resistant infections was found for methicillinresistant S. aureus bloodstream infections (MRSA BSI), ceftazidime-resistant P.
aeruginosa pneumonia (CRPA pneumonia), and ciprofloxacin-resistant E. coli urinary
tract infections (CREC UTI). Their resistance rates increased from 27% to 54.1%, 16.6%
to 22.7%, and 0.9% to 9.8%, respectively (Figure 7) (Klevens, Edwards et al. 2008).

31

% Antimicrobial Resistance

60

54.1

50
40
30

27

1990-1994

22.7

2000-2004

16.6

20

9.8

10
0.9
0
MRSA BSI

CRPA pneumonia

CREC UTI

Figure 7. Comparison of antimicrobial-resistance rates between 1990-1994 and 20002004 (Klevens, Edwards et al. 2008).
2.5.1.2 Other Prevention Measures
Appropriate antibiotic use and hand hygiene are significant parts of infection
control programs. Those related to hands include general washing, antisepsis, and
hygiene techniques, choice of hygiene agents, and surgical hand antisepsis (Schwegman
2008). Other infection prevention and control measures include:
Continual improvement of surveillance so data is increasingly representative and
ensuring the surveillance uses are valid (Weinstein 1998)
Contact precautions such as the use of gloves, gowns, and masks (Bascetta,
Edwards et al. 2008)
Enhanced environmental cleaning (Bascetta, Edwards et al. 2008)
Appropriate handling of laundry (Division of Healthcare Quality Promotion 2007)
Patient placement and transport (Division of Healthcare Quality Promotion 2007)

32

Appropriate handling of patient care equipment and instruments/devices
(Division of Healthcare Quality Promotion 2007).
As the care of devices was previously discussed as a potential source of infection,
here again, there is potential for infection prevention by gaining a greater
understanding of how sterile barriers fail and improving package/device designs which
aid in aseptic technique.
2.5.2 Device and Packaging Design
Improving the design of invasive devices is another option that is particularly
interesting and potentially quite innovative. “Given the choice of improving technology
or improving human behavior, technology is the better choice” (Weinstein 1998). The
theory behind this is, given the choice of changing human behavior (improving aseptic
technique or adhering to procedure) or designing a better device, the device will be the
more successful option (Weinstein 1998).
Devices with better functionality may have the potential to reduce the rates of
HAIs. One approach is to use materials that mitigate the colonization of
microorganisms. Another approach is to design devices which facilitate aseptic
technique. Again, this is an opportunity to change the environment so that the user is
successful in executing infection control procedures. Taking this a step further,
improving the design of device packaging has the potential for the same benefits. Here,
the device and its package could work in tandem, optimizing their functionality, so that
the package aids in device placement (Allen 2010).

33

2.5.3 Successful Prevention Initiatives
There is evidence that infection control programs are working. In a 500-bed
study institution in Arkansas, the introduction of the OSHA Control Plan significantly
lowered (p<.001) nosocomial infection rates from an average of 3.9% to 2.6% during a
three-year study (1991-1993) through the concomitant increase in glove use and
widespread use of barrier hand foam (Malone and Larson 1996).
There is also evidence that even small investments aimed at prevention have
been successful in reducing both the number of infections and associated costs. In
2004, the Jewish Health Foundation and Pittsburg Regional Health Initiative
implemented an evidence-based prevention measure across 40 hospitals to reduce
central line associated bloodstream infections (CLABs) which resulted in an overall
reduction of 63% (results were audited by the CDC). One participant in the initiative,
Allegheny General Hospital, also reduced ventilator-associated pneumonias (VAPs) by
82%. Their two- and one-year efforts to eliminate CLABs and VAPs, respectively,
resulted in a total cost savings of $2.2 million. The initial cost to obtain that savings was
only about $35,000 (Murphy, Whiting et al. 2007).
A similar outcome was noted when BJC Healthcare, a 13-hospital, non-profit
healthcare system based in St. Louis, MO increased resources to eliminate HAIs from
2000-2004. Their excess costs were estimated at $8.2 million for coronary artery bypass
graft (CABG) and spinal surgical site infections, bloodstream infections, and ventilator
associated pneumonia. They invested $350,000 across the system and individual
hospitals within it invested an additional $50,000-$150,000 to increase the number of
34

staff dedicated to infection prevention and/or medical direction during that period. This
reduced CABG SSIs by 18%, spinal SSIs by 61%, BSIs by 82%, and VAP by 52% resulting in
a cost savings of $2.5 million (Murphy, Whiting et al. 2007).
In addition to the cost savings, it is likely that these reduced rates of infection
also resulted in lower mortality rates, which raises the question, “What is the cost of
savings lives?” However, despite the focused efforts and improvements of infection
control programs, nosocomial infections still have not been eliminated. This suggests
that there are unknown contributors to HAIs which require investigation and supports
why examining the sterile barrier as a potential source of infections is necessary.
2.6 Packaging
Despite all of the efforts to prevent nosocomial infections from occurring, they
have not been eliminated and are becoming increasingly difficult to treat. It is also
difficult determine factors of significance in the causal pathway, such as tracing an
infection back to its source. There are widely accepted surveillance methods for
tracking and tracing nosocomial infections, but they are limited in scope, not sensitive,
and applied inconsistently (Brossette, Hacek et al. 2006). Devices have been cited as
sources of infection, both as modes of transmission and reservoirs, but the origin of the
pathogens and their transfer to the device are unknown and not widely studied. This
raises questions as to whether the devices become contaminated when they are
opened, or if their sterile barriers are breached prior to use.
Package integrity is the limiting factor in the maintenance of a product’s sterility
until use (Schneider 1980) and has been identified as the most important objective for
35

the medical device industry (Allen 2009). A sterile barrier system (SBS) is defined as the
“minimum package that prevents ingress of microorganisms and allows aseptic
presentation of the product at the point of use” (ISO 2006; ISO 2006). It is expected,
and even assumed, that the contents of a sterile barrier system remain sterile until the
package is opened or damaged, leading some to define a “perfect wrapper” as being
“impervious to extraneous microbes, liquid-proof, free of holes, free of lint, free of
memory, strong enough to resist punctures and tears, and economical to use” (Belkin
2004).
Several factors would have to occur to facilitate the spread of nosocomial
infections as a result of failed package integrity. Pathogenic microbes would need to
penetrate through or around the sterile barrier, remain viable within the package, and
adhere to the device inside the package. Then the device (or other intermediate object
which came into contact with the contaminated device, including personnel) would
need to transfer the pathogen into the patient’s body (Spitz 1994). The problem is that
there is no way of knowing if a sterility compromising event has occurred (Webster,
Lloyd et al. 2003).
2.6.1 Package Testing
Uncertainty exists as to what size hole is a threat for breaches of sterility. The
common belief is that any hole is a threat, no matter what its size is. As a result,
integrity tests are becoming increasingly sensitive (finding smaller and smaller holes);
however, even if a hole is present, is it possible for a microbe to travel through it, and if
so, what are the variables that facilitate this?
36

The FDA states that products labeled as sterile are expected to be free from
viable microbial contamination and that physical tests may be more useful than sterility
testing in demonstrating the potential for product contamination (U.S. Department of
Health and Human Services 2008). Device manufacturers test their packages to ensure
that there are no defects present and sterility of the product is maintained. Most
commonly, physical test methods are used, which include strength and integrity tests.
Strength tests investigate the strength of seals and resistance to bursting under
pressure, while integrity tests are used to identify pinholes or channel defects. Physical
tests are commonly used because they are established, economical, and convenient;
however, they lack significance because there is no proven correlation between
microbial penetration and sterile barrier properties. They identify if defects are present,
but not if microbes have penetrated, or can penetrate, though them (Placencia, Arin et
al. 1988). They do not specify the defects’ sizes or the accuracy of the measuring
methods (Axelson, Cavlin et al. 1990). It is necessary to know what the target leak size is
that presents a threat before test equipment can be designed and utilized to find them
(Floros 1994).
Additionally, consequences exist with integrity testing. Tests that pass packages
with defects can impact patient health because there is potential for microbial ingress to
occur, resulting in a nosocomial infection. Conversely, tests that fail packages that do
not present a risk drive up costs because those products are destroyed unnecessarily
(Severin, Bix et al. 2007). Between 1980-84, 31.4% of device recalls were due to holes in

37

the packaging (Placencia, Arin et al. 1988), but the percentage of those holes that were
actually a threat for microbial contamination is unknown.
Integrity tests can be destructive or non-destructive, performed on porous and
non-porous barriers, and identify channels and holes. Destructive tests are subjective,
typically with pass/fail results, and include bubble and dye testing. Here, companies
have to waste product in order to perform the tests. Nondestructive quantitative tests
include pressure, force, and vacuum decay methods, which are very sensitive and can
detect small leaks down to about 10 µm. Trace gas methods are even more sensitive,
detecting leaks a fraction of a micron in size. Visual inspection can only identify
channels as small as 75 micron with 60-100% probability and relies on the visual abilities
of the inspectors. Other methods include ultrasound/acoustic micro-imaging (Allen
2002), x-ray and thermal imaging (Allen 2003), and helium mass spectrometry (Franks
2003).
Strength tests are also used to assess the mechanical strength of seals and
ensure that they are properly bonded so that they will maintain integrity throughout life
of the product. These methods include tensile strength and burst or creep testing
(Franks 2002).
In 1995 Hansen et al. published a study comparing microbial challenge and
physical testing to identify channel defects in film pouch seals. They created defects
across the seals using metal shims and wires. Pouches each contained a paper towel
and plastic forceps. For the pouches subjected to microbial challenge, a nebulizer was
used to aerosolize Bacilus subtilis var. niger spores (ATCC #9372) with a starting
38

concentration 1 x 106 cells/ml for 30 minutes within the microbial challenge chamber.
Physical tests included dye penetration and visual inspection. To test for microbial
ingress, the paper towels and forceps were removed from the pouches after microbial
challenge testing and placed in soybean-casein digest broth and observed for growth.
Results indicated that two pouches containing defects, of the 150 pouches that
were microbial challenge tested (30 controls and 120 with seal defects), tested positive
for the test organism. A false positive was found in one of the control samples and
there were four cases of positive findings for organisms other than the test organism.
The same packages were also physically tested and the defects were correctly identified
in all of the samples, with no false positives for samples without defects.
Researchers concluded that physical tests are more effective than wholepackage microbial challenge testing when evaluating whole-package integrity because
of the inconsistency of the microbial challenge method in identifying defects, the
positive findings of organisms that were not the test organism, and the false positive in
the control pouch (Hansen, Jones et al. 1995).
2.6.2 Factors that Can Affect Microbial Ingress
When performing microbial challenge tests in an effort to understand how
sterility is compromised in in-use environments, there are a multitude of variables to
consider. These include packaging characteristics, defects, characteristics related to
microorganisms, torturous path, methods of microbial challenge, and factors related to
transport, storage, and handling.

39

2.6.2.1 Packaging Characteristics
The type and configuration of packaging materials are important choices for
sterile barriers (Webster, Lloyd et al. 2003) and characteristics to consider include
geometry, thickness, rigidity and porosity. Rigid trays and flexible pouches are common
medical device packages. Trays are typically silicon coated or uncoated polyethylene
terephthalate (PET) or glycol-modified polyethylene terephthalate (PETG) (Blocher
2009) that are sealed with porous or nonporous lids. PETG is used for 60% of all
preformed medical device trays and is resistant to the hazards associated with radiation
sterilization processes (Pilchik 1999). Pouches can be manufactured from a variety of
polymers, but a common configuration is a PET/LDPE (low density polyethylene) layer
(forming one side) sealed to porous layer (forming the other side). Other materials that
are common include Nylon and foil (Blocher 2009).
In 2006, Dunkelberg and Rohmann published a study comparing the penetration
of airborne microorganisms into sterile pouches made of transparent plastic film and
sterilization paper. The pouches, containing thermoresistant agar dishes, were
sterilized, placed in an exposure chamber, and subjected to an aerosol of
Saccharomyces cerevisiae (baker’s yeast) with a concentration of 108 cells/ml.
Atmospheric pressure inside the chamber was cycled through 48 pressure changes of 75
mbar. Following treatment, the samples were incubated and inspected for growth.
When agar dishes were placed with their open surfaces under the paper sides of
the pouches, 82 of 237 pouches (35%) had contamination. When the dishes were
placed under the film sides of the pouches, 3 of 166 pouches (2%) had contamination.
40

Researchers concluded that it was the paper, not plastic, sides of the pouches that
allowed the ingress of microorganisms (Dunkelberg and Rohmann 2006).
2.6.2.2 Packaging Defects
The type and size of defects in the packages are also factors for consideration.
Package defects can come in all shapes and sizes and be present anywhere on the
package including surfaces and seals. Defect types include pinholes, cracks, sealing
failures, and defects in barrier layers (Axelson, Cavlin et al. 1990). Holes or defects in
packaging not only increase the risk of microbial contamination and nosocomial
infections as a result, but they can also create economic loss due to the costs of treating
infections and in the event of product recalls (Chen, Harte et al. 1991). Conversely,
microscopic holes may not impair sterile package integrity because a driving force is
needed for contamination and a myriad of other factors could affect this. Overly
stringent sensitivity levels for package integrity testing may actually reject packages that
are effective in maintaining sterility, resulting in economic loss (Jones, Hansen et al.
1995).
Tested defects are commonly pinholes through the sterile barrier or channels
within the package seal and their shapes, diameters, and lengths likely affect the ability
of microorganisms to travel through them. Allen-Wojtas et al. examined flow rate
through microperforations in plastic films. Researchers found that flow rate is
dependent on geometric features of both the entry and exit sides of the holes. As area
decreases, proportionally more gas molecules are affected by the side walls, which
results in a greater than linear reduction in transport rate, essentially increasing
41

viscosity. Perforations >55 µm in diameter demonstrated that air convection
contributed more strongly to gas transfer. Researchers concluded that small holes are
also less likely to transmit contamination and both perimeter and shape may also be
factors in transmission (Allan-Wojtas, Forney et al. 2008).
2.6.2.2.1 Creating and Measuring Hole Defects in Packaging
When performing tests to quantify microbial ingress into sterile packaging, it is
necessary to produce hole defects that are repeatable and consistent. Producing very
small defects (<75 µm) can be difficult (Jones, Hansen et al. 1995). The most common
methods include use of orifices or capillaries, mechanical punctures, and laser drilling.
When producing holes, both cold and hot needles are slow and only create large
perforations (≥1 mm diameter). Hot needles melt the plastic to form holes and then
redeposit it as large rims around the hole edges. Cold needles punch rough holes
through the film where flaps remain that can cover the holes. Lasers use heat energy to
evaporate the plastic (ablation) to produce small, clean holes that are sealed along the
edges. Depending on the type of plastic, it may completely evaporate or some may
redeposit on the film surface. Microelectric discharge machining (micro-EDM), also
referred to as spark machining or spark eroding, is a relatively new technique for
microperforations in plastic films (Allan-Wojtas, Forney et al. 2008).
Several researchers have used a variety of methods to create and measure holes
in packaging. Keller et al. used nickel microtubes to create seal defects in flexible
pouches with internal diameters of 10 and 20 ±2 µm and lengths of 5 and 10 ±1 µm

42

(Keller, Marcy et al. 1996). Chen et al. used nickel alloy orifices measuring 5, 10, and 15
µm to create defects in flattop cartons (Chen, Harte et al. 1991).
Maunder et al. created holes in flexible pouches using a No. 14 sewing needle
held in hemostatic forceps so that 1.5mm of the needle point was exposed. This
resulted in diameters measuring 33-160 µm, with the longest dimension being recorded
for non-round holes (Maunder, Folinazzo et al. 1968).
Hansen et al. created channels across the seals of film pouches using metal
shims and wires that were 0.375 to 0.005 inches in diameter. They found that the wire
channels could become blocked with adhesive and their average widths were from
0.022 to 0.439 inches. Their method of measurement was done by peeling the package
open and measuring the channels at outer edges under 10x and 50x magnification
(Hansen, Jones et al. 1995).
Ahvenainen et al. created holes in plastic cups using touch needles with point
radii varying down to 2 µm. The resulting diameters were 5-100 µm and were measured
using 120x magnification and a scale in the ocular. They examined the hole shapes
using 198x magnification and determined that touch needles can only be used with very
thin packaging material, or the hole will be conically shaped (Ahvenainen, MattilaSandholm et al. 1992).
Axelson et al. also created holes in plastic cups using honed sewing needles and
touch needles with point radii of about 2 µm, resulted in diameters of 5-10 µm.
Additionally, they used capillary columns measuring ≥25 µm in diameter and drilled
holes measuring 300 µm in diameter. All holes were measured with a microscope.
43

Axelson et al. also measured holes using electrolytic conductance. This method
is based on the theory that the electrolytic conductance of a packaging made of an
insulating material is drastically changed by a small hole, provided the hole is filled with
a conducting liquid. It also requires knowledge of length of hole and can measure down
to 0.8 µm if the holes are not longer than 1 mm (Axelson, Cavlin et al. 1990).
Gilchrist et al. used a laser beam to create holes in trilaminate pouches (PE,
aluminum, and PP) that were approximately round with no tears or flaps; they
measured 17-81 µm in diameter. They also punched holes using a stainless steel wire
(0.004 inch diameter) that did result in tears and flaps; these measured 22-175 µm in
diameter (Gilchrist, Shah et al. 1989).
Lampi used a laser to create holes in pouch blanks. Preliminary sizing was
conducted with a flat field microscope. Then the gas flow rates through the holes were
measured, scanning electron microphotographing of defects was conducted, and the
hole diameters were calculated. Lampi also created defects by flexing the pouches
using an MIT Folding Endurance Tester, which resulted in the smallest hole size of 11
µm; however, it was determined that flexing was not a likely cause of failure because
1000-5000 flexes were required to create a failure. Additionally, Lampi punctured the
pouches with a fine fire-hardened tungsten wire resulting in the smallest hole size of
100 µm. Lastly, Lampi created holes using abrasion of fold in the pouches against a
relatively smooth fiberboard surface. This resulted in the smallest hole size of 24 µm
(Lampi 1981).

44

Bix et al. created 50µm holes in PETG trays using an excimer laser. The holes
were measured using SEM and confocal microscopes. Researchers found that
significant differences (p<.0001) existed between the two methods of measurement,
with the confocal images having more detail, resulting in larger measurements than
those from the SEM. Additionally, researchers found that the entry sides of the holes
were larger than the exit sides (p<.0001); the exit sides were closer to 50µm, while the
entry sides were 50-100% larger (Bix, Kassarjian et al. 2005).
Laser drilling is an important method for creating very small hole diameters
ranging from 10-50 µm, because conventional alternative methods become difficult and
cost inefficient. PET is a good candidate for laser drilling because it allows easy
observation with optical microscopes. Shorter wavelengths have the best precision for
lateral diameters, but sometimes the diameters end up being larger than intended. This
is likely the result of liquid film forming on the inner walls during drilling. The liquid
becomes accelerated toward the exit side of the hole by momentum transferred from
the outcoming gas plume, therefore contributing to an increase in hole diameter (Lazare
and Tokarev 2004).
2.6.2.2.2 Microscopy Methods for Measuring Holes
Various microscopic techniques have been employed to measure pinhole defects
in packaging. The main differences in microscopy techniques are resolution and depth
of focus/field (Allan-Wojtas, Forney et al. 2008). Some methods have been successful in
capturing images of the entry and exit sides of the holes on the surfaces of the
packaging materials; however, differences have been found between measurements of
45

the same holes using different types of microscopes. Additionally, differences in hole
diameters were found between the entry and exit sides of the holes (Bix, Kassarjian et
al. 2005). This raises questions as to what shape and size is within the entire hole
tunnel, in between what the microscopes can image at the surfaces.
Ghosh and Anantheswaran used a Ziess Photoscope II to measure
microperforations in film that were created with mechanical sparks. Researchers
printed the magnified images and measured hole diameters in four places 45 degrees
from one another and then averaged the values. A photograph of a 0.1 mm stage
micrometer was used for calibration. The perforations (which included film deposits
around the edges of the holes) were sliced through with a razor blade in preparation for
thickness measurements. Then the samples were coated with gold/palladium in a BALTEC SCD050 sputter coater and SEM images were obtained with a 10kV JOEL 5400 SEM.
Images were recorded on Polapan 100 film and transferred to a Princeton-Gamma Tech
Integrated Micro-analyzer for further imaging and x-ray to generate thickness
measurements. The cut edges were digitized to generate measurements in microns.
Five measurements were taken from random locations within perforations and their
averages were calculated. Their resulting measurements ranged from 96-247 µm in
diameter and 50.7-76.9 µm in thickness. Additionally, perforation thickness increased
as diameters increased, due to larger amounts of the film being deposited around the
microperforation edges when the holes were created (Ghosh and Anantheswaran 2001).
Piergiovanni et al. measured hot needle perforations in film using two imaging
techniques. First, researchers placed the film a on scanner, covered it with a mirror,
46

collected digitalized grey scale images with 1200 pixel/in2 resolution, and stored them in
TIF format to determine the density of the holes per cm2. Second, Dia-Brightfield
microscopy, with a computerized system for image processing and analysis, was used to
collect images for diameter measurements. Here, four measurements were also taken
45 degrees apart from one another and the averages were calculated; this was done for
both the internal and external sides of film. While no statistically significant differences
were identified between sides, the internal sides had more defined borders while the
external sides had uneven borders. Polarized light was then used to identify irregular
crests of melted polymer coming out from the horizontal planes. Researchers stated
that the presence of crests could reduce entry of contaminants through the holes. The
film was also subjected to artificial sweat contamination with a compressive machine
and cotton pads containing sweat simulant, and artificial sneezes with a manual aerosol
dispenser containing saliva simulant. Researchers concluded that the risk of
contamination was highly correlated to hole surface area and it was likely that artificial
saliva and sweat transmission through perforations is proportional to the hole
dimensions and densities (Piergiovanni, Limbo et al. 2003).
Allan-Wojtas et al. compared three microscopy methods (light, SEM, LV-SEM) to
measure microperforations in 1 mil polyethylene and polypropylene films. With light
microscopy, photons illuminate the sample and that interaction provides
microstructural information. Brightfield transmitted light microscopy (BFTLM) requires
minimal sample preparation and is best for thin samples that have some inherent
contrast or can be stained. Differential interference constrast light microscopy (DICLM)
47

can be used to observe local differences in thickness for samples with little contrast
because the edges between two areas with different refractive indices are enhanced.
The light microscopy techniques used included observing the upper and lower surfaces
of the films with 40x magnification under Brightfield conditions and then switching to
DICLM conditions using a Nikon Eclipse E800 light microscope equipped with a Nikon
DXM 1200 digital camera. Grayscale images were captured as digital files with 3600 x
2880 pixels (Allan-Wojtas, Forney et al. 2008).
Scanning electron microscopes (SEM) work by scanning the samples with an
electron beam and creating images from that interaction. SEM works well for thicker
samples because light clarity is better than with light microscopy. The SEM samples
were coated with gold/palladium using a Hummer VII sputter coater and observed with
a JEOL T330A SEM operated at 5kV. Images were recorded on Polaroid Type 55
positive/negative film, and the negatives were scanned into grayscale digital files (1320
x 1056 pixels) using a Polaroid SpringScan45 Ultra scanner (Allan-Wojtas, Forney et al.
2008).
Low-vacuum SEM (LV-SEM) was conducted using a Quanta 200 environmental
SEM (ESEM) under low-vacuum mode (20kV, 133.0 Pa). Both sides of the samples were
observed and images were captured as grayscale digital files (1024 x 884 pixels) (AllanWojtas, Forney et al. 2008).
These methods were used to determine the area, perimeters, and diameters of
the microperforations. Area was defined as the region bound by the plastic rim and
perimeter values were measured from the inside edges of the plastic rims. Hole shape
48

was also observed to determine roundness using an estimate of circularity (AllanWojtas, Forney et al. 2008).
Accurate measurements were not possible with Brightfield microscopy because
distinct hole edges were not visible; they were only in focus at the upper and lower
surfaces of film. Additionally, both sides of the holes varied in structure and size,
indicating the hole tunnels were not cylindrical (Allan-Wojtas, Forney et al. 2008).
DICLM was able to provided visible areas of contrast between the holes and their
rims, such that the edges and deposits of film around them were observed. This
method showed that the rims were thicker than the surrounding plastic and was
determined to be suitable for quick measurements (Allan-Wojtas, Forney et al. 2008).
The SEM showed distinct structural and topographical characteristics that were
clearly visible due to the increased depth of focus. This method was suitable for more
accurate diameter measurements because the hole edges were well defined (AllanWojtas, Forney et al. 2008).
While the LV-SEM offered same depth of focus as the SEM, it did not require
altering the samples to prepare them for imaging. This allowed the samples to be used
for other tests and they could also be quickly removed and remounted so that the other
sides could be observed. Researchers concluded that this was the best technique for
studying microperforations (Allan-Wojtas, Forney et al. 2008).
2.6.2.3 Torturous Path and Layers of Packaging
Torturous path is the physical difficulty of the “route” through the package that a
microorganism needs to travel in order to breach the sterile barrier. This can include
49

the microscopic routes through defects and material surfaces as well as relatively larger
routes across or around multiple layers of materials.
Discrepancies exist on the number of layers that a package needs to ensure
sterility. A single layer of packaging requires less labor and costs less, but it is generally
believed that a double layer is essential for the practice of asepsis. A survey of nurses
indicates that they prefer double layers over single layers because it allows them easier
aseptic presentation of the devices into the sterile field (Blocher, Neid et al. 2010). The
outer layer is intended to protect against dust particles that could contaminate the
contents when the package is opened and is supposed to be removed before the pack is
delivered to the clean zone (Belkin 2004). It has also been suggested that safe storage
time is significantly increased with multiple wraps because an increased number of
layers will provide a protective barrier that creates a torturous path for microorganisms
to travel through as well as prevent tears and punctures (Spitzley 2002). But, “if
something can be wrapped once instead of twice, provide just as good a barrier, and be
delivered to the field without compromising sterility, why take the time and expense to
wrap it twice?” (McCormack 1995).
At least one study suggests that multiple layers of packaging may actually inhibit
aseptic presentation. In 2008, Crick et al. published a study comparing the opening of
multiple, individually double-wrapped, orthopedic implant screws vs. screw banks (one
package containing multiple screws). Five nurses each bathed their hands in Glitterbug
(a cream that is visible under ultraviolet light) and opened 20 double wrapped screws
and one screw bank. Of the 100 double wrapped samples, one had contamination.
50

None of the screw banks had contamination. They concluded that opening individually
wrapped items increased the risk for potential contamination of an operative field as
compared to opening a single package and recommended further investigating a move
towards minimizing the number of separately wrapped sterile packages during a surgery
(Crick, Chua et al. 2008).
This idea runs counter to early research, which suggests that multiple layers of
packaging are necessary to increase storage time. In 1973, Standard et al. published a
study examining microbial penetration of sterile packs with multiple layers. The three
configurations were described as: double-wrap (two layers each) muslin, single-wrap
(two layers) muslin inner covering with single-wrap (one layer) two-way crepe paper
outer covering, and single-wrap (two layers) muslin inner covering with single-layer
BAR-BAC wrappers. These were used to wrap gauze sponges and stored on open
shelves in a central sterile supply department in a hospital. Packs were chosen in groups
of two or four at weekly intervals and assayed for growth.
The packs with double-wrap muslin had microbial penetration at 28 days. The
single-wrap (two layers) muslin inner covering with single-wrap (one layer) two-way
crepe paper outer covering had penetration at 77 days. The single-wrap (two layers)
muslin inner covering with single-layer BAR-BAC wrappers had penetration at 63 days.
With the objective of increasing storage time before sterility is compromised, these
results led to recommendations that include using sterile impervious plastic bags,
closed-cabinet storage, and more than single layers of packaging (Standard, Mallison et
al. 1973).
51

Other research further supports the design of systems composed of multiple
layers. In 2004, Dunkelberg and Wedekind published a study examining contamination
of sterile packages with multiple layers. The packages were described as wire baskets
with sterilization sheets affixed using adhesive tape. There were four configurations:
single-paper packaging (P), double-paper packaging (PP), textile and paper double
packaging (TP), and double packaging with transport packaging (TPP). They were
subjected to repetitive mechanical stress by being pressed approximately 1 to 2 cm five
times per minute with a 1 kg weight for 0, 1, and 3 hours. This was conducted in two
separate rooms with mean ambient airborne bacterial counts of 35 and 440 CFU/m 3.
Following the mechanical stress, the packages were incubated and inspected for
growth. Results indicated that mechanical stress led to time-dependent contamination
and that barrier efficacy increased with the number of layers of packaging (Dunkelberg
and Wedekind 2004).
In 2006, Dunkelberg published another study with Rohmann comparing the
performance of single- and double-layer pouches composed of transparent plastic film
and sterilization paper when atmospheric pressure changes were the driving force. The
pouches were sterilized containing thermoresistant agar dishes, placed into an exposure
chamber, and subjected to an aerosol of Saccharomyces cerevisiae (baker’s yeast) with a
concentration of 108 cells/ml. Twenty-five samples of each were exposed to 4 hours in
the exposure chamber with 12 atmospheric pressure changes of 75 mbars occurring per
hour. Following treatment, they were incubated and inspected for growth. Four
percent of the single-layer pouches had contamination while none of the double-layer
52

pouches did. They concluded that double-layer pouches will provide a sufficient sterility
assurance level in clinical settings (Dunkelberg and Rohmann 2006).
These studies examine the use of multiple layers of packaging in environments
which require aseptic presentation and in storage and transport environments. The
study performed by Crick et al. suggests that the more times that devices are opened,
the greater the probability for a contaminating event to occur. The studies performed
by Standard et al, Dunkelberg, Wedekind, and Rohmann suggest that the greater the
number of layers, the longer the devices will remain sterile in transport or storage
environments. This suggests that more layers of packaging will maintain the devices’
sterility up to the point of use, but it also suggests that increased layers of packaging at
the point of use increase the probability for a patient to acquire an infection.
2.6.2.4 Microorganisms
Characteristics related to microorganisms can also affect microbial ingress of
sterile packaging. Factors to consider include type, size, mode of motility,
concentration, the duration of exposure, and viability.
2.6.2.4.1 Nosocomial Microorganisms and Pathogenicity
Nosocomial infections can be bacterial, viral, or fungal. Bacterial infections are
those that are typically reported and are also becoming more worrisome due to the
rapid increase of drug resistance in these organisms. While there are seven classes of
antibiotics for treatment, there is resistance to at least one antibiotic in at least 70% of
HAIs (Kowalski). The most common bacterial infections in humans tend to be those that
are gaining antimicrobial resistance. Nosocomial infections are among these and
53

examples include: penicillin-resistant Streptococcus pneumoniae, vancomycin-resistant
enterococci, methicillin-resistant Staphylococcus aureus (World Health Organization
2002).
Viral infections also represent a significant fraction of HAIs and, like their
bacterial counterparts, have the potential to evolve rapidly, due to high mutation rates.
There are less vaccines available than there are viruses to treat because each variant of
a virus requires a specific antiviral agent and vaccine. Viral infections are also
worrisome because epidemics can occur before vaccines are developed (Kowalski 2007).
Fungal infections, while on the rise, are not typically considered nosocomial.
They are associated with immunodeficiency diseases that can facilitate their spread.
Like viruses, they have a small number of fungicidal agents that can be used for
treatment (Kowalski 2007).
The number of viable virulent microorganisms required for a 50% probability of
infecting a patient, also referred to as minimum infectivity number (MIN), depends on
the virulence of the microorganism, site of entry into the patient, health status of the
patient, and the normal route of infection upon entry (Allison 1999). Malone and Larsen
suggest that the microorganisms most commonly linked with nosocomial infections are
Enterococci, E. coli, Pseudomonas sp., S. aureus, and Candida sp. (Malone and Larson
1996). Richards suggests the most common pathogens and their corresponding
infections are: S. aureus for nosocomial pneumonia with a rate of 17%, coagulasenegative staphylococci for BSIs with a rate 39%, E. coli for UTIs with a rate of 19%, and
Enterococcus for SSIs with a rate of 17% (Richards, Edwards et al. 2000). Allison
54

suggests that Pseudomonas aeruginosa is of greatest concern for producers of sterile
products, particularly those for eyes (Allison 1999).
2.6.2.4.2 Test Microorganisms and Microbial Concentration
When choosing a test organism, not only does it need to be an accurate model
for a nosocomial pathogen, but its concentration and safety for use also need to be
considered. Like the other factors that are known to impact penetration rates,
researchers have used a variety of microorganisms and concentrations for package
integrity testing.
In 1996, Keller et al. published a study examining the effects of microorganism
motility and microbial concentration on ingress through microtubes in flexible pouches.
They compared motile and nonmotile Pseudomonas fragi at concentrations of 102 and
106 CFU/ml. Six of 127 pouches tested positive for contamination. All of these were
exposed to 106 CFU/ml. They concluded that contamination was affected by both level
of motility and concentration of the test organism (Keller, Marcy et al. 1996).
Chen et al. used Lactobacillus cellobiosus at a 2.5 x 106 cells/ml concentration
(Chen, Harte et al. 1991). Hansen et al. used Bacilus subtilis var. niger spores at a 1 x 106
cells/ml concentration (Hansen, Jones et al. 1995). Dunkelberg and Rohmann used
Saccharomyces cerevisiae (baker’s yeast, which is approximately 10 µm in diameter)
(Dunkelberg and Rohmann 2006). Dunkelberg and Schmelz used Micrococcus luteus at
a 108 cells/ml (Dunkelberg and Schmelz 2009).

55

2.6.2.4.3 Viability on Packaging
In addition to the questions surrounding the ability of microbes to breach the
sterile barrier of a package, there has also been uncertainty regarding whether or not
microbes will remain viable inside the package until the time of use. Evidence suggests
that their viability is a realistic concern, especially with one of the most threatening
bacteria in existance, MRSA.
In 2001, Dietze et al. published a study on the survival of MRSA on sterile goods
packaging. The packaging materials they used were single-use syringe wrappers
composed of a paper side and laminated side consisting of foil, polyethylene, and
polyamide. The samples were contaminated with a MRSA suspension by micropipette
(0.1 mL on the paper sides and 0.05 mL on the laminated sides) and stored in conditions
representative of a storage closet without light, dust-protected, and at 20°C and 20-30%
RH. At regular intervals samples were inspected and researchers found that the MRSA
survival time was greater than 38 weeks. At 50 weeks, there was no further survival.
Paper and foil had similar survival rates (Dietze, Rath et al. 2001).
2.6.2.5 Methods of Microbial Challenge
Microorganisms that may contaminate packages are present in the atmosphere
as both dust particles and water droplets (Floros 1994). When conducting microbial
challenge testing on packaging, there are a few representative options to choose from;
these typically include aerosols, submersions, or talc methods.
Literature indicates that aerosol methods are most commonly used for microbial
challenge of packaging. Understanding how aerosolized particles containing microbes
56

travel through microscopic defects in the packages is intended to be used to set realistic
limits on the critical leak size for package integrity (Floros 1994). Aerosols consist of
small particles suspended in a gaseous medium and the type of microorganism,
concentration and viscosity of the microbial suspension, its contact time with the test
packages, and pressure differences are all factors for consideration (Gnanasekharan and
Floros 1995).
The microbial challenge systems previous researchers have used generally
incorporated some form of chamber with a system for aerosolizing microbial solutions.
Keller et al. utilized a Plexiglass® chamber with two aerosol entry ports. The aerosol was
created using nebulizer kits (Keller, Marcy et al. 1996). Chen et al. also used a
Plexiglass® chamber. This had a recirculation system with two pumps that forced fluid
through 32 nozzles mounted on the all sides of the chamber (Chen, Harte et al. 1991).
2.6.2.6 Transport, Storage, and Handling
It is generally believed (and somewhat intuitive) that the probability of a sterility
compromising event occurring increases with both time and handling (Webster, Lloyd et
al. 2003) and that shelf life depends not only on the quality of the material, but also on
transport, storage, and handling conditions (Belkin 2004). Packages can be subjected to
risk during contact with atmospheres in processing, storage, distribution, and in-use
environments (Gnanasekharan and Floros 1995) and one of the factors of concern here
are pressure differentials across the sterile barrier.
Transportation conditions, such as shock, vibration, and compression, are
routinely simulated to test packages in laboratory settings; however, it is difficult to
57

simulate handling conditions. It is unknown how package handling within the hospital
environment affects sterility, such as when packages are put into bins or nurses’ pockets
(Allen 2010). Additionally, it is also unknown how the number of times a package is
handled before its final use affects sterility (Webster, Lloyd et al. 2003). Factors that
could impact this include conditions within the hospital environment, such as the
concentration of bacteria in ambient air (Dunkelberg and Wedekind 2004), and static
charges (Hansen, Jones et al. 1995).
The type of storage areas and their conditions may also have an effect on
microbial ingress and shelf life. Storage shelving can be open or closed cabinet
(Webster, Lloyd et al. 2003), and researchers suggest closed cabinet storage will offer
better protection for sterile packages (Standard, Mallison et al. 1973). The conditions of
the storage area are also relevant and include levels of cleanliness, temperature, and
humidity (Webster, Lloyd et al. 2003). Again, multiple layers of packaging may also
produce longer sterile storage, such as with device packs that are sealed into sterile
impervious plastic bags (Standard, Mallison et al. 1973).
Another factor that can compromise sterility is pressure differential across the
sterile barrier. This serves as a driving force to pull microorganisms through microscopic
defects in the packaging. Pressure differences can occur during transportation by
aircraft or in vehicles traveling over mountain passes. They can also occur within the
hospital through general handling, when a package is transferred to an area with
positive pressure such as an operating room with hyperbaric pressure, from transport
within an elevator (Dunkelberg and Fleitmann-Glende 2006), by opening and closing of
58

storage closets and drawers (Allen 2010), and from pressure pulses created by heating
and cooling systems (Spitz 1994). Weather changes can also create pressure
differentials (Dunkelberg and Wedekind 2004).
2.6.2.6.1 Pressure Differential
The flow of gas may be the most serious threat to sterility of a packaged device
(Hackett 2001). Microorganisms can be carried by air currents into packages or fall
through defects due to gravity (Hackett, Scholla et al. 2000). Flow is dependent on
pressure differential, headspace within the package, and time (Spitz 1994). When a
pressure differential is induced on a package, it forces headspace gas to flow out or
surrounding gas to flow in. This can change the geometry of the package, including
shape, volume, and rigidity, and the headspace and surrounding gas pressures
(Gnanasekharan and Floros 1995).
A major cause of pressure differential during transport is the ascent and descent
of aircrafts (Hackett 2001). Airborne microorganisms moving at very low velocities, low
pressure differentials induced from handling or air and ground transportation, can be a
threat for sterile packages (Hackett, Scholla et al. 2000). A pressure differential of -3.78
psi simulates and aircraft descent from 8,000 ft (Severin 2006) and researchers suggest
that approximately 3 lb of pressure is required to overcome capillarity in a 10 µm
diameter hole; therefore, a 3-4 psi pressure differential is reasonable for microbial
challenge testing (Lampi 1981).

59

2.6.2.6.1.1 Pressure Differential Research
Researchers have studied the effects of pressure differential on microbial ingress
of various sterile packages. Studies include the effects of pressure levels, the number of
pressure changes, combined effects of pressure and microbial concentration, and
simulated transport environments.
In 2006, Dunkelberg and Fleitmann-Glende published a study examining pressure
differential on the microbial barrier effectiveness of reusable sterilization containers
with filters incorporated in the lids in an effort to prevent nosocomial infections. They
tested 105 standard containers (58 x 28 x 15 cm) and 111 half-size containers (29 x 28 x
25 cm) that were furnished with paper, textile, or permanent plastic filters.
Thermoresistant agar plates (15 x 20 cm) were placed into the containers prior to
sterilization. Test conditions included two hours in an exposure chamber with an
aerosol of baker’s yeast suspension (concentration = 108 cells/ml). There were 12
atmospheric pressure changes per hour that reduced and increased pressure by 1000 Pa
with a maximum level of 7000 Pa. Upon completion of testing, the containers were
incubated and inspected for growth.
Of the standard containers, 9 out of 11 with paper filters, 70 out of 79 with
textile filters, and 1 out of 15 with permanent plastic filters had growth on the agar
plates. Of the half-size containers, 9 out of 11 with paper filters, 70 out of 79 with
textile filters, and 0 out of 15 with permanent plastic filters had growth. Their
recommendation was to use quantitative microbial challenge testing with pressures
between 25 and 70 hPa at one-year intervals to identify containers that are no longer
60

effective barriers in an effort to prevent nosocomial infections (Dunkelberg and
Fleitmann-Glende 2006).
2.6.2.6.1.2 Number of Pressure Changes
In 2006, Dunkelberg and Rohmann published a study examining how the number
of pressure changes affected the contamination of pouches. Researchers developed a
microbial challenge test to detect penetration of airborne microorganisms into sterile
pouches comprised of transparent plastic film and sterilization paper. The pouches
were sterilized containing thermoresistant agar dishes, placed into an exposure
chamber, and subjected to an aerosol of Saccharomyces cerevisiae (baker’s yeast) with a
concentration of 108 cells/ml. Atmospheric pressure inside the chamber was
periodically changed to simulate weather-dependent pressure changes. Following
treatment, samples were incubated and inspected for growth.
Researchers found that contamination of the pouches was dependent on the
number of decreases in atmospheric pressure. When pouches were subjected to
periodic pressure changes in increments of 25 mbar cycling between 0 and 75 mbar
(each cycle lasting 4 to 5 minutes), the number of contaminated pouches increased with
the number of pressure changes. One hundred pouches were used for each test group.
The numbers of pressure changes used for each group of 100 were 0, 6, 12, 24, 36, and
48 with contamination occurring in 0, 1, 2, 8, 9, and 11% of the pouches (Figure 8).

61

Contaminated Pouches, %

12%
10%
8%
6%
4%
2%
0%
0

6

12

24

36

48

Number of Pressure Changes
Figure 8. Effect of the number of pressure changes on the percentage of
contaminated pouches (Dunkelberg and Rohmann 2006).
2.6.2.6.1.3 Pressure Differential and Microbial Concentration
Dunkelberg and Rohmann also examined the combined effects of pressure with
microbial load. When comparing two microbial aerosol concentrations (10 7 and 108
cells/ml) in the presence of simulated atmospheric pressure changes, pouches subjected
to the higher concentration had greater contamination compared to those subjected to
the lower concentration (Dunkelberg and Rohmann 2006).
2.6.2.6.1.4 Pressure Differentials Simulating Transport and Storage
In 2009, Dunkelberg and Schmelz published a study on the efficacy of sterile
barrier systems against microbial challenge under pressure differentials simulating those
in transport and storage. Here, they combined levels of pressure and the number of
pressure changes as test factors. Two styles of flexible peel pouches were tested: paper
with plastic film material and nonwoven (HDPE fibers) with plastic film material. Each
pouch contained an open agar plate so that one test series had the open sides under the
62

paper or nonwoven surfaces and another test series had the plates under the plastic
film surfaces. The pouches were subjected to 20 periodic atmospheric pressure changes
of 50 and 70 hPa in a microbial exposure chamber; each cycle period lasted six minutes
for a total of two hours. A nebulizer was used to aerosolize Micrococcus luteus at a 108
cells/ml concentration within the chamber.
Growth was only observed when the plates were under the paper or nonwoven
sides of the pouches. For the paper pouches, there was contamination in 30 of 99
plates (30%) and 48 of 100 (48%) plates after the 50 and 70 hPa exposures, respectively.
For the nonwoven pouches there was contamination in 3 of 99 (3%) plates and 7 of 99
(7%) plates after the 50 and 70 hPa exposures, respectively (Figure 9).
60%
48%

Contamination, %

50%

40%
30%

30%

20%
10%

3%

7%

0%

Figure 9. Effects of pressure differentials for paper and nonwoven material rates of
contamination (Dunkelberg and Schmelz 2009).

63

2.6.3 Investigation of Microbial Ingress through the Sterile Barrier
HAIs are a significant cause of morbidity and mortality in the United States that
drive up the cost of healthcare. Despite the practices employed in infection control
programs, they are still occurring and have not been eliminated. They are also
becoming more difficult to treat due to increased drug resistance, which adds urgency
to the situation. While multiple causes of HAIs have been identified, it is difficult to
trace infections back to a source. When infections are linked to a device, it is still
unknown how the device became contaminated.
Packaging and the sterile barrier have been largely overlooked as a source of
infection. Healthcare workers appear to trust that the devices they use are sterile at the
point of use and do not suspect breeches of the sterile barrier. Device manufacturers
are diligent in their efforts to ensure sterility; however, there are limitations with the
test methods currently available. Even when a defect can be found in a package, it is
unknown whether it can be penetrated by microorganisms. A limited amount of
research studying factors that affect microbial ingress has been published and the
results, although suggestive, are inconclusive. Variables that have been studied include
hole size, pressure differential across the sterile barrier, and the effects of multiple
layers of packaging.
The purpose of this research is to study the combined effects of these three
variables in realistic packaging configurations in an effort to identify how they affect
microbial ingress across the sterile barrier. Understanding what facilitates the ability of
microbes to traverse the sterile barrier can be applied toward tracing sources of
64

infection and improving package/device design to maintain the sterile barrier and aid in
aseptic technique.

65

PART 1:
EFFECT OF PRESSURE DIFFERENTIAL ON MICROBIAL INGRESS OF STERILE MEDICAL
DEVICE TRAYS WITH 100 µm PINHOLES

3.1 Objectives
Part 1 of this study builds on the work of Severin (2006) entitled “The Effect of
Pressure Differential on Microbial Penetration of a Sterile Medical Device Tray”. Severin
developed a new microbial challenge methodology and examined the effect of pressure
differential on microbial ingress into medical device trays through varying pinhole sizes.
Reported findings indicated that one tray with a 100 µm pinhole out of eight tested had
ingress in the absence of an induced pressure differential. These results suggested that
further investigation was necessary and the following research objectives were
developed:
to further validate that the microbial challenge methodology developed by
Severin (2006) does not produce false positives or negatives.
to confirm with a larger sample size that microbial penetration can occur in trays
with 100 µm pinholes in the absence of pressure differential.
3.2 Materials and Methods
The methods employed to investigate all objectives included the following steps:
1. aseptically filling sealed, sterile device packages with a known volume of
appropriate sterile growth medium,
2. exposing the packages to an aerosolized microbial challenge,

66

3. inducing pressure differential on the packages inside the microbial challenge
chamber,
4. incubating the packages,
5. inspecting for growth,
6. and recording binary (growth or no growth) and variable (number of colony
forming units) data.
A total of thirty-two samples were tested to investigate the study objectives.
Each contained a 100 µm pinhole and was sealed with a nonporous lid. Of these, two
samples were damaged during testing and removed from the study. All samples were
uniquely identified and aseptically injected with sterile growth medium.
The samples were subjected to an aerosolized microbial challenge with a
solution of Escherichia coli K12 ATCC Number 29181 that had a starting concentration of
6

1 x 10 cells/ml. The aerosol was sprayed for 15 seconds and allowed to settle for 30
minutes before the samples were removed from the exposure chamber. Immediately
following aerosolization, a pressure differential was induced at random positions within
the chamber on two of the test samples per run by withdrawing a known volume of air
from the headspace of the tray with a needle and syringe (Severin 2006). When all
testing was completed, of the usable trays, 15 had a pressure differential induced across
the sterile barrier and 15 did not. The 32 samples were exposed to the test conditions
in eight runs of four trays each. Run number and sample position within the chamber
were recorded.

67

Independent variables for model consideration included: sealing run, sealing
position, injection run, injection frame position, chamber run, chamber position, and the
presence or absence of pressure differential.
3.2.1 Holes
The pinholes in the test trays were thermal laser drilled (Lenox Laser, Glen Arm,
MD) with nominal values of 10 µm (+/- 10%) or 100 µm (+/- 10%) in diameter; test trays
that contained holes each had one hole drilled in the bottom of the tray. The holes
were flow calibrated and each tray was labeled with the tray ID, flow rate, and hole
diameter. (See Appendix A for a complete itemized materials list.)
3.2.2 Tray Sealing
The test trays used (Figure 10) were “Medtronic Inc. Outer Tray Part No. 350215001” 0.025 inch (preform thickness) blue tint uncoated polyethylene terephthalate
(PETG) trays (Perfecseal, Mankato, MN).

Figure 10. Test trays. For interpretation of the references to color in this and all other
figures, the reader is referred to the electronic version of this thesis (or dissertation).
68

Trays were sealed with LKF-002 Paper/PE/Foil/PE/HSC die cut lids (Amcor
Flexibles Healthcare, Madison, WI) (Figure 11) using a CeraTek model MD-2420 shuttlestyle heat sealer (SenCorp, Hyannis, MA) fixtured with a Teflon impregnated fiber glass
barrier blanket. There were nine sealing positions that were recorded for each tray per
run (Figure 12).
1

3

4

5

6

7

Figure 11. Tray lids.

2

8

9

Figure 12. Sealing positions.

Sealing parameters were: 300°F, 70 psi, 2.5 seconds of dwell time. After sealing,
trays were visually inspected for defects according to ASTM F 1886 Standard Test
Method for Determining Integrity of Seals for Medical Packaging by Visual Inspection
(ASTM 2004); those with defects were removed from the study. Trays were packed
inside a double-layer of PE bags, twist tied shut, stacked into corrugated shippers, and
shipped via UPS to a sterilization facility for gamma irradiation.
3.2.3 Sterilization
Gamma sterilization was provided by Smith & Nephew, Inc. (Memphis, TN) and
performed by Sterigenics (West Memphis, AR) according to process procedure P000257
Rev. M.
69

3.2.4 Non-destructive Leak Testing
All lidded trays were subjected to non-destructive leak testing. This was done in
addition to visual inspection to verify that the trays did not contain defects (other than
the intended pinholes) across their sterile barriers. (See Appendix B for results tables.)
Testing was done using an ATC, Inc. Model VE2 supported with the Leak-Tek©
data acquisition program (Indianapolis, IN) (Figure 13). The unit functions based on
mass conservation law; once the test specimen is pressurized and reaches steady state
pressure, the amount of mass flow into the specimen is equal to the amount mass flow
that leaks out. Micro-flow sensors, or Intelligent Gas Leak Sensors (IGLS), measure the
mass flow extracted from the test specimen to maintain a steady vacuum condition.

Figure 13. ATC, Inc. Model VE2 non-destructive leak tester.

70

3.2.5 Agar Injection of Lidded Trays
Glass bottles that each contained approximately 30 ml of sterile nutrient growth
agar (enough for a single test sample) were prepared by the Media Prep Lab at Michigan
State University (East Lansing, MI). The bottles were sealed with foam stoppers and
autoclave tape was placed over them to verify the autoclave process was completed
(Figure 14). Once the autoclaving was finished, the autoclave bin containing the bottles
was transferred into an insulated box for transport. The box was designed to keep the
bottles hot so that the agar did not harden before they could be placed in a water bath
at the location for agar injection.

Figure 14. Glass bottles containing sterile nutrient growth agar.
Prior to working in the lab, all table tops were disinfected using Multiterge (5 ml
in 100 ml of water). The water bath that contained the agar-filled bottles was turned on
one hour prior to agar injection. The control and safety thermostats were set between
53-60°C. The water bath was filled to a level slightly above the agar level inside the
71

bottles. If the water level is too low, the agar can begin to solidify; if it is excessively
higher than the agar level, condensation can form on the interior walls of the bottles.
Sterile trays that were injected with sterile agar were each subjected to the
following process: The tray was visually inspected a second time for any defects; trays
2

with defects were removed from the study. A 2 in section in the center of the lid was
disinfected for approximately 30 seconds using a 70% isopropyl alcohol swab (Figure
15). Two pairs of forceps were dipped in 70% isopropyl alcohol and flashed with a
Bunsen burner. A pre-cut piece of self-sealing septum (Mocon, Minneapolis, MN),
2

approximately 2 cm in size, was handled as follows: The backing was removed from
the adhesive side. The septum was dipped into 70% isopropyl alcohol and flashed with
a Bunsen burner (Figure 15). The septum was placed onto the previously disinfected
surface of the tray lid and pressed firmly with the forceps so that it fully adhered (Figure
15).

72

Trays are
sealed and
sterilized.

2

Swab a 6.5 cm
area with 70%
isopropyl alcohol.
Use forceps to
remove backing
material from selfsealing septum.

Aseptically fill
with sterile
growth medium
through septum.

Dip septum in
alcohol and waft
in Bunsen burner
flame to flash off
any remaining
alcohol.

Agitate tray to
ensure even
distribution of
agar on tray
bottom.

Place septum
(adhesive side
down) in
location that
was
disinfected in
Step2.

Challenge trays
with microbes.

Swab the
septum with
70% isopropyl
alcohol.

Incubate at optimal
conditions for the
challenge
microorganism and
inspect for growth.

Figure 15. Microbial challenge methodology (Severin 2006). (Reprinted with
permission.)
The tray was then placed in the injection frame and position was recorded; the
frame had four positions. Injection run was also recorded. Trays were oriented with the
lid-sides down so that trays containing holes had the hole-side at the top to prevent the
holes from becoming blocked with agar (Figures 16 and 17).

73

Figure 16.
Figure 17.
Injection frame with test samples: top view (Figure 16), bottom view (Figure 17).
Next, the septum on the tray lid was disinfected for approximately 30 seconds
using a 70% isopropyl alcohol swab (Figure 15); care was used to ensure that only the
septum, not any of the lid surface, was swabbed because the septum would come loose
if the lid got wet.
An agar bottle was removed from the water bath; one bottle containing 30 ml of
agar was used for each tray. The autoclave tape was carefully pulled off of the foam
stopper so that the stopper remained intact. The stopper was disinfected for
approximately 30 seconds using a 70% isopropyl alcohol swab. Both a 60 ml syringe
(Becton, Dickinson and Company, Franklin Lakes, NJ) and an 18G1½ needle (Becton
Dickinson and Company, Franklin Lakes, NJ) were aseptically removed from their
packages so that the needle could be placed on the syringe; a single syringe was used
for each tray. The needle was gently pushed through the foam stopper (using care to
not push the stopper into the bottle) and the bottle was tipped onto its side so that the
agar could be drawn into the syringe. Once complete, any air in the syringe was pushed

74

out, and the first needle was removed. Then a 16G1 vented needle (Becton, Dickinson
and Company, Franklin Lakes, NJ) was opened placed on the syringe so that it could be
used to inject the agar into the sterile tray. The use of the vented needle allowed for
out gassing from the non-porous package as the agar was injected. The needle was
pushed through the septum on the lid of the tray (oriented so that the lid was on the
bottom) (Figure 17) at an angle so that it did not pierce through the tray (on the top)
(Figure 16) as the agar was injected. The frame was tipped and held forward so that
when the agar was injected, it did not surround the injection site. The agar was slowly
pushed out of the syringe into the tray so that it covered the lid surface on the bottom
and did not touch any of the other surfaces inside the tray. Once the entire volume of
agar was injected, the syringe was removed from the tray and the frame was tipped
back to its resting, flat position. The tray was gently agitated by rocking the frame so
that the agar covered the lid surface evenly. Once a set of four trays was complete, they
were removed from the frame and set aside.
Once all of the trays had undergone agar injection, the septums for pressure
differential were placed on them. This was done using the same procedure as for the
placement of the agar injection septums. The septums were centered on the front sides
of the trays.
If any errors occurred during this procedure (inadvertent puncture, etc.), those
samples were removed from the study. All samples were stacked in corrugated boxes
and held at room temperature until the following day when microbial challenge testing
was conducted.
75

3.2.6 Preparation of Dilution Series Agar Plates
To prepare sterile nutrient growth agar, Difco™ Nutrient Agar powder was mixed
with deionized water. As a benchmark, 23 g of powder can be mixed with 1000 ml of
deionized water to make a batch of agar.
This procedure was used to prepare six dilution series agar plates. Each plate
requires approximately 15 ml of agar; therefore, 6 plates x 15 ml = 90 ml of agar. Since
the amount of agar per plate does not need to be exact, and to simplify the
measurements for the procedure, agar was prepared in 100 ml batches.
A 500 ml Erlenmeyer flask was filled with 100 ml of deionized water. (It is
important that the depth of the liquid does not exceed roughly 1.5 inches to ensure that
the autoclaving conditions reach the center). Using a spatula to transfer the powder
into a boat, 2.3 g of powder was weighed on scale. The powder was then poured into
the flask. The flask was gently swirled to dissolve the powder. The opening of the flask
was sealed with aluminum foil and a piece of autoclave tape was placed on top of the
foil.
The flask was put into an autoclave bin, the bin was placed inside an autoclave,
and the door was closed. The cycle time was set at 25 Sterilize minutes (Dry time = 0).
The orange Fluid Cycle button and then the red “On” button were pressed. The
autoclave pressurized to 15 psi and the cycle began when the green and red arrows
align at 121°C.
Once the 25 minute cycle was completed, there was an additional wait of 10-15
minutes to allow the pressure and heat to come down from the cycle. Then the door
76

was slowly opened (to allow for careful steam release). The bin was then removed
using thermally insulated gloves.
To pour the agar into the petri plates, six plates were stacked on top of one
another. The foil was removed from the top of the flask using care to not contaminate
the sterile opening. Then the lid of the top plates was removed and held with one hand
(maintaining the lid’s orientation as it was lifted) while the other hand poured
approximately 15 ml of agar (enough to coat the bottom) into the plate. The lid was
then replaced and the plate was lifted from the pile and set aside. During this
movement, it was important that the flask was continuously held at the pour angle to
prevent drips from going back into the flask and contaminating the sterile agar. This
process was continued until all six plates were filled with agar. The plates were kept at
room temperature to solidify until use the following day.
3.2.7 Microbial Solution Preparation
The microbial solution used Butterfield’s Solution as a diluent for the test
organism, Escherichia coli K12 ATCC Number 29181 (Figure 18).

Figure 18. Escherichia coli K12 ATCC Number 29181 (Severin 2006).
77

To prepare the Butterfield’s Solution, a 2L bucket was filled with 1L of distilled
water. Using a spatula to transfer 68 g of Potassium Phosphate, Monobasic, Crystal
(KH2PO4) to a boat, the KH2PO4 was weighed on a scale and then poured into the bucket.
A stir bar was added to the bucket and a Corning Stirrer was used to dissolve the
KH2PO4. A pH meter was placed in the solution. Wearing latex gloves, 1N Sodium
Hydroxide (1N NaOH), was added to the solution with a pipette until the pH was
brought up to 7. A pH strip was dipped into the solution to verify the pH. Then distilled
water was added to the solution until the volume reached 2L.
The solution was poured into four 1L Erlenmeyer flasks so that each contained
500 ml of the solution. (It is important that the depth of the liquid does not exceed
roughly 1.5 inches to ensure that the autoclaving conditions reach the center). The
openings of the flasks were sealed with aluminum foil and pieces of autoclave tape were
placed on top of the foil.
The flasks were put into an autoclave bin, the bin was placed inside an autoclave,
and the door was closed tightly. The cycle time was set at 15 Sterilize minutes (Dry time
= 0). The orange Fluid Cycle button and then the red “On” button were pressed. The
autoclave pressurized to 15 psi and the cycle began when the green and red arrows
align at 121°C.
Once the 15 minute cycle was completed, there was an additional wait of 10-15
minutes to allow the pressure and heat to come down from the cycle. Then the door
was slowly opened (to allow for careful steam release). Bins were removed from the
autoclave using thermally insulated gloves.
78

The flasks were set aside at room temperature for inoculation the following day.
3.2.8 Dilution Series
A dilution series was performed to determine and verify the viability of the E.
coli. Three dilution bottles were each filled with 99 ml of Butterfield’s Solution. The lids
were put on the bottles but left loose so that pressure could release during autoclaving.
Autoclave tape was adhered to the tops of the lids. The bottles were placed in a wire
basket and the basket was placed in an autoclave bin. The autoclaving procedure for
Butterfield’s Solution was followed. The bottles were labeled 1, 2, and 3, respectively.
To perform the dilution series, 1 ml of Escherichia coli K12 ATCC Number 29181
stock was needed. The E. coli was stored in vials in a -80°C freezer (Figure 19). Each vial
contained approximately 1 ml of E. coli. One vial was removed from the freezer and
thawed by rolling it in between hands; this also ensured the E. coli was well mixed.

Figure 19. Vials of E. coli stock.

79

A 1 ml pipette with disposable pipette tips was used to transfer the E. coli into
bottle 1. A sterile tip was placed on the pipette and then 1 ml of the E. coli was drawn
up into the pipette from the vial. The lid on the bottle was carefully removed and held
in one hand while maintaining its orientation. With the other hand, the pipette was
lowered into the bottle until its tip was in the solution. Then the E. coli was pushed
from the pipette into the solution. (During this procedure, it is important the sides of
the pipette do not come in contact with the sides of the bottle so that contamination
does not occur). The pipette was lifted from the bottle and the tip disposed of. Then
the cap was placed tightly back on top of the bottle and the bottle was shaken for 30
seconds at a 1 ft length to mix the E. coli in the solution. Bottle 1 had dilution factor of
-2

10 .
Another sterile tip was placed on the 1 ml pipette, and following the same
methodology, 1 ml of solution was pulled into the pipette from bottle 1 and transferred
to bottle 2. Bottle 2 was also shaken for 30 seconds at a 1 ft length to ensure mixing.
-4

Bottle 2 had a dilution factor 10 . This procedure was repeated a third time to transfer
-6

1 ml of solution from bottle 2 to bottle 3 so that bottle 3 had a dilution factor of 10 .
To continue with the dilution series, aliquots of solution from the bottles had to
be placed on agar plates. The six agar plates were arranged in a row and labeled from
-2

left to right so that each one was labeled with one of the following dilution ratios: 10 ,
-3

-4

-5

-6

-7

10 , 10 , 10 , 10 , 10 . It is more efficient to perform the series from right to left (10
7

-2

to 10 ) because the same pipette can be used for the entire process. Using a sterile

80

-

-7

pipette, 0.1 ml of solution was drawn up from bottle 3 and added to plate 10 . Then 1
-6

ml of solution from bottle 3 was drawn up and added to plate 10 . From bottle 2, 0.1
-5

ml was drawn up and added to plate 10 , and 1 ml was drawn up and added to plate 10
4

-3

. From bottle 1, 0.1 ml was drawn up and added to plate 10 , and 1 ml was drawn up
-2

and added to plate 10 (Figure 20).

1 ml

1 ml

1 ml

1
1

Stock

10

2
1

-2

10

1 ml

10

-2

0.1 ml

10

-3

3
1

-4

1 ml

10

10
0.1 ml

-4

10

1 ml

-5

Figure 20. Dilution series.

81

10

-6

0.1 ml

-6

10

-7

-

Lastly, the solution was spread on the plates. Setup for this included a boat
containing 70% isopropyl alcohol and a lit Bunsen burner. A glass spreader (“hockey
stick”) was dipped into the alcohol and flamed with the burner. (It is important to keep
the end of the stick that was dipped in alcohol tipped downward so that alcohol cannot
-7

-2

drip toward the hand). Working again from 10 to 10 , the stick was used to spread
the solution over each agar plate in up-and-down, side-to-side, and around-the-edge
motions to ensure even coating. The plates were incubated at 35°C, 50% RH for 24
hours.
After the incubation period was complete, the colony forming units (CFU) on
each plate were counted to determine the viable concentration of the stock. A Quebec
Colony Counter (Figure 21) was used to light and magnify, and a Sharpie was used to
mark the plate as each colony was counted.

Figure 21. Quebec Colony Counter illuminating a dilution series agar plate.

82

3.2.9 Inoculation
Prior to working in the lab, all table tops were disinfected using Multiterge (5 ml
in 100 ml of water). To inoculate the Butterfield’s Solution, and as determined by the
dilution series, 1 ml of Escherichia coli K12 ATCC Number 29181 was required for each
100 ml volume of solution; therefore, 5 ml of E. coli were added to each flask containing
500 ml of solution. This produced a 1 x 106 cells/ml microbial solution concentration.
The vials of E. coli were removed from the freezer and thawed by rolling them in
between hands. Each vial contained approximately 1 ml of E. coli. A 1 ml pipette with
disposable pipette tips was used to transfer the E. coli into the flasks containing sterile
Butterfield’s Solution. A sterile tip was placed on the pipette and then 1 ml of the E. coli
was drawn up into the pipette from a vial. The foil on the flask was carefully removed
and set aside so that its orientation remained the same; this helps reduce the risk of
contaminating the foil. The pipette was lowered into the flask until its tip was in the
solution. Then the E. coli was pushed from the pipette into the solution. (During this
procedure, it is important the sides of the pipette do not come in contact with the sides
of the flask so that contamination does not occur). The pipette was lifted from the flask
and the tip disposed of. This procedure was repeated until the flask had 5 ml of E. coli
added to it. Then the foil was placed back on top of the opening and the flask was
gently agitated to mix the E. coli in the solution. This procedure was repeated for all
flasks.

83

3.2.10 Microbial Challenge and Induction of Pressure Differential
A microbial challenge chamber consisting of 0.25 in thick Plexiglas was used in
this study (Figures 22 and 23). The internal dimensions of the chamber are 28.5 in x 20
in x 36 in (L x W x H). The chamber was designed with one removable face attached
with wing nuts; this allows entry into the cabinet for sample loading and chamber
cleaning. This face also contains a pair of gloves used for manipulation of objects inside
the chamber while maintaining the seal of the chamber (Figure 23). The interior of the
chamber contains a racking system that holds four samples. This racking system also
holds four syringes (for induction of pressure differential) and includes a retraction bar
that can be used to simultaneously draw back the syringes. The racking system is
approximately 12 inches from the floor of the chamber. Two paint sprayers are affixed
to the top of the chamber to aerosolize the microbial solution (Figure 22). The sprayers
are Central Pneumatic Air Spray Guns Model #43760 with internal mix nozzles
measuring 0.082 inches in diameter. For this study, they were pressurized at 40 psi
(Severin 2006).

84

Figure 22. Microbial challenge chamber.

Figure 23. Interior view of microbial challenge chamber.

85

Prior to microbial challenge testing, the chamber was cleaned using a bleach
solution. Additionally, the two spray canisters were thoroughly rinsed with water,
including spraying water from them to rinse the spray nozzles. Each canister was then
filled with 1L of the microbial solution and affixed to the top of the chamber. The
canisters were attached to air hoses with pressure set at 40 psi.
Next, the 60 ml syringes (maximum of four) used for induction of pressure
differential (Becton, Dickinson and Company, Franklin Lakes, NJ) were loaded into the
chamber. Each was outfitted with an 18G1½ needle (Becton, Dickinson and Company,
Franklin Lakes, NJ) with the cap still intact to maintain sterility. Test samples were then
individually loaded into the chamber in batches of four (Figure 23). First, the pressure
differential septum on the exterior of the test sample was sterilized with an alcohol
swab. Then the cap on the needle was removed from a syringe secured in the chamber
so that it could pierce through the septum on the sample as the sample was placed in its
position in the chamber (Figure 24). It was important to ensure accurate alignment of
the needle through the septum(s) into the tray so that it remained above the agar in the
headspace. It was also important to ensure that the needle did not accidentally
puncture another surface of the package during insertion.

86

Figure 24. Pressure differential syringe placement.
Four samples can fit into the chamber for each test run. Once all samples were
loaded, the retraction bar used to simultaneously and consistently pull the syringe
plungers back to induce pressure differential was placed on the plunger handles. (Figure
25). Lastly, an open agar plate was placed on the bottom of the chamber (Figure 25) to
verify the viability of the microbial solution and spray consistency across the chamber.
Sample position and run were recorded.

87

Retraction
Bar

Agar Plate
Figure 25. Retraction bar and agar plate placement.
Once the samples and agar plate were in place, the chamber was closed and
ready for microbial challenge. The canisters were sprayed simultaneously for 15
seconds. Then the retraction bar on the syringes was pulled to induce pressure
differential. This was done by placing hands in the gloves affixed to the chamber, and
consistently pulling the bar back for approximately one minute until the syringes had
pulled a volume of 62 ml of air from the test samples. This induced a pressure
differential of -3.78 psi, which is equivalent to an aircraft or ground shipment descent of
8,000 ft (Severin 2006).
Thirty minutes from the time the microbial solution was aerosolized into the
chamber, the syringes were removed from the samples and the samples were
transferred to an environmental chamber for incubation. Additionally, the agar plate
for each run was incubated.

88

This entire microbial challenge process was repeated until all samples were
tested. Upon completion, the chamber was cleaned with a bleach solution and the
canisters were thoroughly rinsed with water.
3.2.11 Incubation
After the microbial challenge testing, the samples and agar plates were
incubated in an environmental chamber at 37°C and 50% RH for approximately 48
hours.
3.2.12 Colony Counting of Test Samples
Test samples were inspected for colony growth after the incubation period was
complete. The bases of the inverted trays were cut away using a razor blade to create a
viewing window (Figure 26). If colonies were present, they were counted and recorded.
(See Appendix C for results table.)

Figure 26. Colony viewing window cut from tray.

89

3.2.13 Colony Counting of Agar Plates
After the incubation period, colonies were also counted on the agar plates to
verify the viability of the microbial solution. A Quebec Colony Counter (Figure 27) was
used to light and magnify, and a Sharpie was used to mark the plate as each colony was
counted. The number of CFU was recorded. (See Appendix D for results table.)

Figure 27. Quebec Colony Counter illuminating a microbial challenge agar plate.
3.2.14 Enterotubes
Enterotubes (BD BBL™ Enterotube™ II REF 211832) were used to verify that the
observed CFU were Escherichia coli K12 ATCC Number 29181. The white cap of the tube
was removed to expose a wire that was dipped into the colony being tested and the cap
was replaced. Following this, the blue cap was removed from the opposite end of the
tube to expose a wire handle. The handle was used to pull the wire (while gently
twisting it back and forth) so that the end that touched the colony traveled through
each compartment of the tube. Then the wire was pushed back through the tube
90

towards its starting position until the notch on the wire aligned with the edge of the
tube. The wire was bent back and forth to break it off at the notch and the blue cap was
replaced. The piece of wire that was broken off was used to pierce holes through the air
inlets of the compartments that had them. The tubes were incubated at 35°C, 50% RH
for 18 hours.
After the incubation period was complete, the enterotubes were observed.
Wearing latex gloves, Kovac’s reagent was drawn up in a 1 ml syringe and five drops
were injected into each Indole compartment. Then the tubes were color matched to a
key to verify that the colony was the test organism (Figure 28).

Figure 28. Example of enterotube and color key.
3.3 Results and Discussion
Two dependent variables were considered during analysis: the number of trays
that exhibited growth (binary data: growth or no growth), and the number of CFU
observed per tray (variable data).
91

A total of 30 usable trays were tested for the effect of pressure differential on
microbial ingress through a 100 µm pinhole in sterile trays. Of the fifteen trays that
were exposed to a pressure differential, all of them (100%) exhibited growth (Figure 29).
Of the fifteen trays that were not exposed to a pressure differential, nine (60%)
exhibited growth (Table 5). (See Appendix C for results table).

Figure 29. Example of tray with colony growth.
Table 5. Percent of total trays with CFU.
Pressure
Samples
Differential Sample Size
with
(psi)
Growth
-3.78
15
15
0
15
9

% Growth

CFU Mean

100%
60%

21.0667
1.0667

CFU
Standard
Error Mean
4.8745
0.3584

Question 1: What is the effect of pressure differential on the probability of microbial
penetration of medical device trays? (Binary data: growth or no growth)
The response variable, trays with microbial ingress, was modeled as binary
distribution using a generalized linear mixed model fitted with the GLIMMIX procedure
92

of SAS (SAS version 9.1, SAS Institute Inc., Cary, NC). Due to extreme category problems
(all trays assigned to pressure differential had growth), the effect of pressure was
modeled as random in a Bayesian-type approach. Sensitivity analyses were performed
using starting values for the random variance equal to 1, 10, 100, 1000, and 3000.
Variance estimates and pressure differences did not change for the starting values
considered. Contrasts were used to compare levels of pressure. Other random factors
such as sealing run, sealing position, injection run, injection frame position, chamber
run, and chamber position were evaluated for inclusion in the model. Due to variance
components estimates equal to 0, these random blocks were not included in the final
model. The null and research hypotheses for this analysis are:
H0: Pressure differential will not increase the number of trays with microbial
ingress.
Ha: Pressure differential will increase the number of trays with microbial ingress.
Here, the null hypothesis was not rejected; therefore, conclusions cannot be
made on the research hypothesis. A marginal effect of pressure differential (p=.0694)
on the probability of microbial ingress in trays was identified (α=.05). The predicted
probabilities of microbial penetration were 87.1 ± 26.8% for pressure differential and
12.9 ± 26.8% for no pressure differential (Figure 30).

93

Predicted probability of
microbial penetration, %

120%
100%

87.1%

80%
60%
40%

12.9%

20%
0%
-20%

Pressure
(-3.78 psi)

No Pressure
(0 psi)

Figure 30. Comparison of predicted probabilities of microbial penetration when
pressure differential is the independent variable.
Question 2: What is the effect of pressure differential on the number of CFU in trays?
(Variable data: number of CFU)
A generalized linear mixed model was fitted using a negative binomial
distribution of the response variable, number of CFU, using the GLIMMIX procedure of
SAS (SAS version 9.1, SAS Institute Inc., Cary, NC). The model included the fixed effect of
pressure. Due to convergence problems, it was not possible to include other random
blocking factors such as sealing run, sealing position, injection run, injection frame
position, chamber run, and chamber position in the final model. The null and research
hypotheses for this analysis are:
H0: Pressure differential will not increase the number of CFU in trays.
Ha: Pressure differential will increase the number of CFU in trays.

94

Here, the null hypothesis was rejected and the research hypothesis was
accepted. Data provided evidence for a significant effect of pressure differential (p
<.0001) on the number of CFU in trays (α=.05); there was a significantly greater number
of CFU in trays exposed to a pressure differential than without. The estimated number
of CFU was 21.1 ± 4.9 for trays exposed to a pressure differential, and 1.1 ± 0.3 for no
exposure (Figure 31).

Estimated number of
CFU per tray

30

21.1

25
20
15
10
5

1.1

0
Pressure
(-3.78 psi)

No Pressure
(0 psi)

Figure 31. Comparison of estimated number of CFU per tray when pressure
differential is the independent variable.

95

PART 2
EFFECTS OF PRESSURE DIFFERENTIAL AND SECONDARY PACKAGING ON MICROBIAL
INGRESS OF STERILE MEDICAL DEVICE TRAYS

4.1 Objective
The objective for Part 2 was:
to examine the effects of secondary packaging (pouches and cartons) and
pressure differential on the microbial ingress of sterile medical device trays.
4.2 Materials and Methods
The methods employed to investigate all objectives were the same as in Part 1
(except where indicated) and included the following steps:
1. aseptically filling sealed, sterile device packages with a known volume of
appropriate sterile growth medium,
2. exposing the packages to an aerosolized microbial challenge,
3. inducing pressure differential on the packages inside the microbial challenge
chamber,
4. incubating the packages,
5. inspecting for growth,
6. and recording binary (growth or no growth) and variable (number of colony
forming units) data.
A total of one-hundred and twenty samples were tested to investigate the study
objective. Forty of the samples were unlidded trays sealed in Nylon/LDPE/HDPE

96

pouches, forty were unlidded trays packaged in paperboard cartons, and the remaining
forty were lidded trays, each containing a 100 µm pinhole and sealed with a nonporous
lid, packaged in paperboard cartons. All samples were uniquely identified and
aseptically injected with sterile growth medium. Of these, usable data was obtained
from 118: 39 unlidded trays in pouches, 39 unlidded trays in cartons, and 40 lidded
trays, each containing a 100 µm pinhole, in cartons. The “unlidded tray in pouch”
sample that was removed from the study had the pouch punctured by the syringe used
to induce pressure differential. The “unlidded tray in carton” sample that was removed
from the study had the syringe used to induce pressure differential pull out of the
sample during that procedure.
The samples were subjected to an aerosolized microbial challenge with a
solution of Escherichia coli K12 ATCC Number 29181 that had a starting concentration of
1 x 106 cells/ml. The aerosol was sprayed for 15 seconds and allowed to settle for 30
minutes before the samples were removed from the chamber. Immediately following
aersolization, a pressure differential was induced on all of the test samples by
withdrawing a known volume of air from the headspace of all of the packages with a
needle and syringe. Run number and sample position within the chamber were
recorded.
Independent variables for model consideration included: injection number,
chamber run, chamber position, lidding condition (tray lidded or unlidded), secondary
package type (pouch or carton), and the presence or absence of a 100 μm laser-drilled
hole. The design approach was tiered, first investigating the most “open” designs
97

(unlidded) to determine if penetration could occur with the secondary packages and
then adding lids from there to determine if penetration could occur through both the
secondary and primary packages in conjunction.
The 40 unlidded trays in pouches were exposed to the aerosolized microbial
solution and pressure differential in ten runs of four trays each. No growth was
observed in any of the 39 usable samples; therefore, there was no additional
investigation of lidded trays in pouches.
The 40 unlidded trays in cartons were also exposed to the aerosolized microbial
solution and pressure differential in ten runs of four trays each. Because 37 out of 39
usable samples exhibited growth, an additional tier of testing was added (lidded trays
with 100 µm pinholes in cartons).
The 40 trays sealed with nonporous lids, each containing a 100 µm pinhole and
packaged in paperboard cartons, were also exposed to the aerosolized microbial
solution and pressure differential in ten runs of four trays each.
4.2.1 Sample Preparation for Unlidded Trays in Pouches
The trays each had a self-sealing septum (Mocon, Minneapolis, MN) adhered to
them on the front corner; these septums later served as the sites used for inducing
pressure differential (Figure 32).

98

Label
Pressure
Differential
Septum

Figure 32. Unlidded tray with pressure differential septum.
Test pouches were 100GA Biax Nylon (0.001), 0.0007 LDPE, 0.002 HDPE Coex
pouches (Lot #: H150978/1/A) with dimensions of 7.25 in x 9.50 in (Mangar Industries,
Inc., New Britain, PA) (Figure 33). One unlidded tray was placed inside each pouch.

Figure 33. Example of test pouch.
Pouches were sealed using a CeraTek model 24-AS/1 (Serial No. 06-04236) heat
bar sealer (SenCorp, Hyannis, MA). Sealing parameters were: 275°F, 60 psi, and 1
second of dwell time.
99

After sealing, pouches were visually inspected for defects according to ASTM F
1886 Standard Test Method for Determining Integrity of Seals for Medical Packaging by
Visual Inspection (ASTM 2004); those with identified seal defects were removed from
the study. The remaining samples each had a self-sealing septum (Mocon, Minneapolis,
MN) adhered to the top exterior of the pouch over the open top of the tray; these
septums later served as the agar injection sites (Figure 34). Additionally, each pouch
had a self-sealing septum adhered to the bottom exterior so that it aligned with the
pressure differential septum on the tray (Figure 34). Each pouch was placed in a
paperboard carton (to protect from distribution hazards) and the cartons were packed
into corrugated shippers for sterilization. These cartons were for transportation only
and were removed prior to agar injection and microbial challenge testing.

Agar
Injection
Septum

Label

Pressure
Differential
Septums

Figure 34. Unlidded tray in pouch test configuration with septum placement.

100

4.2.2 Sample Preparation for Unlidded Trays in Cartons
Reverse tuck folding cartons were constructed from 15.5 pt solid bleached
sulfate (SBS) on an Artios Kongsberg Premium Line 1930 cutting table. Carton
dimensions were 5 3/16 in x 15/16 in x 6 3/4 in.
The trays each had a self-sealing septum (Mocon, Minneapolis, MN) adhered to
the front corner; these septums later served as the sites for induction of pressure
differential (Figure 35). A single tray was placed inside each carton so that all samples
were in the same orientation. The cartons were labeled with the tray identification
numbers.
It was imperative that all septums placed on the cartons aligned with the
septums on the trays inside. To ensure this, a template was used to identify placement
points for the septums and injection points on the septums. The template, constructed
from one of the cartons, was used to identify the center points on the tops of the test
cartons. At each center point, a self-sealing septum was adhered to the top exterior of
the carton over the open top of the tray; these septums later served as the agar
injection sites (Figures 35 and 36). Additionally, each carton had a self-sealing septum
adhered to the front corner so that it aligned with the pressure differential septum on
the tray (Figure 36).

101

Label
Agar
Injection
Septum
Pressure
Differential
Septums
Figure 35.
Figure 36.
Unlidded tray in carton test configuration with septum placement: open view (Figure
35), closed view (Figure 36).
Lastly, the template was used to mark the injection points on the pressure
differential septums with ink (Figure 37). Since it was not possible to see the trays
inside, this was done to ensure that the needles would align and pierce through both
septums. The cartons were packed into corrugated shippers for sterilization.

Figure 37. Injection points on pressure differential septums.

102

4.2.3 Sample Preparation for Lidded Trays, with 100 µm Pinholes, in Cartons
Lidded trays, each containing one 100 µm pinhole, were visually inspected for
damage to the trays, lids, and seals. Each tray had a self-sealing septum (Mocon,
Minneapolis, MN or Illinois Instruments, Johnsburg, IL) adhered to the center of the lid
that later served as the agar injection site. The trays also had a self-sealing septum
adhered to them at the front corner to later serve as the site used for inducing pressure
differential (Figure 38). Here, the opposite corner is used because the trays are inverted
(with their lid sides down) inside the microbial challenge chamber; this ensured that the
fixtures for inducing pressure differential inside the chamber would still align with the
septums on the samples.

Agar
Injection
Septum

Pressure
Differential
Septum

Figure 38. Septum placement for lidded tray in carton test configuration.
One tray was placed inside each carton so that all samples were in the same
orientation (lid side down) and the cartons were labeled with the tray identification
numbers (Figure 39).

103

Figure 39. Lidded tray in carton test configuration.
The template was again used to identify the locations for adhering septums to
the cartons and marking the septums with the injection points. Here, the agar injection
septums were placed on the bottom exterior of the cartons because the trays were lid
side down inside the cartons (Figure 40). Lastly, the cartons were packed into
corrugated shippers for sterilization.
Agar
Injection
Septum
Pressure
Differential
Septum

Figure 40.
Figure 41.
Lidded tray in carton septum placement: bottom view (Figure 40), top view (Figure
41).

104

4.2.4 Agar Injection for Unlidded Trays in Pouches and Unlidded Trays in Cartons
The agar injection process for unlidded trays in pouches and unlidded trays in
cartons, including syringe and needle setup and drawing up of the agar into the syringe,
were performed as stated previously with the lidded tray injection. The samples were
visually inspected for any defects; those with defects were removed from the study.
Each sample was placed and oriented on the lab bench so that the opening of the tray
faced up. Injection number was recorded. The septum on the top of the pouch or
carton was disinfected using a 70% isopropyl alcohol swab. The vented needle was
pushed through the septum at an angle so that it did not pierce through the bottom of
the tray inside. The agar was slowly pushed out of the syringe into the tray so that it
filled the bottom surface and did not touch the other surfaces inside the tray or pouch
(Figure 42) or carton. Once the entire volume of agar was injected, the syringe was
removed from the sample. The sample was gently agitated so that the agar covered the
bottom surface of the tray evenly.

Figure 42. Example of agar injection into unlidded tray in pouch configuration.

105

If any errors occurred during this procedure (inadvertent puncture, etc.), those
samples were removed from the study. All samples were stacked in corrugated boxes
and held at room temperature until the following day for the microbial challenge
procedure.
4.2.5 Agar Injection for Lidded Trays, with 100 µm Pinholes, in Cartons
Cartons containing lidded trays with 100 µm pinholes that were injected with
sterile agar each went through the following process: The sample was visually inspected
for any defects; those with defects were removed from the study. The sample was
placed and oriented on top of the injection frame with the lid-side down so that trays
containing holes had the hole-side at the top and the holes could not become blocked
with agar (Figure 43). Injection number was recorded. The septum on the bottom of
the carton was disinfected using a 70% isopropyl alcohol swab.

Figure 43. Test sample placement on agar injection frame (bottom view).

106

The agar injection process, including syringe and needle setup and drawing up of
the agar into the syringe, were performed as stated above with the lidded tray injection.
The vented needle was pushed through the septum on the carton, that aligned with the
septum on the tray lid inside, at an angle so that it did not pierce through the other side
of the tray and carton. The agar was slowly pushed out of the syringe into the tray so
that it filled the bottom surface and did not touch the other surfaces inside the tray.
Once the entire volume of agar was injected, the syringe was removed from the sample.
The sample was gently agitated so that the agar covered the bottom surface of the tray
evenly.
If any errors occurred during this procedure (inadvertent punctures, etc.), those
samples were removed from the study. All samples were stacked in corrugated boxes
and held at room temperature until the following day for the microbial challenge
procedure.
4.2.6 Microbial Challenge and Induction of Pressure Differential
Test samples were subjected to microbial challenge, induction of pressure
differential, and incubation using the methods described in Part 1. The same methods
in Part 1 were also employed for colony counting of agar plates (see Appendix E for
results tables) and use of enterotubes.
4.2.7 Colony Counting of Test Samples
Test samples were inspected for colony growth after the incubation period was
complete. For test sample configurations with open trays in pouches or cartons, the
trays were removed from the secondary package and visually inspected for colony
107

growth on the agar. If colonies were present, they were counted and recorded. (See
Appendix F for results tables).
For test sample configurations that had lidded trays inside cartons, the trays
were removed from the cartons and the bases of the inverted trays were cut away using
a razor blade to create a viewing window. If colonies were present, they were counted
and recorded. (See Appendix F for results table.)
4.3 Results and Discussion
Two dependent variables were considered during analysis: the number of trays
that exhibited growth (binary data: growth or no growth), and the number of CFU
observed per tray (variable data).
A total of 118 usable trays were tested for the effect of secondary packaging
(pouches and cartons) on microbial ingress of sterile trays (lidded and unlidded). Zero
(0%) of the 39 “unlidded tray in pouch” samples exhibited growth. Of the 39 “unlidded
tray in carton” samples, 37 (95%) exhibited growth. Of the 40 “lidded tray in carton”
samples, 6 (15%) exhibited growth (Table 6). (See Appendix F for results tables.)
Table 6. Percent of total trays with CFU.
Sample Type

Sample Size

Unlidded tray in pouch
Unlidded tray in carton
Lidded tray, with 100 µm pinhole, in carton

39
39
40

108

Samples
with Growth
0
37
6

% Growth
0%
95%
15%

Question 1: What is the effect of secondary package type on the probability of
microbial penetration of trays? (Binary data: growth or no growth)
The response variable, microbial penetration, was modeled as binary distribution
using a generalized mixed model fitted with the GLIMMIX procedure of SAS (SAS version
9.2, SAS Institute Inc., Cary, NC). The model included the fixed effect of treatment
(secondary package type) and the random effect of run nested within treatment, in
order to account for technical replication present in the design. The marginal log
likelihood was approximated using an adaptive Gauss-Hermite quadriture method in
order to expedite model convergence. Due to an extreme category problem (quasicomplete separation of data points given by no penetration in any of the trays assigned
to pouches), the effect of treatment was modeled as a random effect in a Bayesian-type
approach. Sensitivity analyses were performed using starting values for the random
variance equal to 1, 10, 100, 1000, and 3000. Variance estimates and treatment
differences were not affected by the starting values considered; therefore, the Bayesian
analysis was considered robust and valid. Contrasts were used to compare treatments.
The null and research hypotheses for this analysis are:
H0: There will be no difference in the probability of microbial penetration
between trays packaged in pouches and trays packaged in cartons.
Ha: Trays packed in pouches will have a lower probability of microbial
penetration than trays packaged in cartons.

109

Here, the null hypothesis was rejected and the research hypothesis was
accepted. Data provided evidence of a significant effect of secondary package type
(p=.0102) on microbial penetration of trays (α=.05). The type of secondary package
(pouch or carton) affected the probability of microbial penetration of trays, such that
penetration was more likely to occur in trays packed inside cartons (99.96% predicted
probability) compared to trays packaged inside pouches (0%) when the trays were
unlidded (Figure 44).

99.96%

Predicted probability of
microbial penetration, %

100%
75%
50%
25%
0%

0.00%
Pouch

Carton

Figure 44. Comparison of predicted probability of microbial penetration of pouches
and cartons.
Question 2: What is the effect of secondary package type on the number of CFU in
trays? (Variable data: number of CFU)
The response variable, number of CFU, was modeled as Poisson distributed with
overdispersion using a generalized linear mixed model fitted with the GLIMMIX
procedure of SAS (SAS version 9.2, SAS Institute Inc., Cary, NC). The model included the
fixed effect of treatment (secondary package type) and the random effect of run nested
110

within treatment in order to account for technical replication present in the design. Due
to all zero-counts in one of the treatments, convergence was impaired and the effect of
treatment was modeled as a random effect in a Bayesian-type approach as with
Question 1. Contrasts were used to compare treatments. The null and research
hypotheses for this analysis are:
H0: There will be no difference in the number of CFU between trays packaged in
pouches and trays packaged in cartons.
Ha: Trays packed in pouches will have less CFU than trays packaged in cartons.
Here, the null hypothesis was rejected and the research hypothesis was
accepted. Data provided evidence of a significant effect of secondary package type
(p=.0279) on the number of CFU in trays (α=.05). Type of secondary package affected
the number of CFU in trays such that the number of CFU was greater in trays packed
inside cartons (18.7) compared to trays packaged inside pouches (0.1) (Figure 45).

18.7

Estimated number of
CFU per tray

20
15
10
5
0

0.1
Pouch

Carton

Figure 45. Comparison of estimated number of CFU per tray in pouches and cartons.
111

Question 3: Given a secondary package of a carton, does the presence of a lid affect
the probability of microbial penetration of trays? (Binary data: growth or no growth)
The response variable, microbial penetration, was modeled as binary distribution
using a generalized mixed model fitted with the GLIMMIX procedure of SAS (SAS version
9.2, SAS Institute Inc., Cary, NC). The model included the fixed effect of treatment
(presence or absence of a lid) and the random effect of run nested within treatment, in
order to account for technical replication present in the design. The marginal log
likelihood was approximated using an adaptive Gauss-Hermite quadrature method in
order to expedite model convergence. The null and research hypotheses for this
analysis are:
H0: When packaged inside cartons, there will be no difference in the probability
of microbial penetration between trays with lids and a 100 µm hole and trays
without lids.
Ha: When packaged inside cartons, trays with lids and a 100 µm hole will have a
lower probability of microbial penetration than trays without lids.
Here, the null hypothesis was rejected and the research hypothesis was
accepted. Data provided evidence of a significant effect of lidding condition (p<.0001)
on the likelihood of microbial penetration in trays packaged inside cartons (α=.05).
Microbial penetration was more likely to occur in trays without a lid (94.9 ± 3%)
compared to those with a lid and a 100 µm hole (15 ± 5%) (Figure 46).

112

Predicted probability of
microbial penetration, %

100%

94.9%

75%
50%

15.0%

25%
0%
No Lid
(Unlidded Tray
in Carton)

Lid
(Lidded Tray
in Carton)

Figure 46. Comparison of predicted probability of microbial penetration of unlidded
and lidded trays in cartons.
Question 4: Given a secondary package of a carton, does the presence of a lid affect
the number of CFU in trays? (Variable data: number of CFU)
The response variable, number of CFU, was modeled as Poisson distributed with
overdispersion using a generalized linear mixed model fitted with the GLIMMIX
procedure of SAS (SAS version 9.2, SAS Institute Inc., Cary, NC). The model included the
fixed effect of treatment (presence or absence of a lid) and the random effect of run
nested within treatment in order to account for technical replication present in the
design. The null and research hypotheses for this analysis are:
H0: When packaged inside cartons, there will be no difference in the number of
CFU between trays with lids and a 100 µm hole than trays without lids.

113

Ha: When packaged inside cartons, trays with lids and a 100 µm hole will have
less CFU than trays without lids.
Here, the null hypothesis was rejected and the research hypothesis was
accepted. Data provided evidence of a significant effect of lidding condition (p=.0048)
on the number of CFU present in trays packaged inside cartons (α=.05). The number of
CFU was greater in trays without a lid (14.6 ± 4.4) compared to those with a lid (0.1 ±
0.2) (Figure 47).

Estimated number of
CFU per tray

14.6
20
15
10
5

0.1
0
No Lid
(Unlidded Tray
in Carton)

Lid
(Lidded Tray
in Carton)

Figure 47. Comparison of estimated number of CFU per tray of unlidded and lidded
trays in cartons.

114

PART 3
EFFECTS OF HOLE SIZE, PRESSURE DIFFERENTIAL, AND SECONDARY PACKAGING ON
MICROBIAL INGRESS OF STERILE MEDICAL DEVICE TRAYS

5.1 Objective
The objective for Part 3 was:
to examine and compare the combined effects of hole size, pressure differential,
and secondary packaging on the microbial ingress of lidded sterile medical
device trays.
5.2 Materials and Methods
The methods employed to investigate the objective were the same as in Parts 1
and 2 (except where indicated) and included the following steps:
1. aseptically filling sealed, sterile device packages with a known volume of
appropriate sterile growth medium,
2. exposing the packages to an aerosolized microbial challenge,
3. inducing pressure differential on the packages inside the microbial challenge
chamber,
4. incubating the packages,
5. inspecting for growth,
6. and recording binary (growth or no growth) and variable (number of colony
forming units) data for use as the dependent response variables.

115

The experimental design is a split plot with three factors containing two levels
each:
Factor A: pressure differential (0 and -3.78 psi)
Factor B: hole size (10 and 100 µm)
Factor C: secondary packaging (presence or absence of a paperboard carton).
Each test sample contained one level of each of the three factors so that there
were eight configurations total (Figure 48). Trays contained either a 10 or 100 µm
pinhole, were packaged inside a carton or without a carton, and were subjected to 0 or
-3.78 psi of pressure differential. The eight configurations were:
1. 10 µm hole, carton, 0 psi
2. 10 µm hole, no carton, 0 psi
3. 10 µm hole, carton, -3.78 psi
4. 10 µm hole, no carton, -3.78 psi
5. 100 µm hole, carton, 0 psi
6. 100 µm hole, no carton, 0 psi
7. 100 µm hole, carton, -3.78 psi
8. 100 µm hole, no carton, -3.78 psi

116

0 psi

-3.78 psi

Carton

No Carton

Carton

No Carton

1

2

3

4

10 µm

5

6

7

8

100 µm

Figure 48. Eight test sample configurations combining levels of hole size, presence or
absence of a carton, and presence or absence of a pressure differential.
The microbial challenge chamber was limited to four samples per run; therefore,
the whole plot factor was pressure differential (Factor A) with the experimental unit
being chamber run, or block. The split-plot factors were hole size and secondary
packaging (Factors B and C, respectively) with the experimental units being each
treatment combination, or observation.
Each run of the microbial challenge chamber was assigned one of the two levels
of pressure differential (Factor A). Within the chamber, test samples contained one
level each of Factors B and C (hole size and presence or absence of a carton). The four
chamber positions were identified as A, B, C, or D (Figures 49 and 50). The position of
the test samples within the chamber was randomized, but it is notable that previous
studies (Severin 2006) and the preliminary research presented herein did not suggest an
effect of chamber position on microbial ingress of the trays.

117

A

B

C

D

10 µm
Carton

10 µm
No Carton

100 µm
Carton

100 µm
No Carton

Figure 49. Microbial challenge chamber (assigned to either 0 or -3.78 psi) containing
four test samples (assigned to combinations of hole size and presence or absence of a
carton).

Figure 50. Microbial challenge chamber (assigned to -3.78 psi) containing four test
samples (assigned to combinations of hole size and presence or absence of a carton).
Power analyses were conservatively performed to identify the necessary sample
sizes for both the predicted probability of microbial ingress and the number of CFU as
response variables when α=.05. Main effects (pressure differential, hole size, and
secondary package), their 3-way interaction, and the contrast of the research
hypotheses extremes (100 µm hole, -3.78 psi pressure differential, and no carton vs. 10
µm hole, 0 psi pressure differential, and carton) were analyzed.
118

To have a power of 80% for both response variables, approximately 80 chamber
runs of four trays each were required for all effects except the 3-way interaction when
predicted probability of microbial ingress is the response variable (Figures 51 and 52).
Here, the analysis indicates that 200 runs of four trays each results in a power of 65%
(Figure 63). As a result, 320 samples (80 runs of four) were required to reach 80%
power for all effects excluding the 3-way interaction.

100%
90%

80%
Pressure

70%
Power, %

60%

Hole

50%

Pkg

40%

3-way interaction

30%
Contrast

20%

10%
0%
0

25

50
75 100 125 150
Number of Chamber Runs

175

200

Figure 51. Power analysis for predicted probability of microbial ingress.

119

100%
90%
Pressure

80%
Power, %

70%

Hole

60%
50%

Pkg

40%

3-way
interaction
Contrast

30%
20%
10%
0%
20 30 40 50 60 70 80 90 100 110 120 130
Number of Chamber Runs
Figure 52. Power analysis for number of CFU.

All samples were uniquely identified. Of the 320 samples, 304 were usable,
resulting in 76 runs of the chamber (see Appendix G for agar plate colony count table).
There were 38 samples for each of the 8 test configurations. Samples that were
removed from the study were damaged during the agar injection process.
5.3 Results and Discussion
Two dependent variables were considered during analysis: the number of trays
that exhibited growth (binary data: growth or no growth), and the number of CFU
observed per tray (variable data).
A total of 304 usable trays were tested for the effects of hole size, pressure
differential, and presence or absence of a carton, with 38 samples for each test
configuration. The number of samples that exhibited growth and the percent growth
120

for each sample type are presented in Table 7 and Figure 53. (See Appendix H for
results table.)
Table 7. Percent of total samples with CFU.
Sample
Samples with
Sample Type
Size
Growth
10 µm, carton, 0 psi
38
0
10 µm, no carton, 0 psi
38
0
10 µm, carton, -3.78 psi
38
1
10 µm, no carton, -3.78 psi
38
3
100 µm, carton, 0 psi
38
2
100 µm, no carton, 0 psi
38
14
100 µm, carton, -3.78 psi
38
29
100 µm, no carton, -3.78 psi
38
35

0 psi

% Growth
0%
0%
3%
8%
5%
37%
76%
92%

-3.78 psi

Carton

No Carton

Carton

No Carton

0%

0%

3%

8%

10 µm

5%

37%

76%

92%

100 µm

Figure 53. Percent growth for the eight test sample configurations.
Question 1: What are the combined effects of hole size, pressure differential, and
cartons on the probability of microbial penetration of trays? (Binary data: growth or
no growth)

121

A generalized linear mixed model was fitted to the binary response variable
indicating whether growth was observed in the test samples (0 = no, 1 = yes). The
statistical model used for analysis included the fixed effects of pressure differential (0
and -3.78 psi), hole size (10 and 100 µm), presence of a carton (1 = yes, 2 = no), and the
2-way interactions between pressure differential and carton, and between hole size and
carton. The remaining interaction terms, the 2-way interaction between pressure
differential and hole size and the 3-way interaction between all three effects, were
considered for inclusion in the model but were dropped due to model overfit and pvalues >.90. The model also included the random effect of chamber nested within
pressure differential in order to recognize the experimental unit for this factor and the
split plot arrangement of the experimental design. The fixed effect of chamber position
(A, B, C, D) was also considered in the model to check the randomization process.
Chamber position was found to be not significant (p = .15) and was dropped from the
model. Degrees of freedom were estimated using Kenward Roger's procedure. The
model was fitted using the GLIMMIX procedure of SAS (Version 9.2, SAS Institute Inc.,
Cary, NC). Estimated probabilities of contamination for combinations of the factors of
interest, as well as differences between factor levels, were reported. Pairwise
comparisons were conducted using Tukey-Kramer adjustment to avoid inflation of Type
I error rate due multiple comparisons.
The null and research hypotheses for this analysis are:

122

H0: There will be no difference in the probability of microbial penetration when
comparing any of the test configurations.
Ha: Trays containing a 10 µm hole, in the presence of a carton, and in the
absence of a pressure differential will have the lowest probability of microbial
penetration while trays containing a 100 µm hole, in the absence of a carton,
and in the presence of a pressure differential will have the highest probability of
microbial penetration.
The data did not support any 3-way or any 2-way interaction between the 3
factors of interest. A main effect of pressure differential was identified (p<.0001,
α=.05). The probability of contamination was significantly increased under conditions of
a negative pressure differential (0.53 ± 0.08) compared to no pressure differential (0.01
± 0.01) (Figure 54).

Predicted probability of
microbial penetration, %

100%
80%
60%

53%

40%
20%

1%
0%
Pressure

No Pressure

Figure 54. Comparison of probabilities of microbial penetration when pressure
differential is the main effect.

123

A main effect of hole size was identified (p<.0001, α=.05). The probability of
contamination was significantly increased with 100 µm holes (0.51 ± 0.07) compared to
10 µm holes (0.007 ± 0.005) (Figure 55).

Predicted probability of
microbial penetration, %

100%
80%
60%

51%

40%
20%

0.7%
0%

100 µm

10 µm

Figure 55. Comparison of probabilities of microbial penetration when hole size is the
main effect.
A main effect of secondary packaging was identified (p=.0146, α=.05). The
probability of contamination was significantly increased in the absence of a carton (0.18
± 0.05) compared to the presence of a carton (0.04 ± 0.02) (Figure 56).

124

Predicted probability of
microbial penetration, %

100%
80%
60%
40%

18%
20%

4%
0%
No Carton

Carton

Figure 56. Comparison of probabilities of microbial penetration when secondary
packaging is the main effect.

Question 2: What are the combined effects of hole size, pressure differential, and
cartons on the number of CFU in trays? (Variable data: number of CFU)
A generalized linear mixed model was fitted to the response "number of CFU",
which was modeled to follow a negative binomial distribution. This distribution was
used to accommodate the proportion of zero CFU observed. The statistical model used
for analysis included the fixed effects of pressure differential (0 and -3.78 psi), hole size
(10 and 100 µm), presence of a carton (1=yes, 2 = no), and the 2-way interactions
between pressure differential and carton, and between hole size and carton. The
remaining interaction terms, the 2-way interaction between pressure differential and
hole size and the 3-way interaction between all three effects, were considered for
inclusion in the model but were dropped due to convergence problems. The model also
included the random effect of chamber nested within pressure differential in order to

125

recognize the experimental unit for this factor and the split plot arrangement of the
experimental design. The model was fitted using the GLIMMIX procedure of SAS
(Version 9.2, SAS Institute Inc., Cary, NC). Estimated number of CFU for combinations of
the factors of interest, as well as differences between factor levels, were reported.
Pairwise comparisons were conducted using Tukey-Kramer adjustment to avoid inflation
of Type I error rate due multiple comparisons.
The null and research hypotheses for this analysis are:
H0: There will be no difference in the number of CFU between any of the test
configurations.
Ha: Trays containing a 10 µm hole, in the presence of a carton, and in the
absence of a pressure differential will have the lowest number of CFU while trays
containing a 100 µm hole, in the absence of a carton, and in the presence of a
pressure differential will have the highest number of CFU.
The data did not support any 2-way interactions between the factors of interest.
Similar patterns of effects were identified as those described for Question #1. Main
effects of pressure differential (p< .0001), hole size (p<.0001), and secondary packaging
(p=.0006) were identified.
5.4 Conclusions
The split plot design did not indicate any two- or three-way interactions between
hole size, pressure differential, or secondary packaging, but it did indicate significant
effects of each main factor. The sample size used was for 80% power; therefore, an
126

increased sample size would increase the level of power for the experimental design
when investigating interaction effects for predicted probability of microbial ingress and
number of CFU as response variables.
Hole size was identified, with significance (p<.0001), as a main effect when
predicted probability of microbial ingress and number of CFU were response variables.
It is intuitive that there is a relationship between hole size and microbe size when
investigating how microorganisms travel through a hole defect. These results suggest
that the larger the hole is, the easier it will be for a microorganism to travel through it.
These results also indicate that it is possible for microorganism to travel through defects
as small as 10 µm, which is supportive of the continued need for integrity tests with high
levels of sensitivity until modeling of the probability of contamination with multiple
variables can be achieved.
Pressure differential was identified, with significance (p<.0001), as a main effect
when predicted probability of microbial ingress and number of CFU were response
variables. The levels of pressure differential used (0 and -3.78 psi) were outer
boundaries of what could be anticipated in a distribution environment; the level of -3.78
psi is a worst case extreme during transit resulting from air craft or vehicle descent from
8,000 feet. Typically, devices are sterilized within their shipping containers, which
provide additional layers of protection; therefore, the primary packages are not exposed
to microorganisms within a typical distribution environment. However, pressure
differentials that occur within in-use environments that are rich with microorganisms
(healthcare facilities) are unknown. While these results identify that pressure
127

differential does serve as a driving force for microbes to travel through pinhole defects,
more research is needed to determine what pressure differentials occur within hospitals
(handling, doors opening and closing, etc.) and if those levels facilitate microbial
penetration.
The presence or absence of a carton was identified, with significance, as a main
effect when predicted probability of microbial ingress (p=.0146) and number of CFU
(p=.0006) were response variables. It is intuitive that a carton will provide some level of
barrier to microorganisms, but it also understood that it is not intended to serve as a
sterile barrier. Significance was also identified when comparing secondary package type
(pouches and cartons) when microbial penetration (p=.0102) and the number of CFU
(p=.0279) were response variables. Here, pouches were identified as a better sterile
barrier, as they did not have any microbial ingress. This supports industry practice, as
pouches are secondary packages that are intended to serve as a sterile barrier.
However, there are advantages to using cartons as a secondary package, as they provide
label billboard, stacking stability in storage, and protection from distribution hazards.
Trays which are packaged in pouches often require a carton as well, which adds a third
layer of protective packaging, and also drive up costs as a result. In practice, the layer in
contact with the device is intended as the sterile barrier; therefore trays packaged inside
cartons have the ability to deliver a sterile device with the advantages of billboard,
stackability, and protection while using a reduced amount of materials layers compared
to trays packed in pouches.

128

5.5 Limitations
The following are identified as limitations for this research:
Sample size – the sample size used was calculated at an 80% power; there is
potential to increase the level of power by increasing the sample size.
Level of pressure differential – the -3.78 psi pressure differential is typical of a
distribution environment, but not necessarily an in-use environment.
Microbial concentration – the 1 x 106 cells/ml concentration that was used is
higher than what is representative of an in-use environment.
5.6 Future Research
The following are suggestions for future research:
Increase the sample size – it is possible that ingress could occur in trays with 10
µm holes without pressure differential, but the sample size was not large
enough.
Add levels to each factor (hole size and pressure differential) – more data is
required to understand the relationship between hole size and ingress (such as if
it is a linear effect), which is necessary to develop a predictive model.
Additionally, pressure differentials that occur during handling within the hospital
environment and how those levels affect microbial ingress are unknown.
Utilize the data for each factor to develop a mathematical model for the
probability of contamination.

129

PINHOLE CHARACTERIZATION
When performing package integrity testing, in this case microbial challenge
testing, it is necessary to create hole defects that are repeatable, consistent, and
accurate; however, this can be challenging, particularly within the microcosm. Previous
research has shown that excimer laser drilling, described as being one of the most
precise and accurate techniques, results in holes with differing diameters at the entry
and exit sides of the hole tunnels (Bix, Kassarjian et al. 2005). This raises questions as to
what the shape and size are of the entire hole tunnel, and potential effects laser setup
may have on microbial challenge testing.
6.1 Objectives
Study objectives were:
to develop a methodology capable of objectively characterizing defects in
PETG trays,
and to use the newly developed methodology to characterize and compare
thermal and excimer laser drilled holes in PETG trays.
6.2 Materials and Methods
While capturing images of the entry and exit sides of the holes on the packaging
surfaces has been successful, capturing images of the entire hole tunnels has proven
challenging. Here, methods were developed to obtain images of the entire hole tunnel
using confocal microscopy.

130

A confocal laser scanning microscope was utilized due to its ability to capture
high resolution fluorescence images, obtain stacks of images through the z-axis, and
render those images three-dimensionally. These microscopes function by focusing a
laser at one point on the sample; the excitation laser, focal point of the sample, and
confocal aperture at the detector are all in conjugate focal locations. An objective
directs in-focus light from the sample through the confocal aperture to the
photomultiplier tube detector (PMT). The confocal aperture prevents out-of-focus light
from reaching the PMT, resulting in optical sectioning (Figure 57). Confocal imaging
provides the ability to look at layers of the sample along its z-axis, creating a stack of
two-dimensional images called optical sections. Accompanying software can then be
used to create three-dimensional renderings from the series of optical sections.

Figure 57. Confocal light pathways (Olympus 2011).
An Olympus FluoView FV1000 Confocal Laser Scanning Microscope and Olympus
FluoView BSW v. 5.0 software (Center Valley, PA) were used to obtain images of thermal
131

and excimer laser drilled holes in the trays. The holes were 100 µm and 50 µm in
diameter, respectively. To prepare each sample, a Nikon SM2645 stereomicroscope was
first used to identify the location of the hole on the tray. Then, the portion of the tray
containing the hole was cut out so that it measured approximately 1.5 cm x 3 cm and
marked with the tray identification number.
The samples required a coating to fill the hole tunnel (or coat the side walls) so
that the light could interact with it to create an image. Initially, the samples were
coated with osmium using an NEOC-AT osmium coater (Meiwafosis Co., Ltd., Osaka,
Japan). This technique, however, was unsuccessful in providing a sufficient coating for
reflected light interaction and image capture. While the osmium coated the surface of
the tray and the side walls of the hole near the entry and exit sides, it did not penetrate
the entire hole (Figure 58).

Figure 58. Osmium coating of an excimer laser drilled hole.

132

Instead, an aqueous solution of acridine orange dye (0.001%) was used to fill the
holes in the samples so that the fluorescence could be used to image the tunnels. Nunc
Lab-Tek II #1.5 Coverglass Chambers were used to hold the samples (Figure 50). First, a
drop of acridine orange dye was added to the inside of the chamber using a pipette.
Then the sample was placed inside the chamber on top of the dye so that the exterior
side of the tray/sample was facing down (Figure 59). Lastly, an additional drop of
acridine orange was placed on top of the sample at the previously identified location of
the hole.

Figure 59. Sample inside a Lab-Tek II #1.5 Coverglass Chamber.
The microscope was configured with a stage plate designed to hold the Lab-Tek
Chamber. The chamber was loaded onto the stage plate. Images were recorded using a
UPlan FLN 20x (0.50) dry objective. The acridine orange fluorescence was detected
through a 505 nm long pass emission filter following excitation by a 488 nm line of the

133

Argon laser. Brightfield images were also simultaneously recorded on a separate PMT
detector using the same 488nm Argon laser line (Figure 60).

Figure 60. Screen capture of imaging parameters.
Brightness and contrast for the fluorescent and Brightfield images were
optimized by adjusting PMT voltage and offset settings (Figures 61 and 62).

134

Figure 61.
Figure 62.
Thermal (Figure 61) and excimer (Figure 62) entry side laser hole images: fluorescent
(top left), Brightfield (top right), and overlay of both (bottom).
Images were collected through the sample thickness with a z-step size of 10 µm.
The thermal laser hole series was 390 µm and the excimer laser hole series was 380 µm
(Figures 63 and 64).

135

Figure 63. Thermal laser hole z-series images shown in stepwise progression.

136

Figure 64. Excimer laser hole z-series images shown in stepwise progression.
6.3 Results and Discussion
Three-dimensional renderings of the laser drilled hole tunnels were created as
animations from the z-series images. The renderings displayed the side views of the
tunnels, enabling a three-dimensional view of the defect in its entirety, so that their
shapes and diameters could be identified (Figures 65 and 66). The acridine orange dye
was successful in filling the hole tunnels and fluorescing; however, the exit sides of the
137

holes appear dimmer than the entry sides. This is due to light scatter which limits the
distance that the laser can travel through the sample. Light scatter occurs as the laser
travels farther into the sample and as the sample emits the light back to the PMT.
Gamma settings were adjusted to increase displayed color intensity so that the dimmer
areas were more visible and the hole tunnel shape was more defined. Neither the
thermal or excimer laser drilled holes exhibited precisely cylindrical shapes; the
diameters within hole tunnels varied, with the entry sides having the largest diameters
(Figures 65 and 66).

Exit

Entry

Exit

Entry

Figure 65.
Figure 66.
Side views of thermal (Figure 65) and excimer (Figure 66) laser hole 3D renderings.
Additionally, the entry sides of the holes were larger than they were certified to
be (proclaimed drilling sizes). The thermal and excimer laser holes were drilled as 100
µm and 50 µm, respectively; however, their entry sides were much larger than those
measurements in diameter (Figures 67 and 68). This is consistent with previous work

138

conducted by the research team using two-dimensional views of the entry and exit sides
of the holes employing a variety of microscopy techniques (Bix, Kassarjian et al. 2005).
To provide an estimate of diameter for each hole type utilizing this methodology,
measurements were taken from each hole and averaged to determine a mean hole size
for the entry sides (Figures 67 and 68). The mean diameters were 229.96 µm for the
thermal hole and 73.47 µm for the excimer hole (Table 8).

243.8 µm
216.12 µm

83.54 µm

76.91 µm
59.97 µm

100 µm

100 µm

Figure 67.
Figure 68.
Entry side images and diameters of thermal (Figure 67) and excimer (Figure 68) laser
drilled holes.
Table 8. Entry side diameters for thermal and excimer laser holes.
Thermal
Short diameter
216.12 µm
Long diameter
243.8 µm
Middle diameter
n/a
Mean diameter
229.96 µm

139

Excimer
59.97 µm
83.54 µm
76.91 µm
73.47 µm

As seen with the 3D renderings, the entry sides of the tunnels are more clearly
visible than the exit sides (Figures 65 and 66). In an effort to further improve these
images, the following techniques can be refined:
increase the objective from the UPlan FLN 20x (0.50) to a PlanN 40x (0.65)
use a smaller z-step size increment (less than 10 µm)
obtain a z-series from each side of the sample (both entry and exit sides).
The 40x objective will provide a higher resolution image because it has a better
numerical aperture. The smaller z-step size will increase the number of optical sections
and eliminate data gaps between slices along the hole tunnel. A z-series from each side
of the sample will provide clearer imaging of both the entry and exit sides of the hole
tunnel. Diameter measurements can then be more accurately obtained from the
improved images so that described versus measured hole sizes can be compared for
each hole type. Additionally, the 3D renderings of the hole tunnel shape will be more
defined, resulting in a more descript comparison of thermal versus excimer hole tunnel
shapes.

140

APPENDICES

141

APPENDIX A
MATERIALS LIST

Trays
Perfecseal, Mankato, MN
Medtronic Inc. Outer Tray Part No. 350215-001
Part material - 0.025 inch blue tint uncoated polyethylene terephthalate (PETG)
Lidstock
Amcor Flexibles Healthcare, Madison, WI
LKF-002 Paper/PE/Foil/PE/HSC Die Cut Lids
Corrugated shippers
PE bags with twist ties
Tray Heat Sealer
SenCorp, Hyannis, MA
CeraTek model MD-2420 dual-shuttle heat sealer with tray 9-up tooling
Teflon impregnated fiber glass barrier blanket
Holes
Lenox Laser, Glen Arm, MD
Laser drilled 10 µm (+/- 10%) and 100 µm (+/- 10%) holes in bottom of PETG
trays
Flow calibrated and labeled with tray ID, flow rate, and hole diameter
Non-destructive Leak Detector
ATC, Inc., Indianapolis, IN
Model VE2 with Leak-Tek© data acquisition program
Gamma Sterilization
Provided by Smith & Nephew, Memphis, TN
Performed by Sterigenics (West Memphis, AR) according to process procedure
P000257 Rev. M.
Shipping/sterilization cases
SBS cartons for trays and pouches (protection for distribution)
Kraft paper to fill gaps in cases
Media Preparation
Media Prep Lab, Michigan State University, East Lansing, MI
30 ml of sterile nutrient growth agar in glass bottles sealed with foam stoppers
Autoclave bin
142

Insulated box constructed from EPS and corrugated
Agar Injection
Multiterge
Water bath
Injection frame
70% isopropyl alcohol swabs
60 ml syringes (Becton, Dickinson and Company, Franklin Lakes, NJ)
18G1½ needles (Becton, Dickinson and Company, Franklin Lakes, NJ)
16G1 vented needles (Becton, Dickinson and Company, Franklin Lakes, NJ)
Self-sealing septums (Mocon, Minneapolis, MN or Illinois Instruments,
Johnsburg, IL)
70% isopropyl alcohol swabs
Boats
Bunsen burner
Forceps
Butterfield’s Solution
Potassium Phosphate, Monobasic, Crystal (KH2PO4)
1N Sodium Hydroxide (1N NaOH)
Latex gloves
Pipette
Distilled water
Spatula
Boats
Scale (g)
2 L bucket
Stir bar
Corning Stirrer
pH meter
pH strips
(4) 1 L Erlenmeyer flasks
Autoclave bin
Aluminum foil
Autoclave tape
Autoclave
Inoculation of Butterfield’s Solution (Microbial Solution)
Escherichia coli K12 ATCC Number 29181
Pipettes (1 ml)
Sterile pipette tips
Incubation
143

Environmental chamber
Colony Counting of Test Samples
Razor blade
Sharpie
Colony Counting of Agar Plates
Quebec Colony Counter
Sharpie
Enterotubes
BD BBL™ Enterotube™ II REF 211832
Environmental chamber
1 ml syringe
Kovac’s reagent
Latex gloves
Pouches
Mangar Industries, Inc., New Britain, PA
100GA Biax Nylon (0.001), 0.0007 LDPE, 0.002 HDPE Coex
Dimensions: 7.25 in x 9.50 in
Lot #: H150978/1/A
Pouch Heat Sealer
SenCorp, Hyannis, MA
CeraTek Model No. 24AS/1, Serial No. 06-04236
Carton Construction
15.5 pt. SBS
Glue
Rubber bands
Artios Kongsberg Premium Line 1930
Sample Preparation of Trays in Cartons
Trays
Cartons
Self-sealing septums
Template for septum alignment
Agar Preparation for Dilution Series Agar Plates
Petri plates
Difco™ Nutrient Agar
Deionized water
Spatula
Boats
144

Scale (g)
500 ml Erlenmeyer flask
Autoclave bin
Aluminum foil
Autoclave tape
Autoclave
Gloves
Dilution Series
3 dilution bottles
1 ml Escherichia coli K12 ATCC Number 29181
Butterfield’s Solution (99 ml in each dilution bottle)
Wire autoclave basket
Autoclave bin
Autoclave tape
Autoclave
Pipettes (1 ml and 0.1 ml)
Sterile pipette tips
Boat
70% isopropyl alcohol
Hockey stick
Bunsen burner
6 sterile agar plates
Microbial Challenge
Microbial Challenge Chamber
Bleach
Wash cloth
70% isopropyl alcohol swabs
60 ml syringes
18G1½ needles (Becton, Dickinson and Company, Franklin Lakes, NJ)
Petri plates with sterile agar
Pinhole Characterization
Olympus FluoView FV1000 Confocal Laser Scanning Microscope and Olympus
FluoView BSW v. 5.0 software
Nikon SM2645 stereomicroscope
Nunc Lab-Tek II #1.5 Coverglass Chambers
Scissors
Sharpie
Pipette
Acridine orange dye 0.001% concentration
Latex gloves
145

APPENDIX B
NON-DESTRUCTIVE LEAK TEST RESULTS

Table 9. Non-destructive leak test results for trays without pinholes.
ID

Pass or Fail

ID

Pass or Fail

ID

Pass or Fail

1000
1001
1004
1006
1007
1008
1009
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1023
1024
1025
1026
1027
1029
1030
1031
1032
1033

Passed
Passed
Failed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed

1036
1037
1038
1039
1041
1042
1043
1044
1045
1047
1048
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065

Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed

1066
1067
1068
1069
1070
1071
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095

Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed
Passed

146

Table 10. Non-destructive leak test results for trays with 10 µm pinholes.
ID

Pass or Fail

ID

Pass or Fail

ID

Pass or Fail

1
2
3
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
27
28
29
30
31
32
33
35
36
37
38
39
40
41

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

68
69
71
73
74
75
76
77
78
79
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

130
131
132
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

147

Table 10 (cont’d).
42
Failed
43
Failed
44
Failed
45
Failed
46
Failed
47
Failed
48
Failed
49
Failed
50
Failed
52
Failed
53
Failed
54
Failed
55
Failed
56
Failed
57
Failed
58
Failed
59
Failed
60
Failed
61
Failed
62
Failed
63
Failed
64
Failed
65
Failed

107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

148

198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220

Failed
Failed
Passed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

Table 11. Non-destructive leak test results for trays with 100 µm pinholes.
ID

Pass or Fail

ID

Pass or Fail

ID

Pass or Fail

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

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
136

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

200
201
202
205
206
207
208
209
210
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

149

Table 11 (cont’d).
39
Failed
40
Failed
41
Failed
42
Failed
43
Failed
44
Failed
45
Failed
46
Failed
47
Failed
48
Failed
49
Failed
50
Failed
54
Failed
55
Failed
56
Failed
57
Failed
58
Failed
59
Failed
60
Failed
61
Failed
62
Passed
63
Failed
64
Failed
65
Failed
66
Failed
67
Failed
68
Failed
69
Failed
70
Failed
71
Failed
72
Failed
73
Failed
74
Failed
75
Failed
76
Failed
77
Failed
78
Failed
79
Failed

139
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
164
165
166
167
168
169
170
172
174
176
177
178
179
180
181
182
183

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
150

240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
261
262
263
264
265
266
267
268
269
270
271
272
273
275
276
277
278
279
280

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

Table 11 (cont’d).
80
Failed
81
Failed
82
Failed
83
Failed
84
Failed
85
Failed
86
Failed
87
Failed
88
Failed
89
Failed
90
Failed
91
Failed
92
Failed
93
Failed
94
Failed
95
Failed

184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

151

281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
297

Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed
Failed

APPENDIX C
PART 1 TEST SAMPLE RESULTS

Table 12. Part 1 test sample colony counts (CFU).
Injection
Chamber
Injection
Chamber
ID
Position
Position
Run
Run
(A,B,C,D)
(A,B,C,D)
159
11
A
8
A
164
12
A
11
D
167
12
D
15
A
168
13
C
9
D
210
16
D
13
D
212
11
C
12
C
213
16
C
14
B
221
15
D
10
C
223
15
C
14
C
224
15
B
15
D
226
17
B
13
B
262
12
B
11
A
273
13
B
10
B
278
13
A
9
B
284
12
C
12
A
136
18
A
15
B
160
15
A
11
C
165
13
D
14
D
200
17
C
12
B
207
17
D
12
D
208
18
C
15
C
218
10
D
9
C
225
16
A
8
B
227
17
A
14
A
263
11
B
10
A
266
11
D
10
D
269
16
B
11
B
279
14
B
8
D
280
14
A
9
A
282
14
D
13
C

152

Hole
Size
(µm)
100
99.41
96.3
96.73
93.63
98.73
108
95.95
97.37
99.82
96.1
101.69
93.71
105.64
94.56
103.8
101.85
92.77
100.72
93.59
93.34
100.15
95.3
98.56
91.93
99.23
105.37
104.65
103.23
106.21

Flow
Pressure Count
Rate
(psi)
(CFU)
(sccm)
174.89
-3.78
69
172.85
-3.78
9
162.18
-3.78
11
163.66
-3.78
1
153.34
-3.78
11
170.49
-3.78
31
204.00
-3.78
15
161.02
-3.78
20
165.83
-3.78
7
174.27
-3.78
7
161.51
-3.78
10
180.87
-3.78
74
153.58
-3.78
2
195.19
-3.78
18
156.40
-3.78
31
188.44
0
2
181.44
0
0
150.53
0
2
177.44
0
3
153.19
0
0
152.37
0
2
175.43
0
0
158.85
0
0
169.91
0
1
147.80
0
0
172.23
0
1
194.18
0
1
191.56
0
3
186.37
0
0
197.28
0
1

APPENDIX D
PART 1 AGAR PLATE COLONY COUNTS

Table 13. Part 1 microbial chamber agar plate colony counts (CFU per plate).
Chamber Run
Count (CFU)
1
4332
2
4284
3
4410
4
4568
5
4316
6
4111
7
3906
8
3946

153

APPENDIX E
PART 2 AGAR PLATE COLONY COUNTS

Table 14. Microbial chamber agar plate colony counts (CFU per plate) for runs testing
unlidded trays in pouches.
Chamber Run Count (CFU)
1
7576
2
7907
3
5607
4
4631
5
3024
6
2016
7
1276
8
1055
9
1150
10
1134

Table 15. Microbial chamber agar plate colony counts (CFU per plate) for runs testing
unlidded trays in cartons.
Chamber Run Count (CFU)
1
6615
2
5654
3
6505
4
5843
5
4473
6
3827
7
3386
8
2615
9
2331
10
2142

154

Table 16. Microbial chamber agar plate colony counts (CFU per plate) for runs testing
lidded trays, with 100 µm pinholes, in cartons.
Chamber Run Count (CFU)
1
9072
2
10584
3
7308
4
6867
5
6930
6
4662
7
3906
8
3024
9
3150
10
3528

155

APPENDIX F
PART 2 TEST SAMPLE RESULTS

Table 17. Colony counts (CFU) of unlidded trays in pouches.
Chamber
Injection Chamber
Count
ID
Position
Number
Run
(CFU)
(A,B,C,D)
2000
24
7
C
0
2001
37
6
D
0
2002
39
5
D
0
2003
40
5
C
0
2004
30
8
C
0
2005
31
7
A
0
2006
32
7
B
0
2007
17
9
D
0
2008
3
2
B
0
2010
5
1
C
0
2011
14
3
B
0
2012
4
1
D
0
2013
26
10
C
0
2014
25
10
D
0
2015
34
6
A
0
2016
33
6
B
0
2017
23
7
D
0
2018
2
2
C
0
2020
12
3
D
0
2021
22
8
A
0
2023
7
1
A
0
2024
41
5
A
0
2026
6
1
B
0
2028
11
4
A
0
2029
1
2
D
0
2030
18
9
C
0
2031
28
10
A
0
2033
38
6
C
0
2034
19
9
B
0
2035
20
9
A
0
2036
9
4
C
0
2037
10
4
B
0
2038
13
3
C
0
2039
21
8
B
0
156

Table 17 (cont’d).
2043
29
2044
8
2045
16
2046
15
2049
27

8
2
4
3
10

D
A
D
A
B

0
0
0
0
0

Table 18. Colony counts (CFU) of unlidded trays in cartons.
Chamber
Injection Chamber
Count
ID
Position
Number
Run
(CFU)
(A,B,C,D)
3000
1
10
D
0
3001
9
9
A
2
3002
8
7
A
10
3005
37
1
A
3
3006
38
2
B
2
3007
30
5
A
124
3008
31
2
D
62
3009
32
1
B
15
3010
40
3
C
88
3012
5
10
C
0
3016
22
4
C
19
3017
2
9
C
1
3018
19
7
B
7
3020
6
9
B
5
3021
33
4
A
8
3024
12
6
A
4
3025
13
10
A
21
3026
14
8
D
65
3027
15
7
C
3
3028
16
6
D
15
3029
17
9
D
3
3030
18
8
C
2
3031
27
3
A
12
3032
10
10
B
3
3033
21
5
C
7
3034
35
2
C
10
3035
3
8
A
1
3036
7
8
B
2
3037
23
3
B
41
3042
4
6
B
3
157

Table 18 (cont’d).
3044
11
3045
39
3046
26
3048
20
3049
34
3050
29
3051
28
3052
24
3053
25

7
4
5
6
5
3
1
2
4

D
D
B
C
D
D
C
A
B

4
2
49
2
5
101
4
8
40

Table 19. Colony counts (CFU) of lidded trays, with 100 µm pinholes, in cartons.
Chamber
Injection
Chamber
Hole Size Flow Rate
Count
ID
Position
Number
Run
(µm)
(sccm)
(CFU)
(A, B, C, D)
66
35
2
B
101.82
181.33
0
69
33
3
B
92.14
148.49
0
71
38
1
B
93.34
152.39
0
92
25
7
D
98.85
170.91
0
93
32
3
A
98.82
170.80
0
94
29
3
C
97.77
167.18
0
95
30
4
C
106.5
198.51
0
152
14
8
D
98.66
170.25
0
153
24
9
A
93.51
152.93
0
154
34
2
C
93.15
151.75
0
187
7
7
B
97.12
164.98
0
188
11
9
D
101.46
180.04
0
189
12
10
C
96.61
163.25
0
191
17
3
D
91.57
146.65
0
192
10
4
A
101.11
178.81
0
194
16
5
C
99.69
173.83
0
197
28
4
B
94.54
156.31
1
228
26
6
B
108.11
204.41
0
231
31
2
D
97.95
167.80
0
240
40
1
A
94.67
156.75
0
245
5
8
C
102.39
183.36
0
246
1
10
B
93.14
151.73
1
247
20
7
A
99.73
173.97
1
248
2
9
B
92.67
150.20
0
249
19
5
A
95.94
160.99
0
251
37
1
C
100.28
175.87
0
158

Table 19 (cont’d).
252
18
254
39
255
15
256
13
257
3
287
36
288
9
289
22
290
4
291
8
292
23
293
6
294
21
295
27

8
1
8
9
7
2
5
4
6
6
10
10
6
5

A
D
B
C
C
A
D
D
C
D
D
A
A
B

159

104.68
90.04
91.15
98.64
101.9
101.14
103.65
106.35
103.16
104.68
106.52
102.82
102.53
104.68

191.66
141.78
145.33
170.16
181.61
178.92
187.92
197.80
186.12
191.67
198.47
184.91
183.88
191.64

0
0
0
1
0
0
0
0
1
0
0
0
0
1

APPENDIX G
PART 3 AGAR PLATE COLONY COUNTS

Table 20. Part 3 microbial chamber agar plate colony counts (CFU per plate).
Chamber
Count
Chamber
Count
Chamber
Count
Run
(CFU)
Run
(CFU)
Run
(CFU)
1
26649
26
23247
51
19278
2
23814
27
20412
52
23247
3
22113
28
22680
53
19278
4
24381
29
19845
54
17577
5
23247
30
16443
55
24381
6
20412
31
19278
56
23247
7
23247
32
21546
57
24948
8
22680
33
19845
58
24381
9
17577
34
17010
59
22113
10
15309
35
19845
60
21546
11
22113
36
18711
61
21546
12
21546
37
17577
62
22113
13
24948
38
20412
63
22680
14
24381
39
17577
64
21546
15
23814
40
15876
65
18144
16
21546
41
11907
66
19278
17
20412
42
13608
67
24948
18
19278
43
23247
68
24381
19
21546
44
23814
69
25515
20
18711
45
20979
70
26082
21
17577
46
23814
71
26649
22
20412
47
20412
72
25515
23
18144
48
23814
73
21546
24
20979
49
22113
74
23247
25
24381
50
19845
75
24381
76

160

22113

APPENDIX H
PART 3 TEST SAMPLE RESULTS

Table 21. Colony counts (CFU) of split plot test samples.
Chamber
Hole
Flow
Carton
Chamber
ID
Position
Size
Rate
(1=yes,
Run
(A, B, C, D)
(µm)
(sccm)
2=no)
229
1
A
104.72
191.784
1
81
1
B
10.18
1.812
2
52
1
C
8.28
1.2
1
20
1
D
103.99
189.128
2
176
2
A
91.06
145.034
2
70
2
B
91
144.846
1
165
2
C
10.77
2.03
2
63
2
D
9.15
1.464
1
190
3
A
10.79
2.035
1
125
3
B
94.42
155.932
2
100
3
C
104.82
192.163
1
92
3
D
9.53
1.589
2
150
4
A
102.88
185.111
1
33
4
B
102.08
182.262
2
203
4
C
11.46
2.297
2
124
4
D
11.11
2.158
1
42
5
A
11.96
2.5
2
149
5
B
107.69
202.825
1
80
5
C
102.82
184.911
2
208
5
D
11.84
2.453
1
121
6
A
99.29
172.426
2
108
6
B
8.25
1.19
1
110
6
C
8.31
1.207
2
230
6
D
100.24
175.755
1
213
7
A
9.19
1.477
2
31
7
B
11.37
2.263
1
39
7
C
107.61
202.521
1
84
7
D
96.89
164.194
2
18
8
A
101.87
181.5
2
97
8
B
8
1.12
2
35
8
C
10.35
1.875
1
161

Pressure
(psi)

Count
(CFU)

-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0

1
0
0
32
0
0
0
0
0
74
2
0
1
0
0
0
0
4
132
0
0
0
0
0
0
0
3
148
0
0
0

Table 21 (cont’d).
42
8
109
9
122
9
32
9
47
9
28
10
195
10
124
10
194
10
43
11
173
11
130
11
47
11
166
12
193
12
116
12
264
12
123
13
190
13
183
13
207
13
82
14
94
14
40
14
113
14
20
15
40
15
184
15
45
15
120
16
81
16
18
16
198
16
133
17
37
17
237
17
191
17
233
18

D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A

96.99
8.25
93.92
9.09
107.97
9.79
94.4
102.16
11.01
93.26
11.14
106.26
10.64
9.17
95.73
9.17
97.13
8
101.31
95.4
10.43
97.98
10.78
107.09
8.25
10.33
11
93.17
99.72
10.02
98.58
10.07
96.92
97.07
9.25
97.14
10.82
104.22

164.539
1.19
154.268
1.446
203.898
1.678
155.861
182.528
2.121
152.109
2.17
197.471
1.98
1.472
160.268
1.472
164.992
1.12
179.51
159.182
1.902
167.901
2.031
200.577
1.19
1.868
2.117
151.81
173.932
1.755
169.966
1.773
164.293
164.794
1.496
165.028
2.047
189.963
162

1
2
2
1
1
1
1
2
2
1
2
2
1
2
1
1
2
1
1
2
2
2
2
1
1
2
1
2
1
1
2
2
1
2
1
1
2
1

0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0

0
0
33
0
16
0
0
3
0
1
0
72
0
0
0
0
1
0
3
152
0
0
0
0
0
0
0
146
0
0
3
0
0
86
0
0
0
0

Table 21 (cont’d).
83
18
107
18
193
18
196
19
121
19
104
19
16
19
127
20
232
20
211
20
192
20
117
21
100
21
110
21
123
21
12
22
28
22
32
22
39
22
114
23
195
23
131
23
234
23
113
24
182
24
27
24
98
24
38
25
10
25
36
25
87
25
60
26
41
26
93
26
33
26
27
27
30
27
108
27

B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C

98.39
10.13
11.87
91.39
11.54
8.25
99.97
98.32
96.19
10.79
10.57
11.74
11.86
99.63
99.43
9.23
97.45
104.19
9
9.13
10.64
107.1
103.75
106.57
10.84
98.21
8.75
94.89
8.68
9.63
105.7
90.19
105.46
11.84
8.78
10.08
100.12
104.4

169.307
1.796
2.462
146.089
2.33
1.19
174.789
169.06
161.82
2.037
1.1954
2.41
2.46
173.624
172.9
1.49
166.11
189.846
1.418
1.459
1.98
200.606
188.272
198.622
2.054
168.703
1.34
157.495
1.318
1.623
195.39
142.278
194.515
2.45
1.347
1.778
175.323
190.619
163

2
1
2
1
1
2
2
2
1
1
2
1
2
1
2
2
2
1
1
1
2
2
1
1
1
2
2
1
2
1
2
2
1
2
1
1
2
1

0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78

0
0
0
0
0
2
41
4
0
0
0
0
0
0
15
0
1
0
0
0
0
171
0
0
0
0
0
3
0
0
84
0
0
0
0
0
0
1

Table 21 (cont’d).
46
27
109
28
132
28
45
28
185
28
44
29
88
29
120
29
64
29
29
30
188
30
43
30
31
30
14
31
29
31
25
31
134
31
101
32
16
32
122
32
64
32
202
33
21
33
62
33
111
33
61
34
74
34
22
34
2
34
105
35
175
35
98
35
132
35
95
36
26
36
180
36
90
36
115
37

D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A

8.01
106.98
107.31
9.66
9.75
8.1
102.58
97.15
8.71
93.11
10.72
9.53
97.98
8.04
8.39
96.54
107
102.14
11.78
8
104.44
9.28
90.16
11.02
103.14
8.82
105.62
90.35
11.25
107.75
11.09
104.26
11.96
11.96
95.07
11.9
99.66
9.03

1.121
200.177
201.416
1.631
1.661
1.148
184.044
165.059
1.327
151.613
2.009
1.589
167.905
1.13
1.23
163.019
200.226
182.468
2.427
1.12
190.779
1.507
142.164
2.124
186.068
1.36
195.13
142.785
2.213
203.063
2.151
190.105
2.503
2.5
158.093
2.477
173.719
1.427
164

2
1
2
2
1
2
1
2
1
2
1
2
1
2
1
1
2
2
2
1
1
2
2
1
1
1
1
2
2
1
2
2
1
2
1
1
2
1

-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78

0
0
0
0
0
0
1
11
0
3
0
0
0
0
0
2
108
0
0
0
0
0
63
0
0
0
0
0
0
3
0
34
0
0
0
0
0
0

Table 21 (cont’d).
236
37
23
37
86
37
58
38
21
38
46
38
87
38
176
39
38
39
78
39
235
39
127
40
49
40
204
40
99
40
77
41
184
41
75
41
219
41
61
42
144
42
119
42
7
42
180
43
58
43
106
43
179
43
145
44
68
44
174
44
216
44
55
45
170
45
182
45
178
45
197
46
13
46
241
46

B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C

98.03
9.77
105.86
94.9
11.82
104.82
10.31
8.13
8
100.38
106.25
8
102.62
10.13
100.42
103.11
8
105.31
10.58
102.19
96.04
11.22
11.96
92.14
10.43
10.73
91.78
96.87
98.66
9.13
8.95
9.32
10.11
90.21
103.95
9.26
95.46
102.33

168.085
1.67
196.013
157.503
2.446
192.163
1.859
1.156
1.12
176.234
197.447
1.12
184.186
1.794
176.367
185.96
1.12
193.968
1.957
182.63
161.336
2.204
2.5
148.478
1.904
2.013
147.331
164.12
170.235
1.457
1.401
1.519
1.788
142.328
188.986
1.499
159.391
183.14
165

1
2
2
2
2
1
1
2
1
2
1
1
2
2
1
2
1
1
2
1
2
1
2
1
1
2
2
2
1
1
2
1
2
1
2
2
2
1

-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0

1
0
86
0
0
0
0
0
0
61
8
0
1
0
0
1
1
1
0
0
1
0
0
2
0
0
0
0
0
0
0
0
0
4
106
0
1
1

Table 21 (cont’d).
50
46
181
47
214
47
12
47
151
47
36
48
169
48
148
48
15
48
35
49
199
49
130
49
147
49
181
50
63
50
26
50
74
50
15
51
23
51
205
51
25
51
24
52
174
52
168
52
56
52
83
53
242
53
14
53
71
53
146
54
178
54
13
54
112
54
103
55
111
55
106
55
6
55
129
56

D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A

10.67
11.95
9.89
91.42
99.85
101.64
11.4
92.65
10.78
102.09
11.66
11.01
93.45
97.59
99.62
10.03
10.09
94.31
96.1
9.23
10.53
92.45
94.82
11.03
10.66
10.32
102.69
101.47
8.05
99.2
11.72
10.13
104.31
8.25
11.86
99.37
96.58
94.65

1.99
2.497
1.709
146.183
174.373
180.672
2.271
150.126
2.034
182.277
2.378
2.12
152.74
166.58
173.588
1.76
1.78
155.552
161.523
1.488
1.939
149.498
157.236
2.127
1.988
1.863
184.421
180.096
1.132
172.105
2.404
1.796
190.3
1.19
2.459
172.709
163.146
156.674
166

1
1
2
2
1
1
1
2
2
1
2
1
2
2
1
2
1
2
1
2
1
1
2
2
1
2
1
2
1
2
1
2
1
2
1
1
2
2

0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0

0
0
4
173
1
0
0
0
0
7
0
0
97
0
0
0
0
0
0
0
0
0
0
0
0
0
15
60
0
2
0
0
0
0
0
4
42
0

Table 21 (cont’d).
102
56
73
56
186
56
9
57
79
57
60
57
102
57
69
58
253
58
59
58
24
58
222
59
212
59
85
59
118
59
220
60
96
60
189
60
7
60
172
61
104
61
11
61
4
61
76
62
82
62
239
62
57
62
41
63
131
63
128
63
243
63
79
64
57
64
218
64
34
64
5
65
68
65
103
65

B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C

8.32
11.4
107.21
8
97.5
8.96
105.16
8.79
101.83
95.61
10.12
92.67
8.34
91.73
9.46
8.84
102.53
11.79
95.85
8
107.26
8.39
103.04
101.48
9.13
107.16
10.04
12
9.41
97.43
97.66
10.48
91.73
10.58
108.04
93.31
9.97
100.65

1.21
2.271
201.015
1.12
166.251
1.404
193.423
1.35
181.346
159.875
1.79
150.189
1.215
147.169
1.567
1.366
183.86
2.43
160.698
1.12
201.204
1.23
185.693
180.103
1.458
200.835
1.761
2.52
1.55
166.025
166.813
1.923
147.176
1.957
204.171
152.268
1.74
177.17
167

2
1
1
2
2
1
1
1
1
2
2
1
2
2
1
2
1
1
2
1
1
2
2
2
2
1
1
2
1
2
1
1
2
2
1
2
1
1

0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78

0
0
0
0
143
0
1
0
0
0
0
0
0
115
0
0
0
0
0
0
6
0
70
0
0
0
0
0
0
232
0
0
1
0
0
2
0
2

Table 21 (cont’d).
96
65
185
66
17
66
77
66
177
66
244
67
217
67
19
67
11
67
8
68
238
68
48
68
89
68
53
69
196
69
250
69
10
69
167
70
50
70
97
70
75
70
65
71
210
71
91
71
44
71
56
72
3
72
48
72
91
72
54
73
22
73
171
73
2
73
177
74
19
74
17
74
76
74
30
75

D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A

8.02
102.64
103.71
9.13
9.29
99.51
9.85
10.07
103.45
92.37
105.26
10.15
8.35
11.71
11.25
106.7
95.25
9.1
98.05
106.06
11.15
11.43
10.82
101.58
95.3
103.44
11.24
101.3
8.42
105.05
11.23
11.87
100.69
104.19
96.36
8.09
10.61
8.96

1.124
184.251
188.121
1.458
1.509
173.194
1.696
1.773
187.191
149.216
193.767
1.803
1.22
2.398
2.215
199.116
158.684
1.45
168.16
196.722
2.174
2.286
2.047
180.454
158.83
187.126
2.208
179.48
1.24
193.004
2.207
2.464
177.309
189.877
162.4
1.144
1.967
1.404
168

2
1
2
1
2
1
1
2
2
2
1
1
2
1
2
1
2
2
2
1
1
1
2
2
1
1
1
2
2
1
2
1
2
2
1
2
1
1

-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78
-3.78
-3.78
-3.78
0
0
0
0
-3.78

0
0
6
0
0
1
0
0
9
0
0
0
0
0
0
3
54
0
0
0
0
0
0
286
2
0
0
1
0
2
0
0
7
0
0
0
0
0

Table 21 (cont’d).
9
75
65
75
85
75
126
76
55
76
112
76
198
76

B
C
D
A
B
C
D

97.47
102.39
10.66
98.01
90.87
8.25
9.61

166.161
183.37
1.986
168.011
144.427
1.19
1.617

169

2
1
2
1
2
2
1

-3.78
-3.78
-3.78
0
0
0
0

36
6
1
0
1
0
0

BIBLIOGRAPHY

170

BIBLIOGRAPHY

Ahvenainen, R., T. Mattila-Sandholm, et al. (1992). "The Effect of Microhole Size and
Foodstuff on the Microbial Integrity of Aseptic Plastic Cups." Packaging
Technology and Science 5: 101-107.
Allan-Wojtas, P., C. F. Forney, et al. (2008). "Structure of Gas Transmission
Characteristics of Microperforations in Plastic Films." Packaging Technology and
Science 21(4): 217-229.
Allen, D. (2002). "Cleaning Up Package Integrity Testing." Pharmaceutical & Medical
Packaging News: 32-38.
Allen, D. (2003). "Testing Traditions." Pharmaceutical & Medical Packaging News: 38-40.
Allen, D. (2009). Reducing Risk through Packaging. Pharmaceutical & Medical Packaging
News. 15.
Allen, D. (2010). "Innovation is Essential." Pharmaceutical & Medical Packaging News
18(5).
Allison, D. G. (1999). "A Review: Taking the sterile out of sterility." Journal of Applied
Microbiology 87: 789-793.
Anderson, D. J., K. B. Kirkland, et al. (2007). "Underresourced Hosptial Infection Control
and Prevention Programs: Penny Wise, Pound Foolish?" Infection Control and
Hospital Epidemiology 28(7): 767-773.
ASTM (2004). F1886 Standard Test Method for Determining Integrity of Seals for
Medical Packaging by Visual Inspection.
Axelson, L., S. Cavlin, et al. (1990). "Aseptic Integrity and Microhole Determination of
Packages by Electrolytic Conductance Measurement." Packaging Technology and
Science 3: 141-162.
Bascetta, C. A., N. Edwards, et al. (2008). Health-Care-Associated Infections in Hospitals:
An Overview of State Reporting Programs and Individual Hospital Initiatives to
Reduce Certain Infections. United States Government Accountability Office.
Bascetta, C. A., L. T. Kohn, et al. (2008). Health-Care-Associated Infections in Hospitals:
Leadership Needed from HHS to Prioritize Prevention Practices and Improve
Data on These Infections. United States Government Accountability Office.
171

Belkin, N. L. (2004). "Wrapping for event-related shelf life." Healthcare Purchasing News
28(3): 78-79.
Bix, L., O. Kassarjian, et al. (2005). "Using Scanning Electron, Confocal and Optical
Microscopes to Measure Microscopic Holes in Trays." Packaging Technology and
Science 18: 311-320.
Blocher, J. (2009). The Effect of Relative Humidity on the Microbial Barrier Properties of
Porous Packaging Materials Used in the Medical Device Industry. Packaging
Science. Clemson, Clemson University: 79.
Blocher, J., J. Neid, et al. (2010). Nurses Survey Results. HealthPack, San Antonio, TX.
Brossette, S. E., D. M. Hacek, et al. (2006). "A Laboratory-Based, Hospital-Wide,
Electronic Marker for Nosocomial Infection." American Journal of Clinical
Pathology 125: 34-39.
Chen, C., B. Harte, et al. (1991). "Assessment of Package Integrity Using a Spray Cabinet
Technique." Journal of Food Protection 54(8): 643-647.
Christensen, M. and O. B. Jepsen (2001). "Reduced rates of hospital-acquired UTI in
medical patients. Prevalence surveys indicate effect of active infection control
programmes." Journal of Hospital Infection 47: 36-40.
Cole, M. (2008). "The true cost of health care associated infection." Journal of
Orthopaedic Nursing 12: 136-138.
Cosgrove, S. E. (2006). "The Relationship between Antimicrobial Resistance and Patient
Outcomes: Mortality, Length of Hospital Stay, and Health Care Costs." Clinical
Infectious Diseases 42: 582-589.
Cosgrove, S. E., Y. Qi, et al. (2005). "The Impact of Methicillin Resistance in
Staphylococcus aureus Bacteremia on Patient Outcomes: Mortality, Length of
Stay, and Hospital Charges." Infection Control and Hospital Epidemiology 26(2):
166-174.
Craig, C. P. and S. Connelly (1984). "Effect of intensive care unit nosocomial pneumonia
on duration of stay and mortality." American Journal of Infection Control 12(4):
233-238.
Craven, D. E., L. M. Kunches, et al. (1988). "Nosocomial Infection and Fatality in Medical
and Surgical Intensive Care Unit Patients." Archives of Internal Medicine 148:
1161-1168.
172

Crick, B., S. Chua, et al. (2008). "Potential for Contamination of Orthopaedic Implants
Using Individually Wrapped Screws." ANZ Journal of Surgery 78: 266-268.
Dietze, B., A. Rath, et al. (2001). "Survival of MRSA on sterile goods packaging." Journal
of Hospital Infection 49: 255-261.
DiGiovine, B., C. Chenoweth, et al. (1999). "The Attributable Mortality and Costs of
Primary Nosocomial Bloodstream Infections in the Intensive Care Unit."
American Journal of Respiratory and Critical Care Medicine 160: 976-981.
Division of Healthcare Quality Promotion, National Center for Preparedness, Detection,
and Control of Infectious Diseases. (2007). "Information About MRSA for
Healthcare Personnel." Retrieved June 1, 2010, from
http://www.cdc.gov/ncidod/dhqp/ar_mrsa_healthcareFS.html.
Division of Healthcare Quality Promotion, National Center for Preparedness, Detection,
and Control of Infectious Diseases. (2007). "MRSA in Healthcare Settings."
Retrieved June 1, 2010, from
http://www.cdc.gov/ncidod/dhqp/ar_mrsa_spotlight_2006.html.
Division of Healthcare Quality Promotion, National Center for Preparedness, Detection,
and Control of Infectious Diseases. (2007). "S. aureus and MRSA Surveillance
Summary 2007." Retrieved June 1, 2010, from
http://www.cdc.gov/ncidod/dhqp/ar_mrsa_surveillanceFS.html.
Division of Healthcare Quality Promotion, National Center for Preparedness, Detection,
and Control of Infectious Diseases. (2010). "Healthcare-associated Methicillin
Resistant Staphylococcus aureus (HA-MRSA)." Retrieved June 1, 2010, from
http://www.cdc.gov/ncidod/dhqp/ar_mrsa.html.
Dunkelberg, H. and F. Fleitmann-Glende (2006). "Measurement of the microbial barrier
effectiveness of sterilization containers in terms of the log reduction value for
prevention of nosocomial infections." American Journal of Infection Control
34(5): 285-289.
Dunkelberg, H. and S. Rohmann (2006). "Test to Determine Sterile Integrity of Wrapped
Medical Products at a Probability of Recontamination of 1:1,000,000." Infection
Control and Hospital Epidemiology 27(4): 367-371.
Dunkelberg, H. and U. Schmelz (2009). "Determination of the Efficacy of Sterile Barrier
Systems Against Microbial Challenges During Transport and Storage." Infection
Control and Hospital Epidemiology 30(2): 179-183.

173

Dunkelberg, H. and S. Wedekind (2004). "Preliminary Results for a New Final Package
Test to Assess the Quality of Sterile Package Systems." Infection Control and
Hospital Epidemiology 25(1): 26-29.
Emori, G. T. and R. P. Gaynes (1993). "An Overview of Nosocomial Infections, Including
the Role of the Microbiology Laboratory." Clinical Microbiology Reviews 6(4):
428-442.
Engemann, J. J., Y. Carmeli, et al. (2003). "Adverse Clinical and Economic Outcomes
Attributable to Methicillin Resistance among Patients with Staphylococcus
aureus Surgical Site Infection." Clinical Infectious Diseases 36: 592-598.
Fagon, J.-Y., A. Novara, et al. (1994). "Mortality Attributable to Nosocomial Infections in
the ICU." Infection Control and Hospital Epidemiology 15(7): 428-434.
Floros, J. D. (1994). "Critical Leak Size and Package Integrity." Packaging Technologies &
Inspection News IV: 3.
Franks, S. (2002). "Strength and Integrity, Part Two: Basics of Seal-Strength Testing."
Pharmaceutical & Medical Packaging News: 52-53.
Franks, S. (2003). "Strength and Integrity, Part Three: Package Integrity Testing."
Pharmaceutical & Medical Packaging News: 41-43.
Gerberding, J. L. (2002). "Hospital-Onset Infections: A Patient Safety Issue." Annals of
Internal Medicine 137(8): 665-670.
Ghosh, V. and R. C. Anantheswaran (2001). "Oxygen transmission rate through mircoperforated films: measurement and model comparison." Journal of Food Process
Engineering 24: 113-133.
Gilchrist, J. E., D. B. Shah, et al. (1989). "Leak Detection in Flexible Retort Pouches."
Journal of Food Protection 52(6): 412-415.
Gilmore, G. K. (2003). "Evolving standards challenge compliance." Nursing Management
34(12): 24-25.
Givens, C. D. and R. P. Wenzel (1980). "Catheter-associated urinary tract infections in
surgical patients: a controlled study on the excess morbidity and costs." Journal
of Urology 124(5): 646-648.
Glabman, M. (2004). "The Top Ten Malpractice Claims." Hospitals & Health Networks
78(9): 61-62, 64, 66.
174

Gnanasekharan, V. and J. D. Floros (1995). A Theoretical Perspective on the Minimum
Leak Size for Package Integrity Evaluation. Plastic Package Integrity Testing Assuring Seal Quality, Institute of Packaging Professionals and Food Processors
Institute55-65.
Graves, N. (2004). "Economics and Preventing Hospital-acquired infection." Emerging
Infectious Diseases 10(4): 561-566.
Hackett, E. T. (2001). Effect of Pinholes on Sterile Barrier Properties. HealthPack. St.
Petersburg, Florida.
Hackett, E. T., M. H. Scholla, et al. (2000). "How Small is Small? The Effects of Pinholes
on Medical Packaging." Medical Device & Diagnostic Industry: 52-66.
Haley, R. W., D. H. Culver, et al. (1985). "Identifying Patients at High Risk of Surgical
Wound Infection." American Journal of Epidemiology 121(2): 206-215.
Haley, R. W., D. H. Culver, et al. (1985). "The Nationwide Nosocomial Infection Rate. A
New Need for Vital Statistics." American Journal of Epidemiology 121(2): 159167.
Hansen, J., L. Jones, et al. (1995). "In Quest of Sterile Packaging: Part 1 Approaches to
Package Testing." Medical Device & Diagnostic Industry 17(8): 56-61.
Harbath, S., H. Sax, et al. (2003). "The preventable proportion of nosocomial infections:
an overview of published reports." Journal of Hospital Infection 54: 258-266.
Heyland, D. K., D. J. Cook, et al. (1999). "The Attributable Morbidity and Mortality of
Ventilator-Associated Pneumonia in the Critically Ill Patient." American Journal of
Respiratory and Critical Care Medicine 159: 1249-1256.
Hollenbeak, C. S., D. Murphy, et al. (2002). "Nonrandom Selection and the Attributable
Cost of Surgical-Site Infections." Infection Control and Hospital Epidemiology
23(4): 177-182.
Hu, K. K., D. L. Veenstra, et al. (2004). "Use of Maximal Sterile Barriers during Central
Venous Catheter Insertion: Clinical and Economic Outcomes." Clinical Infectious
Diseases 39(15): 1441-1445.
ISO (2006). ISO 11607-1:2006 Packaging for terminally sterilized medical devices - Part 1:
Requirements for materials, sterile barrier systems, and packaging systems.
ISO (2006). ISO 11607-2:2006 Packaging for terminally sterilized medical devices - Part 2:
Validation requirements for forming, sealing, and assembly processes.
175

Jarvis, W. R. (1996). "Selected Aspects of the Socioeconomic Impact of Nosocomial
Infections: Morbidity, Mortality, Cost, and Prevention." Infection Control and
Hospital Epidemiology 17(8): 552-557.
Jones, L., J. Hansen, et al. (1995). "In Quest of Sterile Packaging: Part 2 Physical Package
Integrity Test Methods." Medical Device & Diagnostic Industry 17(9): 81-85.
Keehan, S., A. Sisko, et al. (2008) "Health Spending Projections Through 2017: The BabyBoom Generation is Coming to Medicare." Health Affairs 27, 145-155.
Keller, S. W., J. E. Marcy, et al. (1996). "Bioaerosol Exposure Method for Package
Integrity Testing." Journal of Food Protection 59(7): 768-771.
Kilgore, M. L., K. Ghosh, et al. (2008). "The Costs of Nosocomial Infections." Medical
Care 46(1): 101-104.
Kirkland, K. B., J. P. Briggs, et al. (1999). "The Impact of Surgical-Site Infections in the
1990s: Attributable Mortality, Excess Length of Hospitalization, and Extra Costs."
Infection Control and Hospital Epidemiology 20(11): 725-730.
Klevens, R. M., J. Edwards, et al. (2008). "The Impact of Antimicrobial-Resistant Health
Care-Associated Infections on Mortality in the United States." Clinical Infectious
Diseases 47: 927-930.
Klevens, R. M., J. Edwards, et al. (2006). "Changes in Epidemiology of MethicillinResistant Staphylococcus aureus in Intensive Care Units in US Hospitals, 19922003." Clinical Infectious Diseases 42: 389-391.
Klevens, R. M., J. R. Edwards, et al. (2007). "Estimating Health Care-Associated Infections
and Deaths in U.S. Hospitals, 2002." Public Health Reports 122: 160-166.
Klevens, R. M., M. A. Morrison, et al. (2007). "Invasive Methicillin-Resistant
Staphylococcus auerus Infections in the United States." Journal of the American
Medical Association 298(15): 1763-1771.
Kollef, M. H. (1993). "Ventilator-associated pneumonia: a mulitvariate analysis." Journal
of the American Medical Association 270(16): 1965-1970.
Kowalski, W. J. (2007). "Airborne Superbugs: Can Hospital-Acquired Infections Cause
Community Epidemics?" Consulting - Specifying Engineer 42(3): 28.
Kuehnert, M. J., H. A. Hill, et al. (2005). "Methicillin-resistant Staphylococcus aureus
Hospitalizations, United States." Emerging Infectious Diseases 11(6): 868-872.
176

Lampi, R. A. (1981). "Retort Pouch: The Development of a Basic Packaging Concept in
Today's High Technology Era." Journal of Food Process Engineering 4: 1-18.
Lazare, S. and V. Tokarev (2004). Recent Experimental and Theoretical Advances in
Microdrilling of Polymers with Ultraviolet Laser Beams. International Symposium
on Laser Precision Microfabrication. Bellingham, WA. 5662: 221-231.
Leu, H. S., D. L. Kaiser, et al. (1989). "Hospital acquired pneumonia: attributable
mortality and morbidity." American Journal of Epidemiology 129(6): 1258-1267.
Malone, N. and E. Larson (1996). "Factors associated with a significant reduction in
hospital-wide infection rates." American Journal of Infection Control 24(3): 180185.
Maunder, D. T., J. F. Folinazzo, et al. (1968). "Bio-test Method for Determining Integrity
of Flexible Packages of Shelf-Stable Foods." Food Technology 22(615): 81-84.
McCormack, J. (1995). "Put Those Nagging Sterilization Worries to Rest, Once and For
All." Materials Management in Health Care 4(9): 50,52.
Murphy, D., J. Whiting, et al. (2007). "Dispelling the Myths: The True Cost of HealthcareAssociated Infections." Association for Professionals in Infection Control and
Epidemiology February.
National Coalition on Health Care (2009). Facts on Health Care Costs.
National Coalition on Health Care. (2009). "Health Care Facts: Costs." Retrieved August
17, 2010, from http://nchc.org/facts-resources/fact-sheet-cost.
NNIS System, Division of Healthcare Quality Promotion, National Center for Infectious
Diseases, Centers for Disease Control and Prevention, Public Health Service, U.S.
Department of Health and Human Services, Atlanta, Georgia (2001). "National
Nosocomial Infections Surveillance (NNIS) System Report, Data Summary from
January 1992-June 2001, Issued August 2001." American Journal of Infection
Control 29: 404-421.
NNIS System, Division of Healthcare Quality Promotion, National Center for Infectious
Diseases, Centers for Disease Control and Prevention, Public Health Service, U.S.
Department of Health and Human Services, Atlanta, Georgia (2004). "National
Nosocomial Infections Surveillance (NNIS) System Report, data summary from
January 1992 through June 2004, issued October 2004." American Journal of
Infection Control 32(8): 470-485.

177

Noskin, G. A., R. J. Rubin, et al. (2007). "National Trends in Staphylococcus aureus
Infection Rates: Impact on Economic Burden and Mortality over a 6-Year Period
(1998-2003)." Clinical Infectious Diseases 45: 1132-1140.
Olympus. (2011). Retrieved January 23, 2011, from
http://www.olympusfluoview.com/theory/index.html.
Piergiovanni, L., S. Limbo, et al. (2003). "Assessment of the risk of physical
contamination of bread packaged in perforated oriented polypropylene films:
measurements, procedures, and results." Food Additives & Contaminants 20(2):
186-195.
Pilchik, R. (1999). "Materials for Preformed Rigid-Tray Packages." Medical Device
Technology 10(3): 20-24.
Pimentel, D., M. Tort, et al. (1998). "Ecology of Increasing Disease." BioScience 48(10):
817-826.
Pittet, D., D. Tarara, et al. (1994). "Nosocomial Bloodstream Infection in Critically Ill
Patients." Journal of the American Medical Association 271(20): 1598-1601.
Placencia, A. M., M. L. Arin, et al. (1988). "Physical Tests are not Enough." Medical
Device & Diagnostic Industry 10(9): 72, 74-78.
Richards, M. J., J. R. Edwards, et al. (2000). "Nosocomial Infections in Combined
Medical-Surgical Intensive Care Units in the United States." Infection Control and
Hospital Epidemiology 21(8): 510-515.
Rubin, R. J., C. A. Harrington, et al. (1999). "The Economic Impact of Staphylococcus
aureus Infection in New York City Hospitals." Emerging Infectious Diseases 5(1):
9-17.
Schierholz, J. M. and J. Beuth (2001). "Implant infections: a haven for opportunistic
bacteria." Journal of Hospital Infection 49: 87-93.
Schneider, P. M. (1980). "Microbiological Evaluation of Packaging-Material Integrity."
Medical Device & Diagnostic Industry 2(5): 29-37.
Schwegman, D. (2008). "Prevention of Cross Transmission of Microorganisms Is Essential
to Preventing Outbreaks of Hospital-Acquired Infections."
Scott, R. D., II (2009). The Direct Medical Costs of Healthcare-Associated Infections in
U.S. Hospitals and the Benefits of Prevention. Division of Healthcare Quality
Promotion, National Center for Preparedness, Detection, and Control of
178

Infectious Diseases, Coordinating Center for Infectious Diseases, Centers for
Disease Control and Prevention.
Severin, J. (2006). The Effect of Pressure Differential on Microbial Penetration of a
Sterile Medical Device Tray. School of Packaging. East Lansing, Michigan State
University: 160.
Severin, J., L. Bix, et al. (2007). "A New Methodology for Whole-Package Microbial
Challenge Testing for Medical Device Trays." Journal of Testing and Evaluation
35(4): 373-380.
Smith, R. L., S. M. Meixler, et al. (1991). "Excess Mortality in Critically Ill Patients with
Nosocomial Bloodstream Infections." Chest 100: 164-167.
Soule, B. M. (2008). "Avoiding the Drama of Healthcare-Associated Infections." Trustee
61(8): 38-40.
Spitz, J. (1994). "The Relevance of Microbial Challenge Testing to Package Sterility."
Medical Device & Diagnostic Industry 15(1): 142-148.
Spitzley, J. (2002). The State of Sterile Package Integrity Testing in the Medical Device
Industry. WorldPak, East Lansing, Michigan.
Standard, P. G., G. F. Mallison, et al. (1973). "Microbial Penetration Through Three Types
of Double Wrappers for Sterile Packs." Applied Microbiology 26(1): 59-62.
Stone, P. W., D. Braccia, et al. (2005). "Systematic review of economic analysis of health
care-associated infections." American Journal of Infection Control 33(9): 501509.
Stone, P. W., E. Larson, et al. (2002). "A systematic audit of economic evidence linking
nosocomial infections and infection control interventions: 1990-2000." American
Journal of Infection Control 20(3): 145-152.
Townsend, T. R. and R. P. Wenzel (1981). "Nosocomial Bloodstream Infections in a
Newborn Intensive Care Unit." American Journal of Epidemiology 114(1): 73-80.
U.S. Department of Health and Human Services, Food and Drug Administration, Center
for Biologics Evaluation and Research, Center for Drug Evaluation and Research,
Center for Devices and Radiological Health, Center for Veterinary Medicine
(2008). Guidance for Industry: Container and Closure System Integrity Testing in
Lieu of Sterility Testing as a Component of the Stability Protocol for Sterile
Products.
179

U.S. Department of Health and Human Services Agency for Healthcare Research and
Quality. (2010). "Annual Quality and Disparities Reports Include Data on Rates of
Health Care-Associated Infections, Obesity and Health Insurance." Retrieved
August 18, 2010, from http://www.ahrq.gov/news/press/pr2010/qrdr09pr.htm.
Ulger, F., S. Esen, et al. (2009). "Are we aware how contaminated our mobile phones
with nosocomial pathogens?" Annals of Clinical Microbiology and Antimicrobials
8(7).
Vuong, C. and M. Otto (2002). "Staphylococcus epidermidis infections." Microbes and
Infection 4: 481-489.
Warren, D. K., S. J. Shukla, et al. (2003). "Outcome and attributable cost of ventilatorassociated pneumonia among intensive care unit patients in a suburban medical
center." Critical Care Medicine 31(5): 1312-1317.
Webster, J., W. Lloyd, et al. (2003). "Rethinking Sterilization Practices: Evidence for
Event-Related Outdating." Infection Control and Hospital Epidemiology 24(8):
622-624.
Weinstein, R. A. (1998). "Nosocomial Infection Update." Emerging Infectious Diseases
4(3): 416-420.
Wenzel, R. P. (2000). "Nosocomial Infection; Hygiene Key to Cutting Infections." Health
& Medicine Week: 19.
Whitehouse, J. D., D. Friedman, et al. (2002). "The Impact of Surgical-Site Infections
Following Orthopedic Surgery at a Community Hospital and a University Hosptial:
Adverse Quality of Life, Excess Length of Stay, and Extra Cost." Infection Control
and Hospital Epidemiology 23(1): 183-189.
Wilcox, M. H. (2003). "Books: Hospital infection past and present." The Lancet
362(9398): 1863.
Williams, L. (2008). "Is your infection control program effective?" Nursing Homes 57(2):
40-42.
Wisplinghoff, H., T. Bischoff, et al. (2004). "Nosocomial Bloodstream Infections in US
Hospitals: Analysis of 24,179 Cases from a Prospective Nationwide Surveillance
Study." Clinical Infectious Diseases 39: 309-317.
World Health Organization (2002). Fact Sheet No 194 Antimicrobial Resistance.

180

Zhan, C. and M. R. Miller (2003). "Excess Length of Stay, Charges, and Mortality
Attibutable to Medical Injuries During Hospitalization." Journal of the American
Medical Association 290(14): 1868-1874.

181