the skin sore trial: exploring a better treatment option

The Skin Sore Trial: Exploring a better treatment option for

impetigo in Indigenous children living in remote Australia

Dr Asha Bowen

BA, DCH, MBBS

Thesis submitted for the degree of

Doctor of Philosophy

December 2014

Menzies School of Health Research

Charles Darwin University

Darwin, NT, Australia

DECLARATION

I hereby declare that the work herein now submitted as a thesis for the degree of

Doctor of Philosophy of the Charles Darwin University is the result of my own

investigations, and all references to the ideas and work of other researchers have

been specifically acknowledged. I hereby certify that the work embodied in this

thesis has not already been accepted in substance for any degree, and is not being

currently submitted in candidature for any other degree.

I give consent to this copy of my thesis, when deposited in the University

Library, being made available for loan and photocopying, and online via the

University’s Open Access repository eSpace.

Dr Asha Bowen

PhD Scholar

Menzies School of Health Research

December 2014

iii

ivxlcdm

ABSTRACT

Impetigo is prevalent among Australian Indigenous children living in the

Northern Territory. The cause of this endemic skin infection is multifactorial

including household overcrowding, poverty, endemic scabies and fungal skin

infections, tropical climate, insect bites and minor trauma. Although impetigo is

widespread, affecting up to 50% of children in a community at one time, care

seeking for treatment of impetigo does not reflect this burden accurately,

possibly due to awareness of the painful injection prescribed. A better treatment

is needed.

This thesis presents the first randomised controlled trial conducted amongst

remote living Indigenous Australian children on the treatment of extensive

impetigo, and one of the largest impetigo trials worldwide. Methodological gaps

were addressed, with studies published on the susceptibility of Streptococcus

pyogenes to trimethoprim-sulphamethoxazole (SXT), the transport of impetigo

swabs in remote settings and photography protocols for capturing and scoring

sores. The principal findings are:

1. Either of two short courses of SXT are non-inferior to standard treatment of

impetigo with intramuscular benzathine penicillin G (BPG) – published in The

Lancet;

2. S. pyogenes remains the key driver of impetigo in our region; and

3. S. pyogenes is susceptible in vitro to SXT, and SXT is effective in vivo for the

treatment of skin infections caused by S. pyogenes.

These findings will inform the management of children with impetigo in remote

Indigenous communities and have already been incorporated into treatment

algorithms and guidelines regionally and nationally. This provides a simple,

short-course antibiotic regimen that is palatable, inexpensive and has a low side-

effect profile, as an alternative to an injection of intramuscular penicillin, which

has been the standard of care for more than 20 years. Ongoing surveillance will

v

be needed to understand how this new regimen is utilised and whether

widespread use for a common condition will alter antibiotic resistance profiles.

vi

TABLE OF CONTENTS

Declaration iii

Abstract v

Table of Contents vii

Acknowledgements xiii

Dedication xix

Publications xxi

Presentations xxiii

List of Figures xxvii

List of Tables xxix

Abbreviations xxxiii

Chapter 1: Introduction 1

1.1 Thesis overview 1

1.2 A brief overview of the Northern Territory, Australia 2

1.3 What is impetigo? 3

1.4 Why is understanding impetigo important? 7

1.5 How common is impetigo? 8

1.6 Natural history of impetigo resolution 8

1.7 How is impetigo treated? 10

vii

1.8 A randomised controlled trial for extensive impetigo in

Indigenous children

13

1.9 Ethics 14

1.10 Summary of Introduction 14

Chapter 2: The global epidemiology of impetigo 15

2.1 Summary 15

2.2 Statement of contribution to jointly authored work 15

2.3 Introduction 16

2.4 Methods 18

2.5 Results 21

2.6 Discussion 32

2.7 Conclusions 36

Chapter 3: Is Streptococcus pyogenes susceptible to

trimethoprim-sulphamethoxazole?

37

3.1 Chapter overview 37


3.3 Journal article: Is Streptococcus pyogenes resistant or

susceptible to trimethoprim-sulphamethoxazole? Journal

of Clinical Microbiology, 50: 4067 – 72.

40

viii

Chapter 4: The challenges of collecting and transporting

impetigo swabs in a tropical environment

47



4.3 Journal article: Comparison of three methods for the

recovery of skin pathogens from impetigo swabs collected

in a remote community of the Northern Territory,

Australia. Transactions of the Royal Society of Tropical

Medicine and Hygiene, 107:384-9.

49

Chapter 5: Developing a methodology for capturing and scoring

standardised digital images of impetigo

57



5.3 Journal article: Standardising and assessing digital

images for use in clinical trials: A practical, reproducible

method that blinds the assessor to treatment allocation,

PLOS One: e110395

58

Chapter 6: The results of the Skin Sore Trial 69



6.3 Journal article: Short course oral cotrimoxazole versus

intramuscular benzathine benzylpenicillin for impetigo in

a highly endemic region: an open label, randomised,

71

ix

controlled, non-inferiority trial, The Lancet, 384: 2132-

2140.

Chapter 7: The microbiology of impetigo in Indigenous

children: associations between Streptococcus pyogenes,

Staphylococcus aureus, scabies and nasal carriage

81



7.3 Submitted journal article: The microbiology of

impetigo in Indigenous children: associations between

Streptococcus pyogenes, Staphylococcus aureus, scabies

and nasal carriage.

82

Chapter 8: Conclusions 95


8.2 Discussion of main findings 97

8.3 Study limitations 103

8.4 Future work 105

8.5 Final conclusions 109

Bibliography 111

Appendices 147

Appendix 1 149

x

Appendix 2 153

Appendix 3 157

xi

ACKNOWLEDGEMENTS

There are many people to acknowledge for their support throughout the four

years of work that has gone into this thesis. Above all, I thank my Lord and

Saviour, Jesus Christ, through whom I am strengthened each day to accomplish

what he has called me to do (Philippians 4:13).

It would not have been possible without the steadfast love and support of my

husband Darren Westphal. He has been a constant source of encouragement, day

in and day out, even when it all seemed impossible. He has been by my side

throughout this PhD and has been a great sounding board for my thoughts as they

developed. His skills as an amateur photographer and involvement in health

research guided me through the early days of developing a method for

photographing digital images. He has kept me focused on the bigger world

outside my thesis, encouraged me to present my work at international

conferences so that he could come along too and kept our family functioning at

peak times of my thesis. Our children Zachary and Eliana, who have both been

born during my candidature, have grounded me when I felt overwhelmed. They

have kept me laughing, kept me busy and are really looking forward to when

Mummy no longer needs to work all the time! My family both in Australia and

the USA have taken care of my children to give me time to write, encouraged me

even when they weren’t too sure what I was talking about and listened along the

way to all of my challenges. My mum, Narelle Bowen, has been willing on

several occasions to come and care for my family, to give me the opportunity to

write, even arriving from the other side of the country with less than 24 hours

notice when I needed her.

My supervisors have taught me the art and science of research, motivated me

when paving a path forward seemed impossible and corrected endless drafts of

papers. To each of them I offer my sincere thanks for taking on a novice

researcher to lead this trial. I would like to thank Jonathan Carapetis for his

unending belief in my ability to achieve goals that seemed insurmountable; an

always open door, telephone line or email inbox to discuss my questions and

address challenges; and his vision for the big picture. My involvement in

xiii

research was inspired by Jonathan and would not have been possible without him

opening this doorway. Likewise, my regular meetings would not have occurred

without his excellent personal assistants Caroline Sheridan, Kristy Coulston and

Linda Lorimer. Adding Steven Tong as a co-supervisor in my second year has

been an enormous blessing. His wisdom has been an asset, his understanding of

the PhD journey and timeline important, and his research savvy inspiring. Steve

has sat with me on many occasions nutting out statistical analyses that seemed

impossible, suggesting creative ways forward with papers and has been available

for endless questions. Ross Andrews oversaw the trial management team and

answered many questions on epidemiology along the way. He was willing from

the outset to consider my input into changes in the study design. I have learned

and grown under their collective supervision and am inspired by their ongoing

commitment to answering questions that make a meaningful difference to the

health of Indigenous Australians.

Alongside my core supervisors sit the two other chief investigators on this trial.

Bart Currie has inspired me throughout. His enthusiasm, knowledge and clinical

acumen and his endless energy for high quality research to improve the quality of

care for Indigenous people makes him an outstanding clinician researcher, who

was never too busy for my questions. Bart has also been integral in translating

our findings into treatment guidelines in real time and has guided me through this

process. Malcolm McDonald left Darwin early in my candidature, but has

remained a supporter, encourager and most importantly, an excellent critic of my

written work. His ability to edit words out of a manuscript without losing any

meaning is a skill I hope I will continue to develop in my future work. These five

investigators envisioned this trial, pursued it undeterred through three NHMRC

funding rounds and welcomed me to the team. I am humbled by their collective

experience, inspired by their enthusiasm and motivated to continue to conduct

translational research with meaningful outcomes for Indigenous children.

My PhD is a single, large randomised controlled trial, conducted in very remote

communities of the Northern Territory. It is one of the largest impetigo trials ever

conducted, and the first in truly remote settings where the burden of disease is

the highest. It would not have been possible without the large trial team: the chief

xiv

investigators (Jonathan, Steve, Ross, Bart and Malcolm), the project managers

(Irene O’Meara and her predecessor Tammy Fernandes) and the research team

(Jane Nelson, Melita McKinnon, Therese Kearns, Valerie Coomber, Christine

Francais, Colleen Mitchell, Dianne Halliday, Dev Tilikiratne, Neil Kerinaiua,

Lynette Johnson, Simone Baker, Doreen Jinggarrabarra, Duncan Wutumbul,

Roslyn Gundjirrjirr and Elvira Dhumabuy). Alongside this group were residents,

and medical and nursing students, who participated for a brief week or two as

research assistants to gain exposure to research in Indigenous communities. Jane,

TK, Melita and Margaret Landrigan were critical in the establishment of the trial

and oversaw the operations when the team was in the field. The laboratory

scientists both at Menzies (Rebecca Towers, Rachael Lilliebridge, Donna

Woltring, Jacklyn Ng, Wajahat Mahmood, Bianca Hayes and Vanessa Theobold)

and at Royal Darwin Hospital (Jann Hennesey, Greg Haran, Pam Smith, Luke

Tennent, Nicole McMahon, Gloria de Castro and Nu Nu Swe) were tireless in

their commitment to processing thousands of swabs for culture and antibiotic

susceptibility testing.

Irene O’Meara has been an outstanding project manager throughout the trial,

keeping track of trial progress, planning trips to communities and endlessly

recruiting staff for the study. Irene’s willingness to answer any question, clean

data, discuss results and make suggestions has been invaluable. I cannot thank

her enough for her drive, dedication and enthusiasm to see this trial through to

completion. Irene’s willingness to go the extra mile and her dedication to

Indigenous health is highly commendable. I know with certainty that the results

reported in this thesis would not have been possible without Irene’s commitment.

Irene is the face of the Skin Sore Trial in the remote communities. Children and

families were keen to know the results when they saw Irene arrive back in their

communities. In addition to numerous recruitment trips, she coordinated all the

feedback trips on which we had the exciting opportunity of sharing our findings

with families, teachers, clinic staff and community elders. Having Irene in charge

saw recruitment completed in the required time frame when even this seemed

impossible. Irene has become a good friend and is a dedicated project manager.

xv

Robyn Liddle, Melita McKinnon, Tegan Harris and Linda Walsh were all

involved in data management for the study. Robyn’s willingness to teach me

Microsoft Access and to answer numerous queries about the data structure was

admirable. Robyn was also invaluable in teasing out various challenges in

statistical analysis. Her willingness to build databases and modify these until

working smoothly for the photography review was fantastic. Without Robyn’s

expertise in this area, it would not have been possible to organise and score the

thousands of digital images collected over the course of the study. The members

of the Area9 team at Menzies throughout my PhD were also incredibly helpful in

this process, making sure that the technology for scoring digital images by

colleagues in Melbourne, Perth, Auckland, Adelaide or Brisbane worked

smoothly. They problem solved and developed solutions to make the process

work faster when this was also a challenge.

Kara Burns was integral in developing simple protocols for amateur

photographers that gave us sets of digital images of the same sore that could be

scored. I thank my paediatric colleagues who willingly looked at the paired

digital images. Tom Snelling, Andrew Steer, Claire MacVicar, Deena Parbhoo,

Rebecca Cresp, Sarah Cherian, Sarah Martin and Ruth Lennox were all willing

and available when I needed their assistance.

Mark Chatfield’s assistance with statistical analysis for the study was invaluable.

His advice to the chief investigators on various important statistical decisions

strengthened the study. He also guided us through the process of engaging a data

safety and monitoring board. I thank Jim Buttery who chaired this group along

with Keith Edwards, Stephen Lambert and Peter O’Rourke for their commitment

to the study.

During my training to become an infectious diseases specialist, I experienced a

brief period of training in a microbiology laboratory. Whilst not being a

microbiologist, my PhD candidature has extended my knowledge and skills in

microbiology. It would not have been possible to do so without the willingness of

microbiology colleagues to share their insights and to answer my questions.

These included Rob Baird, Malcolm McDonald, Peter Ward, Phil Giffard, Steven

Tong and Jann Hennesey. To the scientists at Menzies and Royal Darwin

xvi

Hospital who have conducted experiments, processed thousands of specimens

and provided a robust microbiological data set, I am truly grateful.

Sharing office space with other PhD students has been a privilege and

encouragement along the way. Many a problem felt less challenging after

discussion with peers, at various stages, on the same journey. The broader higher

degree community at Menzies has also been invaluable with their willingness to

stop for a chat in the tea room to commiserate, encourage or congratulate! The

education office of Caroline Sheridan and Catherine Richardson has helped me

keep the funding bodies and university reports timely. The finance department

was also integral in administering my scholarships.

Menzies has been a welcoming and productive environment in which to begin

my research career. Whilst not being my immediate supervisors, I have been

mentored and supported by Anne Chang and Peter Morris. Their experience in

conducting randomised trials, and their passion for doing excellent research that

makes a difference for Indigenous children, is inspiring. My colleagues in the

Infectious Diseases Department of Royal Darwin Hospital should also be

thanked. This group of world-class clinician researchers encourage and support

junior colleagues, and are in turn able to make ground-breaking discoveries. This

would not be possible without the extremely hard work of the research

administration and ethics teams. Christina Spargo and Maria Scarlett deserve

special mention for their assistance throughout my candidature. Thanks must also

go to the Communications team at Menzies who were instrumental in allowing

our research findings to gain media attention. Richmond Hodgson, Lucy Barnard

and Claire Addinsall were dedicated and their efforts led to local and national

media reports on the day our Lancet paper was released. I was keen for our

findings to be highlighted in the Indigenous media. Lucy coordinated a number

of interviews to ensure our research was communicated in plain language to

those who may one day need to be treated. I thank my mum, Narelle Bowen and

Sue Dibbs for assistance with editing my thesis.

xvii

It would not have been possible to enrol in a PhD without financial support. The

funding from the National Health and Medical Research Council, Australian

Academy of Sciences and Menzies School of Health Research enabled me to

train as a researcher and to present our findings at regional, national and

international meetings.

xviii

DEDICATION

I dedicate this thesis to the participants and their families throughout the

Northern Territory who freely consented to join the Skin Sore Trial in the hope

that a better treatment for skin sores would be found. It is my hope that an

injection will no longer be a deterrent to seeking effective treatment for skin

sores, and one day we will report that Indigenous children have the same burden

of impetigo as their non-Indigenous counterparts.

To my children, Zachary and Eliana, I hope that one day you might understand

the importance of my work in striving to make a difference for Indigenous

children, and that you may live in a world where studies like this are no longer

needed.

xix

PUBLICATIONS

Bowen AC, Lilliebridge RA, Tong SY, Baird RW, Ward P, McDonald MI,

Currie BJ, Carapetis JR. Is Streptococcus pyogenes resistant or susceptible to

trimethoprim-sulfamethoxazole? Journal of Clinical Microbiology. 2012; 50

(12): 4067 – 4072.

Bowen AC, Tong SY, Carapetis JR. Reply to “Susceptibility to Streptococcus

pyogenes to trimethoprim-sulfamethoxazole”. Journal of Clinical Microbiology.

2013; 51( 4): 1351.

Bowen AC, Tong SY, Chatfield MD, Andrews RM, Carapetis JR. Comparison

of three methods for the recovery of skin pathogens from impetigo swabs

collected in a remote community of Northern Territory, Australia. Transactions

of the Royal Society of Tropical Medicine and Hygiene. 2013; 107 (6): 384 –

389.

Bowen AC, Tong SY, Andrews RM, et al. Short-course oral co-trimoxazole

versus intramuscular benzathine benzylpenicillin for impetigo in a highly

endemic region: an open-label, randomised, controlled non-inferiority trial. The

Lancet 2014, 384 (9960) 2132 - 2140.

Bowen AC, Burns K, Tong SYC, Andrews RM, Liddle R, O’Meara IM,

Westphal DW, Carapetis JR. Standardising and assessing digital images for use

in clinical trials: a practical, reproducible method that blinds the assessor to

treatment allocation. PLoS One. 2014, Nov 6; 9(11): e110395

Submitted Manuscripts under review

Bowen AC, Tong SYC, Chatfield MD, Carapetis JR. The microbiology of

impetigo in Indigenous children: associations of Streptococcus pyogenes,

Staphylococcus aureus, scabies and nasal carriage. Submitted to BMC

Microbiology, under review.

xxi

Manuscripts in development

Bowen AC, Tong SYC, Mahe A, Hay R, Steer AC, Andrews RM, Carapetis JR.

The global burden of impetigo: A systematic review.

Tasani M, Tong SYC, Holt D, Currie BJ, Carapetis JR, Bowen AC. Does the

presence of scabies impact treatment efficacy? Further analysis of the Skin Sore

Trial results

xxii

PRESENTATIONS

International Conferences

Bowen AC, Tong SYC, McDonald MI, Carapetis JR. In vitro and in vivo

susceptibility of Streptococcus pyogenes to trimethoprim/sulphamethoxazole in

Northern Australia. European Congress of Clinical Microbiology and Infectious

Diseases, Barcelona, SPAIN, 10 – 13 May 2014 (Oral and ePoster presentation)

Bowen AC, Tong SYC, Carapetis JR. Epidemiology of impetigo in Indigenous

children in remote northern Australia: results from a randomised controlled trial.

Annual Meeting of the European Society of Paediatric Infectious Diseases,

Dublin, IRELAND, 6 – 9 May 2014 (Oral and poster presentation)

Bowen AC, Tong SYC, Andrews RM, O’Meara IM, Chatfield MD, McDonald

MI, Currie BJ, Carapetis JR. Short course oral trimethoprim-sulphamethoxazole

is as effective as intramuscular benzathine penicillin G for impetigo: Results of a

non-inferiority, randomised controlled trial. Australasian Society of Infectious

Diseases Conference, Adelaide, AUSTRALIA, 26 – 29 March 2014 (oral

presentation)

Bowen AC. The infectious disease burden of Australia’s Indigenous children

living in the Northern Territory. World Congress of the World Society of

Paediatric Infectious Diseases, Cape Town, SOUTH AFRICA, 19 - 22

November 2013 (invited speaker)


MI, Currie BJ, Carapetis JR. Trimethoprim-sulphamethoxazole is non-inferior to

benzathine penicillin G for the treatment of impetigo: Results of a randomised

controlled trial. World Congress of the World Society of Paediatric Infectious

Diseases, Cape Town, SOUTH AFRICA, 19 - 22 November 2013. (Poster)

Bowen AC, Tong SYC, Carapetis JR. The microbiology of impetigo in

Indigenous children of tropical Australia. World Congress of the World Society

of Paediatric Infectious Diseases. Cape Town, SOUTH AFRICA, 19 - 22

November 2013 (poster)

xxiii

Bowen AC, Lilliebridge RA, Tong SYC, Baird R, Ward P, Currie BJ, McDonald

MI, Andrews RM, Carapetis JR. Are Group A streptococci susceptible or

resistant to trimethoprim-sulphamethoxazole? World Congress of the World

Society of Paediatric Infectious Diseases, Melbourne, AUSTRALIA November

2011. (Poster)

Bowen AC, Tong SYC, Lilliebridge RA, Andrews RM, Carapetis JR. A practical

and effective method of transporting pyoderma and nose swabs for recovery of

Streptococcus pyogenes and Staphylococcus aureus in remote areas. Lancefield

Symposium XII, Palermo, ITALY 4 – 8 September 2011. (Poster)

Bowen AC, Lilliebridge RA, Tong SYC, Baird R, Ward P, Currie BJ, McDonald

MI, Andrews RM, Carapetis JR. Are Group A streptococci susceptible or

resistant to trimethoprim-sulphamethoxazole? Lancefield Symposium XII,

Palermo, ITALY 4 – 8 September 2011. (Poster)

Oral presentations at local meetings


MI, Currie BJ, Carapetis JR. The Skin Sore Trial. Child and Adolescent Health

Research Symposium 2014, Perth, WA, 22 – 23 October 2014.

Bowen AC. The infectious diseases burden of remote Indigenous children. The

Skin Sore Trial. Child and Adolescent Health Research Symposium 2014, Perth,

WA, 22 – 23 October 2014.


MI, Currie BJ, Carapetis JR. The Skin Sore Trial: results of a non-inferiority

randomised controlled trial on treatment of impetigo in remote Indigenous

children. Wesfarmers Infectious Diseases Seminar, Telethon Kids Institute,

Perth, WA, 3 September 2014.


MI, Currie BJ, Carapetis JR. Is cotrimoxazole a GAS drug? Results of the Skin

Sore Trial. Western Australian Infectious Diseases Evening Meeting

(WAIDEM), Perth, WA, 12 August 2014.

xxiv


MI, Currie BJ, Carapetis JR. The Skin Sore Trial. General Paediatric Research

Meeting, Princess Margaret Hospital, Perth, WA, 31 July 2014.


MI, Currie BJ, Carapetis JR. No more pus: a new paradigm for skin sores in the

Northern Territory. Royal Darwin Hospital Grand Rounds, Darwin, NT, 7 April

2014.


MI, Currie BJ, Carapetis JR. Short course oral trimethoprim-sulphamethoxazole

is as effective as intramuscular benzathine penicillin G for impetigo: Results of a

non-inferiority, randomised controlled trial. Menzies Seminar, Darwin, NT, 28

January 2014.



benzathine penicillin G for the treatment of impetigo: Results of a Randomised

Controlled Trial. Northern Territory Royal Australasian College of Physicians,

Alice Springs, NT, 9 November 2013.



benzathine penicillin G for the treatment of impetigo: Results of a Randomised

Controlled Trial. Paediatric Grand Rounds, Royal Darwin Hospital, Darwin, NT,

1 November 2013.

xxv

LIST OF FIGURES

Chapter 1 1

Figure 1: Map of Australia, highlighting the Northern

Territory in red

2

Figure 2: The visual appearance of healing sores with

treatment

10

Chapter 2 15

Figure 1: Flowchart of systematic review according to

PRISMA statement

22

Figure 2: Map of countries with available population-

based studies of impetigo prevalence, highlighted and

colour coded by median prevalence of impetigo

25

Box: Applying prevalence estimates for impetigo to

remote living Australian Indigenous children

26

Chapter 4 47

Figure 1: Flow diagram of the collection of swabs from

participants with impetigo in study arms comparing the dry

or liquid Amies techniques with the STGGB technique

52

Chapter 5 57

Figure 1: The Panasonic DMC-TZ8 camera demonstrating

zoom and camera settings as indicated in figure 5

60

xxvii

Figure 2: Use of a scale to define the upper and lower

boundaries of image in the landscape position

61

Figure 3: An example of the participant identification card

described in table 1

61

Figure 4: Capturing the image using the study camera,

grey background, and grey scale in a remote context

61

Figure 5: The camera settings and icons as used in the

protocol

63

Figure 6: Cartoon demonstrating the photographer, camera

and sore were in the same plane to optimise reproducibility

of captured images

64

Figure 7: Database form used for Quality Control check by

medical photographer

64

Figure 8: This shows the database format utilised for

scoring digital images pairs. Image A was taken on day 0

and image B on day 7

65

Chapter 6 69

Figure 1: Digital images and scoring algorithm 74

Figure 2: Trial profile 75

Figure 3: Comparison of treatment success between co-

trimoxazole and benzathine benzylpenicillin for various

analyses

77

Figure 4: Microbiological clearance by treatment group 77

xxviii

LIST OF TABLES

Chapter 1 1

Table 1: Summary of key demographic differences

between the Indigenous and overall population of the

Northern Territory, Australia

3

Table 2: The CARPA manual and history of

recommendations for treatment of impetigo in the NT

12

Chapter 2 15

Table 1: Number of studies of impetigo prevalence by

decade, country and region

23

Table 2: Summary statistics of available studies by age

grouping

24

Table 3: Estimates of children with impetigo by regions of

the world with available data*

27

Table 4: Classification of studies by region and World

Bank Development Indicator in 2005

28

Table 5: Median prevalence of impetigo in childhood and

overall, categorised by the World Development Index

29

Table 6: Variability in median impetigo prevalence by

urban and rural study locations

31

Table 7: Regional variation in the use of pyoderma or

impetigo to describe bacterial skin infections

32

xxix

Chapter 3 37

Table 1: Antibiotic susceptibility testing methods and agar

used

41

Table 2: SXT Etest susceptibility results for study 1 42

Table 3: Difference in MIC in study 1a 42

Table 4: SXT Etest susceptibility results for study 2 42

Table 5: Difference in MIC in study 2a 42

Table 6: Summary of published results of S. pyogenes in

vitro susceptibility to SXT

43

Chapter 4 47

Table 1: Streptococcus pyogenes or Staphylococcus aureus

recovery from simultaneously collected swabs from

impetigo lesions, comparing the dry or liquid Amies

techniques versus the STGGB technique

52

Table 2: Overall detection rates of Streptococcus pyogenes

or Staphylococcus aureus from simultaneously collected

swabs from impetigo lesions

53

Chapter 5 57

Table 1: Study equipment chosen including the required

features, advantages and alternatives available or

recommended in the literature

62

Table 2: Quick list for capturing standardised photographs 63

xxx

Table 3: Definitions used for the quality control

assessment of digital images

65

Table 4: Definitions used for final outcome scoring of

digital images of impetigo

66

Chapter 6 69

Table 1: Baseline characteristics of all participants in the

trial at the first visit according to study group

76

Table 2: Modified intention-to-treat outcomes by treatment

group (unless stated)

77

Table 3: Adverse events associated with receipt of either

study drug

78

Chapter 7 81

Table 1: Results from logistic regression models to assess

associations between impetigo pathogens and age, sex,

severity, presence of scabies and region

87

Table 2: Identification of Staphylococcus aureus from any

impetigo lesion and the anterior nares for all children with

at least one skin and nose swab available (n=504 children)

90

Appendices 147

Appendix 1: Table Overall and childhood pyoderma and

scabies prevalence from 89 studies synthesised in Chapter

2 systematic review

149

xxxi

ABBREVIATIONS

APP As per protocol

ARF Acute rheumatic fever

AST Antibacterial susceptibility testing

BHS beta-haemolytic streptococci

BPG Benzathine penicillin G

CARPA Central Australian Rural Practitioners Association

CI Confidence Interval

CLSI Clinical and Laboratory Standards Institute

CNA Collistin and nalidixic acid

CRF Case report form

CTX cotrimoxazole (or trimethoprim-sulfamethoxazole)

DSMB Data Safety and Monitoring Board

EUCAST European Committee on Antimicrobial Susceptibility Testing

GAS Group A streptococcus (Streptococcus pyogenes)

HREC Human Research Ethics Committee

IDSA Infectious Diseases Society of America

IQR Interquartile range

ITT Intention to treat

JPEG Joint Photographic Experts Group

MHA Mueller Hinton agar

xxxiii

MHBA Mueller Hinton agar with horse blood

MHF Mueller Hinton agar containing defibrinated horse blood and

20mg/liter beta NAD

MHLHBA Mueller Hinton agar with lysed horse blood

MHS Mueller Hinton agar with sheep blood

MIC Minimum Inhibitory Concentration

mITT modified Intention to treat

mMRSA multidrug resistant MRSA

MRSA Methicillin resistant Staphylococcus aureus

MSSA Methicillin susceptible Staphylococcus aureus

NHMRC National Health and Medical Research Council

nmMRSA non multidrug resistant MRSA

NT Northern Territory

OR Odds Ratio

PhD Doctor of Philosophy

QC Quality Control

RCT Randomised Controlled Trial

SMGGB Skim milk tryptone glucose glycerol broth

SMZ sulfamethoxazole

SOP Standard Operating Procedure

SSTI Skin and soft tissue infections

STGGB Skim milk tryptone glucose glycerol broth

xxxiv

SXT Trimethoprim/sulphamethoxazole

TMP trimethoprim

WGS Whole Genome sequencing

WHO World Health Organization

xxxv

CHAPTER 1INTRODUCTION

1.1 THESIS OVERVIEW

This chapter will introduce the main themes of the thesis. To understand the

context of this thesis, it will be important to know about both the study setting

and impetigo: what it is, how it occurs, what treatments are available and what is

known about the natural history of impetigo resolution. This will pave the way

for Chapter 2, a systematic review of population-based studies on the prevalence

of impetigo to define the global burden. Methodology chapters will follow,

defining methods for transporting impetigo swabs in remote contexts (Chapter

3), testing for antibiotic susceptibility of Streptococcus pyogenes to

trimethoprim/sulphamethoxazole, an antibiotic which has long been thought to

be ineffective against this organism (Chapter 4) and capturing digital images for

measuring the outcome of a randomised controlled trial (Chapter 5). The results

of the major study will then be presented in Chapter 6, followed by the

microbiology of impetigo in Indigenous children of Northern Australia in

Chapter 7. Chapter 8 will conclude the thesis with future directions for research

translation, and highlight the need for ongoing studies to dernormalise skin

disease in remote Indigenous children of Australia.

1

0123456789

1.2 A BRIEF OVERVIEW OF THE NORTHERN TERRITORY, AUSTRALIA

Figure 1: Map of Australia highlighting the Northern Territory in red.

Source: www.virtualoceania.net/australia/nt_map.png (accessed 19/11/14).

All the work reported in this thesis has been undertaken in the Northern Territory

(NT), a federal territory of Australia (Figure 1). This brief orientation is provided

to place the work in context. The NT includes about one-sixth of the Australian

continent, with an area of 1.35 million square kilometres

(www.ga.gov.au/scientific-topics/geographic-information/dimensions/area-of-

australia-states-and-territories, accessed 17/11/14). The NT is divided into two

regions, the ‘Top End’ where a tropical climate prevails and ‘Central Australia’

with a semi-arid climate. The NT is vast, remote and sparsely populated. It has

five major centres, of which Darwin is the largest, and more than 100 remote

communities. At the most recent census (2011), the NT population was 211,945

of whom 56,776 (27%) were Indigenous (Aboriginal and Torres Strait Islanders)

(www.censusdata.abs.gov.au, accessed 15/11/14). Table 1 provides a limited

overview of the social deprivation disproportionately suffered by the NT

Indigenous population and consequent health impacts. Nurses and Aboriginal

Health Workers, under the guidance of medical practitioners, provide the bulk of

primary health care in remote community clinics. The CARPA standard

treatment manual is widely used as a clinic manual for the delivery of primary

health care (CARPA, 2014).

2

Table 1: Summary of key demographic differences between the Indigenous and overall population of the Northern Territory, Australia

Overall NT residents Indigenous NT residents

Median age 31 years 23 years

Proportion aged < 15 years 23.2% 33.2%

Median household income per week $1, 674 $1,099

Median number of people per household 2.9 4.2

Average number of people per bedroom 1.0 1.7

Unemployment rate 4.3% 13.7%

% Resident remote (outside Darwin) 50% 90%

Life expectancy at birth (2009)* Male 75.7y

Female 81.2y

Male 61.5y

Female 69.2y

Sources: 2011 census data available at www.abs.gov.au, accessed 15/11/14,

*www.health.gov.au/internet/publications/publishing.nsf/Content/oatish-hpf-2012-

toc~tier1~life-exp-wellb~119, accessed 15/11/14.

1.3 WHAT IS IMPETIGO?

Impetigo contagiosa, also known commonly as skin sores, school sores or

shortened to impetigo, is a non-bullous infection of the superficial layers of the

epidermis. (Hay et al., 2006) It was first described in 1864. (Fox, 1864)

Pyoderma is a term often used interchangeably with impetigo, but pyoderma

encompasses a broader definition of superficial bacterial skin infection that

includes bullous and non-bullous impetigo, ecthyma, folliculitis, cellulitis and,

sometimes, tropical ulcers. (Mahe, 2005) Whilst impetigo is a superficial

infection of the epidermis, ecthyma is a deeper infection that extends into the

dermis. (Hay et al., 2006) The distinction between these two conditions in terms

of clinical recognition and treatment recommendations at a primary care level is

poorly understood and as such they are often discussed together, both for

microbiology and treatment algorithms. (Stevens et al., 2014, Mahe et al., 2005b,

3

Steer et al., 2009) This thesis will predominantly use the terms impetigo or skin

sores to describe this bacterial skin infection.

1.3.1 Microbiology of impetigo

The gram-positive bacteria Staphylococcus aureus and Streptococcus pyogenes

commonly cause impetigo. These bacteria may be found independently or in a

mixed growth, with the relative proportions varying over time, region and

climate. (Koning et al., 2012) Previous studies identified early colonisation of the

skin with S. pyogenes resulting in the development of sores in the ensuing days

through minor breaches in skin integrity. (Ferrieri et al., 1972, Dajani et al.,

1972, Dajani et al., 1973) These early studies reported that impetiginous lesions

were later colonised with S. aureus (Ferrieri et al., 1972, Dillon, 1970) and as

such both pathogens were implicated at different time points in the development

of impetigo, with S. pyogenes causing the infection and S. aureus complicating it.

In more recent decades, S. aureus has assumed prominence in the global

microbiology of skin and soft tissue infections, raising treatment challenges, as

this organism has developed resistance mechanisms against the commonly used

antibiotics for skin infections. (Moran et al., 2013, Tong et al., 2008) The

microbiology of impetigo in the Northern Territory also reflected these trends,

with increasing prevalence of methicillin-resistant S. aureus (MRSA) as a

co-pathogen with S. pyogenes, seen in swabs taken from impetigo lesions

throughout the Northern Territory (McDonald et al., 2006) and in hospital

studies. (Tong et al., 2009) Based on this changing microbiology, it was

uncertain whether benzathine penicillin G remained the most effective treatment

for impetigo in our context.

1.3.2 Antecedents of impetigo

Intact skin is usually resistant to infection with S. aureus or S. pyogenes

(Tunnessen, 1985). Intact skin may be disrupted by trauma, abrasions, insect

bites, scratching, eczema or another dermatosis (e.g., scabies infestation, tinea,

pediculosis or varicella zoster). Individuals in contact with others who have a

high burden of skin infections may also be colonised, (Ferrieri et al., 1972)

followed by the development of impetiginous lesions at the site of skin

4

disruption. (Currie and Carapetis, 2000, Wannamaker, 1970) Skin pathogens are

highly transmissible (Ferrieri et al., 1972) (skin-to-skin and skin-to-fomite) and

result in endemic infections where household overcrowding is extreme, (Harris et

al., 1992) with epidemic peaks as new bacterial strains are introduced. (Dajani et

al., 1973)

Skin on any part of the body can be involved. However, due to trauma and insect

bites being the major risk factors in tropical settings, (Leng and Chay, 1982,

Kottenhahn and Heck, 1994, Taplin et al., 1973, Allen and Taplin, 1974) lesions

are most common on the lower limbs, followed by the upper limbs (Mahe, 2005)

and scalp when superinfected pediculosis occurs. (Jinadu, 1985) The predilection

for the limbs in tropical contexts may also be due to the wearing of minimal

clothes in the heat and humidity; as such, little protection is offered against

abrasions and minor trauma. (Belcher et al., 1977) In other populations, lesions

on the face, around the nose and mouth are more common. (Koning et al., 2012,

Kiriakis et al., 2012)

Impetigo is endemic in hot and humid climates, (Taplin et al., 1973) particularly

if there is a lack of ready access to water for cleaning. (Cairncross and Cliff,

1987, Bhavsar and Mehta, 1985, Taplin et al., 1973) Household overcrowding is

a suitable environment for transmission of skin pathogens and is more commonly

found in regions where poverty prevails. (Harris et al., 1992, Accorsi et al., 2009)

Impetigo is a disease of poverty, with burden increasing as socio-economic status

decreases. (Masawe et al., 1975) Impetigo also perpetuates where understanding

and recognition of the disease as a health priority is poor. (Mahe et al., 2005a,

Hay et al., 1994) Where access to dermatologists is limited, primary health care

providers may not recognise this (or other) skin diseases and few studies are

available to guide effective treatment decisions in these contexts. The integrated

management of childhood illness (IMCI) has been developed as an approach to

the recognition and treatment of impetigo and is widely used in Fiji, (Steer et al.,

2009) but in other regions where impetigo is common, treatment algorithms are

less well utilised. (Figueroa et al., 1998, Mahe et al., 1995) Prioritisation of

community dermatology is still desperately needed. (Ryan, 2008)

5

Finally, the normalisation of sores and poor skin health has been raised as a

possible explanation for ongoing disease burden. When the majority of children

in a community have sores at any one time, and these are recognised to

eventually resolve over weeks to months, sores may become an accepted norm in

childhood. This normalisation of sores can result in a failure to seek health care,

(Dogra and Kumar, 2003) and a failure of health-care practitioners to

appropriately treat for impetigo when the child presents with other complaints.

On the other hand, skin disease has been found to be the second most common

reason for health care seeking in childhood in developing countries. (Estrada

Castanon et al., 1992, Figueroa et al., 1996, Walker et al., 2008) This

contradictory information makes estimating the true burden of skin disease,

including impetigo, difficult. Focusing on episodes of health care sought

(through audits of clinic attendance) rather than population prevalence may result

in under-reporting of the true burden of impetigo.

1.3.3 Identification of impetigo

Skin sores are identified by the classic characteristics of purulence, erythema and

golden crusting, which progresses to form a thick scab. (Wannamaker, 1970)

Sores are often painful, itchy and unsightly. Children are aware of sores due to

pain and pruritus, but may not complain excessively. (Dogra and Kumar, 2003)

Training in the recognition of skin lesions has been identified as a priority for the

provision of primary health care in tropical contexts, where impetigo and other

skin diseases are common. (Mahe et al., 2005a, Mahe, 2005) Identification of the

correct diagnosis is a priority to ensure the most appropriate treatment is offered.

For the Skin Sore Trial, research nurses were trained in skin disease recognition

using a clinical diagnosis manual developed by our team in earlier work. This

diagnostic and treatment manual includes high quality images of dermatological

conditions and is widely used throughout the region for education and training of

health-care workers.

http://www.menzies.edu.au/icm_docs/162092RecognisingandTreatingSkin

Conditions.pdf, accessed 17/11/14.

6

1.4 WHY IS UNDERSTANDING IMPETIGO IMPORTANT?

While skin sores are rarely fatal, impetigo lesions cause discomfort,

inflammation and missed school days. (Hay et al., 1994) In addition, the bacteria

are highly transmissible, with early studies showing infection of all other family

members with streptococcal pyoderma occurring at a mean of 21 days from the

index case. (Ferrieri et al., 1972) More recently in the NT, secondary S. pyogenes

transmission within the household was detected in 19% of occupants. (McDonald

et al., 2008) Likewise S. aureus, particularly MRSA, colonisation of family

members of index cases with skin infections is high. (Miller et al., 2012, Fritz et

al., 2012) The break in the skin is the entry point for bacteria that may lead to

both severe infectious and post-infectious sequelae. Skin sores have been

reported as a risk factor in the development of S. aureus (Skull et al., 1999) and

S. pyogenes (Carapetis et al., 1999) bacteraemia, resulting in hospitalisation.

(Engelman et al., 2014) Acute post-streptococcal glomerulonephritis outbreaks

occur regularly where skin sores are endemic (Marshall et al., 2011) and

contribute to a high burden of chronic kidney disease. (Hoy et al., 2012, White et

al., 2001) Finally, the burden of acute rheumatic fever in Indigenous Australians

is amongst the highest reported in the world, (Parnaby and Carapetis, 2010,

Seckeler and Hoke, 2011) with consequent morbidity and mortality. (Lawrence

et al., 2013) This has long been hypothesised to arise from skin infection rather

than pharyngitis, as is seen elsewhere in the world. (McDonald et al., 2004)

These serious sequelae of impetigo are not reported in this thesis, as the design of

the randomised controlled trial and sample size needed were not sufficient to also

document these much rarer events. However, the importance of treating impetigo

stems from decreasing disease burden, reducing interpersonal transmission and

hence reducing these more serious consequences.

7

1.5 HOW COMMON IS IMPETIGO?

1.5.1 Global burden

A systematic review of population-based studies reported in Chapter 2 confirms

the global burden remains high. This updates previous literature that incorporated

impetigo into other GAS diseases (Carapetis et al., 2005) or referenced impetigo

as one of many skin conditions worthy of attention. (Mahe, 2005, Vos et al.,

2012, Hay et al., 2014) The included studies will describe the prevalence

predominantly in resource-poor contexts. Due to a paucity of population-based

prevalence studies of impetigo in developed countries, estimates from general

practitioner surveys (Koning et al., 2006, Rortveit et al., 2011, Shallcross et al.,

2013) confirm that impetigo is also a common bacterial skin infection in

childhood in industrialised settings.

1.5.2 Indigenous children of the NT

There are ten studies included in the systematic review from Australia, all

reporting impetigo prevalence in remote Indigenous children from northern

Australia. Seven are studies from the NT, two from Queensland and one from

Western Australia. There were no population-based, prevalence studies available

for non-Indigenous children of Australia.

Of 4026 children assessed in the seven studies from the NT, the median

prevalence of impetigo was 45.7% (inter quartile range, IQR 2.6% – 69.4%).

These studies were published between 1992 and 2009, and reflect an ongoing

high burden of disease. An estimate of prevalence from the Skin Sore Trial based

on episodes eligible for inclusion is similarly high at 870/1715 (50.7%).

Research assistants in our trial were seeking to recruit participants with impetigo,

which may result in over-estimation, but it does confirm the burden remains high

throughout the NT.

1.6 NATURAL HISTORY OF IMPETIGO RESOLUTION

Few studies of the natural history of untreated impetigo, documenting the process

and time to resolution, are available. Understanding the natural history of

8

impetigo resolution is critical in evaluating the end point of treatment studies.

Could the changes seen in the appearance of a sore be due to natural resolution

alone or is there a treatment effect apparent?

Current knowledge of the natural history of impetigo emerged from observations

during outbreaks of impetigo associated with acute post-streptococcal

glomerulonephritis at the Red Lake reservation in Minnesota, USA in the 1960s.

(Dajani et al., 1973, Ferrieri et al., 1972, Dajani et al., 1972) Professor Dajani

was available for correspondence via email during my candidature, to discuss the

pathogenesis and healing of impetigo lesions. Having him available to answer

my questions expanded my understanding of his observations in children and

hamster studies. (Dajani and Wannamaker, 1970) This was critical in developing

definitions to formalise objective assessments of the digital images used in our

trial, as studies of impetigo resolution have not been ongoing.

Impetigo usually begins with a 1–2cm erythematous macule that rapidly becomes

a vesicle or pustule. (Dajani and Wannamaker, 1970) The vesicle ruptures in the

first 24–48 hours, leaving a thin crust overlying the erosion. (Dajani and

Wannamaker, 1970) Erythema becomes more marked in the initial few days due

to underlying neovascularisation, then gradually resolves. Concurrent with

erythema, the thin crust thickens over several days and as healing progresses, the

crust reduces in size. (Dajani et al., 1973) As the crust dries and contracts, there

may be evidence of traction around the edges of the crust and peeling of the

surrounding layers of skin. Peeling skin is due to the presence of bacterial

proteases that break down the damaged extracellular matrix, providing the

opportunity for regeneration. Eventually the lesion becomes flat and dry with

peeling around the edge, followed by a slight hypo- or hyper-pigmentation of the

skin. This usually resolves with time, and scarring is reportedly uncommon with

impetigo (Dajani and Wannamaker, 1970) but more common with the deeper

lesion ecthyma. Untreated, impetigo lesions can remain unresolved for 30 days.

(Dillon, 1970)

Hamster models showed resolution of impetigo by 9–11 days from the

introduction of bacteria. (Dajani et al., 1971) This is faster than reported in

humans. (Dillon, 1970) The hamster model, while useful to understand

9

pathogenesis, may be a less reliable predictor of the time to resolution in human

impetigo, due to the absence of ongoing trauma in an animal laboratory,

compared to the real-world experience of children. Beyond this, the natural

history of impetigo without treatment remains poorly understood, as few studies

have focused on this aspect; treatment is usually recommended and when treated

outside of intervention studies, follow-up to determine outcome is rarely

included in the clinical algorithm.

Figure 2: The visual appearance of healing sores with treatment.

Note: This series of photographs obtained from a participant in the Skin Sore Trial on

days 0 (A), 2 (B), and 7 (C) highlight some of the features of sore healing described

above. The erythematous macule, pustule and development of thin crusts can be seen in

image A. Thickening of the crusts, traction at the edges of the sore and peeling skin are

evident in image B. Hypopigmentation can be seen in image C.

1.7 HOW IS IMPETIGO TREATED?

1.7.1 What is the evidence base for treatment of impetigo?

In developed country settings, impetigo is a common cause for childhood

presentations to the general practitioner. (Shallcross et al., 2013) Lesions are

often around the mouth and nose, of small diameter (0.5–1cm) and discrete.

(Sladden and Johnston, 2004, Hartman-Adams et al., 2014) The Cochrane review

on treatment of impetigo (Koning et al., 2012) provides strong evidence for the

topical treatment of limited impetigo. These findings have been incorporated into

guidelines, which are accessed throughout the world. (Stevens et al., 2014)

However, of the 68 included trials, only one includes children with severe

C B A

10

impetigo and that was conducted in a tropical, resource-poor context. (Faye et al.,

2007) The remainder studied limited impetigo and were conducted in hospital

outpatient departments or general practices, predominantly in developed

countries. (Koning et al., 2012) Therefore, while topical therapy with fusidic acid

or mupirocin is recommended for limited impetigo based on the synthesis of high

quality randomised controlled trials, (Koning et al., 2012, Shim et al., 2014) a

major evidence gap remains for the treatment of extensive impetigo.

In regions where impetigo is endemic, lesions are widespread and several body

areas are affected at the same time, topical therapies are impractical and

expensive. Neither guidelines from the Infectious Diseases Society of America,

(Stevens et al., 2014) nor Medecins Sans Frontiers in conjunction with the World

Health Organization, (Broek et al., 2013) recommend topical therapy for

extensive or endemic impetigo. Both rely on expert opinion to support treatment

with oral antibiotics that are active against S. pyogenes and methicillin-

susceptible S. aureus for extensive disease. In addition, the widespread use of

topical therapies leads to the rapid emergence of antimicrobial resistance. (Udo et

al., 1994) There remains an evidence gap for the most effective treatment for

impetigo where the burden is highest (tropical, resource-poor contexts).

1.7.2 How is impetigo treated in Indigenous children in the Northern Territory?

The treatment of impetigo in the NT has historically included a single dose of

intramuscular benzathine penicillin G (Table 2). This regimen was recommended

in the second edition of the CARPA manual in 1994, (CARPA, 1994) at the time

intramuscular benzathine penicillin G (LA-Bicillin®) was first registered for use

in Australia. (Currie, 2006) The inclusion of benzathine penicillin G in the NT

treatment algorithms was based on non-randomised studies conducted in the

USA (Esterly and Markowitz, 1970, Dillon, 1970) in children with extensive

impetigo. Cure rates of between 58% and 100% were achieved compared with

20% to 39% for placebo.

11

Table 2: The CARPA manual and history of recommendations for treatment of impetigo in the NT

Edition, date

When to treat?

Treatment Guidelines Notes

1st, 1992

Not stated Povidine iodine topically to clean sores Treat scabies when impetigo resolved Bicillin AP intramuscular

If no improvement in 48–72 hours, flucloxacillin (with probenicid) or erythromycin for 5–10 days

2nd, 1994

Not stated Povidine iodine topically to clean sores Treat scabies when impetigo resolved

Bicillin AP or Bicillin LA intramuscular

3rd, 1997

>6 infected lesions or the lesions appear severe

Povidine iodine topically to clean sores Give advice about scabies treatment

Bicillin LA intramuscular

If no improvement in 48–72 hours, flucloxacillin (with probenicid)

Roxithromycin if penicillin allergic

4th, 2003

>6 infected sores or the sores look severe

Povidine iodine topically to clean sores Treat impetigo and scabies together

Bicillin LA intramuscular Do not use topical mupirocin as resistance develops rapidlyRoxithromycin for 10 days if penicillin

allergic

5th, 2009

“Clearly infected sores”

Clean sores with soap and water, sponge off crusts

Check for scabies and treat together

Benzathine penicillin intramuscular or amoxicillin for 10 days if injection not possible

Do not use topical mupirocin as resistance develops rapidly

Trimethoprim-‐sulphamethoxazole for 5 days if penicillin allergic

Follow up. If not improving, flucloxacillin for 10 days or trimethoprim-‐sulphamethoxazole for 5 days

Due to the pain of the injection, intermittent BPG stock-outs (Currie, 2006) and

uncertainty about the overall number of sores needing treatment, the threshold

12

for treatment in the NT has varied. In the early CARPA editions, no guidance

was provided on when to treat; whereas in the middle editions, more than six

purulent or crusted sores were the treatment threshold, unless the “sores look

severe”. (Table 2) During the East Arnhem Healthy Skin Program (EAHSP,

2004–2006) when the guideline included the six sores threshold for prescribing

treatment, less than 10% of children qualifying for treatment were actually

receiving it. (Andrews et al., 2009) In 2009, the recommendations in CARPA for

treatment changed to “clearly infected sores”, due to concerns that counting the

number of sores was becoming the basis for treatment. (Andrews et al., 2009)

1.8 A RANDOMISED CONTROLLED TRIAL FOR EXTENSIVE IMPETIGO IN INDIGENOUS CHILDREN

Despite some of the highest reported impetigo rates reported in the world in

Indigenous children, (Carapetis et al., 2005) a formal evaluation of the efficacy

of BPG for impetigo had never been conducted in the NT. In addition, the rising

burden of MRSA in skin and soft tissue infections throughout the NT and

worldwide led us to hypothesise that an antibiotic active against MRSA was

needed. The suspicion that the painful needle was a deterrent to receiving

treatment was also important in considering a RCT.

BPG was well established as the standard of care for treatment of impetigo in the

NT (Table 2). A placebo has rarely been used in impetigo studies (Koning et al.,

2012) and it was not considered ethical in the context of high rates of

streptococcal and staphylococcal sequelae seen in children in the NT.

Clindamycin and trimethoprim-sulphamethoxazole were the available, affordable

antibiotics with activity against MRSA. The bitter, unpalatable taste of

clindamycin in syrup formulation, (Steele et al., 2006) resistance rates in

S. aureus at 25% (McDonald et al., 2006) and three times daily dosing, (Group,

2010) meant clindamycin was rejected. Conversely, trimethoprim-

sulphamethoxazole is palatable, has excellent skin penetration, (Krolicki et al.,

2004) resistance rates below 1% (McDonald et al., 2006), had demonstrated

efficacy in a pilot study, (Tong et al., 2010) and is one of the most widely used

antibiotics in the world.

13

Comparing a new treatment to a known standard or active control in a

randomised controlled trial, requires a non-inferiority design seeking to

determine that the new treatment is not worse than the current treatment, by an

acceptable amount. (Piaggio et al., 2012) A robust, measurable outcome is also

required.

1.9 ETHICS

This study was approved by the Human Research Ethics Committee of the

Northern Territory Department of Health and Menzies School of Health

Research (HREC09/08). Ethics approval was current throughout the period of the

study and concluded on 31 December 2013.

1.10 SUMMARY OF INTRODUCTION

Impetigo is common in childhood, with children living in resource-poor contexts

experiencing the highest burden. Despite this, few treatment studies to guide

evidence-based treatment algorithms are available for extensive impetigo.

Evidence for safe, palatable, simple, effective treatment regimens is needed. The

aim of this thesis is to find a better treatment option for impetigo amongst

Indigenous children living in remote communities of Australia, and add to the

limited available evidence for the treatment of extensive impetigo worldwide.

14

CHAPTER 2THE GLOBAL EPIDEMIOLOGY OF

IMPETIGO

2.1 SUMMARY

We conducted a comprehensive, systematic review of the population prevalence

of impetigo and the broader condition pyoderma. PubMed was systematically

searched for impetigo or pyoderma studies published between 1970 and 2014; 66

manuscripts relating to 89 studies met our inclusion criteria. From population-

based surveillance, 82 studies included data on children, giving a total of 145,028

children assessed for pyoderma or impetigo. Median childhood prevalence was

12.3% (IQR 4.2–19.4%). Fifty-eight (65%) of the studies were from low or low-

middle income countries, where median childhood prevalences were 8.4% (IQR

4.2–16.1%) and 14.5% (IQR 8.3–20.9%) respectively. However, the highest

burden was seen in underprivileged children of high-income countries; median

prevalence 19.4%, (IQR 3.9–43.3%). Based on data from studies published since

2000 from low and low-middle income countries, we estimate the global

population of children suffering from impetigo at any one time to be in excess of

162 million, predominantly in tropical, resource-poor contexts. Impetigo is an

under-recognised, neglected tropical disease and in conjunction with scabies,

comprises a major dermatological concern in childhood, with potential lifelong

consequences of untreated infection.

2.2 STATEMENT OF CONTRIBUTION TO JOINTLY AUTHORED WORK

This chapter is in preparation for submission to a peer-reviewed journal and is

authored by myself, Antoine Mahe, Roderick Hay, Ross Andrews, Andrew Steer,

Steven Tong and Jonathan Carapetis. I designed this study based on previous

work by Antoine Mahe, Roderick Hay, Andrew Steer and Jonathan Carapetis, to

15

systematically describe the global burden of impetigo. I wrote the protocol,

conducted the systematic literature review and assessed all papers for inclusion

in the study. Two reviewers, including myself and one of the other co-authors,

who all did a fraction of the reviews, conducted all of the full text reviews. I

synthesised the results, performed the data analysis and have written this chapter.

2.3 INTRODUCTION

Impetigo is a common dermatosis of childhood. Recent estimates of the global

burden of impetigo are in the order of 111 million children from developing

countries (Carapetis et al., 2005) to 140 million (Vos et al., 2012, Hay et al.,

2014) people affected at any one time. However, these estimates were based on a

limited literature review of impetigo in the context of larger studies, have not

been recently updated and acknowledge that impetigo estimates are imprecise

due to the paucity of published literature from the highest prevalence contexts.

This is because impetigo is a disease of poverty, occurring in settings where

systematic collection of accurate disease burden data is a lower priority and most

available data arise from hospital records, which may under-represent the true

population prevalence of skin disease due to selection bias (Hogewoning et al.,

2013, Saw et al., 2001, Lawrence et al., 1979, Bechelli et al., 1981, Gibbs, 1996,

Bissek et al., 2012) and minimal health-care seeking for skin diseases in

resource-poor contexts. (Hay et al., 1994, Behl, 1979, Mahe, 2005) This study

will systematically collate, and hence update understanding of, the population

prevalence of impetigo in children globally from 1970 to 2014. More precise

estimates for impetigo by age group, global region, climate, levels of

urbanisation and development are needed to target intervention studies and

algorithmic management at the primary health care level.

The bacteria Staphylococcus aureus and Streptococcus pyogenes typically cause

superficial bacterial infections of the skin. (Brook et al., 1997) These superficial

skin infections are variously labelled. The term pyoderma is often used to

describe all superficial bacterial skin infections in tropical contexts and is

inclusive of impetigo, ecthyma and furunculosis. (Mahe et al., 2005b) Impetigo

(also known as skin or school sores) is a subset of pyoderma and has both

16

bullous and non-bullous forms. In addition, impetigo is often divided into

primary and secondary impetigo, depending on whether an underlying

dermatological (e.g. eczema) or alternative skin infection (e.g. scabies,

pediculosis or tinea) preceded the bacterial component of the condition. (Mahe,

2005) Primary impetigo is often preceded by minor trauma or insect bites.

(Mahe, 2005) In an attempt to understand the global burden of superficial

bacterial skin infection, this study includes reports on both pyoderma and

impetigo as the major forms of bacterial skin infection under investigation.

Impetigo is more than a nuisance condition. While mortality estimates for

impetigo are low, the frequency of person-to-person transmission in resource-

poor communities maintains a high burden of disease and affects well-being.

(Murgia et al., 2010) In addition, the infectious (staphylococcal and streptococcal

cellulitis, bacteraemia and deep tissue infections) and post-infectious

(glomerulonephritis, rheumatic fever) sequelae are well known and cause high

morbidity (Carapetis et al., 2005, Hoy et al., 2012, Marshall et al., 2011, Skull et

al., 1999) and variable mortality. (Jackson et al., 2011, Tibazarwa et al., 2008)

The relative cost to families of missed school days and procuring treatments that

may or may not work is also high. (Hay et al., 1994, Yamamah et al., 2012)

Skilled diagnosticians are few where the burden is highest. Community health

workers often have poor training in skin disease recognition. (Mahe et al., 2005a)

Few treatment studies have been conducted in high-burden, endemic settings to

guide the development of evidence-based algorithms. (Bowen et al., 2014, van

der Wouden and Koning, 2014)

The aim of this systematic review is to evaluate the prevalence of impetigo from

studies in the general population and to explore variations in the epidemiology of

impetigo. The number of Indigenous Australian children with impetigo in remote

communities will also be estimated.

17

2.4 METHODS

2.4.1 Search Strategy

The systematic review is reported according to PRISMA guidelines. (Moher et

al., 2009) References were identified through searches of PubMed for papers

published in English between January 1970 and September 2014, which reported

population-based studies of skin disorders, with specific reference to impetigo or

pyoderma prevalence. We hand-searched the bibliographies of retrieved papers,

for additional references. Relevant articles published between 1970 and 2005

were identified through searches by AM and RH (Mahe, 2005) and are

synthesised in this manuscript. We used the following search terms ["impetigo"

OR "pyoderma"] AND ["Africa" OR "Asia" OR “Latin America” OR “Pacific”

OR “Oceania” OR “North America” OR “Europe” OR “Russia” OR “China” OR

“India” OR “Developing Country” OR “tropical” OR “Indigenous”]. Duplicates

were removed from the search before titles were reviewed for relevance

(epidemiology, prevalence, impetigo and pyoderma). If insufficient detail was

available in the title, abstracts or entire articles were reviewed to determine

whether the study met pre-determined inclusion criteria.

2.4.2 Selection Criteria

Studies were included if they were population-based, prevalence studies, with

extractable data on children with pyoderma or impetigo. A physical examination

by a clinician was essential for inclusion. Wherever a term was used that inferred

a bacterial skin infection (pyoderma, impetigo or sores) and numerator and

denominator or proportion-affected data were available, these have been

reported. Outpatient dermatology clinic and hospital-based studies from

developing countries are rich sources of data on treatment-seeking behaviours for

skin diseases. However, due to the management of impetigo occurring

predominantly at primary care settings rather than tertiary referral dermatology

clinics, these datasets were excluded. Our aim was to present the available

baseline data derived from prevalence studies performed in the general

population (community or school surveys).

18

2.4.3 Reviewer Assessment

Papers meeting the inclusion criteria were sourced in full-text and data extracted

by two reviewers independently. All papers were assessed by myself. Each of the

other co-authors independently assessed a subset of the studies to confirm

inclusion criteria and extract data. Data fields extracted included date of study,

country, climate, rural/urban environment, study site (e.g. school, household),

study design, sampling method, population, age range, gender of participants,

qualifications of person conducting the screening, case definition, definition of

bacterial skin infection, number of participants, number with impetigo, childhood

and adult prevalence, location of lesions, microbiology and presence of scabies.

2.4.4 Definitions

Where the study date was not reported, the year of publication was used. The

sampling method was defined as exhaustive if ≥85% of available population

were surveyed and non-exhaustive if it was below 85%. Other options for

sampling method included convenience (non-random selection of the available

population), targeted (orphanages or institutions) and random selection of

participants. Details of the definition used for recording a bacterial skin infection

were collected and later categorised as pyoderma (if the text used this term or

indicated that impetigo, folliculitis, ecthyma, furunculosis, cellulitis or tropical

ulcers were included), impetigo (only impetigo or skin sores were studied) and

secondarily infected scabies (the primary focus of the study was scabies with a

secondary focus on bacterial skin infection). We have concentrated on reporting

impetigo prevalence in children. Adult data were also incorporated where

available. If children and adults were included in the study, but separate rates

were not provided, then the community wide prevalence of impetigo was

reported. It was not possible to ascertain from the studies how representative the

study sample was of the broader population.

Countries were categorised into regions according to the United Nations (UN)

Population Division (www.esa.un.org/wpp/excel-Data/country-

Classification.pdf, last accessed 10 November 2014). Childhood was defined as

children aged 0 to 15 years, to calculate population regional and global

19

prevalence. To assess the total population at risk of impetigo based on the

median prevalence estimate, population data from the UN Department of

Economic and Social Affairs Population Division for 2012 were used. These

were available online at www.esa.un.org/unpd/wpp/unpp/panel_indicators.htm,

last accessed 7 December 2014. Each country was also categorised according to

the World Development Index as at July 2005 (www.data.worldbank.org) as high

(>$US10,725 gross national income (GNI) per capita), upper middle ($US3,466–

$10,725 GNI per capita), lower middle ($US876–$3465 GNI per capita) or low

(≤$US 875 GNI per capita) income.

The Koppen Climate Classification System is the most widely used classification

of climates (Peel et al., 2007) and recognises five major climate systems. These

include tropical, arid, temperate, cold and polar. Where data on climate was not

included in the paper, this classification has been used to code the climate by

country and region. (Peel et al., 2007)

2.4.5 Statistical Analysis

The data are synthesized into a narrative summary. Statistical analysis was

performed using Stata13 (Statacorp, Texas, USA). Where prevalence estimates

have been combined to understand regional and global burden of impetigo, the

median prevalence has been used. In order to estimate regional and global burden

of impetigo, the median prevalence has been applied to the 2011 Australian

Census and 2012 United Nations population estimate for less developed

countries. To take into account the variability in study size, the metaregression

command in Stata was used to calculate the pooled prevalence. Impetigo and

pyoderma are both used commonly to report bacterial skin infections. To assess

for any variation in reporting of prevalence based on the chosen term, we

assessed the use of each term, and calculated a statistical difference between the

median prevalence using a chi squared statistic.

20

2.5 RESULTS

Figure 1 summarises the results of the search conducted. Of the 1007 titles

identified, 952 were from database searching and 55 from additional sources.

Two hundred and thirteen duplicate records were removed and 628 papers

excluded, mainly due to an absence of data on impetigo or because the studies

were conducted in dermatology clinics or hospital settings. The 128 remaining

records were subjected to full-text review; 38 papers were excluded with reasons

given in Figure 1 leaving 90 papers, each of which was assessed by two

reviewers. A further 24 were excluded due to insufficient data reported on

impetigo prevalence. The final dataset includes 66 papers reporting on 89 studies

(Figure 1, Table 1, Appendix 1) conducted over a 45-year period.

The studies were predominantly from Africa (30/89, 34%), Asia (20/89, 22%)

and Oceania (19/89, 21%) and represented populations from 31 countries (Table

1, Figure 2). Studies with data available by decade were: 1970s (28, 32%), 1980s

(15, 17%), 1990s (20, 23%), 2000s (23, 26%) and since 2010 (3, 3%).

Data on impetigo prevalence were available for 174,508 individuals of whom

145,028 were children. The study size varied, ranging from 31 to 19,775

participants per study. The median study size was 636 participants (inter-quartile

range [IQR] 305–1817), median prevalence 11.2% (IQR 4.2–19.4%) and pooled

prevalence 15.5% (95% CI 12.1–19.0%) (Table 3). Extractable prevalence data

were available on children in 82 (92%) of the studies. The median impetigo

prevalence in children was 12.3%, (IQR 4.2–19.3%) and pooled prevalence

16.6% (95% CI 12.7–20.5%). The median number of children in each study was

534 (IQR 258–1,729).

21

Figure 1: Flowchart of systematic review according to PRISMA statement

952 Records identified through database searching

55 additional records identified through other sources

794 records after duplicates removed

166 records screened 38 records excluded

128 full-‐text articles assessed for eligibility

38 full-‐text articles excluded Not in English (n=6) Not a community

prevalence study (n=14) Review paper (n=1)

Unable to access full text (n=6)

Microbiology study (n=2) No impetigo prevalence

data (n=9)

90 papers assessed by 2 reviewers for inclusion

66 papers (89 studies) included in quantitative

synthesis

24 papers excluded by 2 reviewers

Data quality insufficient for reporting prevalence data

(n=24)

Scre

enin

g Id

entif

icat

ion

Elig

ibili

ty

Incl

uded

22

Table 1: Number of studies of impetigo prevalence by decade, country and region.

Decade Number of

studies

available

Countries Regions

1970 -‐ 1979 28 Colombia, Ghana, Tanzania,

New Zealand, Brazil, India,

USA, Gambia, Panama

Latin America &

Caribbean, Africa, Asia,

Oceania, North America

1980 -‐ 1989 15 Pakistan, Solomon Islands,

Nigeria, Ethiopia, Vanuatu,

India, Fiji, Canada

Asia, Oceania, Africa,

North America

1990 -‐ 1999 20 Australia, Honduras, Mali,

Malaysia, Ethiopia, Tanzania,

Ecuador, Samoa, Taiwan,

Kenya, Solomon Islands

Oceania, Latin America

& Caribbean, Africa,

Asia

2000 -‐ 2009 23 Nepal, Australia, India, Fiji,

Tanzania, Nigeria, Timor Leste,

Turkey, Mali, Ghana, Gabon,

Rwanda, Egypt

Asia, Oceania, Africa

2010 -‐ 2014 3* Ethiopia, Cameroon, Tanzania Africa

* Two studies were published in 2010 and did not provide a year of data collection in themanuscript.

23

Table 2: Summary statistics of available studies by age grouping Total available

population

(n=89 studies)

Childhood

population

(n=82 studies)

Adult population

(n=11 studies)

Median (IQR)

prevalence of

impetigo

11.2%

(4.2–19.4%)

12.3%

(4.2–19.4%)

4.9%

(3.1–9.6%)

Pooled prevalence

(95% CI)

15.5%

(12.1–19.0%)

16.6%

(12.7–20.5%)

9.7%

(2.2– 7.2%)

Median (IQR)

number of

participants per

study

636

(305–1,817)

534

(258–1,729)

638

(264–1,645)

Population with

impetigo

23,759 19,811 2,427

Total population

studied

(denominator)

174,508 145,028 18,246

24

Figure 2: Map of countries with available population-based studies of impetigo prevalence, highlighted and colour coded by median prevalence of impetigo.

Coding of median prevalence 0–5% blue, 5.1–10% green, 10.1–15% yellow, 15.1–20% orange and >20% red. Source: Map from powerpointworldmaps.com (accessed 14.10.2014).

2.5.1 Impetigo prevalence in children by region

Reported impetigo prevalence in populations under investigation ranged from

0.2% to 90%. The highest median prevalence of childhood impetigo was

reported from Oceania, where from 19 studies, the median prevalence was 40.2%

(IQR 17.2–48.1%). Excluding studies from Australia (see box) and New

Zealand, of the eight remaining studies from Oceania, the median prevalence

remained high at 29.7% (IQR 14.7–42.0%). The median impetigo prevalence in

Africa was 7% (IQR 4.1–12.3%), Asia 7.3% (IQR 3–16.1%), resource poor

populations in North America 13.3% (IQR 2.1–19.4%) and Latin America and

the Caribbean 15.5% (IQR 12.2–20.8%). There was no data available for Europe

or China.

25

Box: Applying prevalence estimates for impetigo to remote living Australian

Indigenous children

There were 10 population prevalence studies available for Australia. All

represented data from children living in remote Indigenous communities of

northern Australia, with no studies available for non-Indigenous children. The

median prevalence reported from these studies was 44.5% (IQR 34.0–49.2%).

Four studies were conducted since 2000, with a median prevalence of 43.0%

(IQR 40.2–45.7%). We estimated the total number of remote Indigenous children

with impetigo at any one time by applying the median prevalence of 44.5% to the

remote living Indigenous population from the states of Western Australia,

Queensland and Northern Territory aged less than 15 years in the 2011

Australian Census (35,272). We estimate 15,696 Indigenous children are

suffering from impetigo at any one time. This is the first time that prevalence

estimates have been used to generate a total number at risk amongst Australian

Indigenous children. This will be important for local health care planning.

2.5.2 Burden of impetigo in low and low-middle income countries

The United Nations Department of Economic and Social Affairs population

division estimated the global population below 15 years of age resident in less

developed countries in the years 2000–2009 to be at least 1.6 billion children.

Utilising the median population prevalence of 9.9%, (IQR 4.1–16.6%) from the

studies conducted in developing economies (WDI2005 index of lower-middle or

low) since 2000, the estimated population at risk of impetigo at any one time is

more than 162 million children. Excluding China, where no studies were

available, reduces the estimate to 137 million children with impetigo in low and

26

low-middle income countries. Table 3 outlines the regional estimates of children

with impetigo at any one time using the available data.

Table 3: Estimates of children with impetigo by regions of the world with available data*

Region Population in 2012 under 15 years

Median impetigo prevalence in

children

Estimated number of children with

impetigo

Africa 424,072,000 7%

(IQR 4.1–12.3%)

29,685,040

Asia 1,060,076,000 7.3%

(IQR 3–16.1%)

77,385,548

Oceania* 3,653,000 29.7%

(IQR 14.7–42.0%)

1,084,941

Latin American & Caribbean

167,654,000 15.5%

(IQR 12.2–20.8%)

25,986,370

*Studies from Australia, New Zealand and North America excluded as all these studies were

conducted in small, impoverished populations within these countries that may not reflect the

overall burden of impetigo for the childhood population.

2.5.3 Childhood population prevalence by age group

Only 13 (15%) studies reported prevalence of impetigo according to age group.

Of those that did, the median prevalence in 0–4 year olds studied was 19% (IQR

15–31%), 5–9 year olds 19% (IQR 12–43%) and 10–14 year olds 10%

(IQR 7–28%).

2.5.4 Childhood impetigo and World Development Index

The majority of studies, 58/89 (65%) (Table 4), were from low or low-middle

income countries. The remainder were from middle income or resource-poor

populations within high-income countries. Table 5 summarises the median

prevalence estimates according to income level of the country, with the highest

27

estimates coming from underprivileged populations within high-income

countries.

Table 4: Classification of studies by region and World Bank Development Indicator in 2005

Region High income Upper Middle income

Low Middle Income

Low Income

Oceania Australia (10) New Zealand (1)

Fiji* (4) Vanuatu (1) Samoa (1)

Solomon Islands* (2)

Africa Gabon (1) Egypt* (2) Ghana* (3) Rwanda (1)

Cameroon* (1) Ethiopia (3) Nigeria* (2) Tanzania (9) Kenya (3) Mali (3)

The Gambia (2) Asia Taiwan (1) Malaysia* (2)

Turkey* (1) India* (12) Nepal (2)

Pakistan* (1) Timor-‐Leste* (1)

Caribbean & Latin

America

Panama* (2) Honduras* (1) Brazil (2)*

Colombia* (1) Ecuador*(1)

North America

Canada (6) USA (7)

TOTAL 25 6 13 45

*Category has shifted rather than remaining stable within period from 1987–2013. Source:

data.worldbank.org/data-catalog/world-development-indicators, accessed 12.11.2014.

28

Table 5: Median prevalence of impetigo in childhood and overall, categorised by the World Development Index. WDI and number of

studies

Median childhood

prevalence (IQR)

N= 82

Median overall prevalence

(IQR)

N=89

High income

(N= 25)

19.4%

(IQR 3.9–43.3%)

19.4%

(IQR 3.9–43.3%)

Middle income

(n=4 [childhood] OR

n=6 [overall])

9.9%

(1.8–18.6%)

9.9%

(IQR 2–15.1%)

Low-‐Middle

(N=12)

14.5%

(IQR 8.3–20.9%)

12.8%

(7.8–18.3%)

Low

(n= 41 [childhood] OR

n= 46 [overall])

8.4%

(IQR 4.2–16.1%)

7.9%

(4.3–16.1%)

2.5.5 Scabies in association with impetigo

Fifty-seven (64%) studies also reported data on the prevalence of scabies. The

median prevalence of scabies from all included studies was 3.3% (IQR 0.7–

12.9%). Twenty-seven (87%) countries had available data on scabies prevalence.

Scabies prevalence varied by region, with the highest median prevalence found

in Oceania (n=14) at 16% (IQR 4.9–25%). The median scabies prevalence in

Africa (n=28) was 2% (IQR 0.7–7%) and in Asia (n=11) was 3.4% (IQR 0.9–

11.9%). The median prevalence of scabies in studies from Latin America and the

Caribbean (n=3) was 3% (IQR 0.6–10%). Scabies prevalence also varied by

decade. In the 1970s (n=12), the median was 3.2% (IQR 1.4–13%), 1980s (n=8)

1.1% (IQR 0.4–1.8%), 1990s (n=14) 9.2% (IQR 4.9–17%), 2000s (n=20) 1.9%

(IQR 0.6–15.1%) and since 2010 (n=3) 1.4% (IQR 0.3–1.8%).

29

2.5.6 Microbiology of impetigo

Thirty-four (38%) of the studies also reported culture-based microbiology results,

predominantly on streptococcal infection (n=31). Only 11/89 (12%) studies

reported on the relative contributions of S. pyogenes and S. aureus from

microbiological culture of skin lesions. S. pyogenes was identified in a median of

74% (IQR 57–95%) of cultures and S. aureus in a median of 64% (IQR 53–80%)

of cultures.

2.5.7 Body distribution of impetigo

Data were reported on body distribution of impetigo in 23 studies. Of these,

21/23 (91%) reported the lower limbs as being the most common site. In 11

studies, the proportion of body regions affected was given. These were re-

classified as lower limbs, upper limbs and other [scalp, face, neck, torso]. The

respective medians for body region distributions (these were not mutually

exclusive) were lower limbs 58% (IQR 44–86%), upper limbs 18% (IQR 14–

54%) and other 38% (IQR 5–43%).

2.5.8 Impetigo in rural versus urban settings

Studies were classified broadly as rural (n=61), urban (n=15) or both (n=13).

Table 6 shows the variation in prevalence when the data were examined by this

factor, with a higher prevalence of impetigo reported from rural locations

compared to urban settings.

30

Table 6: Variability in median impetigo prevalence by urban and rural study locations

Median impetigo prevalence overall

Median impetigo prevalence in children

Rural N=61 studies N=55 studies

13.3%

(6.7–20.9%)

16.1%

(5.9–22.6%)

Urban N=15 studies N=14 studies

4.8%

(2–10%)

4.5%

(2–7.3%)

Both N=13 studies N=13 studies

5.8%

(2.4–13.3%)

5.8%

(2.4–13.3%)

2.5.9 Childhood impetigo by climatic region

Many studies included details on the climatic conditions of the study region. If

not available, these were coded using the Koppen classification (Peel et al.,

2007). The majority of studies were from tropical environments, 67/89 (75%).

The remaining studies were from cold 9/89 (10%), temperate 9/89 (10%) and

arid 4/89 (5%) climates. Median impetigo prevalence in childhood was 12.2%

(IQR 4.8–20.8%), 17.5% (IQR 8.2–42.8%), 15.8% (IQR 2.2–30.2%) and 3.9%

(IQR 1–13.3%) for tropical, arid, temperate and cold climates respectively.

2.5.10 Variability in sampling technique

There was variability in the sampling techniques employed. The most common

technique described was exhaustive sampling. Of the exhaustive population

surveillance studies (n=57/89, 64%), the median number of participants was 636

(IQR 305–2528) and median prevalence 10.8% (IQR 4.3–19.4%). The median

prevalence was 4% (IQR 2–16.6%) in the 7 studies where random sampling was

employed.

31

2.5.11 Impetigo or pyoderma: is the reporting consistent?

Twenty-six (29%) studies reported on impetigo or skin sores as the primary

definition under investigation. Of these, 25 have data available on the prevalence

of impetigo in children with the median prevalence of 13.3% (IQR 2.3–22.6%).

Pyoderma was reported in 62 (70%) studies with a median prevalence of 11.8%

(IQR 5.1–19%). There was no statistical difference in the median prevalence

reported based on these definitions, p=0.85. Both definitions were used

throughout all decades of the study. There was some variability in the use of

definition by region: studies from Africa, Asia and Latin America predominantly

reported on pyoderma, whereas studies from Oceania used either definition and

North American studies were more likely to report on impetigo (Table 7).

Table 7: Regional variation in the use of pyoderma or impetigo to describe bacterial skin infections

Region Pyoderma reported(%) Impetigo reported(%)

Africa (n=30) 27 (90%) 3 (10%)

Asia (n=18) 15 (79%) 4 (21%)

Oceania (n=19) 10 (53%) 9 (47%)

North America (n=13) 3/13 (23%) 10/13 (77%)

Latin America & Caribbean (n=7)

7/7 (100%) 0

2.6 DISCUSSION

This systematic review provides comprehensive data and confirms an ongoing,

high burden of impetigo in childhood, estimating that more than 162 million

children in low and low-middle income countries are affected at any one time.

Our study revises upwards the estimate of children impacted by impetigo at any

one time, using the same methodology but incorporating more studies, compared

to the previous estimate of 111 million children. (Carapetis et al., 2005) The

32

reported burden has remained high throughout the study period, with a median

prevalence of 12.3% (IQR 4.2–19.3%). Our estimate derived from 89 studies

over 45 years is higher than previously published estimates, which were between

5 and 10%. (Mahe, 2005) Impetigo is more than a benign, nuisance conditions

and these numbers demonstrate the priority concern impetigo is for public health.

Each decade since 1970 has seen 15–20 new studies published in the peer-

reviewed literature. Our search strategy represents a greater depth of coverage

and enriches our understanding of the global burden of impetigo. The reported

data is derived from prevalence studies performed in the general population, at

the community or school level, and provides a more complete understanding of

the global burden of impetigo than previous estimates. Each study is valuable in

describing the burden of disease for a local or regional population, but

collectively they tell a far more compelling story of an under-appreciated disease.

These studies are predominantly reported from contexts of poverty or tropical

regions.

The data include large regions of the globe over a 45-year interval, suggesting

the burden of impetigo remains relatively unchanged over this period. Only three

studies were available for the current decade, of which two were published in

2010 (Komba and Mgonda, 2010, Murgia et al., 2010) and likely represent data

collected during the previous decade. The remaining study reports data from

March 2010. (Bissek et al., 2012) As such, the most precise, recent estimates are

from the decade ending in 2009 and have shown no change in the global burden

of impetigo. In view of this, we combined the prevalence estimates from all

studies published since 2000 in order to estimate the current burden of impetigo.

The data describe a context where progress against control of impetigo appears to

be lacking. Including bacterial skin infections in the list of neglected tropical

diseases may be one option to refocus attention on the need for this to be a

priority target. In addition, the burden experienced by impoverished populations

within wealthy countries is high. This is in keeping with version 2.0 of the

neglected tropical diseases agenda that utilises ‘blue marble health’ as a phrase to

describe the importance of poverty in all countries as a key underlying factor of

neglected tropical diseases. (Hotez, 2013) The widespread burden of impetigo

33

has remained unchanged over more than four decades. By highlighting the global

burden, which has previously been estimated at >2% of the global population at

any one time, (Hay et al., 2014) an agenda for screening, treatment and further

work can crystallise. This will inform primary prevention of kidney and heart

disease and gram-positive bacterial sepsis in resource-poor contexts.

A limitation of this systematic combination of datasets is the variability in

clinical skills of those performing cutaneous disease surveillance. Studies

reported using expert dermatologists, dermatologists in training, paediatricians,

general medical officers, nursing staff and community health workers trained in

identification of skin conditions, to screen populations. This variability may have

resulted in under-reporting of impetigo. In addition, many of the studies were

focused on a specific dermatosis e.g., scabies or pediculosis, or conducted in the

context of nutrition or child health surveys. Reporting of bacterial infection was a

secondary endpoint. Surveys of skin disease were conducted for a variety of

reasons, including to demonstrate the substantial burden of disease and

significant unmet need. (Walker et al., 2008) Guidelines for skin surveillance

studies have recently been developed. (Kotloff, 2008) Only one study to date

reported using these in the study design, (Steer et al., 2009) but the availability of

these guidelines is likely to inform future surveillance.

The definitions and clustering of conditions used in the studies was variable, but

using the two dominant terms of impetigo and pyoderma, there was no statistical

difference in the median prevalence. This suggests that our inclusion of both

terms as a descriptor of bacterial skin infections is robust.

Our study broadens our knowledge of the global distribution of impetigo, by

including studies from countries that were not available in previous disease

burden estimates. (Carapetis et al., 2005, Vos et al., 2012) An improved

understanding of the global burden may also help to prioritise treatment studies

in contexts with the highest burden of impetigo. The Cochrane review on the

optimal treatment of impetigo (Koning et al., 2012) includes studies

predominantly from high-income countries and references only one study (out of

68) conducted in a similar setting to those reported in our study. Similarly,

Karimkhani et al. report bacterial skin infections are under-represented in

34

Cochrane systematic reviews when matched with corresponding disability-

adjusted life years. (Karimkhani et al., 2014)

Initially, we hoped to define the burden of impetigo in developing countries by

including studies from low and low-middle income countries using the World

Bank Development Index. However, very few studies were excluded on this

basis, and the study was broadened to include all countries. It is possible that

studies have been conducted in the regions of highest disease burden leading to

an over-estimate of the global burden. (Mahe, 2005) We think this is unlikely,

given the size, consistency and duration of the estimated burden of impetigo that

we have described. However, population-based prevalence studies from Europe,

East and South-East Asia, and more recently North America, are under-

represented. No studies were available in English from the low or low-middle

income countries of Eastern Europe, and this is a significant gap in our global

understanding of the burden of impetigo. Recently, the burden of post-

streptococcal sequelae in these countries has been reported and is amongst the

highest in the world, (Tibazarwa et al., 2008) suggesting that impetigo may also

be common here. Despite these gaps, our findings are consistent with the 2010

global burden of disease estimates for impetigo. (Vos et al., 2012, Hay et al.,

2014)

This study is limited by the exclusion of unpublished data or grey literature

searching. This limitation is more likely to restrict the analysis to developing

countries; none of the three studies from developed countries which have been

included in the global burden of skin diseases (Hay et al., 2014) were found

using the search methods specified, as these studies were all health-care based.

Overall, few studies were identified from high-income countries. It is possible

that the implementation of a skin disease population prevalence study is because

of a pre-specified expectation of high rates of disease. The absence of data may

represent lower disease burden, or may just be an unstudied population. In

addition, hospital or health care seeking studies were not included.

Our study confirms that the greatest burden of impetigo is in children, with

steady decreases in prevalence with increasing age. (Mahe, 2005, Belcher et al.,

1977) Despite our study reflecting predominantly impoverished settings,

35

increases in impetigo have also been reported in children in developed countries

attending general practices for care (Shallcross et al., 2013) and it causes a large

volume of health care consultations in all regions of the world. (Mohammedamin

et al., 2006, Razmjou et al., 2009, Koning et al., 2006) Our summary of data has

consistently shown that impetigo predominantly affects the lower limbs in

children in resource poor contexts and that both S. pyogenes and S. aureus are

present in swabbed lesions.

Within the 89 studies, there was variability in the prevalence of impetigo

reported. Tropical climate, insect bites, scabies, poverty, limited access to water

and household overcrowding are all reported factors contributing to the

development of impetigo (Currie and Carapetis, 2000) but data on these was not

systematically available in the 89 studies. We looked at the burden of impetigo

by age group, level of urbanisation and broad climatic region. The individual

studies reported prevalence across a range of additional contexts including

ethnicity, gender, disability and socio-economic status but these were outside the

scope of our review and it is unlikely they would be amenable to systematic

analysis given the small numbers of studies addressing sub-groups.

2.7 CONCLUSIONS

Prevalence throughout the study period was highest in Oceania in both resource-

poor countries, and underprivileged populations within high-income countries.

This synthesis of studies is important for regional and national targeting of

healthy skin interventions, as the burden of disease is large. At a global level, this

study revises our estimate upwards of the number of children affected with

impetigo at any one time from 111 million (Carapetis et al., 2005) to 162 million;

this finding alone should drive a comprehensive research agenda for the

detection, treatment and prevention of impetigo in resource-poor contexts. It is

also important to appreciate that as antibiotics are the backbone of current

treatment for impetigo, this disease burden may contribute to burgeoning

antibiotic resistance in the absence of evidence-based treatment algorithms.

(Tong et al., 2014) This combination of high prevalence and moderate morbidity

makes impetigo a high population health priority

36

CHAPTER 3IS STREPTOCOCCUS PYOGENES

SUSCEPTIBLE TO TRIMETHOPRIM-SULPHAMETHOXAZOLE?

3.1 CHAPTER OVERVIEW

This chapter challenges the conventional wisdom that Streptococcus pyogenes is

not susceptible to trimethoprim/sulphamethoxazole (SXT). Historically, in vitro

antibiotic susceptibility testing was confounded by the presence of thymidine, an

inhibitor of SXT. As such, S. pyogenes appeared resistant to SXT in many earlier

assays. The long standing susceptibility of S. pyogenes to penicillin obviated the

need for an alternative antibiotic for treatment of S. pyogenes infections, and

exploration of the in vivo efficacy of SXT was deemed unnecessary or not

considered.

S. pyogenes is a key pathogen in impetigo and is often found in conjunction with

Staphylococcus aureus. The rising prevalence of methicillin-resistant S. aureus

(MRSA) in remote communities of northern Australia (Tong et al., 2008) was

concerning. Both hospital (Tong et al., 2009) and community (McDonald et al.,

2006) studies identified MRSA rates above 20%. As such, treatment with

antibiotics effective against both pathogens may be required and the Skin Sore

Trial was designed to address this question. When the Skin Sore Trial was

designed, SXT susceptibility in community S. aureus strains was 100%.

(McDonald et al., 2006) Susceptibility of S. pyogenes was similarly high in

unpublished laboratory work.

The work in this chapter demonstrates the in vitro susceptibility of S. pyogenes to

SXT. It was a necessary step in paving the way for our understanding of the

results of the Skin Sore Trial. In addition, it alerts the global community dealing

with skin and soft tissue infections where both S. pyogenes and MRSA are

37

implicated, that a simple, oral antibiotic regimen may be effective depending on

the local antibiogram.


I designed this study based on preliminary work done by Peter Ward and

Malcolm McDonald during the process of achieving funding for this trial. The

preliminary work was included as Study 1 in this paper; I then extended that

work in Study 2. I wrote the standard operating procedures (SOP) for the conduct

of Study 2 and incorporated the recently released European Committee on

Antibiotic Susceptibility testing (EUCAST) breakpoints for S. pyogenes and

SXT. Whilst I planned to assist with the laboratory work, delay in the arrival of

supplies meant that Rachael Lilliebridge conducted the susceptibility testing,

while I was on maternity leave. I oversaw the microbiology, cleaned the data and

analysed the results with the assistance of Steven Tong. I conducted the literature

review and wrote all drafts of the manuscript. All co-authors contributed to

revisions of the manuscript and approved the final version for publication.

3.3 JOURNAL ARTICLE (FOLLOWING PAGE)

Is Streptococcus pyogenes resistant or susceptible to trimethoprim-sulfamethoxazole?

APPENDIX 2

Gelfand MS, Cleveland KO, Ketterer DC. Susceptibility of Streptococcus

pyogenes to trimethoprim-sulfamethoxazole. Journal of Clinical Microbiology.

2013; 51(4): 1350.

38

Bowen AC, Tong SY, Carapetis JR. Reply to “Susceptibility of Streptococcus

pyogenes to trimethoprim-sulfamethoxazole.” Journal of Clinical Microbiology.

2013; 51(4): 1351.

39

Is Streptococcus pyogenes Resistant or Susceptible to Trimethoprim-Sulfamethoxazole?

Asha C. Bowen,a,b Rachael A. Lilliebridge,a Steven Y. C. Tong,a,b Robert W. Baird,b Peter Ward,d Malcolm I. McDonald,c

Bart J. Currie,a,b and Jonathan R. Carapetisa,e

Menzies School of Health Research, Charles Darwin University, Darwin, Northern Territory, Australiaa; Royal Darwin Hospital, Darwin, Northern Territory, Australiab; JamesCook University, Cairns, Queensland, Australiac; Microbiology Department, Austin Pathology, Austin Health, Victoria, Australiad; and Telethon Institute for Child HealthResearch, Centre for Child Health Research, University of Western Australia, Perth, Western Australiae

Streptococcus pyogenes is commonly believed to be resistant to trimethoprim-sulfamethoxazole (SXT), resulting in reservationsabout using SXT for skin and soft tissue infections (SSTI) where S. pyogenes is involved. S. pyogenes’ in vitro susceptibility toSXT depends on the medium’s thymidine content. Thymidine allows S. pyogenes to bypass the sulfur-mediated inhibition offolate metabolism and, historically, has resulted in apparently reduced susceptibility of S. pyogenes to sulfur antibacterials. Thelow thymidine concentration in Mueller-Hinton agar (MHA) is now regulated. We explored S. pyogenes susceptibility to SXT onvarious media. Using two sets of 100 clinical S. pyogenes isolates, we tested for susceptibility using SXT Etests on MHA contain-ing defibrinated horse blood and 20 mg/liter �-NAD (MHF), MHA with sheep blood (MHS), MHA alone, MHA with horse blood(MHBA), and MHA with lysed horse blood (MHLHBA). European Committee on Antibacterial Susceptibility Testing (EUCAST)breakpoints defined susceptibility (MIC, <1 mg/liter) and resistance (MIC, >2 mg/liter). In study 1, 99% of S. pyogenes isolateswere susceptible to SXT on MHA, MHBA, and MHLHBA, with geometric mean MICs of 0.04, 0.04, and 0.05 mg/liter, respec-tively. In study 2, all 100 S. pyogenes isolates were susceptible to SXT on MHF, MHS, MHA, and MHLHBA with geometric meanMICs of 0.07, 0.16, 0.07, and 0.09 mg/liter, respectively. This study confirms the in vitro susceptibility of S. pyogenes to SXT, pro-viding support for the use of SXT for SSTIs. A clinical trial using SXT for impetigo is ongoing.

Streptococcus pyogenes was one of the first bacterial infections tobe treated with sulfur antibacterials in the 1930s (16) and

proved to be clinically effective in the treatment and prophylaxisof S. pyogenes infections (10, 16, 23, 29). However, when sulfadi-azine, an early short-acting sulfur antibacterial, was used in massprophylaxis programs to prevent S. pyogenes tonsillitis and acuterheumatic fever (ARF) in military recruits in the 1940s, the clinicalefficacy of this antibacterial was limited due to the presumed de-velopment of resistance (13, 15, 27, 31, 42) among some strains.Initial antibacterial susceptibility testing (AST) of S. pyogenes tosulfur antibacterials using a broth dilution method demonstratedthat some strains were resistant (25, 53); however, AST was in itsinfancy and no standardized reference methods existed at thattime. This early experience resulted in the belief that trim-ethoprim-sulfamethoxazole (SXT) is ineffective against S. pyo-genes, and its use has been discouraged in clinical practice fordecades (35).

Subsequent antibacterial susceptibility experiments showedapparently reduced susceptibility of S. pyogenes (and other bacte-ria) (6, 40) to the sulfur antibacterials due to antagonism of theinhibition of folate metabolism. Harper and Cawston discoveredan inhibitory substance in 1945, eventually identified as thymi-dine, which was interfering with the ability of sulfur antibacterialsto kill the organism (25). Because the activity of SXT is determinedby the antibacterial’s ability to deprive an organism of folate co-enzymes (7), there is a direct relationship between the thymidinelevels in culture media and SXT resistance (11). High thymidinecontent in agar provides an exogenous substrate which can beused by an organism to maintain folate metabolism and henceappear resistant to SXT. In early studies, most culture media con-tained sufficient thymidine to antagonize the inhibitory effects of

sulfur drugs and hence produced resistant results when this classof antibacterials was tested (6).

Notably, lysed horse blood was found to contain the enzymethymidine phosphorylase, which neutralized thymidine (46) andovercame this effect (21, 25). No other mammalian blood con-tains thymidine phosphorylase (21). However, the addition oflysed horse blood was not recommended for AST, despite severalauthors (5, 6, 21, 54) advising supplementation with lysed horseblood for any medium used to test sulfur antibacterial susceptibil-ity if the thymidine concentration was above 0.03 �g/ml (5) (be-low which inhibition does not occur). In this context, the notionthat S. pyogenes was resistant to sulfur antibacterials perpetuated.

SXT was introduced in 1968 (3, 28, 43) and has since becomeone of the most widely used antibacterials in the world. However,recommendations against the use of sulfur antibacterials, includ-ing SXT, for S. pyogenes infections continue in the belief that theorganism is intrinsically resistant (33, 47, 51). Two studies (33, 51)have reported S. pyogenes uniformly resistant to SXT, but this wasprior to thymidine content standardization in Mueller-Hintonagar (MHA) and was on agar containing sheep blood. There havealso been reported clinical failures in the use of this agent in erad-icating S. pyogenes from nasopharyngeal carriage (30). However,other centers have demonstrated full in vitro susceptibility of S.

Received 19 August 2012 Returned for modification 10 September 2012Accepted 4 October 2012

Published ahead of print 10 October 2012

Address correspondence to Asha C. Bowen, [email protected].

Copyright © 2012, American Society for Microbiology. All Rights Reserved.

doi:10.1128/JCM.02195-12

December 2012 Volume 50 Number 12 Journal of Clinical Microbiology p. 4067–4072 jcm.asm.org 4067

on Novem

ber 17, 2012 by guesthttp://jcm

.asm.org/

Dow

nloaded from

40

http://dx.doi.org/10.1128/JCM.02195-12

http://jcm.asm.org

http://jcm.asm.org/

pyogenes to SXT (14, 22, 36, 54). Since 2006, when the thymidinecontent of MHA became strictly regulated by the Clinical andLaboratory Standards Institute (CLSI) to maintain a low level ofthymidine and hence avoid inhibition (M6-A2 protocol) (9), ithas no longer been necessary to add lysed horse blood to the me-dium for AST. However, current methods do use agar supple-mented with mammalian blood.

Given the prevailing view that caution should be exercised inusing SXT for infections involving S. pyogenes and the paucity ofclinical data of SXT efficacy against S. pyogenes, we sought to con-firm or disprove the notion that S. pyogenes is resistant to SXT invitro on various antibacterial susceptibility testing media. Coin-fection of S. pyogenes with Staphylococcus aureus in skin and softtissue infections (SSTI) and the rising prevalence of methicillin-resistant S. aureus (MRSA) provide added stimulus to explore theutility of SXT in the treatment of these infections.

MATERIALS AND METHODSSwab collection and identification of S. pyogenes. Skin, throat, or noseswabs were collected using a rayon tipped cotton swab (Copan, InterpathServices, Melbourne, Australia). In study 1, swabs were plated on horseblood agar (HBA) (Oxoid, Basingstoke, United Kingdom) and HBA con-taining colistin and nalidixic acid (HBA � CNA) (Oxoid) within 48 h ofcollection. In study 2, swabs were stored in skim milk-tryptone-glucose-glycerol broth (STGGB) at �70°C prior to plating on the above-describedmedia. Incubation was at 37°C for 16 h in 5% CO2. �-Hemolytic colonieswere identified morphologically, and confirmation of S. pyogenes was withthe Lancefield streptococcal grouping test for group A (Oxoid). Isolateswere stored in glycerol at �70°C until subsequent replating for AST.

Selection of isolates. (i) Study 1. We began by exploring the issue witha preliminary study of 100 skin and throat isolates of S. pyogenes collectedfrom 3 remote Australian Aboriginal communities between 2003 and2005 during surveillance studies (39). We used an SXT Etest strip (bio-Mérieux, France) on 3 different agars: MHA, MHA supplemented withhorse blood (MHBA), and MHA supplemented with lysed horse blood(MHLHBA) (Oxoid). Interpretation was based on European Committeeon Antibacterial Susceptibility Testing (EUCAST) breakpoints (20) re-leased in 2010.

(ii) Study 2. The Skin Sore Trial is a randomized, controlled trial(RCT) comparing benzathine penicillin G (BPG) treatment of impetigo(the standard of care) with oral SXT. The first 100 S. pyogenes isolates fromskin and nasal swabs from participants in the Skin Sore Trial were used instudy 2. The participants were 3 months to 13 years old and were from 4remote Aboriginal communities of the Top End of the Northern Terri-tory, Australia, recruited in 2010. Swabs were collected from the anteriornares and at least 2 purulent or crusted sores from all children. Swabs werecollected from skin sores on day 0 (pretreatment), day 2 (midtreatment),and day 7 (completion of treatment). Consent for participation and col-lection of specimens was obtained from the guardian or parent of eachparticipant. The study was approved by the Northern Territory Top EndHuman Research Ethics Committee (HREC 09/08) and has been regis-tered with the Australian and New Zealand Clinical Trials Registry(ACTRN 12609000858291).

Antibacterial susceptibility testing methods and agar used. The twointernationally accredited, standardized methods for AST (Table 1) arethe CLSI (8) and EUCAST (19) methods. CLSI does not provide referencebreakpoints for S. pyogenes susceptibility to SXT and recommends the useof Mueller-Hinton agar (MHA) supplemented with 5% sheep blood fortesting the susceptibility of S. pyogenes to other antibacterials. In contrast,EUCAST has released breakpoints for the disk diffusion method and MICon appropriate media. The medium recommended for Streptococcusgroups A, B, C, and G is MHA supplemented with 5% defibrinated horseblood � 20 mg/liter �-NAD. This agar is commonly referred to as MHF(http://www.eucast.org, accessed 25 July 2012).

Two experimental agars were also used for susceptibility testing. MHAis routinely used in a number of other antibacterial susceptibility tests notinvolving �-hemolytic streptococci. MHLHBA was used to explore thehypothesis based on historical literature that the lysing of horse bloodreleases thymidine phosphorylase which breaks thymidine down to thy-mine.

Using the EUCAST breakpoints (20), with an MIC of �1 mg/liter assensitive and an MIC of �2 mg/liter as resistant, and an SXT Etest toperform antibacterial susceptibility, these studies compared the variousmedia that have been recommended for AST. The Etest interpretation isbased on the trimethoprim component of the trimethoprim-sulfame-thoxazole combination, in a ratio of 1:19. The EUCAST methodologyusing the Etest was chosen due to the availability of published breakpoints;however, due to the widespread use of CLSI, the medium upon which thisorganism is tested was also used and results obtained were referenced tothe EUCAST breakpoints.

Antibacterial susceptibility testing. Single colonies were isolatedfrom frozen stocks following overnight incubation on HBA in 5% CO2 at37°C. Susceptibility testing for SXT was performed using a 0.5 McFarlandsuspension to create a confluent lawn inoculum and then applying an SXTEtest as per the manufacturer’s instructions. Plates were read by 2 readersfollowing incubation in 5% CO2 at 35°C for 16 to 20 h. The MIC wasrecorded where the inhibition ellipse intersected the scale. Where a differ-ence in results of more than 2 gradations was noted between the 2 readers,a repeat test was performed with a fresh subculture of S. pyogenes. As SXTis a bacteriostatic antibacterial, this mode of action can alter the appear-ance of an MIC endpoint, resulting in hazy zones. Where haze was pres-ent, both the 80% and 100% points of ellipse intersection were recorded.A penicillin, erythromycin, and clindamycin disk diffusion test accordingto CLSI guidelines (8) was conducted concurrently on all 100 strains instudy 2.

Quality control. In study 2, for every 20 S. pyogenes clinical isolates, acontrol strain of S. pyogenes (ATCC 19615) was also tested on the 4 agarsand found to be susceptible. Quality control of the SXT Etests was per-formed throughout the study using Escherichia coli (ATCC 25922) onMHA and Streptococcus pneumoniae (ATCC 49619) on MHS. The refer-ence ranges for each organism were achieved, namely, 0.064 to 0.25 �g/mlfor E. coli and 0.125 to 1 �g/ml for S. pneumoniae.

STATA version 12.0 (STATAcorp, College Station, TX) was used todetermine the geometric means. The data were logarithmically trans-formed to a normal distribution, and paired t tests were used to determinethe difference between agars for studies 1 and 2.

TABLE 1 Antibiotic susceptibility testing methods and agar used

Method Study Agar used Abbreviation

CLSI (8) 2 Mueller-Hinton agarsupplementedwith 5% sheepblood

MHS

EUCAST (19) 2 Mueller-Hinton agarsupplementedwith 5%defibrinated horseblood � �-NAD

MHF

Experimental 1 and 2 Mueller-Hinton agar MHAExperimental 1 and 2 Mueller-Hinton agar

containing 5%lysed horse blood

MHLHBA

Experimental 1 Mueller-Hinton agarcontaining horseblood

MHBA

Bowen et al.

4068 jcm.asm.org Journal of Clinical Microbiology

on Novem


.asm.org/

Dow

nloaded from

41

http://www.eucast.org

http://jcm.asm.org

http://jcm.asm.org/

RESULTSStudy 1. On MHBA and MHLHBA, S. pyogenes isolates from theskin (n � 36) and throat (n � 64) were uniformly susceptible toSXT (Table 2). Ninety-nine isolates tested on MHA were suscep-tible to SXT (Table 2). The single isolate which appeared resistanton MHA was susceptible on both MHBA and MHLHBA.

There was no statistically significant difference between thegeometric mean MIC measurements on MHA and MHBA. How-ever, isolates tested on both MHA and MHBA had lower MICsthan isolates tested on MHLHBA (Table 3).

Study 2. One hundred isolates of S. pyogenes from 43 childrenwere included in this analysis. S. pyogenes isolates utilized werefrom sores (n � 98) and the anterior nares (n � 2). The majorityof isolates (76%) were from day 0 (before antibacterial treatment);19% were from day 2, and 5% were from day 7. Sixty-four of theswabs from which S. pyogenes was identified also cultured S. au-reus. Of these, 14% were methicillin-resistant S. aureus (MRSA)and 86% were methicillin-susceptible S. aureus (MSSA).

All 100 isolates of S. pyogenes were susceptible to SXT on allagars by both readers (Table 4). Interrater reliability was excellent,with 96% of all MIC readings within �1 MIC gradation. In view ofthis, all analyses were done on results from reader 1. All 100 S.pyogenes isolates were also susceptible to penicillin, erythromycin,and clindamycin.

The geometric means were similar for MHA, MHBA, andMHLHBA (Table 4). MHS had a higher geometric mean MICthan the other media. This was statistically significant, with iso-lates tested on MHS having higher geometric mean MICs than thesame isolates tested on all other agars (Table 5). Despite the higherMICs, all isolates tested on MHS were still susceptible to SXT.There was no difference in MIC between isolates tested on MHAand those tested on MHF. As in study 1, isolates tested on MHAhad lower MICs than the same isolates tested on MHLHBA. Thiswas also found for isolates tested on MHF (Table 5).

Ongoing surveillance with in vitro susceptibility testing isneeded to monitor for changes in rates of SXT resistance withincreased use of SXT. To date, we have tested 910 S. pyogenesisolates cultured from impetigo and anterior nares of children

randomized in the Skin Sore Trial on MHF using an SXT Etestaccording to EUCAST guidelines. Only 8 (0.9%) have been foundto be resistant, with MICs of �2 mg/liter (unpublished data).These results are consistent with those reported from the EUCASTgroup.

DISCUSSION

Although SXT is no longer commonly recommended for treat-ment of respiratory tract infections, it remains one of the mostwidely used and cheapest antibacterials in the world and is animportant option for treatment of SSTI, where S. pyogenes and S.aureus are often copathogens (4, 12, 24, 34). In the era of risingMRSA prevalence, antibacterials that are active against both bac-teria are highly valued.

Impetigo is a significant therapeutic problem in remote com-munities in the Northern Territory of Australia (37, 49, 50), withcommunity-associated MRSA having become highly prevalent inthis region (50). Impetigo is also an endemic problem in manyless-developed countries (41, 45), and MRSA is likely to be on therise in these contexts also (50). In patients with MRSA and S.pyogenes coinfection, finding a single oral agent that is effective,affordable, and easy to use would be a significant advance. Peni-cillins and cephalosporins are no longer an option for MRSAtreatment. In the Northern Territory context, clindamycin is notan option, with up to 22% of MRSA isolates resistant (49), asidefrom its poor palatability in young children and the difficulties inmaintaining adherence to a thrice-daily regimen. Tetracyclinesare not recommended in children under 8 years of age (17), andlinezolid is currently too expensive for the empirical treatment ofsuch a common childhood condition. SXT, which is cheap,widely available, and well tolerated and requires only twice-daily dosing, is a potential single agent for treatment of bothMRSA and S. pyogenes infections. Several studies have con-

TABLE 2 SXT Etest susceptibility results for study 1

Medium

% of isolatesGeometricmean MIC(mg/liter) SD (log)

Susceptible (MIC �

1 mg/liter)Resistant (MIC �2 mg/liter)

MHBA 100 0 0.04 1.52MHA 99 1 0.04 2.2MHLHBA 100 0 0.05 1.78

TABLE 3 Difference in MIC in study 1a

Medium

Difference in MICb

MHA MHLHBA

MHBA MHBA lower by 7%(�9% to 20%)

MHBA lower by 25%*(13% to 35%)

MHA MHA lower by 20%*(3% to 33%)

a Pairwise comparisons of geometric mean MICs of various media.b 95% confidence level in parentheses. *, P � 0.05.

TABLE 4 SXT Etest susceptibility results for study 2

Medium

% of susceptibleisolates (MIC �

1 mg/liter)

Geometricmean MIC(mg/liter) SD (log)

MHF 100 0.07 2.05MHS 100 0.16 1.79MHA 100 0.07 2.03MHLHBA 100 0.09 2.3

TABLE 5 Difference in MIC in study 2a

Medium

Difference in MICb

MHF MHA MHLHBA

MHS MHS higher by128%*(107% to150%)

MHS higher by132%*(113% to153%)

MHS higher by92%* (70%to 116%)

MHF MHF higher by2% (�12%to 7%)

MHF lower by16%* (3% to27%)

MHA MHA lower by17%* (6% to27%)

a Pairwise comparisons of geometric mean MICs of various media.b 95% confidence level in parentheses. *, P � 0.05.

S. pyogenes and Trimethoprim-Sulfamethoxazole

December 2012 Volume 50 Number 12 jcm.asm.org 4069

on Novem


.asm.org/

Dow

nloaded from

42

http://jcm.asm.org

http://jcm.asm.org/

firmed the ongoing susceptibility of S. aureus to SXT in thisregion of Australia (38, 49).

The breakpoints utilized for this study were defined byEUCAST using data collated from a wide range of sources on morethan 2,500 isolates of S. pyogenes tested for susceptibility to SXT usinga variety of methods (32). Of the 2,596 tests reported from multiplesources, geographical areas, and time periods, 2,559 were susceptibleto SXT of �1 mg/liter and 23 isolates were resistant (0.9%) (Euro-pean Committee on Antimicrobial Susceptibility Testing, data fromthe EUCAST MIC distribution website, http://mic.eucast.org, last ac-cessed 18 September 2012). On disk diffusion testing for S. pyogenes

using SXT disks in a ratio of 1:19, a zone size of �18 mm indicatessusceptibility and �15 mm indicates resistance. Using these break-points, a total of 358 tests were performed, with 5 confirmed resistantstrains (1.4%) (http://mic.eucast.org, last accessed 18 September2012). This can be contrasted with results found in U.S.-based liter-ature using the CLSI methods, where AST for S. pyogenes is per-formed on agar supplemented with defibrinated sheep blood andSXT is not routinely tested, as S. pyogenes strains are considered uni-versally resistant (51). Defibrinated sheep blood is utilized, as thehemolytic reactions of �-hemolytic streptococci on blood agar con-taining sheep blood are deemed “true” (1).

TABLE 6 Summary of published results of S. pyogenes in vitro susceptibility to SXTa

Publication Yr Country Method and medium used Findings Resistance

Yourassowsky et al. (54) 1974 Belgium AST determined by the agardilution method on Wellcotestagar supplemented with 5%lysed horse blood

59/59 strains of S. pyogenessusceptible to TMP,SMZ, and SXT

0%

Finland et al. (22) 1976 United States Plate dilution method on amodified thymidine-deficientMueller-Hinton mediumcontaining 5% laked blood

35/35 strains of S. pyogenessusceptible to SXT; 32/35 strains of S. pyogenessusceptible to SMZalone

0%

Darrell et al. (14) 1968 United Kingdom MIC determined by the platedilution method on diagnosticsensitivity agar containing 5%lysed horse blood

14/14 strains of S. pyogenessusceptible to TMP withMIC � 1 �g/ml

0%

Liebowitz et al. (36) 2003 South Africa MICs determined by the brothmicrodilution methodaccording NCCLS guidelines

66/66 S. pyogenes strainssusceptible to SXT

0%

Hartman and Hoes (26) 1949 United States Wilson’s method (53, 53a), asemi-solid medium with lowsulfonamide antagonist content

94/96 strains of S. pyogenessusceptible tosulfadiazine

1.9%

Schultz and Frank (44) 1958 United States Wilson’s method (53, 53a), asemi-solid medium with lowsulfonamide antagonist content

84/86 strains of S. pyogenessusceptible to sulfurantibiotics

2.3%

Eliopoulos andWennersten (18)

1997 United States MICs determined by agar dilutionmethods on Mueller-Hinton IIagar � 5% lysed horse blood orwith thymidine phosphorylaseat 0.2 IU/ml added

58/60 S. pyogenessusceptible to SXT

3.3%

Berger-Rabinowitz andDavies (2)

1970 Israel Wilson’s method (53, 53a), asemi-solid medium with lowsulfonamide antagonist content

849/890 S. pyogenes strainssusceptible tosulfadiazine

4.6%

Dhanda et al. (15a) 2011 India Kirby-Bauer disk-diffusion test asper CLSI guidelines; agar notspecified


6.7%

Bushby (6) 1973 United States Disk diffusion using TMP/SMZdisks 1.25/23.75 �g on variousmedia


7.7%

Lakshmy et al. (34a) 2011 India Kirby-Bauer disk-diffusion test asper CLSI guidelines. Agar notspecified


21.8%

Dumre et al. (16a) 2009 Nepal Kirby-Bauer disk-diffusion test asper CLSI guidelines; agar notspecified


71%

Traub and Leonhard (51) 1997 Germany Agar disk diffusion using NCCLScriteria on sheep blood MHA

0/63 strains of S. pyogenessusceptible to SXT

100%

Kaplan et al. (33) 1999 United States Etest performed on MHAcontaining 5% sheep blood

0/169 strains of S. pyogenessusceptible to SXT withMIC � 32 �g/ml

100%

a TMP, trimethoprim, SMZ, sulfamethoxazole.

Bowen et al.


on Novem


.asm.org/

Dow

nloaded from

43

http://mic.eucast.org

http://mic.eucast.org

http://jcm.asm.org

http://jcm.asm.org/

The in vitro results reported in the current study confirmingthe susceptibility of S. pyogenes to SXT suggest that treatment ofSSTI with SXT is worth considering. Our current RCT to assess thenoninferiority of SXT to the standard treatment with benzathinepenicillin G (BPG) for impetigo will provide the necessary clinicalevidence to inform guidelines. It is based on a pilot study of 13participants which indicated that both BPG and SXT were effica-cious in healing impetigo (48). There is one study published com-paring these agents for S. pyogenes infection in tonsillitis, whichreported a 70% treatment efficacy for SXT compared to 88% forpenicillin, a non-statistically significant difference (52).

The infrequent reports of susceptibility of S. pyogenes to SXTdemonstrate resistance rates ranging from 0% to 100% dependingon which medium and testing conditions are used (Table 6). Al-though the variation in results may relate to the particular strainsincluded or the local prescribing patterns of SXT, it is most likelyrelated to the methodology of testing. All of the studies reportinghigh resistance rates either used media known to have high con-centrations of thymidine or did not provide details of the mediumused. As standardization to ensure a low thymidine concentrationof Mueller-Hinton medium was introduced only in 2006, it islikely that, unless low-thymidine media were specified (2, 14, 18,22, 26, 44, 54), studies in publications prior to this may not havecontrolled for thymidine content.

Alongside this, S. pyogenes has remained 100% susceptible invitro to penicillin. Hence there has been no pressing need to un-derstand SXT susceptibility as an alternative antibacterial in thepublic health approach to treatment of S. pyogenes infections.However, this is changing in the context of rising MRSA rates forSSTI.

Our results show that testing of S. pyogenes for susceptibility toSXT on MHS gives a higher MIC than all of the other agars, al-though the organism remains in the susceptible range. This couldpossibly be due to the availability of thymidine or other inhibitorysubstances in this medium. However, thymidine concentrationsof the various media utilized were not assessed. Alternatively, theabsence of an enzyme to reduce the inhibition in sheep bloodcompared to horse blood may be the explanation.

The original paper describing the identification of the Harper-Cawston factor (25) as thymidine (21) reports an interesting ob-servation that our study has partially demonstrated. Only lysedhorse blood contains thymidine phosphorylase to convert thymi-dine to thymine and hence overcome the inhibition of folate me-tabolism that occurs in the presence of thymidine. No other mam-malian blood contains this enzyme, which is a possible reason forthe higher MICs reported on MHS than on those agars containinghorse blood. In the original paper, the presence of thymidine atconcentrations of 1.6 �g/ml was sufficient to completely preventinhibition by the drugs. The inclusion of lysed horse blood re-stored the inhibition. This has also been shown by Coll et al. (11).However, the MIC of S. pyogenes isolates tested on MHLHBA washigher than those of isolates tested on MHF (study 2), MHA(studies 1 and 2), or MHBA (study 1), which suggests other factorsat play.

A limitation of this study is the reliance upon a single methodfor susceptibility testing, the Etest, which is a commercially de-rived method. Further work using broth or agar dilution methodswould add to our understanding of the susceptibility of S. pyogenesto SXT. As shown in Table 6, when these additional methods havebeen assessed, the susceptibility of S. pyogenes ranges from 0%

(54) to 3.3% (18) resistant, similarly low resistances to those re-ported in this study.

Reading MICs for SXT can be challenging due to haze. In par-ticular, only faint growth of S. pyogenes was achieved on MHA(due to the absence of blood), and this made reading endpointsmore difficult. MHLHBA and MHF had problems similar to thoseof MHA with respect to haze. Despite this in study 2, the MICswere reproducible between readers, with a high level of interraterreliability within 1 MIC gradation.

Conclusions. The widespread belief that SXT is ineffective forS. pyogenes infections because of inherent antimicrobial resistanceis a fallacy due to technical limitations in laboratory methodology:namely, the use of media containing high concentrations of thy-midine, which inhibits the action of sulfur antibacterials. Whenmedia containing low concentrations of thymidine and/or highconcentrations of the enzyme thymidine phosphorylase are used,resistance rates are low in most cases, although this must be mon-itored over time and may vary with local epidemiology and anti-bacterial prescribing patterns. This study provides justification toproceed to clinical trials of SXT for S. pyogenes infections. Corrob-oration with clinical trial data may convince clinicians that SXTcan safely and appropriately be used for infections involving S.pyogenes. The Skin Sore Trial will answer the clinical applicabilityof this current in vitro study. In the era of rising MRSA prevalence,more clinical trials of SXT for treatment of SSTI, where S. pyogenesand S. aureus are frequently copathogens, are needed.

ACKNOWLEDGMENTS

The work of the research assistants in study 1 and the Skin Sore Trial team(Irene O’Meara, Jane Nelson, Tammy Fernandes, Melita McKinnon, Di-anne Halliday, Colleen Mitchell, Valerie Coomber, and Christine Fran-cais) in recruiting participants for study 2 has made this research possible.Study scientists who carried out this work include R.A.L., P.W., andVanya Hampton. Mark Chatfield provided assistance with the statisticalanalysis. We also acknowledge all of the participants and their familieswho have participated in the Skin Sore Trial and Rheumatic Heart Diseasestudies.

This work was supported through a National Health and Medical Re-search Council (NHMRC) Project Grant (545234) on which A.C.B.,S.Y.C.T., M.I.M., B.J.C., and J.R.C. are all investigators. A.C.B. is the re-cipient of a NHMRC scholarship for Ph.D. research (605845) as well as anAustralian Academy of Sciences Douglas and Lola Douglas scholarship.S.Y.C.T. is the recipient of a NHMRC Early Career Fellowship (605829).

We have no conflicts of interest to declare.

REFERENCES1. Anand C, Gordon R, Shaw H, Fonseca K, Olsen M. 2000. Pig and goat

blood as substitutes for sheep blood in blood-supplemented agar media. J.Clin. Microbiol. 38:591–594.

2. Bergner-Rabinowitz S, Davies AM. 1970. Sensitivity of Streptococcuspyogenes types to tetracycline and other antibiotics. Isr. J. Med. Sci. 6:393–398.

3. Bernstein LS. 1982. Combination of trimethoprim with sulfonamidesother than sulfamethoxazole. Rev. Infect. Dis. 4:411– 418.

4. Brook I, Frazier EH, Yeager JK. 1997. Microbiology of nonbullousimpetigo. Pediatr. Dermatol. 14:192–195.

5. Brumfitt W, Hamilton-Miller JMT, Kosmidis J. 1973. Trimethoprim-sulfamethoxazole: the present position. J. Infect. Dis. 128:S778 –S791.

6. Bushby SR. 1973. Sensitivity testing with trimethoprim-sulphamethoxazole.Med. J. Aust. 1(Suppl):10–18.

7. Bushby SR. 1973. Trimethoprim-sulfamethoxazole: in vitro microbiolog-ical aspects. J. Infect. Dis. 128(Suppl):442– 462.

8. CLSI. 2011. Performance standards for antimicrobial susceptibility test-ing: twenty-first informational supplement M100-S21, p 1–172. Clinicaland Laboratory Standards Institute, Wayne, PA.

S. pyogenes and Trimethoprim-Sulfamethoxazole

December 2012 Volume 50 Number 12 jcm.asm.org 4071

on Novem


.asm.org/

Dow

nloaded from

44

http://jcm.asm.org

http://jcm.asm.org/

9. CLSI. 2006. Protocols for evaluating dehydrated Mueller-Hinton agar:approved standard—2nd ed. Clinical and Laboratory Standards Institute,Wayne, PA.

10. Coburn AF. 1944. The prevention of respiratory tract bacterial infectionsby sulfadiazine prophylaxis in the United States navy. JAMA 126:88 –93.

11. Coll PF, Ausina VR, Vernis JV, Mirelis BO, Prats GP. 1984. Exogenousthymidine and reversal of the inhibitory effect of sulfamethoxazole-trimethoprim on streptococci. Eur. J. Clin. Microbiol. 3:424 – 426.

12. Dajani AS, Ferrieri P, Wannamaker LW. 1972. Natural history of impe-tigo. II. Etiologic agents and bacterial interactions. J. Clin. Invest. 51:2863–2871.

13. Damrosch D. 1946. Chemoprophylaxis and sulfonamide resistant Strep-tococci. JAMA 130:124 –128.

14. Darrell JH, Garrod LP, Waterworth PM. 1968. Trimethoprim: labora-tory and clinical studies. J. Clin. Pathol. 21:202–209.

15. Delameter ED, Jennings R, Wallace AW. 1946. Preliminary report of anoutbreak of Streptococcal disease caused by a sulfadiazine resistant groupA, type 17 hemolytic Streptococcus. J. Infect. Dis. 78:118 –127.

15a.Dhanda V, et al. 2011. Invasive group A streptococcus (GAS) from NorthIndia during 2007–2010: surveillance and antibacterial susceptibility, ab-str 5, p 18. Abstr. XVIII Lancefield Symp., Palermo, Italy.

16. Domagk G. 1935. Ein Beitrag zur Chemotherapie der Bakteriellen Infek-tionen. Dtsch. Med. Wochenschr. 61:250 –253.

16a.Dumre SP, et al. 2009. Asymptomatic throat carriage rate and antimicro-bial resistance pattern of Streptococcus pyogenes in Nepalese school chil-dren. Kathmandu Univ. Med. J. 7:392–396.

17. Eisen DT. 2010. Tetracycline, p 843– 894. In Grayson ML, et al (ed), Theuse of antibacterials, sixth ed, vol 1. ASM Press, Washington, DC.

18. Eliopoulos GM, Wennersten CB. 1997. In vitro activity of trimethoprimalone compared with trimethoprim-sulfamethoxazole and other antimi-crobials against bacterial species associated with upper respiratory tractinfections. Diagn. Microbiol. Infect. Dis. 29:33–38.

19. EUCAST. 29 June 2012, posting date. Antimicrobial susceptibility test-ing. http://www.eucast.org/antimicrobial_susceptibility_testing/disk_diffusion_methodology/.

20. EUCAST. 5 January 2011, posting date. EUCAST MIC and zone diameterbreakpoint tables. http://www.eucast.org/antimicrobial_susceptibility_testing/breakpoints/.

21. Ferone R, Bushby SR, Burchall JJ, Moore WD, Smith D. 1975. Identi-fication of Harper-Cawston factor as thymidine phosphorylase and re-moval from media of substances interfering with susceptibility testing tosulfonamides and diaminopyrimidines. Antimicrob. Agents Chemother.7:91–98.

22. Finland M, Garner C, Wilcox C, Sabath LD. 1976. Susceptibility ofbeta-hemolytic streptococci to 65 antibacterial agents. Antimicrob.Agents Chemother. 9:11–19.

23. Freis ED. 1944. The treatment of tonsillitis with small doses of sulfon-amides. JAMA 126:93–94.

24. Geria AN, Schwartz RA. 2010. Impetigo update: new challenges in the eraof methicillin resistance. Cutis 85:65–70.

25. Harper GJ, Cawston WC. 1945. The in-vitro determination of the sul-phonamide sensitivity of bacteria. J. Pathol. Bacteriol. 57:59 – 66.

26. Hartman TL, Hoes JS. 1949. The occurrence of a type five sulfadiazineresistant hemolytic streptococcus. J. Pediatr. 34:331–337.

27. Hartman TL, Weinstein L. 1948. The problem of sulfonamide-resistanthemolytic streptococci. N. Engl. J. Med. 238:560 –563.

28. Hitchings GH. 1973. Mechanism of action of trimethoprim-sulfamethoxazole. I. J. Infect. Dis. 128(Suppl):433– 436.

29. Holbrook WC. 1944. The Army Air Forces Rheumatic Fever ControlProgram. JAMA 126:84 – 87.

30. Hoskins TW, Bernstein LS. 1981. Trimethoprim/sulphadiazine com-pared with penicillin V in the treatment of streptococcal throat infections.J. Antimicrob. Chemother. 8:495– 496.

31. Johnson RD, Hartman TL. 1947. Sulfadiazine resistant streptococcalinfections in a civilian community. J. Clin. Invest. 26:325–328.

32. Kahlmeter G, et al. 2003. European harmonization of MIC breakpointsfor antimicrobial susceptibility testing of bacteria. J. Antimicrob. Che-mother. 52:145–148.

33. Kaplan EL, Johnson DR, Del Rosario MC, Horn DL. 1999. Susceptibilityof group A beta-hemolytic streptococci to thirteen antibiotics: examina-tion of 301 strains isolated in the United States between 1994 and 1997.Pediatr. Infect. Dis. J. 18:1069 –1072.

34. Koning S, et al. 2012. Interventions for impetigo. Cochrane DatabaseSyst. Rev. 1:CD003261. doi:10.1002/14651858.CD003261.pub3.

34a.Lakshmy A, Malhotra VL, Dutta R. 2011. Antibacterial resistance patternof Streptococcus pyogenes isolated from clinical samples, abstr 36, p 23.Abstr. XVIII Lancefield Symp., Palermo, Italy.

35. Laurell GL, et al. 1968. Streptococcal and staphylococcal infections.World Health Organization technical report series no. 394. World HealthOrganization, Geneva, Switzerland.

36. Liebowitz LD, Slabbert M, Huisamen A. 2003. National surveillanceprogramme on susceptibility patterns of respiratory pathogens in SouthAfrica: moxifloxacin compared with eight other antimicrobial agents. J.Clin. Pathol. 56:344 –347.

37. Maguire GP, Arthur AD, Boustead PJ, Dwyer B, Currie BJ. 1998.Clinical experience and outcomes of community-acquired and nosoco-mial methicillin-resistant Staphylococcus aureus in a northern Australianhospital. J. Hosp. Infect. 38:273–281.

38. McDonald M, et al. 2006. Recovering streptococci from the throat, apractical alternative to direct plating in remote tropical communities. J.Clin. Microbiol. 44:547–552.

39. McDonald MI, et al. 2006. Low rates of streptococcal pharyngitis andhigh rates of pyoderma in Australian Aboriginal communities where acuterheumatic fever is hyperendemic. Clin. Infect. Dis. 43:683– 689.

40. Najjar A, Murray BE. 1987. Failure to demonstrate a consistent in vitrobactericidal effect of trimethoprim-sulfamethoxazole against enterococci.Antimicrob. Agents Chemother. 31:808 – 810.

41. Okafor OO, Akinbami FO, Orimadegun AE, Okafor CM, Ogunbiyi AO.2011. Prevalence of dermatological lesions in hospitalized children at theUniversity College Hospital, Ibadan, Nigeria. Niger. J. Clin. Pract. 14:287–292.

42. Roberg N. 1946. An epidemic caused by a sulfadiazine resistant strain ofthe streptococcus hemolyticus (group A, type 17). J. Infect. Dis. 78:135–146.

43. Rubin RH, Swartz MN. 1980. Trimethoprim-sulfamethoxazole. N. Engl.J. Med. 303:426 – 432.

44. Schultz I, Frank PF. 1958. The prophylactic use of sulfamethoxypyrida-zine (kynex) during a streptococcal epidemic. Am. J. Med. Sci. 236:779 –785.

45. Steer AC, et al. 2009. High burden of impetigo and scabies in a tropicalcountry. PLoS Negl. Trop. Dis. 3 :e467. doi:10.1371/journal.pntd.0000467.

46. Stokes A, Lacey RW. 1978. Effect of thymidine on activity of trim-ethoprim and sulphamethoxazole. J. Clin. Pathol. 31:165–171.

47. Swedberg G, Ringertz S, Skold O. 1998. Sulfonamide resistance in Strep-tococcus pyogenes is associated with differences in the amino acid sequenceof its chromosomal dihydropteroate synthase. Antimicrob. Agents Che-mother. 42:1062–1067.

48. Tong SY, et al. 2010. Trimethopim-sulfamethoxazole compared withbenzathine penicillin for treatment of impetigo in Aboriginal children:a pilot randomised controlled trial. J. Paediatr. Child Health 46:131–133.

49. Tong SY, et al. 2009. Community-associated strains of methicillin-resistant Staphylococcus aureus and methicillin-susceptible S. aureus inIndigenous Northern Australia: epidemiology and outcomes. J. Infect.Dis. 199:1461–1470.

50. Tong SY, McDonald MI, Holt DC, Currie BJ. 2008. Global implicationsof the emergence of community-associated methicillin-resistant Staphy-lococcus aureus in Indigenous populations. Clin. Infect. Dis. 46:1871–1878.

51. Traub WH, Leonhard B. 1997. Comparative susceptibility of clinicalgroup A, B, C, F, and G beta-hemolytic Streptococcal isolates to 24 antimi-crobial drugs. Chemotherapy 43:10 –20.

52. Trickett PCD, Mogabgab PW. 1973. Trimethoprim-sulfamethoxazoleversus penicillin G in the treatment of group A beta-hemolytic streptococ-cal pharyngitis and tonsillitis. J. Infect. Dis. 128:S693–S695.

53. Wilson AT. 1945. Method for testing in vitro resistance of group Ahemolytic streptococci to sulfonamides. Proc. Soc. Exp. Biol. Med.58:130 –133.

53a.Wilson AT. 1954. A simplified method for testing sulfonamide resis-tance of group A streptococci. J. Lab. Clin. Med. 43:280 –285.

54. Yourassowsky E, Vanderlinden MP, Schoutens E. 1974. Sensitivity ofStreptococcus pyogenes to sulphamethoxazole, trimethoprim, and cotri-moxazole. J. Clin. Pathol. 27:897–901.

Bowen et al.


on Novem


.asm.org/

Dow

nloaded from

45

http://www.eucast.org/antimicrobial_susceptibility_testing/disk_diffusion_methodology/

http://www.eucast.org/antimicrobial_susceptibility_testing/disk_diffusion_methodology/

http://www.eucast.org/antimicrobial_susceptibility_testing/breakpoints/

http://www.eucast.org/antimicrobial_susceptibility_testing/breakpoints/

http://jcm.asm.org

http://jcm.asm.org/

CHAPTER 4THE CHALLENGES OF COLLECTING ANDTRANSPORTING IMPETIGO SWABS IN A

TROPICAL ENVIRONMENT


Being able to effectively collect and transport impetigo swabs in a tropical

environment has been critical to the work undertaken in my thesis. This chapter

describes the method developed and validated for the transportation of impetigo

swabs from very remote contexts for processing in a central microbiology

laboratory. It also succinctly describes the methods used to culture and identify

Staphylococcus aureus and Streptococcus pyogenes. This supplements the

methods for antibiotic susceptibility testing described in Chapter 3.

The primary objective of the Skin Sore Trial was to determine non-inferiority of

trimethoprim/sulphamethoxazole to benzathine penicillin G for the treatment of

impetigo. This objective was complemented by the secondary outcomes, which

were to determine the relative abundance of S. aureus and S. pyogenes at days 0,

2, and 7 of the study. To be confident in these secondary outcomes, it was

necessary to develop and test robust methods for collecting and transporting skin

swab specimens.

The remote communities involved in the study were up to 1500 kilometres away

from the laboratory and did not have daily flights connecting the community to

the laboratory. As such, transportation of swabs was by either road or air.

Previous research protocols for detecting S. pyogenes recommended either:

1) immediate plating of the collected swab in the community and transportation

within 24 hours for incubation in the laboratory or; 2) plating and incubation of

the swab within 48 hours of sampling. (Johnson, 1996, Kotloff, 2008) These

recommendations were developed for contexts similar to the remote

communities of the Northern Territory and had been previously utilised in

47

smaller studies. (McDonald et al., 2006, Carapetis et al., 1995) However, the

size and complexity of recruiting 663 participants over three years from multiple

remote communities with varying available transportation logistics resulted in

the need to develop and test an alternative solution. Transport media have been

recommended for this situation, although little data was available on freezing

such specimens.

An additional benefit of the method reported in this chapter was that swabs

could be processed in batches. This was important for the laboratory scientists

working on the study to manage workflow as more than 4500 swabs of impetigo

lesions and the anterior nares were collected and processed during the course of

the RCT. This methodology also permitted freezing of the swabs for later

molecular and metagenomic studies that have the potential to further our

understanding of the mechanisms of transmission of bacterial skin pathogens

within and between communities.

The transport medium validated in this study is cheap, easy to produce, effective

in transporting skin pathogens, can be frozen and may be utilised in research in

other remote contexts.


Together with Steven Tong, I conceived this experiment, with oversight from

Jonathan Carapetis. I wrote the SOP for this experiment and oversaw the data

collection. I, Steven Tong and Mark Chatfield conducted the data analysis. I

performed the literature review and wrote all drafts of the manuscript. All co-

authors contributed to revisions of the manuscript and approved the final version

for publication. Steven Tong agreed to be the corresponding author while I was

on maternity leave.

48


Comparison of three methods for the recovery of skin pathogens from impetigo swabs collected in a remote community of the Northern Territory, Australia

49

CHAPTER 5DEVELOPING A METHODOLOGY FOR

CAPTURING AND SCORING STANDARDISED DIGITAL IMAGES OF

IMPETIGO


This chapter describes the novel methods needed for assessing and blinding the

primary end point within our RCT, which has the potential to be a standardised

approach for similar studies in the future.

Prior impetigo studies synthesized in the Cochrane review of impetigo treatment

(Koning et al., 2012) have predominantly been conducted in urban physician

offices or hospital outpatient settings. This allows more ready access to experts

who can immediately assess the participant for the primary outcome. Due to the

remoteness of the research locations in our trial, it was not feasible to transport

an expert to assess the outcome of every one of the 663 episodes of impetigo

under investigation. As such, building on the work of tele-dermatology, digital

images were determined to be the best method available to capture and evaluate

the primary endpoint of this trial. However, no published protocols for capturing

standardised digital images by amateur photographers were available.

As an open label RCT, we needed a blinded, robust, reproducible and defendable

primary endpoint. This chapter describes in detail the development and

evaluation of our approach. This was primarily work which I led. Expert

photographic input was sought from a professional photographer to guide this

process. Once standardised digital images of all enrolled sores from days 0, 2 and

7 were available, a method for scoring these images that minimised bias and

provided readily analysable data was needed. This chapter will also describe the

57

method developed for scoring outcomes using assessors blinded to treatment

allocation.

Having published our approach, these methods are now available for future

impetigo treatment studies, and could be implemented in either remote or office-

based research. It may be possible to compare the results from an expert who is

immediately able to objectively assess the process of sore healing with results

from other experts scoring the digital images distant from the participant. This

translation will add value to future studies of impetigo, and facilitate high quality

treatment studies where the burden of impetigo is the highest.


I designed and wrote all the standard operating procedures for collecting digital

images in the Skin Sore Trial. Assistance was sought from Kara Burns, a medical

photographer, for expert advice. I developed the training packages for the

research assistants and oversaw the quality checks of all digital images. Ross

Andrews proposed the concept for scoring the digital images in a randomly

allocated order, which I then revised and developed into a feasible protocol. I

oversaw all aspects of digital image scoring. Robyn Liddle developed the

databases within which digital images were scored. I cleaned and analysed the

data, developed the quick lists and drafted all versions of the manuscript. All co-

authors contributed to revisions of the manuscript and approved the final version

for publication.


Standardising and assessing digital images for use in Clinical Trials: A practical, reproducible method that blinds the assessor to treatment allocation

58

Standardising and Assessing Digital Images for Use inClinical Trials: A Practical, Reproducible Method ThatBlinds the Assessor to Treatment AllocationAsha C. Bowen1,2,4,5*, Kara Burns2,3, Steven Y. C. Tong1,2, Ross M. Andrews1, Robyn Liddle1,

Irene M. O9Meara1, Darren W. Westphal1,4, Jonathan R. Carapetis4,5

1 Menzies School of Health Research, Charles Darwin University, Darwin, NT, Australia, 2 Royal Darwin Hospital, Darwin, NT, Australia, 3 School of Business, Queensland

University of Technology, Brisbane, QLD, Australia, 4 Telethon Kids Institute for Child Health Research, University of Western Australia, Perth, WA, Australia, 5 Princess

Margaret Hospital for Children, Perth, WA, Australia

Abstract

With the increasing availability of high quality digital cameras that are easily operated by the non-professionalphotographer, the utility of using digital images to assess endpoints in clinical research of skin lesions has growingacceptance. However, rigorous protocols and description of experiences for digital image collection and assessment are notreadily available, particularly for research conducted in remote settings. We describe the development and evaluation of aprotocol for digital image collection by the non-professional photographer in a remote setting research trial, together witha novel methodology for assessment of clinical outcomes by an expert panel blinded to treatment allocation.

Citation: Bowen AC, Burns K, Tong SYC, Andrews RM, Liddle R, et al. (2014) Standardising and Assessing Digital Images for Use in Clinical Trials: A Practical,Reproducible Method That Blinds the Assessor to Treatment Allocation. PLoS ONE 9(11): e110395. doi:10.1371/journal.pone.0110395

Editor: H. Peter Soyer, The University of Queensland, Australia

Received June 17, 2014; Accepted September 13, 2014; Published November 6, 2014

Copyright: � 2014 Bowen et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and itssupporting information files.

Funding: This work was supported by a National Health and Medical Research Council (NHMRC) of Australia project grant (545234) for which AB, ST, RA and JCare all investigators and IMO is funded as the trial project manager. AB is the recipient of an NHMRC scholarship for PhD research (605845) and an AustralianAcademy of Sciences Douglas and Lola Douglas scholarship. ST is the recipient of a NHMRC Career Development Fellowship (1065736). The funders had no role instudy design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests: The authors have declared that no competing interests exist.

* Email: [email protected]

Introduction

Telemedicine is increasing in popularity, particularly for

dermatologists and other specialities where a clinician is not

always onsite for direct patient care [1]. With the increasing

availability of cheap, simple, high quality digital cameras for the

non-professional photographer, often a clinician or auxiliary [2,3],

the appeal of utilising digital images to diagnose or monitor

treatment progress is growing. An extension of this application is

the use of digital images of cutaneous disease to assess outcomes

for intervention studies including randomised controlled trials

(RCTs) [4]. Advantages include: maintaining blinding of the

outcome assessor to treatment allocation; scoring digital images in

batches to improve work flow; and utilising expert outcome

assessors remote from the site of data collection. These advantages

have particular appeal for the conduct of research in remote

regions or for multicentre studies where standardisation of clinical

outcome measures is critical. However, published methodologies

to guide such use of digital images are lacking. In addition, unlike

hospital or clinic based health photography which is usually

performed in a dedicated setting with instruments and lighting

operated by a professional photographer [5,6], this ideal may not

be achievable in remote, field-based research. Therefore we aimed

to develop a method to facilitate the acquisition of consistent, high

quality digital images of superficial skin lesions in the context of

conducting a RCT in a remote setting. A medical photographer

(KB) guided this process [4,7]. The digital images were taken by

field research staff and allowed the assessment of outcomes in a

blinded manner.

The four characteristics of an excellent clinical photograph are

correct perspective, use of a scale aligned with the image frame,

even lighting and a neutral background [8]. To achieve these

characteristics, it is important to standardise the equipment,

camera settings, participant positioning and photography tech-

nique so that reproducible images are captured [9,10]. Once

images of satisfactory quality have been captured, consideration

must be given to storing these confidential images for future use.

Standardised protocols that address these priorities for collecting

digital images are not available in the peer-reviewed literature.

Impetigo trials are an example of cutaneous disease research

where digital images can improve the objectivity of outcome

assessment. Impetigo is a common, non-benign cutaneous

infection that mostly occurs in resource-limited contexts [11],

affecting .2% of the global population at any one time [12].

Impetigo also regularly affects school-aged children in industrial-

ised settings [13]. Treatment of impetigo is a public health priority

to prevent severe sequelae including streptococcal and staphylo-

coccal sepsis, focal invasive disease, post-streptococcal glomerulo-

nephritis (which is in turn linked with chronic renal failure) [14],

and a postulated causative link with rheumatic fever and

rheumatic heart disease [15,16]. These sequelae mainly occur in

PLOS ONE | www.plosone.org November 2014 | Volume 9 | Issue 11 | e1103951

59

http://creativecommons.org/licenses/by/4.0/

http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0110395&domain=pdf

resource-limited settings and are responsible for hundreds of

thousands of deaths each year globally [11].

The authors of a meta-analysis on interventions for impetigo

recommended the use of clear and objective outcome measures for

future impetigo research [17]. As the burden of impetigo is in

resource-limited contexts where ready access to clinicians for

immediate end-point assessment may not be feasible, digital image

end-points are appealing and objective. The only RCT from a

resource-limited context included in the meta-analysis reported the

use of photographs for outcome assessment. In this RCT,

successful treatment was defined as clinical cure or marked

improvement with an additional measure using photographs, but

they did not report the methodology of image collection or how

they determined the outcome based on this end-point [18]. Well-

defined, reproducible endpoints are needed and blinded outcomes

are the cornerstone of good clinical trial design [19].

In conducting a RCT on impetigo treatment in a remote

setting, we developed a standardised, reproducible method for

collecting images of skin sores at different time points [6].

Reviewers blinded to treatment allocation assessed the images

using a simple, reproducible, quick method that provided readily

analysable data. The image capture protocol and novel method-

ology for blinded assessment may be useful for other trials in

cutaneous disease research.

Method

EthicsThis study and all consent documentation were approved by the

Human Research Ethics Committee of the Northern Territory

Department of Health and Menzies School of Health Research

(HREC 09/08). Indigenous elders provided community consent

before recruitment commenced. The parent or legal guardian for

all participants provided written informed consent. The study was

explained by a local interpreter or by using a talking book in the

participant’s first language. Written consent was itemised for all

study procedures including the collection of digital images.

SettingWe conducted a non-inferiority RCT in 7 remote Indigenous

communities of the Northern Territory of Australia between 2009

and 2012 [20]. The research team was based in Darwin and

travelled via plane or road to the remote communities up to 1500

kilometres away. There were 663 episodes of impetigo (in 508

children) enrolled in the trial and each had either one or two sores

under investigation. Research assistants trained in the photogra-

phy protocol captured digital images of all trial participants’ sores

on days 0, 2 and 7. Images were stored electronically. A panel of

paediatricians specialising in the care of Indigenous Australian

children externally reviewed these digital images at a later date,

according to a standardised scoring system as reported below.

Details of the protocol for capturing digital images ofimpetigo

Equipment. All images were captured using standardised

equipment (table 1, figures 1–4).

Camera Settings. Due to data collection occurring simulta-

neously in more than one community, we purchased three

identical cameras. All study camera settings were programmed

by the study manager (figure 5) and checked against the standard

operating procedure (SOP) by the research assistants before each

use. Training in the programming of settings and operation of the

camera was conducted with each new research assistant and image

quality reviewed at the completion of most field work trips. A

quick list to summarise the process for image capture in the field

was developed (table 2).

Lighting Conditions. The main light source was the flash as

force flash was always on. This provided a standard light source for

all three cameras in all settings. In addition, preference was given

to capturing digital images outdoors in the shade to improve the

ability of the camera to focus on the subject and avoid distracting

shadows [10]. Direct sunlight was avoided to minimise the

potential of a stronger light source resulting in an over-exposed

image. For uniformity of conditions, when it was not possible to

photograph participants outdoors, maximal ambient light was

achieved by turning on all lights and opening any curtains or

doors. Night time photography was avoided by returning to

photograph the participant first thing the following day.

Positioning of study participant. As we were working with

children, prior to taking any photographs, the participants were

reminded to remain still and the carer was engaged in reassuring

the child. The participant was positioned comfortably in a chair or

on the floor with a neutral grey background beneath the limb or

site to be photographed (figures 4 and 6). Jewellery, clothing or

hair that might obscure the area of interest were removed or tied

back. Any dressings covering the lesion were also removed.

Once settings were rechecked, a 5 cm neutral grey scale was

placed in a vertical position, in the same plane as the sore, as close

as possible to the left of the sore without obscuring any edges of the

lesion (figures 2 and 6). The upper limit (0) of the scale was

positioned at the top of the frame and the lower limit (5) at the

bottom. This ensured all images were captured at the same scale so

that when paired images were reviewed the sores were comparable

and any reduction in size could be assessed as a measure of sore

healing.

Photography techniqueTo maximise sharpness by depth of field, the camera lens plane

was positioned parallel to the sore plane with the photographer

standing above the sore (figure 6). The sore was centred using the

white square corners at the centre of the camera screen. The

shutter was depressed halfway to focus the lesion prior to capturing

the image.

A minimum of three images of each sore were taken at each

time point, to ensure that at least one adequate image was

available for outcome assessment, a technique known as bracket-

ing [21]. The research assistant was instructed to check each image

for SOP conformity and to take additional photographs if a clear,

focussed image showing all details of the sore had not been

Figure 1. The Panasonic DMC-TZ8 camera demonstrating zoomand camera settings as indicated in figure 5. http://www.digitalcamerawarehouse.com.au/prod6699.htm, last accessed 28 Sep-tember 2014).doi:10.1371/journal.pone.0110395.g001

Standardising Digital Images in Clinical Trials


60

http://www.digitalcamerawarehouse.com.au/prod6699.htm

http://www.digitalcamerawarehouse.com.au/prod6699.htm

obtained. Each photograph number was recorded in the respective

participant’s case report form (CRF).

Once all images had been captured, brief notes were made in

the participant’s CRF to describe the positioning of the participant

so that whenever possible the same position could be used for

future images. Further follow up images of the same sore were

required on day 2 and day 7 from enrolment. For consistency of

orientation, previous images of the same sore were checked on the

study camera before capturing the next image.

Image download and storageStandardisation of image storage is critical to the meticulous

utilisation of this method [9,22]. At the completion of each day,

the image files were downloaded from the camera memory card to

a password-protected laptop. The laptop files served as a data

backup, which was important given that study visits lasted between

two and three weeks and internet was not reliable enough in the

remote context to upload numerous large files every day. Upon

return of the team to the research centre in Darwin; all images

were downloaded from the camera to the main computer server

where daily backups occur. All images were taken and stored in

high quality.JPEG (Joint Photographic Experts Group) format for

convenience as our chosen camera did not shoot in an

uncompressed (‘raw’ or ‘ loss-less’) format. There are limitations

to using a compressed or ‘lossy’ format but the benefits of using a

compact camera and storage of images for comparison outweighed

these and is in line with other clinical studies [23]. Three copies of

each unmodified image were saved: one in the participant’s folder

labelled with participant number; one in the generic backup folder

labelled with the camera-generated photograph number as

recorded in the participant CRF; and one labelled with a

randomly generated number between 1 and 15 000. Only once

this had occurred were images deleted from the memory card.

These re-identifiable images are to be stored on a secure server for

up to 25 years, in keeping with ethical requirements for research in

children [24]. From the three available digital images of each sore,

the best quality image (key criteria were focus, exposure and

magnification) was selected for outcome assessments.

Quality control processAs all images were collected by amateur photographers, images

were regularly checked by the study doctor (AB) and feedback

provided if the image did not conform to the SOP. In addition, prior

to commencing primary outcome reviews, a quality control (QC)

check of all available digital images was performed mid-way

through the study (collected from the first 200 study participants) by

a medical photographer (KB) experienced in capturing digital

images of skin conditions. The QC was a priority in this study as the

method described had not been previously used or evaluated and we

were unsure whether digital image manipulation might be needed

for image scoring. Digital images can be manipulated to overcome

flaws in image capture [21], however this has limitations. Our apriori hypothesis was that digital image manipulation would not be

needed. To confirm this after recruitment of 200 participants, 1 300

images were scored as either adequate or unable to be interpreted

using the definitions in table 3. If ‘‘unable to be interpreted’’ was

chosen, the reviewer was instructed to provide reasons (figure 7).

Figure 2. Use of scale to define the upper and lower boundaries of image in the landscape position. This series of images of the samesore on days 0 (A), 2 (B) and 7 (C) utilise the scale well with the 0 at the top of the image and the 5 at the bottom, are clear and focussed anddemonstrate sore healing over time. Limitations include different availability of light as captured during different parts of the day using outdoor light.doi:10.1371/journal.pone.0110395.g002

Figure 3. An example of the participant identification carddescribed in table 1. This card contains participant number, date ofimage, study day, and whether it is sore A or B as up to two-thirds ofstudy participants had two sores enrolled in the study.doi:10.1371/journal.pone.0110395.g003

Figure 4. Capturing the image using the study camera, greybackground, and grey scale in a remote context. The individualsin this image have given written informed consent to publish thisimage.doi:10.1371/journal.pone.0110395.g004



61

Methods for digital image assessmentWe developed a method for scoring digital image pairs that was

simple, limited bias, quick and afforded readily analysable data. As

the methodology was novel, a paper-based pilot was conducted

employing 13 clinicians and researchers to confirm usability prior

to building an automated database for scoring. In the pilot, 22

paired digital images from 10 participants of either day 0 and 2 (5

participants) or day 0 and 7 (6 participants) were reviewed in

random order (days 0/2 or 2/0 and 0/7 or 7/0) by the 13

reviewers. The initial definition of healing or improved included

both a visual description of the sore pair and a clinical decision as

to whether further antibiotic treatment was needed.

The primary outcome for the RCT was treatment success at day

7 according to paired digital image scoring. After the successful

pilot, the digital image pairs were organised in random order in a

purpose built database (figure 8) using the randomly assigned

number between 0 and 15 000 as the only identifying information.

Scoring of the digital image pairs was on non-standardised

computer screens at locations remote from the primary study site.

Scoring was by a group of eight paediatricians with expertise in

caring for Indigenous children with impetigo. Primary outcome

reviewers blinded to treatment allocation were provided pairs of

images from day 0 and 7 (or day 0 and 2) in random order. Each

reviewer was unaware of which image (A or B) was pre- or post-

treatment and was asked to decide if image A, compared to image

B, was healed, improved, the same, worse, or unable to be

determined using the definitions (table 4) and vice versa (i.e.,

image B compared to image A). To expedite this process, an auto-

fill was used in the database. For example, when image A was

scored as ‘‘worse’’, auto-fill made available the options of ‘‘healed’’

or ‘‘improved’’ only for the comparison of image B to image A.

Where ‘‘healed’’ or ‘‘improved’’ were selected for image A, auto-

fill completed the scoring with ‘‘worse’’ for image B. The use of

auto-fill made the scoring process as rapid as possible. Thus

reviewers were blinded to both treatment allocation and the

chronological order of sores. Every image pair was evaluated by

two independent reviewers from the panel of eight. Where

disagreements occurred, an expert panel of three determined the

final result by consensus.

Following un-blinding of the chronological order of sores and to

produce readily analysable results from the blinded scoring system,

if image B was the day 2 or 7 sore, treatment success was deemed

to have occurred if image B was healed or improved compared to

image A (day 0 sore). If image B was the day 0 sore, then success

Table 1. Study equipment chosen including the required features, advantages and alternatives available or recommended in theliterature.

Item Specific Choice Features Advantages Alternatives

Digital Camera Panasonic LUMIXDMC-TZ8 14.5 megapixeldigital camera (figure 1)

Macro to 3 cm from a 12Xzoom lens equivalent to a25–300 mm lens on an SLR

Inexpensive Digital single lens reflex (SLR)camera [32]

Built in flash Readily available

Rugged metal casing Automatic

Lithium ion rechargeablebattery

Pre-specified settingsto achieve uniform, reproducibleimages whilst using amateurphotographers

4GB memory card

Background Grey background(figure 4)

A small grey board positionedbehind the body part beingassessed

Transportable and light weight Green or grey surgical drapes[33]

Manoeuvrable

Non-reflective

Neutral

Scale [34–36] 5cm, non-reflectivegrey scale (figure 2)

Vertical scale Defined the upper and lowerboundaries of the photograph(figure 2)

Paper or commercially availablescale e.g. the ABFO scale [34]

Single use sticker Easily removed and discarded

Good infection control

Identification Card Cardboard handwrittencard (figure 3)

Pre-formatted to add individualrandomisation number,study visit day, sore number,date and time of image capture

Maintained study blinding

Photographed prior to impetigoimage capture to avoid inclusionof any identifying details inthe image for scoring by blindedreviewers[37]

When forgotten, identification card wasphotographed upside downimmediately after the series of soreswas captured [37]

doi:10.1371/journal.pone.0110395.t001



62

Figure 5. The camera settings and icons as used in the protocol (these icons and settings are standard across other popular cameramodels). In addition the rationale is provided on why these settings were adopted.doi:10.1371/journal.pone.0110395.g005

Table 2. Quick List for capturing standardised photographs.

Step Action

1 Confirm all camera settings are correct (figure 5), participant, paperwork and previous image for orientation (if available)

2 Position participant in the shade: comfort, lesion exposed, neutral background, scale in the same plane as the lesion

3 Photograph participant ID (table 1) prior to capturing series to preserve blinding.

4 Position camera in same plane as sore (figure 6). Centre the sore and focus camera. Take minimum of 3 photos. Take additional photographs if none are clearand focussed.

5 Record photograph number and notes in participant’s file

6 Save digital images in secure location and delete from camera

This could be printed on a small card to be carried with the camera as a reminder to research assistants capturing images.doi:10.1371/journal.pone.0110395.t002



63

was deemed to have occurred if image B was worse compared to

image A (day 2 or 7 sores) and image A was either healed or

improved (table 4).

Results

Overall image collectionOver the 3-year study, almost 10 000 digital images were

collected and stored by more than 20 research assistants who

collected data in 7 remote communities covering an area of 1.35

million km2. From all of the images collected, the best available

image of the bracketed set was selected for outcome assessment.

Approximately 3 300 digital images were required to determine

the primary and secondary outcomes of the RCT by analysing the

results of paired comparisons. The project manager (IO) reviewed

all images and selected the best available image for the paired

comparison. The best available image (determined by focus,

exposure and magnification) was most often the second or third

image captured. This was consistent by study visit day.

Of the 9 944 digital images collected, only 17 (0.17%) were not

available for assessment due to staff error. These errors were the

wrong site photographed on subsequent days to the original site

(n = 6), the photograph not saved or filed (n = 4) and the

photograph not taken due to staff error (n = 7).

Quality control checkFor the QC check, 1 300 digital images from the first 200

participants that conformed to the described methodology were

reviewed. 1 258 (96.8%) were deemed adequate using the

definitions provided. Of the 42 images (3.2%) deemed unable to

be interpreted there was some overlap in categorisation: 29 were

due to incorrect exposure (8 too light, 21 too dark), 16 were due to

lack of focus, 2 had incorrect magnification due to lack of focus

and 1 had incorrect magnification (Table S1). Results of the QC

review were reassuring and consequently digital image modifica-

tion was not required.

Figure 6. Cartoon demonstrating the photographer, cameraand sore were in the same plane to optimise reproducibility ofcaptured imagesdoi:10.1371/journal.pone.0110395.g006

Figure 7. Database form used for Quality Control check by medical photographer.doi:10.1371/journal.pone.0110395.g007



64

Pilot for digital image scoringThirteen reviewers piloted the digital image scoring process. All

pilot reviewers agreed the process was quick and manageable with

image quality being adequate. Initially definitions for healing,

improved, same or worse included a 2-armed definition with a

description of both sore healing and a clinical judgement as to

whether treatment with further antibiotics was indicated. The

reviewers reported that the need for additional treatment was

difficult based on images alone and as the decision had no timely

clinical impact, we removed this decision from the definitions.

Digital image scoring resultsOutcome scorers reported 98.3% of digital images as able to be

interpreted using the quality codes shown in figure 8. Of these,

89.9% were adequate and 8.4% suboptimal but still able to be

interpreted (Table S2). The inter-rater reliability of digital image

scoring was moderate. When assessing for treatment success

(pooled healed and improved, table 4) versus treatment failure

(unchanged or worse), there was 86% agreement between

reviewers with a kappa score of 0.4. When assessing using the 5

Table 3. Definitions used for the quality control assessment of digital images.

Assessment Definition

Adequate The entire sore was seen in enough detail to determine the margin and most of the interior; AND the scale was seenin sufficient detail to determine the approximate size.

Unable to be interpreted The image of the sore could not be interpreted due to incorrect exposure (too light or too dark), focus (lack of focusor depth of field) or incorrect magnification.

doi:10.1371/journal.pone.0110395.t003

Figure 8. This shows the database format utilised for scoring digital image pairs. Image A was taken on day 0 and image B on day 7.Image B shows erythema and as such was scored as improved using the definitions in table 4.doi:10.1371/journal.pone.0110395.g008



65

available definitions (figure 8, table 4) there was 64% agreement

between reviewers, with a kappa score of 0.3 (Table S3).

Discussion

This is the first description of a method for capturing and

scoring comparative digital images of skin lesions in clinical

research. The methods outlined were practical even in remote

contexts, robust, reproducible and simple enough for non-

professional photographers to consistently follow. Strengths of

the described methodology include the quality control check and

more than 98% of captured images being interpretable. The gold

standard for needing to retake orthodontic photographs for poor

quality was set at 90% [3] and our findings of adequacy were at

this level, but when ‘suboptimal but still able to be interpreted’ was

included exceeded this gold standard. In addition, the described

method was followed by more than 20 study staff in remote

contexts resulting in ,0.2% of images being unavailable for

assessment. Digital images were the only available form of

documentary evidence for this blinded, clinical trial so it was

essential to have a robust process. Our results support that the

process outlined works.

The adoption of a standard set of image settings (figure 5) and a

Quick List that guided training in the methodology (table 2)

facilitated a uniform set of images that did not require any digital

manipulation. Guidelines on the manipulation of digital images

specify that while ‘‘it is acceptable practice to adjust the overall

brightness and contrast of a whole image’’ [25] it is best practice if

a group of images are to be compared to each other, that the

processing of individual images should be identical. The question

of ‘‘what constitutes a ‘‘reasonable’’ adjustment of image settings

such as brightness and contrast, etc.’’ has become important for

publication in scientific journals and is now included in

instructions to authors [26]. For example, the instructions to

authors in the Journal of Cell Biology outline that, if manipulation

of a digital image is undertaken, these manipulations must not

obscure, eliminate, or misrepresent any information present in the

original [25]. Forensic guidelines also emphasise this rigorous

approach for reproducibility [27]. As the method for capturing

digital images reported above had not been previously validated

and the outcome was based on a comparison of image pairs, the

QC check by a professional medical photographer was an

important step in determining whether our images would meet

this industry standard. Based on these results, we did not permit

the use of photo-editing software to modify any of the images [27].

We recommend following a protocol such as ours that has been

subjected to rigorous QC checks for future skin disease research

which should largely obviate the need for any digital image

manipulation.

High staff turnover when working in remote settings [28]

resulted in frequent training and re-training sessions in capturing

digital images using the methods described. Educational Power-

Point slides were developed for this purpose and supplemented by

the quick reference guides developed (tables 1 and 2, figures 5 and

6). In addition, real time review of captured digital images with

feedback to the research assistants was useful. Despite the use of a

standard protocol and ongoing training, occasional human errors

did occur as described above.

A limitation when using digital images for endpoint assessment

is the inability of reviewers to make a clinical decision using the

additional senses of hearing (patient feedback on pain and

pruritus), touch (warmth, fluctuance) and smell, when provided

only with the image. For cutaneous diseases where the appearance

of a lesion is the primary determinant of outcome, this limitation

can be partly addressed with a robust protocol for capturing

reproducible, diagnostic images for outcome assessment. This

known limitation impacted upon the inter-rater reliability agree-

ment as physicians were asked as reviewers to use a novel

diagnostic modality to score outcomes. To overcome this, we used

a consensus panel of three to adjudicate any discrepant scoring.

The consensus panel discussed all image pairs until consensus was

reached. The possibility that the use of the project manager to

select the best available image introduced bias is a possible

limitation. However, the QC check by a professional photogra-

pher confirms that the perceived bias was minimal with high

quality images consistently being provided to reviewers.

We suggest that this protocol could also be adapted from the

research setting for use in clinical care. In settings where

specialised clinicians are not readily available, standardised digital

photography of cutaneous lesions could be used in telemedicine to

allow highly skilled clinicians to assist local health staff to manage

patients in remote locations.

A unique feature of this protocol for standardising the

comparison between image pairs of the same sore where the only

detectable difference was changes in the appearance of the lesion

[6], was the requirement for all research assistants to check the

orientation of the image on the study camera before capturing the

next image. Guidelines on doing this were provided. Previous

expert advice has been for image capture to be performed

consistently by the same photographer [21]. This was not possible

within the remote research environment and overall ,5% of

participants had all 3 days of images collected by the same person.

Nonetheless, .99% of image pairs were assessable for the primary

outcome. This finding adds to the photography literature.

Providing amateur photographers with simple instructions and

guidance for collecting digital images using standardised camera

settings results in digital images that are of a high quality and can

Table 4. Definitions used for final outcome scoring of digital images of impetigo.

Assessment Criteria Outcome

Healed Lesions no longer evident or flat, with no evidence ofcrusting, erythema or purulence,but possibly with evidence of hyper- or hypo-pigmentationwhere the original sore was located.

Success

Improved Lesions reduced in sore diameter and erythema; ANDprogression from blister tocrusting and flattening of the sore. Purulence not evident.

Same No appreciable change in diameter, erythema or purulenceof lesion.

Failure



66

be assessed by blinded, independent reviewers for outcome

determination.

A non-inferiority RCT comparing two treatments of a common

condition such as impetigo requires rigorous, blinded, objective

endpoints for assessment. Here, we have described the method

developed based on the available evidence and expertise, for

capturing digital images. We report these here for use in

subsequent research trials. This method was simple, reproducible

and from the QC check provided 97% of images that were

adequate for assessment. Whilst other RCTs have used a digital

image of the skin as a primary outcome, such as in pyoderma

gangrenosum [29], wound healing [30], or pressure sores [31], this

is the first report of a standardised, reproducible protocol that has

been subjected to a rigorous QC assessment for an impetigo RCT.

Our results confirm the reproducibility of the simple resources

developed and published herein which will further enhance the

rigour of trials using a photographic end-point.

Conclusions

This is the first report of a standardised, reproducible protocol

that has been subjected to a rigorous QC assessment for research

involving impetigo and could be adapted for other skin disease

research. Our study confirms non-professional photographers are

able to capture high quality digital images of skin for this purpose.

We present a simple method for capturing high quality digital

images of skin sores in a RCT and the methods used to score

digital image pairs. Future trials for management of skin

conditions, particularly in remote contexts, may benefit from

adopting this protocol.

Supporting Information

Table S1 Results of quality control (QC) check. When the

QC check was adequate, all other fields were automatically

completed as not applicable. Where the QC check score was not

interpretable, the subsequent fields of exposure, focus and

magnification were provided for the professional photographer

to give reasons for the decision.

(XLSX)

Table S2 Results of the quality assessment conductedby the primary outcome reviewers of the trial. Results

reported for simplicity are the combined quality result, as if either

one of the images were suboptimal, the image pair decision was

difficult. There were 8 reviewers in the study and quality

assessments from all reviewers are included in this table.

(XLSX)

Table S3 Dataset used to calculate inter-rater reliabil-ity and the kappa scores provided. Reviewers were

numbered 5, 6, 7, 9, 10, 12, 14 and 16. When all reviewers

scored all image pairs, the number of the reviewers selected for the

calculation is listed in columns revA_num and revB_num.

(XLSX)

Acknowledgments

We acknowledge the participants and families who participated in this trial

in the hope of finding a better treatment option for impetigo. Jane Nelson,

Deb Taylor-Thomson, Bart Currie, Malcolm McDonald and Mark

Chatfield provided critical evaluation of the methodology as it was

developed.

Author Contributions

Conceived and designed the experiments: AB ST RA RL JC. Performed

the experiments: AB KB IO. Analyzed the data: AB ST. Wrote the paper:

AB KB ST RL IO DW JC. Reviewed and approved the final version of the

manuscript: AB KB ST RA RL IO DW JC. Built databases: RL.

References

1. Thomas J, Kumar P (2013) The scope of teledermatology in India. Indian

Dermatol Online J 4: 82–89.

2. Sandler J, Gutierrez RJ, Murray A (2012) Clinical photographs: the gold

standard, an update. Prog Orthod 13: 296–303.

3. Sandler J, Dwyer J, Kokich V, McKeown F, Murray A, et al. (2009) Quality of

clinical photographs taken by orthodontists, professional photographers, and

orthodontic auxiliaries. Am J Orthod Dentofacial Orthop 135: 657–662.

4. Katugampola R, Lake A (2012) The role of photography in dermatology

research. J Vis Commun Med 35: 5–10.

5. Nirmal B, Pai SB, Sripathi H (2013) A simple instrument designed to provide

consistent digital facial images in dermatology. Indian J Dermatol 58: 194–196.

6. Witmer WK, Lebovitz PJ (2012) Clinical photography in the dermatology

practice. Semin Cutan Med Surg 31: 191–199.

7. Young S (2008) Research for medical illustrators: designing photographic

protocols. J Vis Commun Med 31: 99–102.

8. Nayler JR (2003) Clinical photography: a guide for the clinician. J Postgrad Med

49: 256–262.

9. Sheridan P (2013) Practical aspects of clinical photography: part 1—principles,

equipment and technique. ANZ J Surg 83: 188–191.

10. Jakowenko J (2009) Clinical photography. J Telemed Telecare 15: 7–22.

11. Carapetis JR, Steer AC, Mulholland EK, Weber M (2005) The global burden of

group A streptococcal diseases. Lancet Infect Dis 5: 685–694.

12. Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, et al. (2012) Years

lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–

2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet

380: 2163–2196.

13. Shallcross LJ, Petersen I, Rosenthal J, Johnson AM, Freemantle N, et al. (2013)

Use of primary care data for detecting impetigo trends, United kingdom, 1995–

2010. Emerg Infect Dis 19: 1646–1648.

14. Hoy WE, White AV, Dowling A, Sharma SK, Bloomfield H, et al. (2012) Post-

streptococcal glomerulonephritis is a strong risk factor for chronic kidney disease

in later life. Kidney Int 81: 1026–1032.

15. McDonald M, Currie BJ, Carapetis JR (2004) Acute rheumatic fever: a chink in

the chain that links the heart to the throat? Lancet Infect Dis 4: 240–245.

16. Parnaby MG, Carapetis JR (2010) Rheumatic fever in indigenous Australian

children. J Paediatr Child Health 46: 527–533.

17. Koning S, van der Sande R, Verhagen AP, van Suijlekom-Smit LW, Morris AD,

et al. (2012) Interventions for impetigo. Cochrane Database Syst Rev 1:

CD003261.

18. Faye O, Hay RJ, Diawara I, Mahe A (2007) Oral amoxicillin vs. oral

erythromycin in the treatment of pyoderma in Bamako, Mali: an open

randomized trial. Int J Dermatol 46 Suppl 2: 19–22.

19. Baumgarten M, Margolis DJ, Selekof JL, Moye N, Jones PS, et al. (2009)

Validity of pressure ulcer diagnosis using digital photography. Wound Repair

Regen 17: 287–290.

20. Bowen A C, Tong SYC, Andrews RM, O’Meara IM, McDonald MI, et al.

(2014) Short-course oral trimethoprim/sulphamethoxazole is non-inferior to

intramuscular benzathine penicillin G for impetigo in a highly endemic region:

results of a randomised controlled trial. The Lancet, August 26.

21. Humphrey CD, Kriet JD (2010) Intraoperative photography. Facial Plast Surg

Clin North Am18: 329–334, Table of Contents.

22. Murdoch J (2005) Standardization in clinical photography at Addenbrooke’s

Hospital. J Vis Commun Med 28: 68–71.

23. Gulkesen KH, Akman A, Yuce YK, Yilmaz E, Samur AA, et al. (2010)

Evaluation of JPEG and JPEG2000 compression algorithms for dermatological

images. J Eur Acad Dermatol Venereol 24: 893–896.

24. The National Health and Medical Research Council tARCatAV-CsC (2007,

updated 2013)National Statement on Ethical Conduct in Human Research.

Canberra: Commonwealth of Australia.

25. Rossner M, Yamada KM (2004) What’s in a picture? The temptation of image

manipulation. J Cell Biol 166: 11–15.

26. Cromey DW (2010) Avoiding twisted pixels: ethical guidelines for the

appropriate use and manipulation of scientific digital images. Sci Eng Ethics

16: 639–667.

27. O’Malley T, Bails D, Chadwick Don behalf of the Electronic Evidence Specialist

Advisory Group (2004) Australasian Guidelines for Digital Imaging Processes,

version 2.

28. Bar-Zeev SJ, Kruske SG, Barclay LM, Bar-Zeev N, Kildea SV (2013) Adherence

to management guidelines for growth faltering and anaemia in remote dwelling



67

Australian Aboriginal infants and barriers to health service delivery. BMC

Health Serv Res 13: 250.

29. Craig FF, Thomas KS, Mitchell EJ, Williams HC, Norrie J, et al. (2012) UK

Dermatology Clinical Trials Network’s STOP GAP trial (a multicentre trial of

prednisolone versus ciclosporin for pyoderma gangrenosum): protocol for a

randomised controlled trial. Trials 13: 51.

30. Flanagan M (2003) Wound measurement: can it help us to monitor progression

to healing? J Wound Care 12: 189–194.

31. Jesada EC, Warren JI, Goodman D, Iliuta RW, Thurkauf G, et al. (2013)

Staging and defining characteristics of pressure ulcers using photographs by staff

nurses in acute care settings. J Wound Ostomy Continence Nurs 40: 150–156.

32. Khavkin J, Ellis DA (2011) Standardized photography for skin surface. Facial

Plast Surg Clin North Am 19: 241–246.33. Institute of Medical Illustrators (2007) IMMI National Guidelines Clinical

Photography in Wound Management.

34. Shapter M (2004) Placement considerations when using linear scales inphotographs. J Audiov Media Med 27: 54–57.

35. Ainslie GM, Reilly J (2003) The use of linear scales in the photography of skinlesions. J Audiov Media Med 26: 15–22.

36. Young S (2009) Research for medical illustrators: Using linear scales in

photographs. Journal of Visual Communication in Medicine 32: 4–7.37. Nayler J, Geddes N, Gomez-Castro C (2001) Managing digital clinical

photographs. J Audiov Media Med 24: 166–171.



68

CHAPTER 6THE RESULTS OF THE SKIN SORE

TRIAL


This chapter presents the results of the Skin Sore Trial, one of the largest

impetigo trials ever performed and the first to be conducted in remote Aboriginal

communities, where the disease burden is the highest worldwide (Chapter 2).

The trial results, now published in The Lancet, have already been incorporated

into local treatment guidelines.

It was possible to conduct a high quality, randomised controlled trial in these

remote communities due to a number of factors. These included the support from

the staff at the government and non-government remote clinics where extensive

consultation with health clinic managers and staff occurred. In all the

communities, elders or traditional owners provided consent to conduct the trial.

This facilitated participation. Finally, the commitment and dedication of the

senior trial staff, who trained and led a diverse and often transient group of

research assistants in the remote communities, was essential.

We initially proposed to recruit children presenting to the clinic for treatment of

skin sores. However, it was quickly realised that the high activity and acuity of

the clinics and the limited number of children presenting for treatment of sores,

made this approach unlikely to be successful. Ultimately, only 5% of participants

came from clinic referrals. Importantly, teachers at the local schools recognised

the need for sores to be treated and invited us in. The interest, support and active

participation of the school in each community made recruitment feasible. More

than 65% of the study participants were recruited from school settings, with

another 25% being siblings of those recruited in the school.

69

Each study trip had complicated logistics to arrange, including: coordination with

the school (avoiding holiday intervals) and with the clinic; arranging access to

accommodation and transportation within the community (both of which are in

short supply); coordinating, training and employing study staff (who had limited

longevity due to the demands of the trial) and; arranging transport schedules for

microbiology specimens to arrive back for processing at the laboratory in

Darwin. In total, 33 study trips were completed over the course of three years.

The NT is vast and sparsely populated. More than 50,000 kilometres were

travelled by road and air by the study team during the course of the study.

Recruitment commenced on 26 November 2009 and concluded on 20 November

2012. The proportion of participants recruited per study year was: 2009 (1%),

2010 (11%), 2011 (40%) and 2012 (48%). Due to challenges of conducting

research in these settings and previous experience, we allowed for up to 10% loss

to follow up in calculating our sample size. Pleasingly, more than 96% of study

participants completed all visits and of these, 97% received all doses of the study

drug.

The key findings were that: treatment success was achieved in 85% of

participants in both groups; SXT was non-inferior to BPG for the treatment of

impetigo; impetigo was driven by S. pyogenes even in the presence of S. aureus

and; BPG had a concerning rate of adverse events when administered

intramuscularly. These findings are important for Indigenous children from

remote communities. An alternative to the intramuscular BPG injection is now

available that is evidence-based, palatable, simple for adherence purposes and

pain-free. This study adds to the available evidence for treatment of extensive

impetigo.


The chief investigators designed and secured funding for the trial before I joined

as a PhD student. I critically revised and oversaw the protocol, participated in

data collection, oversaw the scientific direction of the study, wrote SOPs, and

cleaned and analysed all the data. I was assisted in data cleaning by Irene

70

O’Meara and Robyn Liddle. I led the statistical analysis with input from Mark

Chatfield and Steven Tong. I conducted the literature review and wrote all drafts

of the manuscript. All co-authors contributed to revisions of the manuscript and

approved the final version for publication.


Short course oral cotrimoxazole versus intramuscular benzathine benzylpenicillin for impetigo in a highly, endemic region: an open label, randomised, controlled, non-inferiority trial

APPENDIX 3

van der Wouden JC and Koning S. Treatment of impetigo in resource-limited

settings. Lancet 2014, August 26.

71

CHAPTER 7THE MICROBIOLOGY OF IMPETIGO IN

INDIGENOUS CHILDREN: ASSOCIATIONSBETWEEN STREPTOCOCCUS PYOGENES,STAPHYLOCOCCUS AUREUS, SCABIES,

AND NASAL CARRIAGE


This chapter broadens our understanding of the microbiology of impetigo. It

reinforces the ongoing dominance of Streptococcus pyogenes as a key pathogen

in impetigo in tropical contexts, which differs to the microbiology reported from

developed contexts. Staphylococcus aureus, and increasingly methicillin-

resistant S. aureus (MRSA), are reported from these contexts where the majority

of impetigo trials have been conducted. Few high-quality microbiology studies

are conducted where the burden is the highest and it is important to know, in

2014, that treatment directed against S. pyogenes for impetigo remains a priority.

This was assumed in the only other RCT conducted in a similar resource-limited

context where severe disease was also prevalent, but no microbiology specimens

were included in that study. (Faye et al 2009)

This chapter explores the associations of the microbiology with severe disease,

the presence of scabies, age groups, sex and region. There was no association

found between children who were in the severe strata of our study (>= 2 purulent

or crusted sores and > 5 overall body sores) and the detection of S. pyogenes,

S. aureus or both. This was surprising, as our initial hypothesis was that severe

sores might be more likely to be co-infected with both pathogens. The presence

of scabies was associated with detection of S. pyogenes. Whilst this finding was

not unexpected, this is the first study to confirm the association. Swabbing of the

anterior nares was included in the study design to assess for S. aureus nasal

81

carriage. We found that there was no association between the presence of

S. aureus in the anterior nares and detecting S. aureus in impetiginous lesions. In

fact, participants without S. aureus in their skin sores were more likely to

harbour S. aureus in the anterior nares. This suggests that nasal decolonisation

strategies are unlikely to be effective in reducing the burden of impetigo in

remote Indigenous communities.

The results of antibiotic susceptibility testing for both S. pyogenes and S. aureus

are reported in more detail in this chapter. Susceptibility to trimethoprim-

sulphamethoxazole was above 99% for both skin pathogens in this study.

However, increasing use of SXT may drive increasing resistance rates, and

ongoing surveillance is a priority. Whilst performing antibiotic susceptibility

testing of S. pyogenes to SXT is not routine in Australian microbiology

laboratories, the methods to do so are available and interest is growing in

incorporating this test.


I wrote all of the standard operating procedures for the collection, transportation,

storage, culture and antibiotic susceptibility testing of isolates from swabs

collected during the trial. I oversaw the scientists working in the laboratories at

the Menzies School of Health Research and Royal Darwin Hospital. I did all the

statistical analysis with assistance from Steven Tong and Mark Chatfield. I

drafted all versions of this manuscript. All co-authors contributed to revisions of

the manuscript and approved the final version for publication.

7.3 SUBMITTED JOURNAL ARTICLE, UNDER REVIEW (FOLLOWING PAGE)

The microbiology of impetigo in Indigenous children: associations of

Streptococcus pyogenes, Staphylococcus aureus, scabies and nasal carriage.

BMC Microbiology, under review.

82

Authors: Bowen AC, Tong SYC, Chatfield MD, Carapetis JR

7.4 ABSTRACT

Background: Impetigo is caused by both Streptococcus pyogenes and

Staphylococcus aureus; the relative contributions of each have been reported to

fluctuate with time and region. While S. aureus is reportedly on the increase in

most industrialised settings, S. pyogenes is still thought to drive impetigo in

endemic, tropical regions. However, few studies have utilised high quality

microbiological culture methods to confirm this assumption. We report the

prevalence and antimicrobial resistance of impetigo pathogens recovered in a

randomised, controlled trial of impetigo treatment conducted in remote

Indigenous communities of northern Australia.

Results: From 508 children, we collected 872 swabs of sores and 504 swabs

from the anterior nares prior to commencement of antibiotic therapy. S. pyogenes

and S. aureus were identified together in 503/872 (58%) of sores; with an

additional 207/872 (24%) sores having S. pyogenes and 81/872 (9%) S. aureus,

in isolation. Skin sore swabs taken during episodes with a concurrent diagnosis

of scabies were more likely to culture S. pyogenes (OR 2.2, 95% CI 1.1–4.4,

p=0.03). Eighteen percent of children had nasal carriage of skin pathogens. There

was no association between the presence of S. aureus in the nose and skin.

Methicillin-resistance was detected in 15% of children who cultured S. aureus

from either a sore or their nose. There was no association found between the

severity of impetigo and the detection of a skin pathogen.

Conclusions: S. pyogenes remains the principal pathogen in tropical impetigo;

the relatively high contribution of S. aureus as a co-pathogen has also been

confirmed. Children with scabies were more likely to have S. pyogenes detected.

While clearance of S. pyogenes is the key determinant of treatment efficacy, co-

infection with S. aureus warrants consideration of treatment options that are

effective against both pathogens where impetigo is severe and prevalent.

Trial Registration: This trial is registered; ACTRN12609000858291.

83

7.5 BACKGROUND

Impetigo is an epidermal infection caused by Staphylococcus aureus and

Streptococcus pyogenes. It is common in Indigenous children of northern

Australia, with prevalence as high as 70% (Currie and Carapetis, 2000). The

reported relative abundance of S. aureus and S. pyogenes has varied over time

(Koning et al., 2012). In recent decades, S. aureus and increasingly, methicillin-

resistant S. aureus (MRSA), has been the dominant reported pathogen in

impetigo studies worldwide, most of which have taken place in temperate-

climate regions, usually in affluent countries (Geria and Schwartz, 2010) where

there is a low burden of disease. By contrast, in tropical regions, impetigo is far

more common, and carries the greatest burden of sequelae (Carapetis et al.,

2005). S. pyogenes is assumed to have remained the dominant pathogen (Parks et

al., 2012) but some reports are emerging on skin and soft tissue infections caused

by S. aureus (Abdel Fattah and Darwish, 2012, Alvarez-Uria and Reddy, 2012).

There is limited microbiological surveillance of causative impetigo pathogens

from high burden contexts and rates of antimicrobial resistance are often

unknown (Parks et al., 2012). Impetigo is strongly associated with scabies

infestation in tropical environments (Steer et al., 2009), but the influence of

scabies on the microbiology of impetigo has not previously been described. We

report here on the microbiology of impetigo in a high-burden setting, and explore

the associations of this microbiology with age, sex, region, severity, presence of

scabies and nasal carriage of skin pathogens. Our dataset derives from a large,

non-inferiority randomised controlled trial (RCT) comparing trimethoprim-

sulphamethoxazole (SXT) with benzathine penicillin G (BPG) for the treatment

of impetigo in Indigenous children (Bowen et al., 2014).

7.6 METHODS

Study Design

Indigenous children aged 3 months to 13 years were participants in the RCT.

Children were eligible to participate on more than one occasion if at least 90 days

had elapsed since their last involvement. As such, 508 children from 12 remote

84

communities of the Northern Territory were enrolled for 663 episodes of

impetigo; all analysis presented here has been restricted to a child’s first episode

only. Six communities (comprising 463/508 children) were located in the tropical

climatic region, commonly referred to as the ‘Top End’. The remaining

communities were in Central Australia where a desert climate prevails. Children

were stratified by impetigo severity. The severe stratum included children with

≥2 purulent or crusted sores and ≥5 overall body sores.

Swabbing, transportation and culture methods

Each child had swabs taken from one or two sores (according to whether the

episode was classified as mild or severe) prior to commencing antibiotics. A

swab of the anterior nares was obtained to determine carriage of impetigo

pathogens in the context of infection. Swabs were collected between

26 November 2009 and 20 November 2012. Rayon tipped cotton swabs (Copan,

Italy) were transported at 4°C in 1 mL of skim milk tryptone glucose glycogen

broth (STGGB) and frozen at −70°C within 5 days of collection. Swabs were

defrosted, vortexed and an aliquot plated on horse blood agar and incubated for

48 hours at 37°C (Bowen et al., 2013). S. aureus and S. pyogenes were identified

morphologically and confirmed with latex agglutination. All S. aureus isolates

were staphytect (Oxoid, UK) and deoxyribonuclease (DNase, BD Diagnostics,

USA) positive. S. pyogenes agglutinated with group A Lancefield antisera

(Oxoid).

Antimicrobial susceptibility testing

Antimicrobial susceptibility testing for S. aureus was determined on the Vitek2

platform using 22359 VITEK AST-P612 cards (bioMerieux, France) with

Clinical and Laboratory and Standards Institute (CLSI, 2011) breakpoints utilised

(CLSI, 2011). MRSA was defined as any S. aureus with a positive cefoxitin

screen. Non-multidrug resistant MRSA (nmMRSA) was defined as MRSA

resistant to <3 additional non beta-lactam antibiotics (Tong et al., 2009).

Multidrug-resistant MRSA (mMRSA) was defined as MRSA that was resistant

to ≥3 non beta-lactam antibiotics (Tong et al., 2009).

85

We performed susceptibility testing for S. pyogenes with penicillin, erythromycin

and clindamycin discs using CLSI disc diffusion standards. SXT susceptibility

for S. pyogenes was determined with an E test® (bioMerieux) according to the

European committee on antimicrobial susceptibility testing (EUCAST) standards

(www.eucast.org, last accessed 15 November 2014). SXT susceptible strains had

a MIC ≤1mg/L and resistant isolates had a MIC >2mg/L.

Ethics statement

This study was approved by the Northern Territory Department of Health and

Menzies School of Health Research human research ethics committee (09/08).

Written informed consent was obtained from a child’s parent or guardian for all

study procedures.

Statistical analysis

Mixed-effects logistic regression (random effects accounting for correlated data

due to multiple sores for the children in the severe impetigo strata) using Stata 13

(Statacorp, Texas, USA) was performed to assess associations between the

growth of skin pathogens and the presence of scabies, severe impetigo, age, sex,

region and nasal carriage.

7.7 RESULTS

Baseline Microbiology Results

Sores

We obtained 872 swabs of sores from 508 children with untreated impetigo

(median age 7 years, interquartile range [IQR] 5–9 years) at baseline. Seventy-

two percent of children were in the severe impetigo stratum, all of whom had two

sores swabbed. An impetigo pathogen was identified in 488/508 (96%) of

children. From the 872 sores swabbed, S. aureus and S. pyogenes were identified

together in 503/872 (58%), S. pyogenes alone in 207/872 (24%), and S. aureus

alone in 81/872 (9%) of sores. Swabs from children in the severe impetigo

stratum were not more likely to detect one or both skin pathogens compared with

86

swabs from children in the mild impetigo stratum (Table 1). S. aureus was less

likely to be detected in older children (OR 0.6, 95% CI 0. –0.9 for children ≥ 5

years, Wald test on 2 degrees of freedom p = 0.04) or in children from Central

Australia (OR 0.5, 95% CI 0.3–0.9, p=0.02). Children from Central Australia

were less likely to have sores co-infected with both S. aureus and S. pyogenes

than children from the Top End (OR 0.5, 95% CI 0.3–0.9, p=0.01)

Table 1: Results from logistic regression models to assess associations between impetigo pathogens and age, sex, severity, presence of scabies and region.

Variable S. pyogenes in sores S. aureus in sores Both S. aureus

and S. pyogenes

in sores

MRSA in sores

positive for S.

aureus

OR 95% CI OR 95% CI OR 95% CI OR 95% CI

Female 1.3 0.8–2.1 1.0 0.7–1.4 1.1 0.8–1.5 1.0 0.6–1.7

0–4 years 1 (ref) 1 (ref) 1 (ref) 1 (ref)

5–9 years 1.1 0.7–2.0 0.6 0.4–0.9 0.8 0.5–1.1 0.9 0.5–1.6

10–13 years 1.2 0.6–2.4 0.5 0.3–0.9 0.7 0.4–1.2 0.7 0.3–1.5

Severe strata 1.4 0.8–2.6 0.7 0.4–1.1 1.1 0.7–1.7 0.6 0.3–1.1

Scabies present 2.2 1.1–4.4 0.8 0.5–1.3 0.9 0.6–1.4 1.4 0.7–2.6

Central

Australia

1.2 0.5–2.8 0.5 0.3–0.9 0.5 0.3–0.9 1.1 0.4–2.7

Scabies

Scabies was diagnosed at baseline in 84/508 (17%) of children: by age group: 0-

4 years (28/136, 21%); 5–9 years (40/271, 15%) and 10–13 years (16/101, 16%).

Those with scabies present were more likely to also have S. pyogenes detected

from sores (OR 2.2, 95% CI 1.1–4.4, p = 0.03) (Table 1).

87

Other beta haemolytic streptococci

Seven hundred and fifty-four beta haemolytic streptococci were cultured from

skin and nose swabs of 508 children with impetigo. Of these, 740/754 (98%)

were S. pyogenes with 710 from skin swabs and 30 from the nose swabs. The

remaining beta haemolytic streptococci were group C (2 skin, 2 nose) and group

G (7 skin, 3 nose).

Antibiotic susceptibility

One isolate each of S. pyogenes and S. aureus was selected per child for

reporting of the antibiotic susceptibility profile. Where possible, isolates cultured

from skin sores were selected, with the remainder coming from swabs of the

anterior nares.

S. pyogenes

There were 455 children with at least one S. pyogenes isolate available for

antibiotic susceptibility assessment. All S. pyogenes isolates were susceptible to

penicillin and erythromycin. Clindamycin resistance was detected in 9/455 (2%)

S. pyogenes isolates. SXT resistance was detected in 4/455 (<1%) isolates at

baseline with a MIC >2mg/L. (Bowen et al., 2012) The median SXT MIC for S.

pyogenes was 0.094 (interquartile range: 0.094–0.125) mg/L.

S. aureus

There were 435 children with at least one S. aureus isolate available for

antibiotic susceptibility assessment. Methicillin-resistance in S. aureus was

detected in 65/435 (15%) isolates, all of which were nmMRSA. There was no

detection of mMRSA. The respective resistance rates for other antibiotics by

child were penicillin 413/435 (95%), SXT 3/435 (<1%), erythromycin 60/435

(14%) and fusidic acid 9/435 (2%). Inducible clindamycin resistance was

reported in 60/435 (14%), 86% of which were MSSA. All 3 SXT-resistant

isolates were nmMRSA. Other antibiotics included on the VITEK card for

S. aureus susceptibility testing all had resistance rates below 0.02%. For those

88

with S. aureus detected, the presence of MRSA did not reach significance for

sex, age group, severe strata, the presence of scabies or region (table 1).

Nasal carriage of skin pathogens

S. aureus

A nasal swab was taken at baseline for 504/508 (99%) of children. Before

treatment, 91/504 (18%) children had confirmed carriage of a skin pathogen in

the anterior nares. Both S. aureus and S. pyogenes were recovered from 16/91

(18%), S. pyogenes alone from 14/91 (15%) and S. aureus alone from 61/91

(67%). Of children with nasal carriage of S. aureus, 66/77 (86%) had MSSA and

11/77 (14%), MRSA. Nasal carriage of S. aureus and S. pyogenes was

respectively 77/504 (15%) and 30/504 (6%).

We investigated the association of nasal carriage of S. aureus and the presence of

S. aureus in impetigo lesions. There were 504 children with both skin and nose

swabs available (Table 2). Surprisingly, of the 410 children culturing S. aureus

on the skin, only 54 (13%) also harboured S. aureus in the nose. Based on the

antibiogram, of the 54 with S. aureus in both the nose and skin sores, four had

discordant S. aureus strains (i.e., MSSA at one site and MRSA at the other). Of

the 94 children without S. aureus on the skin, there were 23 (24%) children who

harboured S. aureus in the nose. Of 424 children harbouring S. pyogenes on the

skin, 28 (7%) had concurrent growth of S. pyogenes from the anterior nares.

There were 2 children with isolated growth of S. pyogenes from the anterior

nares. Extending a model of Table 2, nasal colonisation with S. aureus was

associated with less, not more, S. aureus on the skin (OR 0.6, 95% CI 0.4–1.0,

p = 0.04), suggesting a separate epidemiology of S. aureus at these two sites.

89

Table2: Identification of Staphylococcus aureus from any impetigo lesion and the anterior nares for all children with at least one skin and nose swab available (n=504 children)

Anterior Nares Total

Positive Negative

Impetigo At least one sore

positive

54 (13%) 356 (87%) 410 (100%)

Negative 23 (24%) 71 (76%) 94 (100%)

Total 77 (15%) 427 (85%) 504 (100%)

7.8 DISCUSSION

S. pyogenes remains the key impetigo pathogen in Indigenous children of remote

Australia. Co-infection with S. aureus is also highly prevalent. In addition, our

study extends understanding of the microbiology of impetigo where scabies is

endemic. Where scabies is present, we found that S. pyogenes is more likely to

be recovered from impetigo lesions. We also found no positive correlation

between nasal carriage of S. aureus and recovery of S. aureus from impetigo

lesions, and no association between the severity of impetigo and recovery of

either S. pyogenes, S. aureus or both.

The stronger association of S. pyogenes (than S. aureus) with scabies presence

concurs with earlier work that linked scabies outbreaks with subsequent

epidemics of post-streptococcal glomerulonephritis (Svartman et al., 1972) and

recommended treatment of scabies at a community level to reduce the

complications of streptococcal pyoderma (Taplin et al., 1991, Carapetis et al.,

1997). Scabies association with impetigo in urban Indigenous children was

similarly high to that found in our study at a rate of 25/111 (23%) (Valery et al.,

2008). No prior studies have reported on the association between scabies and

MRSA detection. While the rates of both were high, we were unable to detect a

significant association between these.

90

Both S. pyogenes and S. aureus have been reported as key impetigo pathogens,

however the reported relative contributions of each have fluctuated over time and

region. Previous microbiology studies of impetigo in both urban and remote

Australian Indigenous children have shown high rates of co-infection (Valery et

al., 2008, McDonald et al., 2006b). Valery et al detected co-infection with skin

pathogens in 54% of urban Indigenous children. Of those children who had both

impetigo and scabies, S. pyogenes was recovered from 82% and S. aureus from

77% (Valery et al., 2008). S. pyogenes remains an important skin pathogen in our

region, as has also been reported from Fiji (Steer et al., 2009, Jenney et al., 2014)

and supports the ongoing prescription of antibiotics active against S. pyogenes

for the effective treatment of impetigo in tropical regions. Results of our clinical

trial confirmed the importance of clearing S. pyogenes in order to achieve healing

of impetigo lesions (Bowen et al., 2014).

The isolated dominance of S. aureus and rising rates of MRSA reported

elsewhere in skin and soft tissue infections (Abdel Fattah and Darwish, 2012,

Iovino et al., 2011, Phakade et al., 2012) were not found in our study.

Methicillin-resistance was detected in 15% of S. aureus isolates. This is lower

than previously reported for our remote tropical context (McDonald et al., 2006a)

and lower than recent rates found in a 20 year survey of S. aureus in our region

(Tong et al., 2014) but consistent with a study of impetigo in urban Indigenous

children of northern Australia (Valery et al., 2008). We were uncertain whether

MRSA varies by age, as this had not previously been explored, but we found no

evidence that it does. The observation that MRSA rates were stable across age

groups suggests early colonisation of children in remote communities with

MRSA strains.

We found that children in the severe strata, who had more than five purulent or

crusted sores, were not more likely than children in the mild strata to be infected

with S. pyogenes and/or S. aureus. Although those stratified as mild, indicating

fewer overall sores, may have had a less severe phenotype for each individual

sore, the microbiology of these sores was no different. We thus provide robust

evidence to refute earlier observational data in patients with impetigo where the

91

association of co-infection with a more severe phenotype was suspected but not

confirmed (Barrow, 1955, Hughes and Wan, 1967).

Whilst throat carriage of S. pyogenes is commonly reported and is not thought to

be the reservoir for skin infection, the anterior nares are thought to harbour S.

pyogenes only rarely. (Masters et al., 1958) As such, throat swabs were not

included in our study design as we focused on the characterisation of nasal

carriage of S. aureus. However, our findings of nasal carriage of S. pyogenes in

7% of children provide external validation of results from military recruits where

nasal carriage of S. pyogenes was found in 8% of those with ecthyma (Wasserzug

et al., 2009). There was no correlation between S. aureus nasal carriage and the

recovery of S. aureus from impetigo lesions. Previous studies have concluded

that when impetigo predominantly affects the lower limbs, there is either an

absence of S. aureus nasal carriage or a different genotype (Barth, 1987). Our

data supports this observation in that 67% of impetigo episodes involved only the

lower limbs (Bowen et al., 2014). As has previously been shown in military

recruits (Ellis et al., 2007), nasal decolonisation is also unlikely to be a useful

strategy in reducing the burden of impetigo in our tropical, endemic setting.

One study from Ghana in the 1970s with a similar tropical climate found

impetigo to be dominated by Lancefield groups C and G streptococci (Belcher et

al., 1975). These findings were not reproduced in our study and have not been

confirmed in other published microbiology studies from our region (McDonald et

al., 2006b) or other tropical contexts (Lawrence et al., 1979). Non-group A

streptococci do not appear to play a significant role in the pathogenesis of

impetigo.

A strength of this study is the standardisation of procedures for screening,

swabbing and microbiological culture in the context of a clinical trial conducted

according to International Conference on Harmonisation Good Clinical Practice

guidelines. In addition, the large number of children recruited from 12 distinct

communities from two regions of the Northern Territory suggests that the

conclusions are rigorous. Genotypic assessment of the isolates was not conducted

to determine the correlation of skin and nose isolates. The inability to correlate

the molecular epidemiology with the phenotype is a limitation of this study.

92

While the cost of molecular typing overall has become more affordable, the large

number of isolates in this study made further molecular analysis cost prohibitive.

7.9 CONCLUSIONS

S. pyogenes remains a key pathogen in impetigo in tropical contexts, despite the

rise in S. aureus found in many industrialised and tropical settings. Our findings

are in keeping with those reported from Fiji (Steer et al., 2009) and confirm that

impetigo in tropical contexts continues to be driven by S. pyogenes. In the

absence of local microbiology for impetigo, treatment algorithms should remain

focused on the treatment of S. pyogenes. However, we have demonstrated that

co-infection with both S. pyogenes and S. aureus is likely. As such, consideration

of treatment of impetigo with an agent that is effective against both S. pyogenes

and S. aureus is important. We have also concluded that in the context of

impetigo, there is no association between nasal colonisation and skin infection

with S. aureus.

Competing Interests

The authors declare that they have no competing interests.

Authors Contributions

AB, ST and JC participated in the design of the study. AB analysed the data and

drafted the manuscript with assistance from ST and MC. All authors read and

approved the final version of the manuscript.

Acknowledgements

This work was supported through a National Health and Medical Research

Council of Australia (NHMRC) project grant (545234) on which AB, ST, and JC

are all investigators. The funder had no role in the design, collection, analysis or

interpretation of the data, writing of the manuscript or decision to submit the

manuscript for publication. AB is the recipient of a NHMRC scholarship for PhD

research (605845) as well as an Australian Academy of Sciences Douglas and

93

Lola Douglas scholarship. ST is the recipient of an NHMRC Career

Development Fellowship (1065736).

We acknowledge the children and families who contributed to this trial in the

hope of finding a better treatment option for their sores, and the staff of the

Menzies School of Health Research and Royal Darwin Hospital microbiology

laboratory who contributed to data collection, culturing and antibiotic

susceptibility testing of all isolates. We acknowledge the co-investigators on this

study, Bart Currie, Ross Andrews and Malcolm McDonald. Malcolm McDonald

and Rob Baird provided feedback on the manuscript in development.

94

CHAPTER 8CONCLUSIONS


In this final chapter, the overall findings of the thesis are presented, conclusions

drawn and recommendations for future research outlined. I have reviewed the

global prevalence of impetigo and pyoderma using available population-based

studies. This has extended our understanding of the prevalence of impetigo at a

community level and, once published, should enhance the global understanding

of this high burden, but relatively unstudied, problem in the world’s poor. The

review also highlighted the ongoing problem that, despite living in a high-

income, developed economy, Australia’s Indigenous children have the highest

published rates of impetigo in the world.

There has been a long-standing dogma that trimethoprim-sulphamethoxazole

(SXT) is not active or effective in the treatment of Streptococcus pyogenes

infections. It was therefore important to re-look at the microbiology from which

this dogma arose, and use standardised susceptibility testing to confirm in vitro

susceptibility of S. pyogenes to (SXT). Chapter 3 describes this work and

confirms that S. pyogenes isolates from the NT are susceptible to SXT. The

accompanying literature review in Chapter 3 outlines the variability worldwide in

susceptibility of S. pyogenes to SXT, ranging from 0 to 100%. Much of this

variability can be attributed to the presence of thymidine, in media used for

susceptibility testing, acting as an inhibitor to SXT and thus appearing resistant.

The standardisation of thymidine content at a very low level, or addition of lysed

horse blood to susceptibility testing media, are two strategies that have been used

to overcome this appearance of resistance. This work was needed to better

understand the results of our clinical trial, and to legitimise the use of SXT as an

effective agent in the treatment of impetigo. Likewise, ongoing monitoring of

S. pyogenes susceptibility to SXT is needed, as this antibiotic is more widely

prescribed for impetigo.

95

In addition, we addressed methodological issues that were critical to the

successful operation of the Skin Sore Trial in a very remote context. We

compared a cheap, simple and effective transport medium that is likely to be

cost-effective and efficient in resource-limited contexts for future impetigo

research, to a known standard. Previous guidelines have recommended

immediate plating of swabs. (Kotloff, 2008) This approach has microbiological

merit, but is difficult to achieve in remote settings with long distances to the

laboratory. As such, the microbiological description of impetigo where the

burden is highest is under-reported. The ability to perform high-quality

microbiology studies by transporting stored swabs back to a central laboratory

for later processing builds our understanding of the local microbiology and

antibiogram in endemic settings and, thus, makes uncovering the molecular

epidemiology achievable.

Digital images are useful for capturing and storing an endpoint when it is not

feasible to transport an expert to the point of care. These rigorous, standardised

methods for scoring digital images will be useful in future impetigo and other

skin disease trials.

Pleasingly, our in vivo results confirmed the in vitro finding of S. pyogenes being

susceptible to SXT, with the demonstration of non-inferiority of SXT to BPG for

the treatment of impetigo. In addition, clearance of S. pyogenes was equivalent

between these two antibiotics, with penicillin having known excellent efficacy

against S. pyogenes. No isolate of S. pyogenes has ever been reported that is

resistant to penicillin. The demonstration of equivalent clearance with both

antibiotics upholds the efficacy of SXT for S. pyogenes.

Surprisingly, in the MRSA era, our study confirmed the historical observations

made over four decades ago that S. pyogenes is the dominant pathogen driving

impetigo and that S. aureus is in fact a co-coloniser. We have shown that

benzathine penicillin G (BPG), an antibiotic with limited S. aureus activity but

very active against S. pyogenes, was effective in the treatment of impetigo, even

in the presence of high rates of S. aureus including MRSA. As such, MRSA-

targeted treatment for impetigo is not essential, but using an MRSA-active

antibiotic such as SXT may have the additional benefit of reducing S. aureus and

96

MRSA skin carriage in a region where the incidence of sequelae of both gram-

positive pathogens is high. The increased use of SXT for treatment of impetigo

may see a reduction in the consequences of both pathogens on the short- and

long-term morbidity in Indigenous children.

Many questions remain and future studies, described below, are underway that

will continue to illuminate our understanding of this ubiquitous problem of

Indigenous childhood. The ultimate goal is that sores are not ignored, and are

effectively treated, with the outcome of this being a dramatic decline in the

prevalence of impetigo such that inter-personal transmission is no longer self-

sustaining.

8.2 DISCUSSION OF MAIN FINDINGS

8.2.1 Trimethoprim/sulphamethoxazole is effective against Streptococcus pyogenes in vitro

The current recommendation by the Infectious Diseases Society of America

(IDSA) when using SXT for treatment of skin and soft tissue infections involving

both S. pyogenes and S. aureus, is to include a second agent that is active against

S. pyogenes e.g. penicillin, cephalexin or amoxicillin. (Stevens et al., 2014) Our

findings dispute the need for a second antibiotic and are important. The use of a

single antibiotic when appropriate, rather than combination therapy, is a key

principle in antimicrobial stewardship. (SHEA et al., 2012) It is also a critical

factor in adherence to an oral treatment regimen, where the simplest, shortest

regimen, that is directly associated with symptoms, has the best correlation with

adherence. (Perez-Gorricho and Ripoll, 2003)

8.2.2 Trimethoprim/sulphamethoxazole is non-inferior to benzathine penicillin G for impetigo

The work in Chapter 3 paved the way for a complete understanding of the results

of our trial outlined in Chapter 6. We concluded, using the robust digital image

methodology described in Chapter 5, that SXT is non-inferior to BPG for the

treatment of skin sores. In both treatment arms, treatment success was achieved

in 85% of participants, absolute difference 0.5% (95% CI -6.2 to +7.3%).

97

(Bowen et al., 2014) We set the non-inferiority margin at 10%, to detect a

clinically relevant difference, if one existed. There were two SXT dosing arms in

the study, one treating children twice a day for three days (six doses) and the

other once a day for five days (five doses). The regimens, pooled together and

also analysed separately, both reached significance with respect to the non-

inferiority margin and as such either can be recommended with confidence.

This provides a palatable, effective, short course, non-painful alternative

treatment to intramuscular BPG for children with impetigo in remote Indigenous

communities. The challenge now is to see these findings implemented into

practice. As discussed in Chapter 1, the CARPA manual is well utilised

throughout our region as the standard treatment manual. The 6th edition has

recently been published (October 2014) and the results of our study have been

included. (CARPA, 2014) The results have also been included in the national

antibiotic reference, Therapeutic Guidelines: Antibiotic, released in November

2014 (Antibiotic Expert Group, 2014) and in treatment guidelines in the

Kimberley, the northern region of Western Australia with a large, remote,

Indigenous population. SXT now sits in regional and national guidelines as an

effective treatment for impetigo, based on the results of our trial.

Our study also confirmed, for the first time in our region, that BPG remains an

effective treatment for impetigo. This is important, as there are circumstances

where BPG will remain the preferred treatment for impetigo. Clinic nurses and

clinicians in the region are familiar with prescribing and administering BPG, and

the simplicity of a single dose without the need to address adherence, has appeal.

This was the feedback received when our results were presented in health-care

settings in remote communities. As such, BPG remains as first-line therapy in

both the guidelines mentioned above, with the introduction of SXT as an

alternate, evidence-based option. Future evaluation of the uptake of both

antibiotics will be needed to understand the full impact of our study on

community-based health care.

98

8.2.3 Trimethoprim/sulphamethoxazole is effective against S. pyogenes in vivo

Supporting our findings of in vitro efficacy outlined in Chapter 3, are the in vivo

results reported in Chapter 6. As outlined above, SXT is an effective antibiotic in

the treatment of impetigo caused by S. pyogenes. To be confident of any results

reported from our trial (and allay concerns that adherence would be poor), all

doses of antibiotics were directly observed. More than 96% of children had all

doses of study drug observed. This strengthens our findings. Likewise, we have

shown that S. pyogenes is cleared from infected sores with either BPG or SXT.

Culture positive rates before treatment were above 85% for S. pyogenes and by

day 7, S. pyogenes could be detected in less than 7% of sores in any of the three

treatment arms. (Bowen et al., 2014) The clearance of S. pyogenes from impetigo

lesions with BPG treatment is well known from earlier studies. (Ferrieri et al.,

1973) The declining detection of S. pyogenes on days 2 and 7 with both

antibiotics, with no difference in the rate of decline between BPG and SXT,

provides supportive data to confirm the in vivo efficacy of SXT.

8.2.4 S. pyogenes remains the driver of impetigo even in the MRSA era

We have confirmed that impetigo is driven by S. pyogenes, with clearance of

S. pyogenes being the only factor associated with sore healing in a step-wise

backwards logistic regression (OR 5.2, 95% CI 1.8–14.1). However, S. aureus is

frequently found in association with S. pyogenes as reported in Chapter 7. Whilst

severe outcomes can ensue if S. aureus is untreated, the focus for the interruption

of transmission of impetigo is on S. pyogenes. Our trial was designed mindful of

the rising prevalence of MRSA in both the hospital and community in the NT.

Our results suggest the significance of MRSA, as an impetigo driver, is lower

than previously thought.

8.2.5 Trimethoprim/sulphamethoxazole has fewer side effects than benzathine penicillin G

Our study confirmed very few side effects from SXT. In the 343 children treated

with SXT, we documented one case of urticarial rash without mucosal

99

involvement that resolved within 24 hours. Three older children were unable to

tolerate the large volumes of SXT syrup administered using weight-based dosing

(40ml per day). When switched to tablets, they continued in the trial without

further problems. One child had persistent vomiting and was discontinued. SXT

has been identified in the global literature as an antibiotic with a concerning side-

effect profile, (Goldman et al., 2013) and we were closely monitoring for these

events. We did not see any rash heralding the potentially life-threatening Stevens

Johnson syndrome. However the study was not powered to detect uncommon

side effects. Given the high prevalence of impetigo in remote communities in our

region, ongoing education about and monitoring for this rare, but life-

threatening, side effect will be needed.

The high rate of adverse events in the BPG arm of the trial was somewhat of a

surprise. Whilst anecdotal reports abound of the painful intramuscular injection

being a deterrent to treatment, the high rate of reported pain at 48 hours –

affecting one in three children randomised to the BPG arm – is concerning.

Likewise, treatment refusal by four children randomised to the BPG arm

confirms the impression that the painful needle is a deterrent to treatment. It is

possible that the 22 children who declined to participate in the trial did so for

similar reasons. Early experience of the painful intramuscular injection for sores

and a recurring need for treatment, with a median of three visits per year in

infants, (Kearns et al., 2013) in addition to the other childhood needles for

immunisation, might have conditioned children to these responses. By one year

of age, almost all infants in five remote communities had documented skin sores.

(Kearns et al., 2013) Having an oral regimen that circumvents the painful

injection is important to preserve intramuscular antibiotic injections for

circumstances where there is currently no alternative e.g. rheumatic heart disease

secondary prophylaxis. (Wyber, 2013)

The early studies that introduced BPG to the arena of treatment for impetigo

were in the context of acute post-streptococcal glomerulonephritis (APSGN)

outbreaks, where treatment that was easy to deliver and long acting was a

priority. (Ferrieri et al., 1974) APSGN epidemics are also problematic in the NT,

occurring every 5–7 years (Marshall et al., 2011). In the context of the high

100

burden of impetigo in the NT, and a limited evidence base for the treatment of

extensive impetigo, treatment that was effective during and possibly effective

between APSGN epidemics was adopted. Our study has confirmed that BPG is

an effective antibiotic for the treatment of impetigo, and confirmed the anecdotal

reports that the injection is a painful limitation to the use of this antibiotic. Where

an alternative, effective antibiotic exists with fewer adverse events, we would

recommend it in preference with antimicrobial resistance surveillance in place.

However, there will remain circumstances where health care providers or parents

prefer to administer BPG, and our study has confirmed that this remains an

effective option.

8.2.6 Australian Indigenous children continue to have the highest reported rates of impetigo in the world

Chapter 2 summarises the prevalence of impetigo globally. We searched for

population-based studies from anywhere in the world, however all available

studies were conducted in resource-poor countries or contexts. This reinforces

the connection between poverty and high rates of impetigo in children. In this

context, it is unacceptable that Australian Indigenous children have the highest

reported prevalence of impetigo in the world. Australia is a high-income country

and must do more to reduce this burden of disease.

In the NT, epidemiological surveys of bacterial skin infection and scabies have

been occurring for more than 20 years, with no real change in the prevalence.

This imposes a considerable, ongoing burden on primary health care (Kearns et

al., 2013, McMeniman et al., 2011) and we have estimated almost 16,000

children are in need of treatment in remote communities at any one time. Geo-

political borders do not restrict impetigo in remote Indigenous children; the

burden of disease is equally high in Western Australia and Queensland. The next

steps will involve developing a coordinated strategy for recognising and treating

skin infections across these jurisdictions, with routine evaluations.

Clearly, effective treatment that is simple, palatable and has few adverse events

is needed in our region. Communicating these findings is well underway and

translation into regional guidelines has occurred. The responsibility for future

101

research lies in continuing to evaluate both the prevalence and utilisation of this

new regimen, with the hope that this population will not be world leaders for

impetigo prevalence in the future.

8.2.7 Methodologies to take impetigo research to high burden contexts

Previous impetigo trials have mostly been conducted in physician offices or

hospital outpatient settings. (Koning et al., 2012) In order to conduct this trial in

the remote communities of Australia, several methods were developed and

published. This contribution to the global literature on the methodologies for

impetigo research reported in Chapters 4 and 5 will hopefully facilitate adoption

of treatment studies in similar contexts, where the burden of disease is the

highest. We intentionally published these papers in journals that are more likely

to come to the attention of researchers in these contexts; either open-access or

tropical medicine journals. In the digital era, experts are no longer needed at the

point of care for assessing end-points, and could easily be located in Darwin,

New York or Bamako to confirm findings of studies using digital images. We

have developed a method for collecting and scoring digital images. The next step

is to validate this method. As most impetigo trials to date have been conducted in

hospital outpatient or office-based settings where experts are readily available for

end-point assessment, it may be possible to validate this method in a future

impetigo trial. This would involve comparing the point-of-care expert assessment

with the remote review of digital images by a similarly qualified expert. All

reviewers would remain blinded to the prescribed treatment.

Likewise, the transport of skin swabs from remote contexts to a central

laboratory benefits from the use of a transport medium. We used one that was

familiar in our laboratory, and validated it against the commercially-available

Amies transport medium. The benefit of STGGB as a transport medium is the

ability to freeze for longer periods in order to facilitate batching of specimens

and preserve bacteria for future molecular work.

102

8.3 STUDY LIMITATIONS

8.3.1 Antiseptic or topical body washes?

In designing our trial, serious consideration was given to having a non-antibiotic

arm, one that used only topical antiseptic. Partly because of concerns around the

safety of not treating bacterial skin infection with antibiotics, but mainly to keep

our sample size feasible and our findings simple and easily interpretable, this arm

was not incorporated into the study design. Hygiene is very important in

managing impetigo as was shown in Mozambique, where access to water

influenced the burden of disease. (Cairncross and Cliff, 1987) In addition, hand-

washing studies have shown a significant decline in impetigo using soap and

water. (Luby et al., 2002, Luby et al., 2005) However, it is still unclear whether

antiseptics or topical body washes would add considerable value as an adjunct to

hygiene measures and treatment, (Koning et al., 2012) but these are unlikely to

cause harm.

Neither were traditional, Indigenous methods for controlling skin disease

evaluated in our trial. Bush medicine is still practised and many grandmothers

and elders asked questions about these treatments when we visited their

communities. Incorporating both worldviews into a regional treatment strategy to

reduce the burden of skin disease will be an important step to consider in

engaging the community.

8.3.2 Placebo control

Our trial lacked a placebo arm. It was not considered ethical with the high

disease burden and complications of impetigo seen in the NT. Few placebo-

controlled trials for impetigo exist, (Koning et al., 2012) and we deemed that the

best design was a non-inferiority study, comparing the new treatment with the

standard of care.

8.3.3 Adherence concerns

There has always been uncertainty about adherence to oral antibiotics and other

prescribed medicines in Indigenous Australians. (Burns, 1992, Kemp K, 1994,

103

McConnel, 2003) This is a challenge when prescribing treatment for any patient,

but has an added complexity for Indigenous patients, due to language and

worldview differences between prescriber and patient. (McConnel, 2003) The

known strategies for achieving treatment adherence such as prescribing a simple,

short regimen, temporally linked to the disease process, are the most likely to be

successful. (Perez-Gorricho and Ripoll, 2003) In childhood, palatability of syrup

formulations is of equal importance. (Liu et al., 2014) These principles guided

the selection of an oral regimen to compare with an intramuscular injection. As

such, both oral regimens trialled were simple, palatable, short and easy to

administer. In our study, directly observed treatment was incorporated to

reinforce the message that adherence is important and to ensure our results were

valid.

The once daily regimen has the benefit of being simple enough to administer via

a directly observed approach at the school or health clinic in the remote

community, if there are ongoing concerns about treatment adherence. This would

be possible with a five-day course that coincides with the standard working week

when both the clinic and school would be open, but may not be feasible or

acceptable to families. This highlights an avenue for future qualitative research to

understand whether uptake of these strategies is occurring and if not, what

obstacles need to be overcome.

SXT is stable at room temperature (20–25ºC), but temperatures well above this

occur in most remote communities of the NT on a daily basis and where access

to a refrigerator is not universal, concerns abound about the efficacy of oral

treatment. SXT was stored in air-conditioned environments monitored by the

research assistants throughout the trial, and as such further work is needed to

understand the impact this may have on treatment efficacy.

8.3.4 Poverty and overcrowding

Whilst poverty, racism, household overcrowding and social disadvantage exist, it

is unclear whether a simple treatment regimen will on its own improve skin

disease. These primordial factors need to be addressed in an ongoing program of

104

reducing the gap in health outcomes between Indigenous and non-Indigenous

Australians.

8.3.5 Antibiotic resistance

Recent work in synthesizing the antibiotic resistance patterns of S. aureus from

swabs collected in remote communities across the NT has shown a concerning

trend. During the period under investigation (1993–2012), more than 20,000

S. aureus isolates were collected. The more recent years showed a doubling in

the SXT resistance, albeit from a very low baseline value. (Tong et al., 2014b)

This requires ongoing evaluation, and studies are being developed to validate

both pharmaceutical prescriptions and changes in resistance phenotypes as

guidelines incorporate the SXT regimen into the standard of care.

8.4 FUTURE WORK

8.4.1 Does the presence of scabies impact treatment efficacy?

In Chapter 7 we showed that the presence of scabies in children with impetigo

was associated with an increased likelihood of detecting S. pyogenes from sores.

As both regimens trialed are active against S. pyogenes, we hypothesised that

treatment efficacy when analysed by the presence or absence of scabies would be

similar. Preliminary analysis suggests the opposite, in that children with scabies

co-infection had fewer treatment successes scored than children without scabies.

Surprisingly, this also revealed that the absolute difference in treatment success

for impetigo-scabies with BPG was 18.4% (95% CI -1 to 38%), whereas with

SXT it was only 7% (-4 to 19%). The lower overall treatment success for

impetigo-scabies and smaller difference between groups with and without

scabies, treated with SXT (rather than BPG), warrants further investigation. It is

possible that SXT might have added benefits in the treatment of impetigo when

scabies is also present. Current animal models could be used to determine

whether activity of SXT against scabies might explain some of these findings.

(Holt and Fischer, 2013)

105

8.4.2 Whole genome sequencing to detect resistance mechanisms

Antibiotic resistance monitoring is an essential component of further research to

understand the changes in microbiology that occur with increasing antibiotic

pressure. However, we currently do not know the resistance mechanisms for

either organism against SXT for isolates from our region. There are several

possible molecular techniques that can be used to determine resistance

mechanisms. Of these, we assessed whole genome sequencing (WGS) to be the

most cost effective and efficient. In addition, the bioinformatics computer

platform and collaborators with experience in detecting resistance mechanisms

are also based at Menzies. We plan to use WGS to illuminate the possible

genotypic mechanisms of SXT resistance in skin sore isolates that are

phenotypically resistant. All SXT-resistant S. pyogenes and S. aureus isolates

have been sub-cultured and DNA extracted, then submitted for WGS on the

Illumina HiSeq2000 platform by Macrogen in Korea. Resulting data will be used

to determine the mechanism of resistance in these isolates. This question has

importance, as it will impact on the likely selective pressure of introducing SXT

into routine treatment of skin sores in the NT, based on the results of our trial.

There were 8 (<0.5%) S. pyogenes isolates resistant to SXT detected in the

overall study. The S pyogenes resistance was identified using an SXT E test

according to the EUCAST methodology reported in Chapter 3. Factors that cause

trimethoprim resistance in S. pyogenes have recently been described in isolates

from India and Germany. (Bergmann et al., 2014) We will explore whether these

horizontally transferrable, trimethoprim-insensitive dihydrofolate reductase (dfr)

genes are also found in our SXT resistant S. pyogenes isolates using WGS.

Likewise, S. aureus resistance to SXT was at a low level in our study with only

24 (1%) S. aureus isolates resistant to SXT. The S. aureus resistance was

identified using VITEK2 platform, with CLSI breakpoints as reported in Chapter

7. The correlation between the resistance phenotype and genotype will guide

future assessments of antibiotic resistance.

106

8.4.3 Understanding the dynamics of skin sore transmission using WGS

The burden of impetigo suggests transmission of impetigo pathogens is common

in Indigenous communities. However, little work has been conducted on defining

the transmission dynamics in an endemic setting. We hypothesised that

transmission within a household is more common than elsewhere in the

community, in an endemic setting. If so, interventions to reduce household

crowding are most likely to be effective in reducing the burden of impetigo in

remote Indigenous communities. To explore this hypothesis, we have selected 54

isolates of S. pyogenes and S. aureus from 11 household clusters within the same

community on the same recruitment trip in November 2011 to sequence. As

such, we hope to understand how closely related these strains are and to build on

this by developing methods for inferring transmission. These models are needed

to determine the most appropriate interventions to reduce transmission, without

conducting expensive interventional studies. This work on transmission

dynamics can be built on the wealth of isolates, epidemiological data points and

outcomes of studies conducted in remote NT communities by Menzies

researchers over the past 20 years.

8.4.4 The niche for Staphylococcus argenteus?

Researchers at Menzies School of Health Research have recently described a

new lineage of Staphylococcus in the Northern Territory of Australia. (Tong et

al., 2013, Tong et al., 2014a) In light of this, we hypothesised that S. argenteus

might be better adapted to skin infection rather than nasal colonisation. To

explore this, multi-locus sequence typing (MLST) was performed on paired

samples of S. aureus from the nose and skin of participants from the Skin Sore

Trial. This hypothesis might explain why S. argenteus was more likely to be

found in community rather than hospital samples in previous work. Preliminary

work showed the rate of S. argenteus in our study samples to be very low, with

no significant difference between skin and nasal specimens.

107

8.4.5 Evaluations of uptake

Our study was designed to find a better treatment option for impetigo, based on

an assumption that pain may be a limitation in the acceptance of treatment for

skin sores. However, health-care workers in the NT express a strong preference

for the use of an injection for treatment of sores due to concerns about adherence.

In a recent education session with remote doctors, comments about the injection

ranged from “the mothers come wanting the needle as they don’t want to muck

around with syrup or tablets” to “some mums won’t come to the clinic because

they don’t want their kids to get the needle”. These assumptions have not been

explored with parents or children in remote communities. There is a gap in our

knowledge that affects the translation of this research into practice. To explore

the uptake of our findings, qualitative interviews and monitoring of pharmacy

dispensing records will be helpful. While children and parents do not decide the

prescribed treatment when they seek care, qualitative research will guide the

implementation of guidelines and development of policy from our results. Future

studies are needed to evaluate the acceptance of this new regimen by both

prescribers and families.

It was important for clinicians in the region to have data to confirm BPG is

effective for the treatment of impetigo, despite high rates of S. aureus and

MRSA. Concerns about non-adherence to a short course antibiotic may drive the

ongoing prescription of BPG. However, as was seen in our trial, children may

vote with their feet, and run away if and when the needle is mentioned. Oral SXT

now has a strong evidence base for its efficacy and can be recommended with

confidence.

8.4.6 Ongoing surveillance

The increased use of any antibiotic causes increasing selection pressure and may

drive resistance. Likewise, the introduction of a resistant clone could also

compromise this regimen. (Nickerson et al., 2011) Ongoing surveillance is

necessary, both to assess whether a more palatable treatment is able to reduce the

burden of impetigo and what happens to antibiotic resistance profiles when this is

widely prescribed. Surveillance for the development of antimicrobial resistance

108

in S. aureus and S. pyogenes to SXT will occur through routine microbiological

channels, using both hospital and community pathology specimens throughout

the region.

Our findings have highlighted the efficacy of SXT in S. pyogenes skin infections,

have outlined the complex history that resulted in the myth being perpetuated and

highlighted an available methodology with breakpoints for antibiotic

susceptibility testing to facilitate ongoing surveillance for the emergence of SXT

resistance.

Included in this will be ongoing surveillance of the incidence and prevalence of

APSGN and RHD over the long term to see whether there is any impact on these

severe complications of impetigo.

8.4.7 The ‘denormalisation’ of sores: A coordinated approach for recognition and treatment of skin disease in remote Indigenous children throughout Australia

Skin sores are common, rarely deemed serious enough to seek or obtain

treatment, highly transmissible and hence found at epidemic levels in many

remote Indigenous communities. The current paradigm is of skin infection being

normal rather than the exception. A paradigm shift is needed to achieve the de-

normalisation of skin sores. The next steps will involve developing a coordinated

strategy across jurisdictions for recognising and treating skin infections that is

routinely evaluated.

8.5 FINAL CONCLUSIONS

The studies described in this thesis have contributed to our understanding of the

burden, microbiology and efficacy of treatment of extensive impetigo. This fills a

gap in the impetigo literature on evidence-based treatments for extensive

impetigo. Our finding that impetigo is driven by S. pyogenes concurs with the

only other trial on the treatment of extensive impetigo, which subjectively

concluded that S. pyogenes was important. The methodology for conducting

impetigo studies in remote circumstances where disease burden is often the

highest has also been reported. This will provide a framework for future studies

109

on the treatment of impetigo for children who need it the most. As is evident in

the systematic review (Koning et al., 2012), there is no generally agreed upon

standard treatment for impetigo and as such, treatments that are appropriate for

different contexts should continue to be trialled, perhaps using some of the

methodologies we have reported.

Chapter 2 summarises the prevalence of impetigo in many resource-poor

communities throughout the world, and highlights the ongoing burden suffered

by remote Indigenous children of Australia. The association of impetigo with

poverty make these rates unacceptable in a high-income country like Australia.

Clearly, effective treatment that is simple, palatable and has few adverse events

is needed in our region. These findings have been communicated throughout the

region and translated into treatment guidelines. In support of this, ongoing health

education and health promotion activities are needed to maintain a focus on skin

disease. In conjunction, surveillance of changes in the prevalence, microbiology

and antibiotic resistance profile as utilisation of a new regimen increases are

needed. These studies are in development, with the hope that Indigenous children

will not be world leaders for impetigo prevalence in the future.

110

BIBLIOGRAPHY

ABDEL FATTAH, N. S. & DARWISH, Y. W. 2012. Antibiogram testing of

pediatric skin infections in the era of methicillin-resistant Staphylococci aureus:

an Egyptian University Hospital-based study. Int J Dermatol, 51, 1441-7.

ACCORSI, S., BARNABAS, G. A., FARESE, P., PADOVESE, V.,

TERRANOVA, M., RACALBUTO, V. & MORRONE, A. 2009. Skin disorders

and disease profile of poverty: analysis of medical records in Tigray, northern

Ethiopia, 2005-2007. Trans R Soc Trop Med Hyg, 103, 469-75.

AGRESTI, A. & MIN, Y. 2005. Simple improved confidence intervals for

comparing matched proportions. Stat Med, 24, 729-40.

AINSLIE, G. M. & REILLY, J. 2003. The use of linear scales in the photography

of skin lesions. J Audiov Media Med, 26, 15-22.

ALLEN, A. M. & TAPLIN, D. 1974. Skin infections in eastern Panama. Survey

of two representative communities. Am J Trop Med Hyg, 23, 950-6.

ALVAREZ-URIA, G. & REDDY, R. 2012. Prevalence and Antibiotic

Susceptibility of Community-Associated Methicillin-Resistant Staphylococcus

aureus in a Rural Area of India: Is MRSA Replacing Methicillin-Susceptible

Staphylococcus aureus in the Community? ISRN Dermatol, 2012, 248951.

ANAND, C., GORDON, R., SHAW, H., FONSECA, K. & OLSEN, M. 2000.

Pig and goat blood as substitutes for sheep blood in blood-supplemented agar

media. J Clin Microbiol, 38, 591-4.

ANDREWS, R. M., KEARNS, T., CONNORS, C., PARKER, C., CARVILLE,

K., CURRIE, B. J. & CARAPETIS, J. R. 2009. A regional initiative to reduce

skin infections amongst aboriginal children living in remote communities of the

Northern Territory, Australia. PLoS Negl Trop Dis, 3, e554.

ANTIBIOTIC EXPERT GROUP, 2010. Therapeutic guidelines: antibiotic, 14th

ed. Melbourne, Therapeutic Guidelines Limited.

111

ANTIBIOTIC EXPERT GROUP, 2014. Therapeutic guidelines: antibiotic, 15th

ed. Melbourne, Therapeutic Guidelines Limited.

BAR-ZEEV, S. J., KRUSKE, S. G., BARCLAY, L. M., BAR-ZEEV, N. &

KILDEA, S. V. 2013. Adherence to management guidelines for growth faltering

and anaemia in remote dwelling Australian Aboriginal infants and barriers to

health service delivery. BMC Health Serv Res, 13, 250.

BARROW, G. I. 1955. Clinical and bacteriological aspects of impetigo

contagiosa. J Hyg (Lond), 53, 495-508.

BARTH, J. H. 1987. Nasal carriage of staphylococci and streptococci. Int J

Dermatol, 26, 24-6.

BAUMGARTEN, M., MARGOLIS, D. J., SELEKOF, J. L., MOYE, N., JONES,

P. S. & SHARDELL, M. 2009. Validity of pressure ulcer diagnosis using digital

photography. Wound Repair Regen, 17, 287-90.

BECHELLI, L. M., HADDAD, N., PIMENTA, W. P., PAGNANO, P. M.,

MELCHIOR, E., JR., FREGNAN, R. C., ZANIN, L. C. & ARENAS, A. 1981.

Epidemiological survey of skin diseases in schoolchildren living in the Purus

Valley (Acre State, Amazonia, Brazil). Dermatologica, 163, 78-93.

BEHL, P. N. M. K. C. B. S. 1979. Ecological study of skin diseases in Delhi

area. Indian Journal of Dermatology, Venereology and Leprology, 50, 213 - 4.

BELCHER, D. W., AFOAKWA, S. N., OSEI-TUTU, E., WURAPA, F. K. &

OSEI, L. 1975. Letter: Non-group A streptococci in Ghanaian patients with

pyoderma. Lancet, 2, 1032.

BELCHER, D. W., AFOAKWA, S. N., OSEI-TUTU, E., WURAPA, F. K. &

OSEI, L. 1977. Endemic pyoderma in Ghana: a survey in rural villages. Trans R

Soc Trop Med Hyg, 71, 204-9.

BERGMANN, R., VAN DER LINDEN, M., CHHATWAL, G. S. & NITSCHE-

SCHMITZ, D. P. 2014. Factors That Cause Trimethoprim Resistance in

Streptococcus pyogenes. Antimicrob Agents Chemother, 58, 2281-8.

112

BERGNER-RABINOWITZ, S. & DAVIES, A. M. 1970. Sensitivity of

Streptococcus pyogenes types to tetracycline and other antibiotics. Isr J Med Sci,

6, 393-8.

BERNSTEIN, L. S. 1982. Combination of trimethoprim with sulfonamides other

than sulfamethoxazole. Rev Infect Dis, 4, 411-8.

BHALLA, K. K. 1984. Pattern of skin diseases in a semi-urban community in

Delhi area. Indian Journal of Dermatology, Venereology and Leprology, 50, 213

- 4.

BHATIA, V. 1997. Extent and pattern of paediatric dermatoses in rural areas of

central India. Indian J Dermatol Venereol Leprol, 63, 22-5.

BHAVSAR, B. S. & MEHTA, R. N. 1985. Pyoderma--an epidemiological study

through school survey in Surat District (South Gujarat). Indian J Public Health,

29, 233-7.

BINKS, M. J., TEMPLE, B., KIRKHAM, L. A., WIERTSEMA, S. P., DUNNE,

E. M., RICHMOND, P. C., MARSH, R. L., LEACH, A. J. & SMITH-

VAUGHAN, H. C. 2012. Molecular surveillance of true nontypeable

Haemophilus influenzae: an evaluation of PCR screening assays. PLoS One, 7,

e34083.

BISSEK, A. C., TABAH, E. N., KOUOTOU, E., SINI, V., YEPNJIO, F. N.,

NDITANCHOU, R., NCHUFOR, R. N., DEFO, D., DEMA, F., FONSAH, J. Y.,

NJAMNSHI, A. K. & MUNA, W. F. 2012. The spectrum of skin diseases in a

rural setting in Cameroon (sub-Saharan Africa). BMC Dermatol, 12, 7.

BOWEN, A. C., LILLIEBRIDGE, R. A., TONG, S. Y., BAIRD, R. W., WARD,

P., MCDONALD, M. I., CURRIE, B. J. & CARAPETIS, J. R. 2012. Is

Streptococcus pyogenes Resistant or Susceptible to Trimethoprim-

Sulfamethoxazole? J Clin Microbiol, 50, 4067-72.

113

BOWEN, A. C., TONG, S. Y., CHATFIELD, M. D., ANDREWS, R. M. &

CARAPETIS, J. R. 2013. Comparison of three methods for the recovery of skin

pathogens from impetigo swabs collected in a remote community of Northern

Territory, Australia. Trans R Soc Trop Med Hyg, 107, 384-9.

BOWEN, A. C., TONG, S. Y. C., ANDREWS, R. M., O'MEARA, I. M.,

MCDONALD, M. I., CHATFIELD, M. D., CURRIE, B. J. & CARAPETIS, J. R.

2014. Short-course oral co-trimoxazole versus intramuscular benzathine

benzylpenicillin for impetigo in a highly endemic region: an open-label,

randomised, controlled, non-inferiority trial. The Lancet, 384: 2132 - 2140.

BRAHMADATHAN, K. N. & KOSHI, G. 1988. Epidemiology of streptococcal

pyoderma in an orphanage community of a tropical country. J Trop Med Hyg, 91,

306-14.

BROEK, I., HARRIS, N., HENKENS, M., MEKAOUI, H., PALMA, P. P.,

SZUMILIN, E. & GROUZARD, V. 2013. Clinical Guidelines diagnosis and

treatment manual for curative programmes in hospitals and dispensaries, Paris,

Medecins Sans Frontieres - Guidelines.

BROOK, I., FRAZIER, E. H. & YEAGER, J. K. 1997. Microbiology of

nonbullous impetigo. Pediatr Dermatol, 14, 192-5.

BRUMFITT, W. H.-M., JMT. KOSMIDIS, J. 1973. Trimethoprim-

Sulfamethoxazole: The present position. The Journal of Infectious Diseases, 128,

S778-S791.

BUREAU OF METEREOLOGY, 2011. Point Fawcett (Bathurst Island)

November 2011 Daily Weather Observations [Online]. Darwin: Bureau of

Meteorology. http://www.bom.gov.au/climate/.

BURNS, C. 1992. A Pilot Study on compliance in Aboriginal paediatric patients

receiving oral antibiotic medication. Australian Journal of Hospital Pharmacy,

22, 217-221.

BUSHBY, S. R. 1973a. Sensitivity testing with trimethoprim-

sulphamethoxazole. Med J Aust, 1, Suppl:10-8.

114

BUSHBY, S. R. 1973b. Trimethoprim-sulfamethoxazole: in vitro

microbiological aspects. J Infect Dis, 128, Suppl:442-62 p.

CAIRNCROSS, S. & CLIFF, J. L. 1987. Water use and health in Mueda,

Mozambique. Trans R Soc Trop Med Hyg, 81, 51-4.

CARAPETIS, J., GARDINER, D., CURRIE, B. & MATHEWS, J. D. 1995a.

Multiple strains of Streptococcus pyogenes in skin sores of aboriginal

Australians. J Clin Microbiol, 33, 1471-2.

CARAPETIS, J. R. 1998. Ending the heartache: the epidemiology and control of

acute rheumatic fever and rheumatic heart disease in the Top End of the Northern

Territory [PhD Thesis]. Sydney, University of Sydney.

CARAPETIS, J. R., CONNORS, C., YARMIRR, D., KRAUSE, V. & CURRIE,

B. J. 1997. Success of a scabies control program in an Australian aboriginal

community. Pediatr Infect Dis J, 16, 494-9.

CARAPETIS, J. R., JOHNSTON, F., NADJAMERREK, J. & KAIRUPAN, J.

1995b. Skin sores in Aboriginal children. J Paediatr Child Health, 31, 563.

CARAPETIS, J. R., STEER, A. C., MULHOLLAND, E. K. & WEBER, M.

2005. The global burden of group A streptococcal diseases. Lancet Infect Dis, 5,

685-94.

CARAPETIS, J. R., WALKER, A. M., HIBBLE, M., SRIPRAKASH, K. S. &

CURRIE, B. J. 1999. Clinical and epidemiological features of group A

streptococcal bacteraemia in a region with hyperendemic superficial

streptococcal infection. Epidemiol Infect, 122, 59-65.

CARPA 1992. CARPA Standard Treatment Manual, Alice Springs, Central

Australian Rural Practitioners Association.

CARPA 1994. CARPA Standard Treatment Manual, 2nd ed., Alice Springs,

Central Australian Rural Practitioners Association.

CARPA 1997. CARPA Standard Treatment Manual, 3rd ed., Alice Springs,


115

CARPA 2003. CARPA Standard Treatment Manual, 4th ed. , Alice Springs,


CARPA 2009. CARPA Standard Treatment Manual, 5th ed., Alice Springs,


CARPA 2014. CARPA Standard Treatment Manual, 6th ed. , Alice Springs,

Centre for Remote Health.

CHUTTANI, C. S. & PAL, S. C. 1971. An epidemiological study of

streptococcal skin infection. Indian J Med Res, 59, 1822-6.

CLSI 2003. Quality Control of Microbiological Transport Systems: Approved

Standard M40-A. Wayne, Pennsylvania: Clinical and Laboratory Standards

Institute.

CLSI 2006. Protocols for evaluating dehydrated Mueller-Hinton Agar;

Approved standard- second edition, Wayne, PA, Clinical and Laboratory

Standards Institute

CLSI 2011. Performance Standards for Antimicrobial Susceptibility Testing;

Twenty-First Informational Supplement M100-S21. Wayne, PA: Clinical and

Laboratory Standards Institute.

COBURN, A. F. 1944. The prevention of respiratory tract bacterial infections by

sulfadiazine prophylaxis in the United States navy. JAMA, 126, 88-93.

COLL, P. F., AUSINA, V. R., VERNIS, J. V., MIRELIS, B. O. & PRATS, G. P.

1984. Exogenous thymidine and reversal of the inhibitory effect of

sulfamethoxazole-trimethoprim on streptococci. Eur J Clin Microbiol, 3, 424-6.

COPAN eSwab collection and preservation. Product insert and how to use swab

guide. Eswab product information. 1st ed, Brescia, Italy: Copan; 2010.

CRAIG, F. F., THOMAS, K. S., MITCHELL, E. J., WILLIAMS, H. C.,

NORRIE, J., MASON, J. M. & ORMEROD, A. D. 2012. UK Dermatology

Clinical Trials Network's STOP GAP trial (a multicentre trial of prednisolone

116

versus ciclosporin for pyoderma gangrenosum): protocol for a randomised

controlled trial. Trials, 13, 51.

CROMEY, D. W. 2010. Avoiding twisted pixels: ethical guidelines for the

appropriate use and manipulation of scientific digital images. Sci Eng Ethics, 16,

639-67.

CUMMING, C. G., ROSS, P. W. & LOUGH, H. 1981. Optimal methods for the

isolation of groups A, B, C and G streptococci. J Laryngol Otol, 95, 377-84.

CURRIE, B. 2006. Benzathine penicillin - down but not out. The Northern

Territory Disease Control Bulletin, 13, 1 - 4.

CURRIE, B. J. 1996. Are the currently recommended doses of benzathine

penicillin G adequate for secondary prophylaxis of rheumatic fever? Pediatrics,

97, 989-91.

CURRIE, B. J. & CARAPETIS, J. R. 2000. Skin infections and infestations in

Aboriginal communities in northern Australia. Australas J Dermatol, 41, 139-43;

quiz 144-5.

DAGNEW, M. B. & ERWIN, G. 1991. Epidemiology of common transmissible

skin diseases among primary school children in north-west Ethiopia. Trop Geogr

Med, 43, 152-5.

DAJANI, A. S., FERRIERI, P. & WANNAMAKER, L. 1973. Endemic

superficial pyoderma in children. Arch Dermatol, 108, 517-22.

DAJANI, A. S., FERRIERI, P. & WANNAMAKER, L. W. 1972. Natural

history of impetigo. II. Etiologic agents and bacterial interactions. J Clin Invest,

51, 2863-71.

DAJANI, A. S., HILL, P. L. & WANNAMAKER, L. W. 1971. Experimental

infection of the skin in the hamster simulating human impetigo. II. Assessment of

various therapeutic regimens. Pediatrics, 48, 83-90.

DAMROSCH, D. 1946. Chemoprophylaxis and Sulfonamide resistant

streptococci. Journal of American Medical Association 130, 124-128.

117

DARRELL, J. H., GARROD, L. P. & WATERWORTH, P. M. 1968.

Trimethoprim: laboratory and clinical studies. J Clin Pathol, 21, 202-9.

DELAMETER, E. D., JENNINGS, R., WALLACE, A.W. 1946. Preliminary

report of an outbreak of streptococcal disease cuased by a sulfadiazine resistant

group A, type 17 hemolytic streptococcus. The Journal of Infectious Diseases,

78, 118 - 127.

DERRICK, C. W. & DILLON, H. C., JR. 1970. Further studies on the treatment

of streptococcal skin infection. J Pediatr, 77, 696-700.

DHANDA V. CHAUDHARY, P. R., P. CHAKRABORTI, A. TALWAR, K.

KUMAR, R. 2011. Invasive Group A Streptococcus (GAS) from North India

during 2007 - 2010: surveillance and antibiotic susceptibility. Abstract 5, page

18. XVIII Lancefield Symposium. Palermo, Italy.

DILLON, H. C., JR. 1970. The treatment of streptococcal skin infections. J

Pediatr, 76, 676-84.

DOGRA, S. & KUMAR, B. 2003. Epidemiology of skin diseases in school

children: a study from northern India. Pediatr Dermatol, 20, 470-3.

DOMAGK, G. 1935. Ein Beitrag zur Chemotherapie der Bakteriellen

Infektionen. Deutsche Medizinische Wochenschrift, 61, 250 - 253.

DOS SANTOS, M. M., AMARAL, S., HARMEN, S. P., JOSEPH, H. M.,

FERNANDES, J. L. & COUNAHAN, M. L. 2010. The prevalence of common

skin infections in four districts in Timor-Leste: a cross sectional survey. BMC

Infect Dis, 10, 61.

DUMRE, S. P., SAPKOTA, K., ADHIKARI, N., ACHARYA, D., KARKI, M.,

BISTA, S., BASANYAT, S. R. & JOSHI, S. K. 2009. Asymptomatic throat

carriage rate and antimicrobial resistance pattern of Streptococcus pyogenes in

Nepalese school children. Kathmandu Univ Med J (KUMJ), 7, 392-6.

118

EASON, R. J. & TASMAN-JONES, T. 1985. Resurgent yaws and other skin

diseases in the Western Province of the Solomon Islands. P N G Med J, 28, 247-

50.

EISEN, D. T. 2010. Tetracycline, p.843 - 894. In: GRAYSON, M. L., Crowe,

S.M.M., MILLS, J.S., MOUTON, J.W., NORRBY, R., PATERSON, D.L.,

PFALLER, M.A. (ed.) The use of Antibacterials. Sixth edition ed., vol 1. ASM

Press, Washington DC.

ELIOPOULOS, G. M. & WENNERSTEN, C. B. 1997. In vitro activity of

trimethoprim alone compared with trimethoprim-sulfamethoxazole and other

antimicrobials against bacterial species associated with upper respiratory tract

infections. Diagn Microbiol Infect Dis, 29, 33-8.

ELLIS, M. W., GRIFFITH, M. E., DOOLEY, D. P., MCLEAN, J. C.,

JORGENSEN, J. H., PATTERSON, J. E., DAVIS, K. A., HAWLEY, J. S.,

REGULES, J. A., RIVARD, R. G., GRAY, P. J., CEREMUGA, J. M.,

DEJOSEPH, M. A. & HOSPENTHAL, D. R. 2007. Targeted intranasal

mupirocin to prevent colonization and infection by community-associated

methicillin-resistant Staphylococcus aureus strains in soldiers: a cluster

randomized controlled trial. Antimicrob Agents Chemother, 51, 3591-8.

ENGELMAN, D. HOFER A, DAVIS, J.S., CARAPETIS, J.R., BAIRD, R.W.,

GIFFARD, P.M., HOLT, D.C., & TONG, S.Y.C. 2014. Invasive Staphylococcus

aureus Infections in children in tropical Northern Australia. J Ped Infect Dis,

3(4): 304 - 311.

ESTERLY, N. B. & MARKOWITZ, M. 1970. The treatment of pyoderma in

children. JAMA, 212, 1667-70.

ESTRADA CASTANON, R., ANDERSSON, N. & HAY, R. 1992. Community

dermatology and the management of skin diseases in developing countries. Trop

Doct, 22 Suppl 1, 3-6.

119

EUCAST. 5 January 2011, posting date. EUCAST MIC and zone diameter

breakpoint tables. http://www.eucast.org/antimicrobial_susceptibility_testing

/breakpoints/.

EUCAST. 29 June 2012, posting date. Antimicrobial Susceptibility Testing.

http://www.eucast.org/antimicrobial_susceptibility_testing/disk_diffusion_metho

dology/.

FACKLAM, R. R. 1976. A review of the microbiological techniques for the

isolation and identification of streptococci. CRC Crit Rev Clin Lab Sci, 6, 287-

317.

FATHY, H., EL-MONGY, S., BAKER, N. I., ABDEL-AZIM, Z. & EL-

GILANY, A. 2004. Prevalence of skin diseases among students with disabilities

in Mansoura, Egypt. East Mediterr Health J, 10, 416-24.

FAYE, O., HAY, R. J., DIAWARA, I. & MAHE, A. 2007. Oral amoxicillin vs.

oral erythromycin in the treatment of pyoderma in Bamako, Mali: an open

randomized trial. Int J Dermatol, 46 Suppl 2, 19-22.

FERIE, J., DINKELA, A., MBATA, M., IDINDILI, B., SCHMID-

GRENDELMEIER, P. & HATZ, C. 2006. Skin disorders among school children

in rural Tanzania and an assessment of therapeutic needs. Trop Doct, 36, 219-21.

FERONE, R., BUSHBY, S. R., BURCHALL, J. J., MOORE, W. D. & SMITH,

D. 1975. Identification of Harper-Cawston factor as thymidine phosphorylase

and removal from media of substances interfering with susceptibility testing to

sulfonamides and diaminopyrimidines. Antimicrob Agents Chemother, 7, 91-8.

FERRIERI, P., DAJANI, A. S., WANNAMAKER, L. W. & CHAPMAN, S. S.

1972. Natural history of impetigo. I. Site sequence of acquisition and familial

patterns of spread of cutaneous streptococci. J Clin Invest, 51, 2851-62.

FERRIERI, P., DAJANI, A. S. & WANNAMAKER, L. W. 1973. Benzathine

penicillin in the prophylaxis of streptococcal skin infections: a pilot study. J

Pediatr, 83, 572-7.

120

FERRIERI, P., DAJANI, A. S. & WANNAMAKER, L. W. 1974. A controlled

study of penicillin prophylaxis against streptococcal impetigo. J Infect Dis, 129,

429-38.

FIGUEROA, J. I., FULLER, L. C., ABRAHA, A. & HAY, R. J. 1996. The

prevalence of skin disease among school children in rural Ethiopia--a preliminary

assessment of dermatologic needs. Pediatr Dermatol, 13, 378-81.

FIGUEROA, J. I., FULLER, L. C., ABRAHA, A. & HAY, R. J. 1998.

Dermatology in southwestern Ethiopia: rationale for a community approach. Int J

Dermatol, 37, 752-8.

FINLAND, M., GARNER, C., WILCOX, C. & SABATH, L. D. 1976.

Susceptibility of beta-hemolytic streptococci to 65 antibacterial agents.

Antimicrob Agents Chemother, 9, 11-9.

FLANAGAN, M. 2003. Wound measurement: can it help us to monitor

progression to healing? J Wound Care, 12, 189-94.

FOX, W. T. 1864. On Impetigo Contagiosa, or Porrigo. British Medical Journal,

1, 467 - 469.

FREIS, E. D. 1944. The treatment of tonsillitis with small doses of sulfonamides.

JAMA, 126, 93 - 94.

FRITZ, S.A., HOGAN, P.G., HAYEK, G., EISENSTEIN, K.A., RODRIGUEZ,

M., KRAUSS, M., GARBUTT, J. & FRASER, V.J. 2012. Staphylococcus aureus

colonization in children with community-associated Staphylococcus aureus skin

infections and their household contacts. Arch Pediatr Adolesc Med, 166, 551 - 7.

FRY, A. M., JHA, H. C., LIETMAN, T. M., CHAUDHARY, J. S., BHATTA, R.

C., ELLIOTT, J., HYDE, T., SCHUCHAT, A., GAYNOR, B. & DOWELL, S.

F. 2002. Adverse and beneficial secondary effects of mass treatment with

azithromycin to eliminate blindness due to trachoma in Nepal. Clin Infect Dis,

35, 395-402.

121

GELFAND, M. S., CLEVELAND, K. O. & KETTERER, D. C. 2013.

Susceptibility of Streptococcus pyogenes to trimethoprim-sulfamethoxazole. J

Clin Microbiol, 51, 1350.

GERIA, A. N. & SCHWARTZ, R. A. 2010. Impetigo update: new challenges in

the era of methicillin resistance. Cutis, 85, 65-70.

GIBBS, S. 1996. Skin disease and socioeconomic conditions in rural Africa:

Tanzania. Int J Dermatol, 35, 633-9.

GULKESEN, K. H., AKMAN, A., YUCE, Y. K., YILMAZ, E., SAMUR, A. A.,

ISLEYEN, F., CAKCAK, D. S. & ALPSOY, E. 2010. Evaluation of JPEG and

JPEG2000 compression algorithms for dermatological images. J Eur Acad

Dermatol Venereol, 24, 893-6.

GOLDMAN, J. L., JACKSON, M. A., HERIGON, J. C., HERSH, A. L.,

SHAPIRO, D. J. & LEEDER, J. S. 2013. Trends in adverse reactions to

trimethoprim-sulfamethoxazole. Pediatrics, 131, e103-8.

HARE, K. M., GRIMWOOD, K., LEACH, A. J., SMITH-VAUGHAN, H.,

TORZILLO, P. J., MORRIS, P. S. & CHANG, A. B. 2010a. Respiratory

bacterial pathogens in the nasopharynx and lower airways of Australian

indigenous children with bronchiectasis. J Pediatr, 157, 1001-5.

HARE, K. M., STUBBS, E., BEISSBARTH, J., MORRIS, P. S. & LEACH, A. J.

2010b. Swab transport in Amies gel followed by frozen storage in skim milk

tryptone glucose glycerol broth (STGGB) for studies of respiratory bacterial

pathogens. J Microbiol Methods, 81, 253-5.

HARE, K. M., SMITH-VAUGHAN, H. C. & LEACH, A. J. 2011. Viability of

respiratory pathogens cultured from nasopharyngeal swabs stored for up to 12

years at -70 degrees C in skim milk tryptone glucose glycerol broth. J Microbiol

Methods, 86, 364-7.

HARPER, G. C., WC. 1945. The in-vitro determination of the sulphonamide

sensitivity of bacteria. The Journal of Pathology and Bacteriology, 57, 59 - 66.

122

HARRIS, M., NAKO, D., HOPKINS, T., POWELL, D. M., KENNY, C.,

CARROLL, C. & CARROLL, K. 1992. Skin infections in Tanna, Vanuatu in

1989. P N G Med J, 35, 137-43.

HARTMAN-ADAMS, H., BANVARD, C. & JUCKETT, G. 2014. Impetigo:

diagnosis and treatment. Am Fam Physician, 90, 229-35.

HARTMAN, T. L., & HOES, J.S. 1949. The occurrence of a type five

sulfadiazine resistant hemolytic streptococcus. Journal of Pediatrics, 34, 331-

337.

HARTMAN, T. L. & WEINSTEIN, L. 1948. The problem of sulfonamide-

resistant hemolytic streptococci. N Engl J Med, 238, 560-3.

HAY, R., BENDECK, S. E., CHEN, S., ESTRADA, R., HADDIX, A.,

MCLEOD, T. & MAHE, A. 2006. Skin Diseases. In: JAMISON, D. T.,

BREMAN, J. G., MEASHAM, A. R., ALLEYNE, G., CLAESON, M., EVANS,

D. B., JHA, P., MILLS, A. & MUSGROVE, P. (eds.) Disease Control Priorities

in Developing Countries. 2nd ed. Washington (DC).

HAY, R. J., ESTRADA CASTANON, R., ALARCON HERNANDEZ, H.,

CHAVEZ LOPEZ, G., LOPEZ FUENTES, L. F., PAREDES SOLIS, S. &

ANDERSSON, N. 1994. Wastage of family income on skin disease in Mexico.

BMJ, 309, 848.

HAY, R. J., JOHNS, N. E., WILLIAMS, H. C., BOLLIGER, I. W.,

DELLAVALLE, R. P., MARGOLIS, D. J., MARKS, R., NALDI, L.,

WEINSTOCK, M. A., WULF, S. K., MICHAUD, C., C, J. L. M. & NAGHAVI,

M. 2014. The Global Burden of Skin Disease in 2010: An Analysis of the

Prevalence and Impact of Skin Conditions. J Invest Dermatol.134 (6): 1527 - 34.

HENDERSON, C. A. 1996. Skin disease in rural Tanzania. Int J Dermatol, 35,

640-2.

HITCHINGS, G. H. 1973. Mechanism of action of trimethoprim-

sulfamethoxazole. I. J Infect Dis, 128, Suppl:433-6 p.

123

HOGEWONING, A., AMOAH, A., BAVINCK, J. N., BOAKYE, D.,

YAZDANBAKHSH, M., ADEGNIKA, A., DE SMEDT, S., FONTEYNE, Y.,

WILLEMZE, R. & LAVRIJSEN, A. 2013. Skin diseases among schoolchildren

in Ghana, Gabon, and Rwanda. Int J Dermatol, 52, 589-600.

HOLBROOK, W. C. 1944. The Army Air Forces Rheumatic Fever Control

Program. JAMA, 126, 84-87.

HOLT, D. C. & FISCHER, K. 2013. Novel insights into an old disease: recent

developments in scabies mite biology. Curr Opin Infect Dis, 26, 110-5.

HOSKINS, T. W. & BERNSTEIN, L. S. 1981. Trimethoprim/sulphadiazine

compared with penicillin V in the treatment of streptococcal throat infections. J

Antimicrob Chemother, 8, 495-6.

HOTEZ, P.J. 2013. NTDs V 2.0: "blue marble health"--neglected tropical disease

control and elimination in a shifting health policy lanscape. PLoS Negl Trop Dis,

7, e2570.

HOY, W. E., WHITE, A. V., DOWLING, A., SHARMA, S. K., BLOOMFIELD,

H., TIPILOURA, B. T., SWANSON, C. E., MATHEWS, J. D. & MCCREDIE,

D. A. 2012. Post-streptococcal glomerulonephritis is a strong risk factor for

chronic kidney disease in later life. Kidney Int, 81, 1026-32.

HUGHES, W. T. & WAN, R. T. 1967. Impetigo contagiosa: etiology,

complications, and comparison of therapeutic effectiveness of erythromycin and

antibiotic ointment. Am J Dis Child, 113, 449-53.

HUMAN, R. P. & JONES, G. A. 2004. Evaluation of swab transport systems

against a published standard. J Clin Pathol, 57, 762-3.

HUMPHREY, C. D. & KRIET, J. D. 2010. Intraoperative photography. Facial

Plast Surg Clin North Am, 18, 329-34, Table of Contents.

INSTITUTE OF MEDICAL ILLUSTRATORS, 2007. IMMI National

Guidelines Clinical Photography in Wound Management.

124

INANIR, I., SAHIN, M. T., GUNDUZ, K., DINC, G., TUREL, A. &

OZTURKCAN, S. 2002. Prevalence of skin conditions in primary school

children in Turkey: differences based on socioeconomic factors. Pediatr

Dermatol, 19, 307-11.

IOVINO, S. M., KRANTZ, K. D., BLANCO, D. M., FERNANDEZ, J. A.,

OCAMPO, N., NAJAFI, A., MEMARZADEH, B., CELERI, C., DEBABOV,

D., KHOSROVI, B. & ANDERSON, M. 2011. NVC-422 topical gel for the

treatment of impetigo. Int J Clin Exp Pathol, 4, 587-95.

JACKSON, S. J., STEER, A. C. & CAMPBELL, H. 2011. Systematic Review:

Estimation of global burden of non-suppurative sequelae of upper respiratory

tract infection: rheumatic fever and post-streptococcal glomerulonephritis. Trop

Med Int Health, 16, 2-11.

JAKOWENKO, J. 2009. Clinical photography. J Telemed Telecare, 15, 7-22.

JENNEY, A., HOLT, D., RITIKA, R., SOUTHWELL, P., PRAVIN, S.,

BUADROMO, E., CARAPETIS, J., TONG, S. & STEER, A. 2014. The clinical

and molecular epidemiology of Staphylococcus aureus infections in Fiji. BMC

Infect Dis, 14, 160.

JESADA, E. C., WARREN, J. I., GOODMAN, D., ILIUTA, R. W.,

THURKAUF, G., MCLAUGHLIN, M. K., JOHNSON, J. E. & STRASSNER, L.

2013. Staging and defining characteristics of pressure ulcers using photographs

by staff nurses in acute care settings. J Wound Ostomy Continence Nurs, 40, 150-

6.

JINADU, M. K. 1985. Pediculosis humanus capitis among primary school

children in Ile-Ife, Nigeria. J R Soc Health, 105, 25-7.

JOHNSON, D. R., KAPLAN, E.L., SRAMEK, J., BICOVA, R., HAVLICEK, J.,

HAVLICKOVA, H., MOTLOVA, J., KRIZ, P. 1996. Laboratory diagnosis of

group A streptococcal infections, Geneva, World Health Organisation.

JOHNSON, R. D. & HARTMAN, T. L. 1947. Sulfadiazine resistant

streptococcal infections in a civilian community. J Clin Invest, 26, 325-8.

125

KAHLMETER, G., BROWN, D. F., GOLDSTEIN, F. W., MACGOWAN, A. P.,

MOUTON, J. W., OSTERLUND, A., RODLOFF, A., STEINBAKK, M.,

URBASKOVA, P. & VATOPOULOS, A. 2003. European harmonization of

MIC breakpoints for antimicrobial susceptibility testing of bacteria. J Antimicrob

Chemother, 52, 145-8.

KAPLAN, E. L., JOHNSON, D. R., DEL ROSARIO, M. C. & HORN, D. L.

1999. Susceptibility of group A beta-hemolytic streptococci to thirteen

antibiotics: examination of 301 strains isolated in the United States between 1994

and 1997. Pediatr Infect Dis J, 18, 1069-72.

KARIMKHANI, C., BOYERS, L.N., PRESCOTT, L., WELCH, V.,

DELAMERE, F.M., NASSER, M., ZAVERI, A., HAY, R.J., VOS, T.,

MURRAY, C.J., MARGOLIS, D.J., HILTON, K., MACLEHOSE, J.,

WILLIAMS, H.C. & DELLAVILEE, R.P. 2014. Global burden of skin disease

as reflected in Cochrane database of systematic reviews. JAMA Dermatol,

150(9): 945 - 51.

KATUGAMPOLA, R. & LAKE, A. 2012. The role of photography in

dermatology research. J Vis Commun Med, 35, 5-10.

KEARNS, T., CLUCAS, D., CONNORS, C., CURRIE, B. J., CARAPETIS, J.

R. & ANDREWS, R. M. 2013. Clinic attendances during the first 12 months of

life for Aboriginal children in five remote communities of northern Australia.

PLoS One, 8, e58231.

KEMP K, N. T., BOSWELL J, LEACH A, KANTILLA C,

TIPUAMANTAMIRRI M, MATTHEWS J 1994. Strategies for and Problems

associated with maximising and monitoring compliance with antibiotic treatment

for otitis media with effusion in a remote Aboriginal Community. The Australian

Journal of Rural Health, 2, 25 - 31.

KHAVKIN, J. & ELLIS, D. A. 2011. Standardized photography for skin surface.

Facial Plast Surg Clin North Am, 19, 241-6.

126

KIRIAKIS, K. P., TADROS, A., DIMOU, A., KARAMANOU, M., BANAKA,

F. & ALEXOUDI, I. 2012. Case detection rates of impetigo by gender and age.

Infez Med, 20, 105-7.

KOMBA, E. V. & MGONDA, Y. M. 2010. The spectrum of dermatological

disorders among primary school children in Dar es Salaam. BMC Public Health,

10, 765.

KONING, S., MOHAMMEDAMIN, R. S., VAN DER WOUDEN, J. C., VAN

SUIJLEKOM-SMIT, L. W., SCHELLEVIS, F. G. & THOMAS, S. 2006.

Impetigo: incidence and treatment in Dutch general practice in 1987 and 2001--

results from two national surveys. Br J Dermatol, 154, 239-43.

KONING, S., VAN DER SANDE, R., VERHAGEN, A. P., VAN SUIJLEKOM-

SMIT, L. W., MORRIS, A. D., BUTLER, C. C., BERGER, M. & VAN DER

WOUDEN, J. C. 2012. Interventions for impetigo. Cochrane Database Syst Rev,

1, CD003261.

KONING, S. & VAN DER WOUDEN, J. C. 2004. Treatment for impetigo. BMJ,

329, 695-6.

KONING, S., VERHAGEN, A. P., VAN SUIJLEKOM-SMIT, L. W., MORRIS,

A., BUTLER, C. C. & VAN DER WOUDEN, J. C. 2004. Interventions for

impetigo. Cochrane Database Syst Rev, CD003261.

KOSHI, G. & BENJAMIN, V. 1977. Surveillance of streptococcal infections in

children in a south Indian community--a pilot survey. Indian J Med Res, 66, 379-

88.

KOSHI, G., BRAHMADATHAN, K. N., THANGAVELU, C. P. & PANDIAN,

R. 1979. Evaluation of different methods for the transport of swabs for

streptococci. Indian J Med Res, 69, 26-31.

KOTLOFF, K., VAN BENEDEN, C. 2008. Standardization of epidemiologic

protocols of surveillance of acute diseases caused by Streptococcus pyogenes:

Pharyngitis, Impetigo, and Invasive diseases. Bethesda, MD: National Institute

127

of Allegry and Infectious Diseases. http://www.niaid.nih.gov/topics/

streptococcal/pages/protocols.aspx

KOTTENHAHN, R. K. & HECK, J. E. 1994. Prevalence of paediatric skin

diseases in rural Honduras. Trop Doct, 24, 87-8.

KROLICKI, A., KLIMOWICZ, A., BIELECKA-GRZELA, S., NOWAK, A. &

MALESZKA, R. 2004. Penetration of cotrimoxazole components into skin after

a single oral dose. Theoretical versus experimental approach. Pol J Pharmacol,

56, 257-63.

KUMAR, R., CHAKRABORTI, A., AGGARWAL, A. K., VOHRA, H.,

SAGAR, V., DHANDA, V., SHARMA, Y. P., MAJUMDAR, S., HOE, N. &

KRAUSE, R. M. 2012. Streptococcus pyogenes pharyngitis & impetigo in a rural

area of Panchkula district in Haryana, India. Indian J Med Res, 135, 133-6.

KUMAR, R., VOHRA, H., CHAKRABORTY, A., SHARMA, Y. P.,

BANDHOPADHYA, S., DHANDA, V., SAGAR, V., SHARMA, M., SHAH, B.

& GANGULY, N. K. 2009. Epidemiology of group A streptococcal pharyngitis

& impetigo: a cross-sectional & follow up study in a rural community of

northern India. Indian J Med Res, 130, 765-71.

LAKSHMY, A. M., V. L. DUTTA, R. 2011. Antibiotic resistance pattern of

Streptococcus pyogenes isolated from clinical samples. XVIII Lancefield

Symposium. Palermo, Italy.

LAURELL, G. L., I.M. MIYAMOTO, Y. MOODY, M.D. MOZZICONACCI, P.

PARKER, M.T. RAMMELKAMP, C.H. ROTTA, J. 1968. Streptococcal and

Staphylococcal Infections. World Health Organization Technical Report Series

no. 394. Geneva: World Health Organization.

LAWRENCE, D. N., FACKLAM, R. R., SOTTNEK, F. O., HANCOCK, G. A.,

NEEL, J. V. & SALZANO, F. M. 1979. Epidemiologic studies among

Amerindian populations of Amazonia. I. Pyoderma: prevalence and associated

pathogens. Am J Trop Med Hyg, 28, 548-58.

128

LAWRENCE, G., LEAFASIA, J., SHERIDAN, J., HILLS, S., WATE, J.,

WATE, C., MONTGOMERY, J., PANDEYA, N. & PURDIE, D. 2005. Control

of scabies, skin sores and haematuria in children in the Solomon Islands: another

role for ivermectin. Bull World Health Organ, 83, 34-42.

LAWRENCE, J. G., CARAPETIS, J. R., GRIFFITHS, K., EDWARDS, K. &

CONDON, J. R. 2013. Acute rheumatic fever and rheumatic heart disease:

incidence and progression in the Northern Territory of Australia, 1997 to 2010.

Circulation, 128, 492-501.

LEHMANN, D., TENNANT, M. T., SILVA, D. T., MCAULLAY, D.,

LANNIGAN, F., COATES, H. & STANLEY, F. J. 2003. Benefits of swimming

pools in two remote Aboriginal communities in Western Australia: intervention

study. BMJ, 327, 415-9.

LENG, T. & CHAY, S. O. 1982. A three-year streptococcal survey among

Singapore school children: Part II. Streptococcal infections. Ann Acad Med

Singapore, 11, 101-9.

LIEBOWITZ, L. D., SLABBERT, M. & HUISAMEN, A. 2003. National

surveillance programme on susceptibility patterns of respiratory pathogens in

South Africa: moxifloxacin compared with eight other antimicrobial agents. J

Clin Pathol, 56, 344-7.

LINES, D. R. 1977. An Auckland high school health survey. Aust N Z J Med, 7,

143-7.

LIU, C., BAYER, A., COSGROVE, S. E., DAUM, R. S., FRIDKIN, S. K.,

GORWITZ, R. J., KAPLAN, S. L., KARCHMER, A. W., LEVINE, D. P.,

MURRAY, B. E., M, J. R., TALAN, D. A. & CHAMBERS, H. F. 2011. Clinical

practice guidelines by the infectious diseases society of america for the treatment

of methicillin-resistant Staphylococcus aureus infections in adults and children.

Clin Infect Dis, 52, e18-55.

LIU, F., RANMAL, S., BATCHELOR, H.K., ORLU-GUL, M., ERNEST, T.B.,,

THOMAS, I.W., FLANAGAN, T., & TULEU, C. 2014 Patient-centred

129

pharmaceutical design to improve acceptability of medicines: similarities and

differences in paediatric and geriatric populations. Drugs, 74, 1871 - 89.

LUBY, S., AGBOATWALLA, M., SCHNELL, B. M., HOEKSTRA, R. M.,

RAHBAR, M. H. & KESWICK, B. H. 2002. The effect of antibacterial soap on

impetigo incidence, Karachi, Pakistan. Am J Trop Med Hyg, 67, 430-5.

LUBY, S. P., AGBOATWALLA, M., FEIKIN, D. R., PAINTER, J.,

BILLHIMER, W., ALTAF, A. & HOEKSTRA, R. M. 2005. Effect of

handwashing on child health: a randomised controlled trial. Lancet, 366, 225-33.

MAGUIRE, G. P., ARTHUR, A. D., BOUSTEAD, P. J., DWYER, B. &

CURRIE, B. J. 1998. Clinical experience and outcomes of community-acquired

and nosocomial methicillin-resistant Staphylococcus aureus in a northern

Australian hospital. J Hosp Infect, 38, 273-81.

MAHE, A., FAYE, O., N'DIAYE, H. T., KONARE, H. D., COULIBALY, I.,

KEITA, S., TRAORE, A. K. & HAY, R. J. 2005a. Integration of basic

dermatological care into primary health care services in Mali. Bull World Health

Organ, 83, 935-41.

MAHE, A., FAYE, O., N'DIAYE, H. T., LY, F., KONARE, H., KEITA, S.,

TRAORE, A. K. & HAY, R. 2005b. Definition of an algorithm for the

management of common skin diseases at primary health care level in sub-

Saharan Africa. Trans R Soc Trop Med Hyg, 99, 39-47.

MAHE, A., PRUAL, A., KONATE, M. & BOBIN, P. 1995. Skin diseases of

children in Mali: a public health problem. Trans R Soc Trop Med Hyg, 89, 467-

70.

MAHE, A. H., R.J. 2005. Epidemiology and Management of common skin

diseases in children in developing countries. Discussion Papers on Child Health.

1 ed. Geneva: World Health Organisation

MARGOLIS, H. S., LUM, M. K., BENDER, T. R., ELLIOTT, S. L.,

FITZGERALD, M. A. & HARPSTER, A. P. 1980. Acute glomerulonephritis and

streptococcal skin lesions in Eskimo children. Am J Dis Child, 134, 681-5.

130

MARSHALL, C. S., CHENG, A. C., MARKEY, P. G., TOWERS, R. J.,

RICHARDSON, L. J., FAGAN, P. K., SCOTT, L., KRAUSE, V. L. & CURRIE,

B. J. 2011. Acute post-streptococcal glomerulonephritis in the Northern Territory

of Australia: a review of 16 years data and comparison with the literature. Am J

Trop Med Hyg, 85, 703-10.

MARTIN, D. R., STANHOPE, J. M. & FINCH, L. A. 1977. Delayed culture of

group-A streptococci: an evaluation of variables in methods of examining throat

swabs. J Med Microbiol, 10, 249-53.

MASAWE, A. E. & NSANZUMUHIRE, H. 1975. Scabies and other skin

diseases in pre-school children in Ujamaa villages in Tanzania. Trop Geogr Med,

27, 288-94.

MASAWE, A. E., NSANZUMUHIRE, H. & MHALU, F. 1975. Bacterial skin

infections in preschool and school children in coastal Tanzania. Arch Dermatol,

111, 1312-6.

MASTERS, P. L., BRUMFITT, W., MENDEZ, R. L. & LIKAR, M. 1958.

Bacterial flora of the upper respiratory tract in Paddington families. 1952-4. Br

Med J, 1, 1200-5.

MCCONNEL, F. 2003. Compliance, culture, and the health of Indigenous

people. Rural Remote Health, 3, 190.

MCDONALD, M., CURRIE, B. J. & CARAPETIS, J. R. 2004. Acute rheumatic

fever: a chink in the chain that links the heart to the throat? Lancet Infect Dis, 4,

240-5.

MCDONALD, M., DOUGALL, A., HOLT, D., HUYGENS, F., OPPEDISANO,

F., GIFFARD, P. M., INMAN-BAMBER, J., STEPHENS, A. J., TOWERS, R.,

CARAPETIS, J. R. & CURRIE, B. J. 2006a. Use of a single-nucleotide

polymorphism genotyping system to demonstrate the unique epidemiology of

methicillin-resistant Staphylococcus aureus in remote aboriginal communities. J

Clin Microbiol, 44, 3720-7.

131

MCDONALD, M., TOWERS, R., FAGAN, P., MCKINNON, M., BENGER, N.,

ANDREWS, R., CURRIE, B. J. & CARAPETIS, J. 2006b. Recovering

streptococci from the throat, a practical alternative to direct plating in remote

tropical communities. J Clin Microbiol, 44, 547-52.

MCDONALD, M. I. 2006. The Epidemiology of pyoderma and pharyngitis in

Aboriginal communities with a high prevalence of rheumatic heart disease. [PhD

Thesis]. Melbourne: University of Melbourne.

MCDONALD, M. I., TOWERS, R. J., ANDREWS, R., BENGER, N., FAGAN,

P., CURRIE, B. J. & CARAPETIS, J. R. 2008. The dynamic nature of group A

streptococcal epidemiology in tropical communities with high rates of rheumatic

heart disease. Epidemiol Infect, 136, 529-39.

MCDONALD, M. I., TOWERS, R. J., ANDREWS, R. M., BENGER, N.,

CURRIE, B. J. & CARAPETIS, J. R. 2006c. Low rates of streptococcal

pharyngitis and high rates of pyoderma in Australian aboriginal communities

where acute rheumatic fever is hyperendemic. Clin Infect Dis, 43, 683-9.

MCMENIMAN, E., HOLDEN, L., KEARNS, T., CLUCAS, D. B.,

CARAPETIS, J. R., CURRIE, B. J., CONNORS, C. & ANDREWS, R. M. 2011.

Skin disease in the first two years of life in Aboriginal children in East Arnhem

Land. Australas J Dermatol, 52, 270-3.

MILLER, L. G., EELLS, S. J., TAYLOR, A. R., DAVID, M. Z., ORTIZ, N.,

ZYCHOWSKI, D., KUMAR, N., CRUZ, D., BOYLE-VAVRA, S. & DAUM, R.

S. 2012. Staphylococcus aureus Colonization Among Household Contacts of

Patients With Skin Infections: Risk Factors, Strain Discordance, and Complex

Ecology. Clin Infect Dis.

MOHAMMEDAMIN, R. S., VAN DER WOUDEN, J. C., KONING, S., VAN

DER LINDEN, M. W., SCHELLEVIS, F. G., VAN SUIJLEKOM-SMIT, L. W.

& KOES, B. W. 2006. Increasing incidence of skin disorders in children? A

comparison between 1987 and 2001. BMC Dermatol, 6, 4.

132

MOHER, D., LIBERATI, A., TETZLAFF, J. & ALTMAN, D. G. 2009.

Preferred reporting items for systematic reviews and meta-analyses: the PRISMA

statement. PLoS Med, 6, e1000097.

MORAN, G. J., ABRAHAMIAN, F. M., LOVECCHIO, F. & TALAN, D. A.

2013. Acute bacterial skin infections: developments since the 2005 Infectious

Diseases Society of America (IDSA) guidelines. J Emerg Med, 44, e397-412.

MURDOCH, J. 2005. Standardization in clinical photography at Addenbrooke's

Hospital. J Vis Commun Med, 28, 68-71.

MURGIA, V., BILCHA, K. D. & SHIBESHI, D. 2010. Community dermatology

in Debre Markos: an attempt to define children's dermatological needs in a rural

area of Ethiopia. Int J Dermatol, 49, 666-71.

NAJJAR, A. & MURRAY, B. E. 1987. Failure to demonstrate a consistent in

vitro bactericidal effect of trimethoprim-sulfamethoxazole against enterococci.

Antimicrob Agents Chemother, 31, 808-10.

NAYLER, J., GEDDES, N. & GOMEZ-CASTRO, C. 2001. Managing digital

clinical photographs. J Audiov Media Med, 24, 166-71.

NAYLER, J. R. 2003. Clinical photography: a guide for the clinician. J Postgrad

Med, 49, 256-62.

NELSON, K. E., BISNO, A. L., WAYTZ, P., BRUNT, J., MOSES, V. K. &

HAQUE, R. 1976. The epidemiology and natural history of streptococcal

pyoderma: an endemic disease of the rural southern United States. Am J

Epidemiol, 103, 270-83.

NICKERSON, E. K., WUTHIEKANUN, V., KUMAR, V., AMORNCHAI, P.,

WONGDEETHAI, N., CHHENG, K., CHANTRATITA, N., PUTCHHAT, H.,

THAIPADUNGPANIT, J., DAY, N. P. & PEACOCK, S. J. 2011. Emergence of

community-associated methicillin-resistant Staphylococcus aureus carriage in

children in Cambodia. Am J Trop Med Hyg, 84, 313-7.

133

NICOLLE, L. E., POSTL, B., URIAS, B., LAW, B. & LING, N. 1990. Group A

streptococcal pharyngeal carriage, pharyngitis, and impetigo in two northern

Canadian native communities. Clin Invest Med, 13, 99-106.

NIMMO, G. R., TINNISWOOD, R. D., NUTTALL, N., BAKER, G. M. &

MCDONALD, B. 1992. Group A streptococcal infection in an aboriginal

community. Med J Aust, 157, 521-2.

NIRMAL, B., PAI, S. B. & SRIPATHI, H. 2013. A simple instrument designed

to provide consistent digital facial images in dermatology. Indian J Dermatol, 58,

194-6.

NORAZAH, A., NORMAZNAH, Y., KAMEL, A. G. & ROHANI, M. Y. 1995.

Streptococcal impetigo among aboriginal children in Malaysia. Southeast Asian J

Trop Med Public Health, 26, 803-4.

NORHAYATI BINTI MOKTAR, M., NOOR HAYATI, M. I., NOR FARIZA,

N., ROHANI, A. K., HALIMAH, A. S., SHAROM, M. Y. & ZAINAL ABIDIN,

A. H. 1998. Health status of Orang Asli (aborigine) community in Pos Piah,

Sungai Siput, Perak, Malaysia. Southeast Asian J Trop Med Public Health, 29,

58-61.

O'BRIEN, K. L., BRONSDON, M. A., DAGAN, R., YAGUPSKY, P., JANCO,

J., ELLIOTT, J., WHITNEY, C. G., YANG, Y. H., ROBINSON, L. G.,

SCHWARTZ, B. & CARLONE, G. M. 2001. Evaluation of a medium (STGG)

for transport and optimal recovery of Streptococcus pneumoniae from

nasopharyngeal secretions collected during field studies. J Clin Microbiol, 39,

1021-4.

O'MALLEY, T., BAILS, D., CHADWICK, D. ON BEHALF OF THE

ELECTRONIC EVIDENCE SPECIALIST ADVISORY GROUP. 2004.

Australasian Guidelines for Digital Imaging Processes. Version 2. Available:

http://www.nifs.com.au/2004%20Digital%20Imaging%20Guidelines.pdf.

134

OGUNBIYI, A. O., OMIGBODUN, Y. & OWOAJE, E. 2009. Prevalence of

skin disorders in school children in southwest Nigeria. Int J Adolesc Med Health,

21, 235-41.

OKAFOR, O. O., AKINBAMI, F. O., ORIMADEGUN, A. E., OKAFOR, C. M.

& OGUNBIYI, A. O. 2011. Prevalence of dermatological lesions in hospitalized

children at the University College Hospital, Ibadan, Nigeria. Niger J Clin Pract,

14, 287-92.

PARKS, T., SMEESTERS, P. R. & STEER, A. C. 2012. Streptococcal skin

infection and rheumatic heart disease. Curr Opin Infect Dis, 25, 145-53.

PARNABY, M. G. & CARAPETIS, J. R. 2010. Rheumatic fever in indigenous

Australian children. J Paediatr Child Health, 46, 527-33.

PATEL, R. B., UDANI, R. H. & KHANNA, S. A. 1982. Pediatric dermatoses

and eradication in slums. Indian J Pediatr, 49, 135-9.

PEEL, M. C., Finlayson, B. L., & MCMAHON T.A. 2007. Updated world map

of the Koppen-Geiger climate classification. Hydrology and Earth System

Sciences Discussion, 11, 1633-1644.

PEREZ-GORRICHO, B. & RIPOLL, M. 2003. Does short-course antibiotic

therapy better meet patient expectations? Int J Antimicrob Agents, 21, 222-8.

PHAKADE, R. S., NATARAJ, G., KUYARE, S. S., KHOPKAR, U. S. &

MEHTA, P. R. 2012. Is methicillin-resistant Staphylococcus aureus involved in

community acquired skin and soft tissue infections? Experience from a tertiary

care centre in Mumbai. J Postgrad Med, 58, 3-7.

PIAGGIO, G., ELBOURNE, D. R., POCOCK, S. J., EVANS, S. J., ALTMAN,

D. G. & GROUP, C. 2012. Reporting of noninferiority and equivalence

randomized trials: extension of the CONSORT 2010 statement. JAMA, 308,

2594-604.

135

PORTER, M. J. 1980. Seasonal change and its effect on the prevalence of

infectious skin disease in a Gambian village. Trans R Soc Trop Med Hyg, 74,

162-8.

PORTER, M. J., MACK, R. W. & CHAUDHARY, M. A. 1984. Pediatric skin

disease in Pakistan. A study of three Punjab villages. Int J Dermatol, 23, 613-6.

PROCTOR, R. A. 2008. Role of folate antagonists in the treatment of

methicillin-resistant Staphylococcus aureus infection. Clin Infect Dis, 46, 584-93.

RAZMJOU, R. G., WILLEMSEN, S. P., KONING, S., ORANJE, A. P.,

SCHELLEVIS, F. & VAN DER WOUDEN, J. C. 2009. Determinants of

regional differences in the incidence of impetigo. Environ Res, 109, 590-3.

ROBERG, N. 1946. An epidemic caused by a sulfadiazine resistant strain of the

streptococcus hemolyticus (group A, type 17). The Journal of Infectious

Diseases, 78, 135 - 146.

RORTVEIT, S., SKUTLABERG, D. H., LANGELAND, N. & RORTVEIT, G.

2011. Impetigo in a population over 8.5 years: incidence, fusidic acid resistance

and molecular characteristics. J Antimicrob Chemother, 66, 1360-4.

ROSSNER, M. & YAMADA, K. M. 2004. What's in a picture? The temptation

of image manipulation. J Cell Biol, 166, 11-5.

RUBBO, S. D. & BENJAMIN, M. 1951. Some observations on survival of

pathogenic bacteria on cotton-wool swabs; development of a new type of swab.

Br Med J, 1, 983-7.

RUBIN, R. H. & SWARTZ, M. N. 1980. Trimethoprim-sulfamethoxazole. N

Engl J Med, 303, 426-32.

RUOFF KL, W. R., BEIGHTON D 2003. Streptococcus, In: Murray, P.R., Baron

E.J., Jorgensen J.H., Pfaller, M.A., & Yolken, R.H. (eds) Manual of Clinical

Microbiology. Washington, ASM Press.

RYAN, T. J. 2008. One of the greatest of health needs without effective

advocacy and shamefully neglected! Br J Dermatol, 158, 205-7.

136

SAEGEMAN, V., FLAMAING, J., MULLER, J., PEETERMANS, W. E.,

STUYCK, J. & VERHAEGEN, J. 2011. Clinical evaluation of the Copan ESwab

for methicillin-resistant Staphylococcus aureus detection and culture of wounds.

Eur J Clin Microbiol Infect Dis, 30, 943-9.

SAN SEBASTIAN, M. & SANTI, S. 1999. The health status of rural school

children in the Amazon basin of Ecuador. J Trop Pediatr, 45, 379-82.

SANDLER, J., DWYER, J., KOKICH, V., MCKEOWN, F., MURRAY, A.,

MCLAUGHLIN, R., O'BRIEN, C. & O'MALLEY, P. 2009. Quality of clinical

photographs taken by orthodontists, professional photographers, and orthodontic

auxiliaries. Am J Orthod Dentofacial Orthop, 135, 657-62.

SANDLER, J., GUTIERREZ, R. J. & MURRAY, A. 2012. Clinical photographs:

the gold standard, an update. Prog Orthod, 13, 296-303.

SAW, S. M., KOH, D., ADJANI, M. R., WONG, M. L., HONG, C. Y., LEE, J.,

CHIA, S. E., MUNOZ, C. P. & ONG, C. N. 2001. A population-based

prevalence survey of skin diseases in adolescents and adults in rural Sumatra,

Indonesia, 1999. Trans R Soc Trop Med Hyg, 95, 384-8.

SCHMELLER, W. 1998. Community health workers reduce skin diseases in

East African children. Int J Dermatol, 37, 370-7.

SCHMELLER, W. & DZIKUS, A. 2001. Skin diseases in children in rural

Kenya: long-term results of a dermatology project within the primary health care

system. Br J Dermatol, 144, 118-24.

SCHULTZ, I. & FRANK, P. F. 1958. The prophylactic use of

sulfamethoxypyridazine (kynex) during a streptococcal epidemic. Am J Med Sci,

236, 779-85.

SECKELER, M. D. & HOKE, T. R. 2011. The worldwide epidemiology of acute

rheumatic fever and rheumatic heart disease. Clin Epidemiol, 3, 67-84.

SHALLCROSS, L. J., PETERSEN, I., ROSENTHAL, J., JOHNSON, A. M.,

FREEMANTLE, N. & HAYWARD, A. C. 2013. Use of primary care data for

137

detecting impetigo trends, United kingdom, 1995-2010. Emerg Infect Dis, 19,

1646-8.

SHAPTER, M. 2004. Placement considerations when using linear scales in

photographs. J Audiov Media Med, 27, 54-7.

SHARMA, N. K., GARG, B. K. & GOEL, M. 1986. Pattern of Skin Diseases in

Urban School Children. Indian Journal of Dermatology, Venereology and

Leprology, 52, 330 - 331.

SHARMA, R. S., MISHRA, R. S., PAL, D., GUPTA, J. P., DUTTA, M. &

DATTA, K. K. 1984. An epidemiological study of scabies in a rural community

in India. Ann Trop Med Parasitol, 78, 157-64.

SHEA, IDSA & PIDS 2012. Policy statement on antimicrobial stewardship by

the Society for Healthcare Epidemiology of America (SHEA), the Infectious

Diseases Society of America (IDSA), and the Pediatric Infectious Diseases

Society (PIDS). Infect Control Hosp Epidemiol, 33, 322-7.

SHELBY-JAMES, T. M., LEACH, A. J., CARAPETIS, J. R., CURRIE, B. J. &

MATHEWS, J. D. 2002. Impact of single dose azithromycin on group A

streptococci in the upper respiratory tract and skin of Aboriginal children.

Pediatr Infect Dis J, 21, 375-80.

SHERIDAN, P. 2013. Practical aspects of clinical photography: part 1--

principles, equipment and technique. ANZ J Surg, 83, 188-91.

SHIM, J., LANIER, J. & QUI, M. K. 2014. Clinical inquiry: what is the best

treatment for impetigo ? J Fam Pract, 63, 333-5.

SKULL, S. A., KRAUSE, V., COOMBS, G., PEARMAN, J. W. & ROBERTS,

L. A. 1999. Investigation of a cluster of Staphylococcus aureus invasive infection

in the top end of the Northern Territory. Aust N Z J Med, 29, 66-72.

SLADDEN, M. J. & JOHNSTON, G. A. 2004. Common skin infections in

children. BMJ, 329, 95-9.

138

SMITH-VAUGHAN, H., BYUN, R., NADKARNI, M., JACQUES, N. A.,

HUNTER, N., HALPIN, S., MORRIS, P. S. & LEACH, A. J. 2006. Measuring

nasal bacterial load and its association with otitis media. BMC Ear Nose Throat

Disord, 6, 10.

SPEERT, D. P., WANNAMAKER, L. W., GRAY, E. D. & CLAWSON, C. C.

1979. Bactericidal effect of oleic acid on group A streptococci: mechanism of

action. Infect Immun, 26, 1202-10.

STEELE, R. W., RUSSO, T. M. & THOMAS, M. P. 2006. Adherence issues

related to the selection of antistaphylococcal or antifungal antibiotic suspensions

for children. Clin Pediatr (Phila), 45, 245-50.

STEER, A. C., ADAMS, J., CARLIN, J., NOLAN, T. & SHANN, F. 1999.

Rheumatic heart disease in school children in Samoa. Arch Dis Child, 81, 372.

STEER, A. C., JENNEY, A. W., KADO, J., BATZLOFF, M. R., LA

VINCENTE, S., WAQATAKIREWA, L., MULHOLLAND, E. K. &

CARAPETIS, J. R. 2009a. High burden of impetigo and scabies in a tropical

country. PLoS Negl Trop Dis, 3, e467.

STEER, A. C., TIKODUADUA, L. V., MANALAC, E. M., COLQUHOUN, S.,

CARAPETIS, J. R. & MACLENNAN, C. 2009b. Validation of an Integrated

Management of Childhood Illness algorithm for managing common skin

conditions in Fiji. Bull World Health Organ, 87, 173-9.

STENDERUP A, S. E., MATAIKA J 1983. Dermatomycosis and other skin

diseases in a Fijian village. Fiji Medical Journal, 18 - 19.

STEVENS, D. L., BISNO, A. L., CHAMBERS, H. F., DELLINGER, E. P.,

GOLDSTEIN, E. J., GORBACH, S. L., HIRSCHMANN, J. V., KAPLAN, S. L.,

MONTOYA, J. G. & WADE, J. C. 2014. Practice guidelines for the diagnosis

and management of skin and soft tissue infections: 2014 update by the infectious

diseases society of America. Clin Infect Dis, 59, 147-59.

STOKES, A. & LACEY, R. W. 1978. Effect of thymidine on activity of

trimethoprim and sulphamethoxazole. J Clin Pathol, 31, 165-71.

139

STREETON, C. L., HANNA, J. N., MESSER, R. D. & MERIANOS, A. 1995.

An epidemic of acute post-streptococcal glomerulonephritis among aboriginal

children. J Paediatr Child Health, 31, 245-8.

SVARTMAN, M., FINKLEA, J. F., EARLE, D. P., POTTER, E. V. & POON-

KING, T. 1972. Epidemic scabies and acute glomerulonephritis in Trinidad.

Lancet, 1, 249-51.

SWEDBERG, G., RINGERTZ, S. & SKOLD, O. 1998. Sulfonamide resistance

in Streptococcus pyogenes is associated with differences in the amino acid

sequence of its chromosomal dihydropteroate synthase. Antimicrob Agents

Chemother, 42, 1062-7.

TAPLIN, D., LANSDELL, L., ALLEN, A. M., RODRIGUEZ, R. & CORTES,

A. 1973. Prevalence of streptococcal pyoderma in relation to climate and

hygiene. Lancet, 1, 501-3.

TAPLIN, D., PORCELAIN, S. L., MEINKING, T. L., ATHEY, R. L., CHEN, J.

A., CASTILLERO, P. M. & SANCHEZ, R. 1991. Community control of

scabies: a model based on use of permethrin cream. Lancet, 337, 1016-8.

THE NATIONAL HEALTH AND MEDICAL RESEARCH COUNCIL. 2007

(updated 2013). National Statement on Ethical Conduct in Human Research

2006. Canberra: Commonwealth of Australia.

THOMAS, J. & KUMAR, P. 2013. The scope of teledermatology in India.

Indian Dermatol Online J, 4, 82-9.

THOMAS, M., WOODFIELD, G., MOSES, C. & AMOS, G. 2005. Soil-

transmitted helminth infection, skin infection, anaemia, and growth retardation in

schoolchildren of Taveuni Island, Fiji. N Z Med J, 118, U1492.

TIBAZARWA, K. B., VOLMINK, J. A. & MAYOSI, B. M. 2008. Incidence of

acute rheumatic fever in the world: a systematic review of population-based

studies. Heart, 94, 1534-40.

140

TONG, S. Y., ANDREWS, R. M., KEARNS, T., GUNDJIRRYIRR, R.,

MCDONALD, M. I., CURRIE, B. J. & CARAPETIS, J. R. 2010. Trimethopim-

sulfamethoxazole compared with benzathine penicillin for treatment of impetigo

in Aboriginal children: a pilot randomised controlled trial. J Paediatr Child

Health, 46, 131-3.

TONG, S. Y., BISHOP, E. J., LILLIEBRIDGE, R. A., CHENG, A. C.,

SPASOVA-PENKOVA, Z., HOLT, D. C., GIFFARD, P. M., MCDONALD, M.

I., CURRIE, B. J. & BOUTLIS, C. S. 2009. Community-associated strains of

methicillin-resistant Staphylococcus aureus and methicillin-susceptible S. aureus

in indigenous Northern Australia: epidemiology and outcomes. J Infect Dis, 199,

1461-70.

TONG, S. Y., MCDONALD, M. I., HOLT, D. C. & CURRIE, B. J. 2008. Global

implications of the emergence of community-associated methicillin-resistant

Staphylococcus aureus in Indigenous populations. Clin Infect Dis, 46, 1871-8.

TONG, S. Y., SHARMA-KUINKEL, B. K., THADEN, J. T., WHITNEY, A. R.,

YANG, S. J., MISHRA, N. N., RUDE, T., LILLIEBRIDGE, R. A., SELIM, M.

A., AHN, S. H., HOLT, D. C., GIFFARD, P. M., BAYER, A. S., DELEO, F. R.

& FOWLER, V. G., JR. 2013. Virulence of endemic nonpigmented northern

Australian Staphylococcus aureus clone (clonal complex 75, S. argenteus) is not

augmented by staphyloxanthin. J Infect Dis, 208, 520-7.

TONG, S. Y., STEER, A. C., JENNEY, A. W. & CARAPETIS, J. R. 2011.

Community-associated methicillin-resistant Staphylococcus aureus skin

infections in the tropics. Dermatol Clin, 29, 21-32.

TONG, S. Y., VARRONE, L., CHATFIELD, M. D., BEAMAN, M. &

GIFFARD, P. M. 2014. Progressive increase in community-associated

methicillin-resistant Staphylococcus aureus in Indigenous populations in

northern Australia from 1993 to 2012. Epidemiol Infect, 1-5.

TONG, SY., SCHAUMBERG, F., ELLINGTON, M.J., CORANDER, J.,

PICHON, B., LEENDERTZ, F., BENTLEY, S.D., PARKHILL, J., HOLT, D.C.,

PETERS, G. & GIFFARD, P.M. 2014a. Novel staphylococcal species that form

141

part of a Staphylococcus aureus related complex: the non-pigmented S. argenteus

sp. nov. and the non-human primate associated s. schweitzeri sp. nov. Int J Syst

Evol Microbiol.

TRAUB, W. H. & LEONHARD, B. 1997. Comparative susceptibility of clinical

group A, B, C, F, and G beta-hemolytic streptococcal isolates to 24 antimicrobial

drugs. Chemotherapy, 43, 10-20.

TRICKETT, P. C. D., P. MOGABGAB, W. 1973. Trimethoprim-

Sulfamethoxazole versus Penicillin G in the Treatment of Group A Beta-

Hemolytic Streptococcal Pharyngitis and Tonsillitis. The Journal of Infectious

Diseases, 128, S693-S695.

TUNNESSEN, W. W., JR. 1985. Practical aspects of bacterial skin infections in

children. Pediatr Dermatol, 2, 255-65.

TURNIDGE, J. 2010. Trimethoprim, Co-trimoxazole (Co-T) and Related

Agents. In: GRAYSON, M. L., CROWE, S. M., MCCARTHY, J.S., MILLS, J,

MOUTON, J.W., RAGNAR NORRBY, S., PATERSON, D.L., PFALLER, M.A.

(ed.) Kucers' The Use of Antibiotics. Sixth ed. Washington: ASM Press

UDO, E. E., PEARMAN, J. W. & GRUBB, W. B. 1994. Emergence of high-

level mupirocin resistance in methicillin-resistant Staphylococcus aureus in

Western Australia. J Hosp Infect, 26, 157-65.

VALERY, P. C., WENITONG, M., CLEMENTS, V., SHEEL, M.,

MCMILLAN, D., STIRLING, J., SRIPRAKASH, K. S., BATZLOFF, M.,

VOHRA, R. & MCCARTHY, J. S. 2008. Skin infections among Indigenous

Australians in an urban setting in far North Queensland. Epidemiol Infect, 136,

1103-8.

VAN BUYNDER, P. G., GAGGIN, J. A., MARTIN, D., PUGSLEY, D. &

MATHEWS, J. D. 1992. Streptococcal infection and renal disease markers in

Australian aboriginal children. Med J Aust, 156, 537-40.

VAN DER WOUDEN, J. C. & KONING, S. 2014. Treatment of impetigo in

resource-limited settings. The Lancet.

142

VAN HORN, K. G., AUDETTE, C. D., SEBECK, D. & TUCKER, K. A. 2008a.

Comparison of the Copan ESwab system with two Amies agar swab transport

systems for maintenance of microorganism viability. J Clin Microbiol, 46, 1655-

8.

VAN HORN, K. G., AUDETTE, C. D., TUCKER, K. A. & SEBECK, D. 2008b.

Comparison of 3 swab transport systems for direct release and recovery of

aerobic and anaerobic bacteria. Diagn Microbiol Infect Dis, 62, 471-3.

VOS, T., FLAXMAN, A. D., NAGHAVI, M., et al. 2012. Years lived with

disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a

systematic analysis for the Global Burden of Disease Study 2010. Lancet, 380,

2163-96.

WALKER, S. L., SHAH, M., HUBBARD, V. G., PRADHAN, H. M. &

GHIMIRE, M. 2008. Skin disease is common in rural Nepal: results of a point

prevalence study. Br J Dermatol, 158, 334-8.

WANNAMAKER, L. W. 1970. Differences between streptococcal infections of

the throat and of the skin. I. N Engl J Med, 282, 23-31.

WASSERZUG, O., VALINSKY, L., KLEMENT, E., BAR-ZEEV, Y.,

DAVIDOVITCH, N., ORR, N., KORENMAN, Z., KAYOUF, R., SELA, T.,

AMBAR, R., DERAZNE, E., DAGAN, R. & ZARKA, S. 2009. A cluster of

ecthyma outbreaks caused by a single clone of invasive and highly infective

Streptococcus pyogenes. Clin Infect Dis, 48, 1213-9.

WHITE, A. V., HOY, W. E. & MCCREDIE, D. A. 2001. Childhood post-

streptococcal glomerulonephritis as a risk factor for chronic renal disease in later

life. Med J Aust, 174, 492-6.

WIERTSEMA, S. P., CHIDLOW, G. R., KIRKHAM, L. A., CORSCADDEN,

K. J., MOWE, E. N., VIJAYASEKARAN, S., COATES, H. L., HARNETT, G.

B. & RICHMOND, P. C. 2011a. High detection rates of nucleic acids of a wide

range of respiratory viruses in the nasopharynx and the middle ear of children

with a history of recurrent acute otitis media. J Med Virol, 83, 2008-17.

143

WIERTSEMA, S. P., KIRKHAM, L. A., CORSCADDEN, K. J., MOWE, E. N.,

BOWMAN, J. M., JACOBY, P., FRANCIS, R., VIJAYASEKARAN, S.,

COATES, H. L., RILEY, T. V. & RICHMOND, P. 2011b. Predominance of

nontypeable Haemophilus influenzae in children with otitis media following

introduction of a 3+0 pneumococcal conjugate vaccine schedule. Vaccine, 29,

5163-70.

WILSON, A. T. 1945. Method for testing in vitro resistance of group A

hemolytic streptococci to sulfonamides. Proceedings of the Society for

Experimental Biology and Medicine, 58, 130 - 133.

WILSON, A. T. 1954. A simplified method for testing sulfonamide resistance of

group A streptococci. J Lab Clin Med, 43, 280-5.

WITMER, W. K. & LEBOVITZ, P. J. 2012. Clinical photography in the

dermatology practice. Semin Cutan Med Surg, 31, 191-9.

WONG, L. C., AMEGA, B., CONNORS, C., BARKER, R., DULLA, M. E.,

NINNAL, A., KOLUMBOORT, L., CUMAIYI, M. M. & CURRIE, B. J. 2001.

Outcome of an interventional program for scabies in an Indigenous community.

Med J Aust, 175, 367-70.

WU, Y. H., SU, H. Y. & HSIEH, Y. J. 2000. Survey of infectious skin diseases

and skin infestations among primary school students of Taitung County, eastern

Taiwan. J Formos Med Assoc, 99, 128-34.

WYBER, R. T., K. MARKO, S. KAPLAN, E.L. 2013. Benzathine Penicillin G

for the management of RHD: Concerns about quality and access, and

opportunities for intervention and improvement. Global Heart, 8, 227 - 234.

YAMAMAH, G. A., EMAM, H. M., ABDELHAMID, M. F., ELSAIE, M. L.,

SHEHATA, H., FARID, T., KAMEL, M. I. & TAALAT, A. A. 2012.

Epidemiologic study of dermatologic disorders among children in South Sinai,

Egypt. Int J Dermatol, 51, 1180-5.

YOUNG, S. 2008. Research for medical illustrators: designing photographic

protocols. J Vis Commun Med, 31, 99-102.

144

YOUNG, S. 2009. Research for medical illustrators: Using linear scales in

photographs. Journal of Visual Communication in Medicine, 32, 4-7.

YOURASSOWSKY, E., VANDERLINDEN, M. P. & SCHOUTENS, E. 1974.

Sensitivity of Streptococcus pyogenes to sulphamethoxazole, trimethoprim, and

cotrimoxazole. J Clin Pathol, 27, 897-901.

YRIOS, J. W., BALISH, E., HELSTAD, A., FIELD, C. & INHORN, S. 1975.

Survival of anaerobic and aerobic bacteria on cotton swabs in three transport

systems. J Clin Microbiol, 1, 196-200.

ZERBINO, D. R. & BIRNEY, E. 2008. Velvet: algorithms for de novo short read

assembly using de Bruijn graphs. Genome Res, 18, 821-9.

145

APPENDICES

147

APPENDIX 1

Table: Overall and childhood pyoderma and scabies prevalence from 89 studies synthesised in the Chapter 2 systematic review.

Overall Children (0—15 years)

Country Year Number studied

Number with

pyoderma

Pyoderma prevalence

(%) Number studied

Number with

pyoderma Pyoderma prevalence

Scabies prevalence

India 1967 488 49 10.0 488 49 10.0 NS

USA* 1969 39 16 41.0 39 16 41.0 NS

USA* 1970 444 235 52.9 444 235 52.9 NS

USA* 1970 31 6 19.4 31 6 19.4 NS

Colombia 1971 1,269 169 13.3 1,269 169 13.3 NS

USA* 1971 80 12 15.0 80 12 15.0 NS

Panama 1972 892 135 15.1 628 111 17.7 0.60%

Panama 1972 131 17 13.0 NS NS NS NS

India 1973 6,641 392 5.9 6,641 392 5.9 NS

Tanzania 1973 532 37 7.0 532 37 7.0 31%

New Zealand* 1974 453 6 1.3 453 6 1.3 0

India 1974 1,038 176 17.0 1,038 176 17.0 Common

Brazil 1974 9,955 1211 12.2 9,955 1,211 12.2 3%

Ghana 1975 3,770 731 19.4 NS NS NS NS

Tanzania 1975^ 794 38 4.8 794 38 4.8 12.90%

Tanzania 1975^ 305 22 7.2 305 22 7.2 0.70%

Tanzania 1975^ 120 8 6.7 120 8 6.7 3.30%

Tanzania 1975^ 636 60 9.4 636 60 9.4 31.60%

Brazil 1976 775 87 11.2 307 64 20.8 NS

India 1976 89 11 12.4 89 11 12.4 NS

India 1976 19,775 586 3.0 19,775 586 3.0 NS

Gambia 1976 974 87 8.9 332 58 17.5 2.50%

USA* 1976 593 79 13.3 593 79 13.3 NS

USA* 1976 434 63 14.5 434 63 14.5 NS

India 1977 1,890 79 4.2 1,890 79 4.2 8.20%

Gambia 1977 945 68 7.2 387 39 10.1 2.00%

USA* 1977 535 104 19.4 535 104 19.4 NS

India 1978 2,771 59 2.1 1,509 48 3.2 13%

Pakistan 1980 444 86 19.4 444 86 19.4 1.90%

India 1982 4,133 431 10.4 NS NS NS 0.12%

Nigeria 1983 2,407 87 3.6 2,407 87 3.6 NS

Fiji 1983^ 61 14 23.0 61 14 23.0 33%

Canada* 1984 258 2 0.8 258 2 0.8 NS

149



Number with

pyoderma

Pyoderma prevalence

(%) Number studied

Number with


Scabies prevalence

Canada* 1984 378 8 2.1 378 8 2.1 NS

Solomon Islands 1984 10,224 4,370 42.7 5,160 2678 51.9 1.30%

Canada* 1985 239 10 4.2 239 10 4.2 NS

Canada* 1985 202 2 1.0 202 2 1.0 NS

Canada* 1985 336 13 3.9 336 13 3.9 NS

Canada* 1985 334 2 0.6 334 2 0.6 NS

India 1986^ 3,697 39 1.1 3,697 39 1.1 0.50%

Indi 1988 666 107 16.1 666 107 16.1 0.90%

Vanuatu 1989 18,223 2133 11.7 9,569 1537 16.1 16%

Ethiopia 1989 1,842 417 22.6 1,842 417 22.6 1.70%

Australia* 1990 120 52 43.3 120 52 43.3 NS

Tanzania 1991 936 178 19.0 NS NS NS 5.98%

Honduras 1992 206 43 20.9 206 43 20.9 10%

Ethiopia 1992 112 6 5.4 112 6 5.4 17%

Australia* 1992^ 180 31 17.2 180 31 17.2 NS

Australia* 1993 583 198 34.0 583 198 34.0 NS

Mali 1993 1,817 224 12.3 1,817 224 12.3 4.30%

Kenya 1993 5,780 735 12.7 5,780 735 12.7 8.30%

Ecuador 1993 495 22 4.4 495 22 4.4 NS

Tanzania 1994 1,114 18 1.6 NS NS NS 4.50%

Australia* 1994 81 39 48.1 81 39 48.1 NS

Australia* 1994 126 62 49.2 62 43 69.4 28.80%

Kenya 1995 4,358 470 10.8 4,358 470 10.8 10.50%

Nepal 1995 458 76 16.6 458 76 16.6 NS

Australia* 1995 79 71 89.9 79 71 89.9 23%

Malaysia 1995^ 41 8 19.5 41 8 19.5 NS

Solomon Islands

1997 226 91 40.3 226 91 40.3 25%

Taiwan 1998 3,029 72 2.4 3,029 72 2.4 1.40%

Malaysia 1998^ 356 24 6.7 NS NS NS 11.90%

Samoa 1998^ 8,767 3,822 43.6 8,767 3822 43.6 4.90%

Kenya 1999 4,961 563 11.3 4,961 563 11.3 8.00%

Australia* 2000 217 49 22.6 217 49 22.6 35%

Australia* 2000 121 80 66.1 121 80 66.1 5.0%

India 2000 4,249 28 0.7 4,249 28 0.7 NS

Egypt 2001 636 99 15.6 636 99 15.6 0.50%

Egypt 2001 720 6 0.8 720 6 0.8 0.10%

India 2001 12,586 925 7.3 12,586 925 7.3 5%

Mali 2001 1,729 310 17.9 1,729 310 17.9 1.30%

150



Number with

pyoderma

Pyoderma prevalence

(%) Number studied

Number with


Scabies prevalence

Mali 2002 1,632 304 18.6 1,632 304 18.6 0.70%

Turkey 2002^ 785 16 2.0 785 16 2.0 2.16%

Tanzania 2003 820 69 8.4 820 69 8.4 1.50%

Fiji 2004 258 6 2.3 258 6 2.3 32.6%

Australia* 2004 582 266 45.7 582 266 45.7 16.10%

Ghana 2004 463 20 4.3 463 20 4.3 0.00%

Gabon 2005 454 7 1.5 454 7 1.5 0.70%

Fiji 2006 3,462 1259 36.4 3,462 1259 36.4 18.50%

Fiji 2006 451 60 13.3 451 60 13.3 14%

Timor Leste 2007 1,535 112 7.3 728 82 11.3 17%

Ghana 2007 1,394 81 5.8 1,394 81 5.8 0.10%

Rwanda 2007 2,528 32 1.3 2,528 32 1.3 0.04%

Nepal 2008^ 878 54 6.2 NS NS NS 3.40%

Nigeria 2009^ 1,415 14 1.0 1,415 14 1.0 0.60%

Australia* 2009^ 2,001 804 40.2 2,001 804 40.2 NS

Cameroon 2010 400 43 10.8 48 2 4.2 1.80%

Ethiopia 2010^ 1,104 29 2.6 1,104 29 2.6 0.30%

Tanzania 2010^ 420 17 4.0 393 16 4.1 1.40%

*Resource-poor populations within high-income OECD countries, ^reflectsyear of publication when year study was commenced is unknown.

151

APPENDIX 2

GELFAND MS, CLEVELAND KO, KETTERER DC. SUSCEPTIBILITY OF STREPTOCOCCUS PYOGENES TO TRIMETHOPRIM-SULFAMETHOXAZOLE. JOURNAL OF CLINICAL MICROBIOLOGY. 2013; 51(4): 1350.

BOWEN AC, TONG SY, CARAPETIS JR. REPLY TO “SUSCEPTIBILITY OF STREPTOCOCCUS PYOGENES TO TRIMETHOPRIM-SULFAMETHOXAZOLE.” JOURNAL OF CLINICAL MICROBIOLOGY. 2013; 51(4): 1351.

153

Susceptibility of Streptococcus pyogenes to Trimethoprim-Sulfamethoxazole

Michael S. Gelfand, Kerry O. Cleveland, Daniel C. Ketterer

Department of Medicine, Division of Infectious Diseases, University of Tennessee Health Science Center, Memphis, Tennessee, USA

With interest we read the recent paper by Bowen et al. in whichthey propose that detection of in vitro susceptibility of Strep-

tococcus pyogenes to trimethoprim-sulfamethoxazole (TMP-SMX) is enhanced by testing on media containing low concentra-tions of thymidine (1). Whether their data on in vitrosusceptibility can be extrapolated to the clinical use of TMP-SMXin clinical infections due to this organism is unclear.

Traditional teaching in infectious diseases and microbiologyhas suggested that S. pyogenes is largely resistant to TMP-SMX,and, in fact, the drug has been incorporated into selective mediafor isolation of S. pyogenes from throat cultures (2–4).

Recently we observed two patients with skin and soft tissueinfections due to S. pyogenes who were treated with oral TMP-SMX and suffered adverse outcomes.

The first patient, a 40-year-old previously healthy man, pre-sented with a hand wound related to an occupational cut. After thewound was cleaned, he was discharged on TMP-SMX (160 mg ofTMP component twice daily). He returned 48 h later with a handinfection that had progressed to necrotizing fasciitis. Wound andblood cultures grew S. pyogenes. Susceptibility testing with TMP-SMX was not done.

The second patient, a 38-year-old woman with multiple skinabscesses, was treated with surgical debridement and tigecycline(50 mg intravenous [i.v.] administration twice daily) for 3 daysfollowed by oral TMP-SMX (160 mg of TMP component twicedaily) for 6 days. A wound culture grew S. pyogenes (no suscepti-bility testing for TMP-SMX was performed) and methicillin-sen-sitive Staphylococcus aureus (MRSA) susceptible to TMP-SMX.She expired after presenting in septic shock 24 h after discharge.

In vitro susceptibility of S. pyogenes to TMP-SMX is less rele-vant than the clinical efficacy of this agent in treating infectionsdue to that organism. Thymidine may become available to infect-ing microorganisms from damaged host tissues and bacteria, al-lowing microorganisms to overcome the metabolic block (5).Clinical failures in the treatment of MRSA infections have beenattributed to this mechanism, and animal model data in MRSAinfections support this explanation (5).

Human sera and urine contain detectable concentrations ofthymidine and folates (6, 7). Even when an organism is demon-strated to have in vitro susceptibility to the drug, use of TMP-SMXhas been discouraged for enterococcal urinary tract infections be-cause of concern about exogenous thymidine and folates beingavailable to the infecting organism (7). As pointed out, TMP-SMX

has not proven to be an effective agent for treatment of strepto-coccal pharyngitis (1).

We urge caution in the use of TMP-SMX for S. pyogenes skinand soft tissue infections before the availability of human clinicalstudies. The study of impetigo being conducted by the authorsmay not resolve the question of clinical utility of TMP-SMX forthese infections, as spontaneous resolution of impetigo is not un-common (8). Data from animal models of infection would bedesirable in helping to resolve this question.

ACKNOWLEDGMENTS

We declare that we have no conflicts of interest.No financial support was received for this work.

REFERENCES1. Bowen AC, Lilliebridge RA, Tong SY, Baird RW, Ward P, McDonald MI,

Currie BJ, Carapetis JR. 2012. Is Streptococcus pyogenes resistant or sus-ceptible to trimethoprim-sulfamethoxazole? J. Clin. Microbiol. 50:4067–4072.

2. Bisno AL, Stevens DL. 2009. Streptococcus pyogenes, p 2593–2610. In Man-dell GL, Bennett JE, Dolin R (ed), Mandell, Douglas, and Bennett’s princi-ples and practice of infectious diseases, 7th ed, vol 2. Churchill LivingstoneElsevier, Philadelphia, PA.

3. Milatovic D. 1981. Comparison of five selective media for beta-haemolyticstreptococci. J. Clin. Pathol. 34:556 –558.

4. Mirrett S, Monahan JS, Reller LB. 1987. Comparative evaluation of me-dium and atmosphere of incubation for isolation of Streptococcus pyogenes.Diagn. Microbiol. Infect. Dis. 6:217–221.

5. Proctor RA. 2008. Role of folate antagonists in the treatment of methicillin-resistant Staphylcoccus aureus infection. Clin. Infect. Dis. 46:584 –593.

6. Holden L, Hoffbrand AV, Tattersall MH. 1980. Thymidine concentra-tions in human sera: variations in patients with leukaemia and megaloblas-tic anaemia. Eur. J. Cancer 16:115–121.

7. Wisell KT, Kahlmeter G, Giske CG. 2008. Trimethoprim and enterococciin urinary tract infections: new perspectives on an old issue. J. Antimicrob.Chemother. 62:35– 40.

8. Koning S, van der Sande R, Verhagen AP, van Suijlekom-Smit LW,Morris AD, Butler CC, Berger M, van der Wouden JC. 2012. Interven-tions for impetigo. Cochrane Database Syst. Rev. 1:CD003261. doi:10.1002/14651858.CD003261.pub3.

Address correspondence to Kerry O. Cleveland, [email protected].

For the author reply, see doi:10.1128/JCM.03373-12.


doi:10.1128/JCM.03329-12

LETTER TO THE EDITOR

1350 jcm.asm.org Journal of Clinical Microbiology p. 1350 April 2013 Volume 51 Number 4

on May 1, 2014 by guest

http://jcm.asm

.org/D

ownloaded from

154

http://dx.doi.org/10.1002/14651858.CD003261.pub3

http://dx.doi.org/10.1002/14651858.CD003261.pub3



http://jcm.asm.org

http://jcm.asm.org/

Reply to “Susceptibility of Streptococcus pyogenes to Trimethoprim-Sulfamethoxazole”

Asha C. Bowen,a Steven Y. C. Tong,a Jonathan R. Carapetisb

Menzies School of Health Research, Darwin, Northern Territory, Australiaa; Telethon Institute for Child Health Research, Perth, Western Australia, Australiab

We thank Gelfand et al. (1) for their interest in our article onthe in vitro susceptibility of Streptococcus pyogenes to trim-

ethoprim-sulfamethoxazole (SXT) (2). We agree that the twocases outlined by Gelfand et al. act as cautionary points for clinicalpractice, as there are currently no good clinical trial data to sup-port the use of SXT for the treatment of S. pyogenes skin and softtissue infections (SSTI). Our aim in publishing the in vitro data isto challenge the misconception that all S. pyogenes isolates areinherently resistant to SXT and hence open the way for clinicaltrials to more precisely determine the role for SXT in the treat-ment of S. pyogenes infections.

In the two clinical scenarios presented by Gelfand et al., theabsence of SXT susceptibility data for the S. pyogenes isolates isconcerning. Other reasons for treatment failure that are not elu-cidated include incomplete adherence, poor absorption, and thereality that the conditions of both of these patients might havedeteriorated regardless of the antibacterial used. One wonderswhy the authors prescribed SXT for these SSTI despite their un-derstanding of the traditional teaching that S. pyogenes is largelyresistant to SXT. Perhaps it was due to the need for a simple,palatable, oral antibacterial regimen that would treat undifferen-tiated SSTI, in the era of highly methicillin-resistant Staphylococ-cus aureus (MRSA) prevalence. We are in agreement with Gelfandet al. that human clinical studies are urgently needed to respond tothis challenge.

The results of our clinical trial comparing oral SXT with intra-muscular benzathine penicillin for impetigo treatment will in-form treatment decisions and generate confidence in the use ofSXT for impetigo treatment in our region and possibly beyond. Todate, with recruitment of 660 participants completed, despite two-thirds of the participants presenting with severe impetigo, nonehave been hospitalized with invasive S. pyogenes infection due topresumed treatment failure. Although the numbers in our studyare not large enough to detect a trend for bacteremia or othercomplications of inadequate treatment, we might have seen clin-ical outcomes similar to those Gelfand et al. have experienced if

SXT really has no role in the treatment of impetigo. Spontaneousresolution of mild impetigo may occur; however, the natural his-tory of impetigo, particularly when severe, is neither completelyunderstood nor benign (3).

We agree that the role of thymidine and its availability in thecontext of damaged host tissues (4) are unknown with respectto the treatment of SSTI with SXT. There are now numerous trialsunder way that involve the use of SXT for SSTI (http://clinicaltrials.gov/), and the results are eagerly awaited to betterdefine the clinical utility of SXT in such settings.

Rather than claiming in vitro data being less relevant than clin-ical efficacy, we would argue that in vitro data should be used toinform the design of appropriate clinical trials to determine theclinical role for SXT, particularly in an era with a truncated anti-microbial pipeline and high rates of mixed SSTI due to MRSA andS. pyogenes.

REFERENCES1. Gelfand MS, Cleveland KO, Ketterer DC. 2013. Susceptibility of Strepto-

coccus pyogenes to trimethoprim-sulfamethoxazole. J. Clin. Microbiol. 51:1350.

2. Bowen AC, Lilliebridge RA, Tong SYC, Baird R, Ward P, MacDonald M,Currie B, Carapetis JR. 2012. Is Streptococcus pyogenes resistant or sus-ceptible to trimethoprim-sulfamethoxazole? J. Clin. Microbiol. 50:4067–4072.

3. Koning S, van der Sande R, Verhagen AP, van Suijlekom-Smit LW,Morris AD, Butler CC, Berger M, van der Wouden JC. 2012. Interven-tions for impetigo. Cochrane Database Syst. Rev. 1:CD003261. doi:10.1002/14651858.pub3.

4. Proctor RA. 2008. Role of folate antagonists in the treatment of methicillin-resistant Staphylococcus aureus infection. Clin. Infect. Dis. 46:584 –593.

Address correspondence to Asha C. Bowen, [email protected].

This is a response to a letter by Gelfand et al. (doi:10.1128/JCM.03329-12).


doi:10.1128/JCM.03373-12

AUTHOR REPLY

Journal of Clinical Microbiology p. 1351April 2013 Volume 51 Number 4 jcm.asm.org 1351

155


http://clinicaltrials.gov/

http://clinicaltrials.gov/

http://dx.doi.org/10.1002/14651858.pub3

http://dx.doi.org/10.1002/14651858.pub3



http://jcm.asm.org

APPENDIX 3

VAN DER WOUDEN J AND KONING S. TREATMENT OF IMPETIGO IN RESOURCE-LIMITED SETTINGS. LANCET 2014, AUGUST 26.

157

Comment

1

Treatment of impetigo in resource-limited settingsWith an estimated worldwide prevalence of more than 140 million cases in 2010, impetigo is one of the seven skin disorders ranking in the 50 most common causes of disease.1 After dermatitis or eczema and viral warts, impetigo is the third most common skin disorder in children, and the most common skin infection in young children presenting in general practice in western Europe.2,3 In tropical ar eas, impetigo can be endemic and aff ect a large proportion of the population.4,5 Several Lancet articles have reported on the treatment of impetigo in endemic populations, notably in soldiers in World Wars 1 and 2, including two controlled (although not randomised) clinical trials.6–8

Indigenous Australian people, like other resource-poor populations in tropical areas, have a raised risk of severe complications associated with impetigo, such as post-streptococcal glomerulonephritis, sepsis, and bone and joint infections.9 In The Lancet, Asha Bowen and colleagues10 present the results of their randomised, controlled non-inferiority trial in Indigenous children in remote communities in the Northern Territory of Australia, in which they compared a short course of oral co-trimoxazole to the standard treatment with intramuscular benzathine benzylpenicillin.10 Unlike most previous impetigo studies, more than 70% of the children had severe impetigo (fi ve or more sores), and the lower limbs were the main aff ected area.

508 patients were randomly assigned to a study treatment: 165 to benzathine benzylpenicillin, 175 to twice-daily co-trimoxazole for 3 days, and 168 to once-daily co-trimoxazole for 5 days. The trial’s primary outcome was treatment success at day 7, objectively assessed by comparison of photographs of the sores at day 0 and day 7 by assessors who were masked to treatment and to the order of the photographs. Treatment success was achieved in 133 (85%) of 156 children who received benzathine benzylpenicillin, and in 283 (85%) of 334 who received co-trimoxazole (absolute diff erence 0·5% [95% CI –6·2 to 7·3]), falling within the prespecifi ed margin of non-inferiority (±10%). Results for the two treatment schedules of co-trimoxazole were similar. The study also underlines the key pathogenic role of Streptococcus pyogenes in impetigo in remote communities in Australia, instead of Staphylococcus aureus as seen over past decades in moderate climates. Adverse event profi les diff ered

substantially between the treatment groups, with 30% of children in the benzylpenicillin group reporting injection-site pain, and one hospital admission due to an injection-site abscess, compared with vomiting and rash reported by less than 1% of children who received co-trimoxazole.

Bowen and colleagues’ study,10 with 508 randomised patients, is one of the largest impetigo studies done so far. Our Cochrane review11 included 68 randomised trials, with an average of only 82 patients per study, and only one of the included studies had randomised more than 500 patients with impetigo. Additionally, this new study is one of the few randomised impetigo trials done in a tropical setting for the benefi t of an underprivileged population. Adherence to the prescribed treatment regimens, and follow-up to day 7 with very low attrition, are also commendable.

Cure rates were compared with those in the study done in Mali by Faye and colleagues,12 because Mali is a country where impetigo is also endemic and severe. However, the prevalence of disease might be less important for cure rates than for the risk of reinfection and development of complications.

A question not addressed by Bowen and colleagues’ study10 is whether successful treatment of severe impetigo reduces the risk of developing severe complications such as glomerulonephritis and sepsis, as is often believed to be the case, since this reduction in development of secondary complications is an important motive for treatment.

In conclusion, we believe that Bowen and colleagues’ new study10 in a resource-poor area makes a valuable contribution to the body of evidence about treatment of an important but under-investigated disease. The option of simple, palatable, pain-free, practical treatment of severe impetigo with co-trimoxazole is a welcome result for Indigenous children living in remote communities in Australia, and provides clinicians and patients with a simple regimen for treatment of impetigo in tropical regions where S pyogenes is the main causative agent.

*Johannes C van der Wouden, Sander KoningDepartment of General Practice and Elderly Care Medicine, EMGO Institute for Health and Care Research, VU University Medical Center, 1007 MB, Amsterdam, Netherlands (JCvdW); and Department of General Practice, Erasmus MC, University Medical Center Rotterdam, Netherlands (SK)[email protected]

Published OnlineAugust 27, 2014http://dx.doi.org/10.1016/S0140-6736(14)61395-7

See Online/Articleshttp://dx.doi.org/10.1016/S0140-6736(14)60841-2

Scie

nce

Phot

o Li

brar

y

Streptococcus pyogenes

www.thelancet.com Published online August 27, 2014 http://dx.doi.org/10.1016/S0140-6736(14)61395-7

158

Comment

2 www.thelancet.com Published online August 27, 2014 http://dx.doi.org/10.1016/S0140-6736(14)61395-7

We declare no competing interests.

1 Hay R, Johns N, Williams H, et al. The global burden of skin disease in 2010: an analysis of the prevalence and impact of skin conditions. J Invest Dermatol 2014; 134: 1527–34.

2 Fleming DM, Cross KW, Barley MA. Recent changes in the prevalence of diseases presenting for health care. Br J Gen Pract 2005; 55: 589–95.

3 Mohammedamin RSA, van der Wouden JC, Koning S, et al. Increasing incidence of skin disorders in children? A comparison between 1987 and 2001. BMC Dermatol 2006; 6: 4.

4 Steer AC, Jenney AW, Kado J, et al. High burden of impetigo and scabies in a tropical country. PLoS Negl Trop Dis 2009; 3: e467.

5 Currie BJ, Carapetis JR. Skin infections and infestations in Aboriginal communities in northern Australia. Australas J Dermatol 2000; 41: 139–43.

6 Harvey F. The treatment of impetigo at the front. Lancet 1917; 190: 582. 7 Bigger JW, Hodgson GA. Impetigo contagiosa treated with microcrystalline

sulphathiazole. Lancet 1944; 244: 78–80.

8 Mackenna RM, Cooper-Willis ES. Impetigo contagiosa in the army, treated with microcrystalline sulphathiazole. Lancet 1945; 246: 357–88.

9 Carapetis JR, Walker AM, Hibble M, Sriprakash KS, Currie BJ. Clinical and epidemiological features of group A streptococcal bacteraemia in a region with hyperendemic superfi cial streptococcal infection. Epidemiol Infect 1999; 122: 59–65.

10 Bowen AC, Tong SYC, Andrews RM, et al. Short-course oral co-trimoxazole versus intramuscular benzathine benzylpenicillin for impetigo in a highly endemic region: an open-label, randomised, controlled, non-inferiority trial. Lancet 2014; published online Aug 27. http://dx.doi.org/10.1016/S0140-6736(14)60841-2.

11 Koning S, van der Sande R, Verhagen AP, et al. Interventions for impetigo. Cochrane Database Syst Rev 2012; 1: CD003261.

12 Faye O, Hay RJ, Diawara I, Mahé A. Oral amoxicillin vs. oral erythromycin in the treatment of pyoderma in Bamako, Mali: an open randomized trial. Int J Dermatol 2007; 46: 19–22.

159

the skin sore trial: exploring a better treatment option

Documents