OR

IGIN

AL

AR

TIC

LE

Original Article

Apparent Contrasting Rates of Pharyngitis andPyoderma in Regions where Rheumatic Heart

Disease is Highly PrevalentMalcolm McDonald, MBBS a,∗, Alex Brown, MBBS b, Tahnia Edwards, RN c,

Alex Hope, MBBS d, Maryanne Amu, RN d, Fran Morey, MAppSci e,Bart J. Currie, MBBS a and Jonathan R. Carapetis, MBBS, PhD a

a Menzies School of Health Research, Charles Darwin University, Darwin, Northern Territory, Australiab Menzies School of Health Research, Alice Springs Campus, Charles Darwin University, Northern Territory, Australia

c Centre for Remote Health, Alice Springs, Northern Territory, Australiad The Ltyentye Apurte Health Centre, Ltyentye Apurte, Northern Territory, Australia

e Medvet Science, Alice Springs, Northern Territory, Australia

Background: The aim of the study was to describe the epidemiology of pharyngitis and pyoderma in a Central AustralianAboriginal community with a high prevalence of rheumatic heart disease (RHD) and compare it to communities in theTop End of the Northern Territory.

T

Methods: Following ethics approval and community consultation, selected households were enrolled and visited overa 13-month period. People were asked if they had a sore throat and/or skin sores and asked about current or recent use ofantibiotics; all throats and any pyoderma lesions were swabbed for bacterial culture. Beta-haemolytic streptococci (BHS),including group A streptococcus (GAS), were identified in the central laboratory using standard methods. Householdcrowding was also assessed. Results were then compared to those from the Top End study.

Results: Sore throat was relatively common (480 episodes per 100 person years), although there was only one case ofGAS pharyngitis in 326 consultations. Only 5.5% of children <15 years had pyoderma during the course of the study. Thisis the opposite picture to that reported in the Top End where symptomatic pharyngitis is rare and pyoderma is common.

Conclusions: Although the data are limited, the epidemiology of pharyngitis and pyoderma in this Central AustralianAboriginal community appears to be more akin to that seen in temperate climates rather than tropical Top End commu-nities. In this community, RHD preventative measure should continue to include aggressive treatment of pharyngitisaccording to recommendations.

(Heart, Lung and Circulation 2007;16:254–259)© 2007 Australasian Society of Cardiac and Thoracic Surgeons and the Cardiac Society of Australia and

New Zealand. Published by Elsevier Inc. All rights reserved.

Keywords. Rheumatic fever; Rheumatic heart disease; Beta-haemolytic streptococcus; Pharyngitis; Pyoderma; Aboriginal;Central Australia

Introduction

he Northern Territory (NT) covers a vast area, fromthe monsoonal Top End to the desert and semi-desert

country of Central Australia. The arid zone of Central Aus-tralia covers about 1 million km2. It extends from northof Tennant Creek to the Anangu-Pitjantjatjara lands ofSouth Australia and the Ngaanyatjarra lands of WesternAustralia and does not correspond to state and territoryboundaries (Fig. 1). The population is about 50,000, 40%

Received 23 October 2006; received in revised form 18 December2006; accepted 7 February 2007

∗ Corresponding author at: Menzies School of Health Research,PO Box 41096, Casuarina, NT 0811, Australia. Tel.: +61 8 89228196;fax: +61 8 89227876.E-mail address: malcolm@menzies.edu.au (M. McDonald).

of whom are Aboriginal Australians.1 The majority of theAboriginal population lives outside urban areas there isa high level of mobility with up to 60% of the populationmoving each year.2,3 Across the NT there are marked dif-ferences in Aboriginal language and culture. There is alsogreat variability in quality of community infrastructure,adult literacy and household size, both between regionsand within each region.1

Non-urban Aboriginal communities in both Top Endand Central Australian regions have exceptionally highreported prevalence rates of rheumatic heart disease(RHD) with about 16 per 1000 population (all ages) inCentral Australian and 17 per 1000 in the Top end (datafrom the NT Top End and Central Australian ARF/RHDRegisters).4,5 In contrast, the prevalence rate of RHD inthe non-Aboriginal population is less than 1 per 10,000

© 2007 Australasian Society of Cardiac and Thoracic Surgeons and the Cardiac Society ofAustralia and New Zealand. Published by Elsevier Inc. All rights reserved.

1443-9506/04/$30.00doi:10.1016/j.hlc.2007.02.087

OR

IGIN

AL

AR

TIC

LEHeart, Lung and Circulation McDonald et al. 255

2007;16:254–259 Pharyngitis and pyoderma in a Central Australian Aboriginal community

Figure 1. Map of the Northern Territory showing Central Australia and the Top End regions.

people.6 The reported rates of RHD in Aboriginal com-munities are also likely to be underestimates because asubstantial proportion probably goes undetected.7 Detec-tion is expected to improve with increasing awareness ofacute rheumatic fever (ARF) and RHD and wider use ofsophisticated diagnostic tools, such as portable echocar-diogram machines. On the other hand, the reported ratesof ARF have been falling in the region over the past fouryears (Fig. 2); this is probably due to the activities of theCentral Australian Rheumatic Heart Disease Program andbodes well for the projected prevalence of RHD.

The hypothesis that, in some settings, streptococcalskin infection plays a role in the pathogenesis of acute

FApe

rheumatic fever has been primarily based on data col-lected from Aboriginal communities in the Top End of theNT.8,9 The aim of the whole project, initially conductedin the Top End, was to undertake intensive, prospectivesurveillance of skin and throat infection and colonisa-tion with beta-haemolytic streptococci (BHD) in familiesof ARF/RHD patients in remote Aboriginal communities,combined with active surveillance for ARF, to documentthe site of streptococcal infections preceding cases ofARF.10

There has been scant published information report-ing rates of streptococcal pharyngitis or pyoderma fromCentral Australian Aboriginal communities. A study pub-lished by Lehmann and others in 2003 reported datacollected in semi-arid Western Australian communitiesand found rates of pyoderma similar to those in theTop End.11 However, anecdotal information from nurses,medical officers and Aboriginal health workers (AHW)in some Central Australia communities suggested thatsymptomatic pharyngitis was a common problem whilstpyoderma was much less common than in communitiesto the west and in the Top End. If so, the epidemiology ofstreptococcal infection, and possibly ARF/RHD, in CentralAustralian communities could be more akin to that seentemperate climates; that is, primarily driven by streptococ-cal pharyngitis.12 This observation prompted the additionof a Central Australian community to the project in anar

hsn(t

igure 2. The reported incidence of acute rheumatic fever in Centralustralia (data from the Central Australian ARF/RHD register-byermission). ARF: acute rheumatic fever; recurrences: recurrentpisodes of ARF.

ttempt to document the apparent disparity between theegions.

The relative burdens of streptococcal and pyodermaave practical implications for local treatment and pos-ible primary prevention of ARF/RHD. In the NT, mosturses, AHW and medical practitioners use a guideline

CARPA Standard Treatment Manual13) that recommendsreatment for sore throat in all people aged 2–25 years

OR

IGIN

AL

AR

TIC

LE

256 McDonald et al. Heart, Lung and CirculationPharyngitis and pyoderma in a Central Australian Aboriginal community 2007;16:254–259

and for people of any age with existing RHD or history ofARF. It also recommends intramuscular benzathine peni-cillin G for routine treatment of pyoderma. In order toinvestigate the epidemiology of symptomatic pharyngi-tis and pyoderma in a Central Australian community withknown high rates of RHD, a prospective surveillance studywas conducted and the results compared to those of theTop End communities where surveillance was done usingessentially the same methods.10

Methods

The study community is 80 km from Alice Springs and waschosen because:

(i) It had a high rate of RHD (information from the Cen-tral Australian Rheumatic Heart Disease Program).

(ii) In a previous needs assessment carried out by theCentre for Remote Health (Alice Springs), the com-munity had identified ARF and child health asprinciple priorities for action.

(iii) Members of the research team had a longstandingrelationship with the community.

The climate is arid with high temperatures (>35 ◦C)during the day, often with cool to cold temperatures atnight. There are four distinct seasons. The population isabout 600 people with a prevalence of RHD at 25 per 1000

study in the Top End, households were similarly selectedbecause one or more occupants had a history of ARF/RHD;this was done to increase the chances of encounteringcases of ARF during the period of surveillance and deter-mining whether the preceding infection was streptococcalpharyngitis or pyoderma.

Researchers endeavoured to visit households on amonthly basis and asked occupants if they had a sorethroat and/or skin sores at the time of the visit. Theyasked about current or recent (within one week) use ofantibiotics, including benzathine penicillin G prophylaxisfor ARF/RHD. They were not asked about symptoms oruse of antibiotics over the whole of the previous monthbecause of concerns about recall. Keeping a family symp-tom diary was also considered impractical because of thehigh month-to-month turnover of household occupants.2

Researchers then examined all throats and took swabsfor culture, and examined limbs and exposed areas forpyoderma. Swabs were taken (up to five) from any activeskin lesions according to the study protocol. If childrenwere missing from households during the day, they weresought at the local school. The questioning of each per-son plus the examination was defined as a consultation.Educational messages about ARF/RHD were reinforcedby the study team at each family visit; people with pyo-derma or other medical conditions that required treatmentwere referred (and usually transported) to the Community

population, as calculated using the Community HealthCentre’s RHD register list and from the number of peo-ple receiving benzathine penicillin prophylaxis. There is awell-established Community Health Centre with an on-site medical practitioner during the week, experiencednursing staff and highly competent AHW, two of whomwere born in the community and have worked with theclinic for more than 10 years. The community has all-weather road access, relatively modern housing and aswimming pool next to the school; it is mainly used by chil-dren in the warmer months of the year (November–April).Overall school attendance is high.

The Central Australian ethics committee approved thestudy protocol prior to commencement. As far as possible,researchers took care to ensure that the study proposalhonoured the primary values (reciprocity, respect, equal-ity, responsibility, survival and protection, and spirit andintegrity) outlined in the National Health and Medi-cal Research Council 2003 document, ‘Values and ethics:guidelines for ethical conduct in Aboriginal and TorresStrait Islander health research’.14 Aboriginal research offi-cers from Alice Springs and researchers from Darwinthen conducted extensive community consultation. Doc-umentation was provided to the Community Council andpresentations were made to the school (teachers and stu-dents), clinic staff and the local Health Board. The HealthBoard required detailed information about all aspects ofthe study and assurances about the likely community ben-efits and use of the data obtained. The research team alsoundertook to provide additional specialist medical ser-vices to the clinic. After a consultation period of almostnine months, we were given the go-ahead to approachindividual families for consent. As with the corresponding

Health Centre.Swabs were transported in a cold box to a central labo-

ratory in Alice Springs on the day of collection, althoughthere was some variability as to how long they remained inthe study vehicle before being put into the cold box. Oncein the laboratory, they were inoculated onto selective andnon-selective culture media, and incubated at 37 ◦C in 5%CO2 and beta-haemolytic streptococci identified accord-ing to the study protocol.15

The incidence density for sore throats was calculated inthe same manner in the Top End study using the equation:incidence per 100 person years = ST/CO/3.5 × 365 × 100(where ST was number of episodes of sore throat andCO was the number of consultations).10 This assumes thatthe duration of a sore throat is 3.5 days.16 Informationon household size was obtained from a housing surveyconducted in 2004 and provided by the Northern Terri-tory Government.17 The results from this community werethen compared to those obtained from the study con-ducted in the three Top End communities.10

Results

Surveillance was conducted from September 2004 toSeptember 2005, although we were asked not to visitthe community for three of those months because of‘sorry business’ (mourning periods following communitydeaths). The study period overlapped the Top End studyfor 9 of the 13 months. On almost 30% of occasions, thefamily was absent from the house when the researcherscalled, even several days in a row. One study visit day wasabruptly halted by a late morning sandstorm.

OR

IGIN

AL

AR

TIC

LEHeart, Lung and Circulation McDonald et al. 257

2007;16:254–259 Pharyngitis and pyoderma in a Central Australian Aboriginal community

Table 1. Comparison of Throat Infection and Carriage, Pyoderma and Household Size Between the Central AustralianCommunity and the Top End Communities

Central Australian Community Top End Communities10

Sore throat: episodes/100 person year (adults andchildren)

480 (95% CI, 290–780) 19 (95% CI, 12–28)

Proven GAS pharyngitis: episodes/100 person year(adults and children)

32 (95% CI, 8–77) 4 (95% CI, 1.4–15.5)

Throat carriage ratesa

GAS 3.7% (95% CI, 1.9–6.3) 4.5% (95% CI, 3.9–5.1)GCS 2.1% (95% CI, 0.9–4.4) 1.7% (95% CI, 1.3–2.1)GGS 2.1% (95% CI, 0.9–4.4) 5.4% (95% CI, 4.8–6.1)

Proportion of children <15 years with one or moreepisodes of pyoderma per year of the study

5.5% (95% CI, 1.8–12.3) 26.5% (95% CI, 23.1–30.1)

Household size: median no. people per bedroom(inter-quartile range)

1.7 (IQR 0.8–2.7) 6.9 (IQR 5.2–8.2)

a Period prevalence.

Nevertheless, we approached 156 people and enrolled145, including 91 children (63%) <15 years, from 13 house-holds. Data were collected from 326 consultations, 205 ofthem in children <15 years. Over the course of the studythere were 15 occasions when people said they had a sorethroat, 9 in children <15 years. On 5 of the 15 occasions(33%) people with sore throat were taking antibiotics atthe time or had taken them in the previous week, as com-pared to 16 of 311 (5%) consultations in people withoutsore throat who had taken antibiotics. There were no sorethroats or skin sores in 11 study participants receivingantibiotic prophylaxis for ARF/RHD, as determined fromthe CHC prophylaxis list.

Using the above equation, the overall incidence den-sity of sore throat was calculated at 480 per 100 personyears (95% CI, 290–780). Of those people with sore throat,four were observed to have purulent pharyngitis. Beta-haemolytic streptococci were recovered from four peoplecomplaining of sore throat, two of whom had purulentpharyngitis; they included group A beta-haemolytic strep-tococcus (GAS) (one isolate from a purulent throat), groupC streptococcus (GCS) (one isolate) and group G strep-tococcus (GGS) (two isolates). The calculated incidencedensity for GAS pharyngitis was 32 episodes per 100 per-son years (95% CI, 8–77). The overall community throatcarriage rates (period prevalence) of BHS were 3.7% forGAS, 2.1% for GCS and 2.1% for GGS.

dassf

mTb

D

T1a

munity almost 20 years ago (1987),18 and 50% higher thanthat reported for the whole region. However, it is aboutthe same as the RHD prevalence in the Top End com-munities used for comparison.19 The high reported rateof sore throat in the Central Australian community (480per 100 person years) contrasts markedly with that of theTop End communities and is more in keeping with ratesfrom cooler European climates.20,21 Although it was alsoin keeping with studies in temperate climates, little elsecan be inferred about the incidence of GAS pharyngitis inchildren because there were so few cases. Nevertheless, inthe Top End there were no confirmed of sore through andGAS from 2694 consultations in children <15 years.10 Bycomparison, the incidence of sore throat in non-AboriginalAustralian children in the southern mainland city of Mel-bourne, where ARF has all but disappeared, is just over 70per 100 person years and that of proven GAS pharyngitisis 14 per 100 person years.22

It could be that most of the sore throats encounteredin the Central Australian community were of viral ori-gin, but then it is puzzling as to why the rates shouldstill be so much higher than in children from the TopEnd. Alternatively, poor recovery of GAS from throatswabs may have been due to concurrent or recent useof antibiotics (in a third of cases) and/or failure of thepathogens to survive the trip to Alice Springs. It couldbe argued that sampling of households where a fam-ibtppitmgtnAcepm

There were only five children <15 years who had pyo-erma over the course of the study, 5.5% (95% CI, 1.8–12.3),nd GAS was recovered from skin sores on three occa-ions. There were no new cases of ARF identified in thetudy households during the course of the surveillance; inact no new cases were reported for the whole community.

The comparison between the Central Australian com-unity and the Top End communities is shown in Table 1.

he differences in household crowding are shown at theottom of the table.

iscussion

he prevalence of RHD in this community (25 cases per000 population) is about twice that reported by Brennannd Patel in another Central Australian Aboriginal com-

ly member is receiving benzathine penicillin may evenias against recovery of GAS if it impedes dissemina-

ion of GAS strains within the household. The benzathineenicillin delivery rate to study household members withast or known ARF/RHD was >85% of scheduled doses

n the previous 12 months (Community Health Cen-re, unpublished data). Alternatively, GCS and/or GCS

ay also be significant causes of streptococcal pharyn-itis in this community; GGS was recovered from two ofhe four documented cases. These organisms are usuallyot recognised as being involved in the pathogenesis ofRF/RHD; however, they do cause other classic strepto-

occal and post-streptococcal diseases23 and frequentlyxchange genetic determinants with GAS, including Mrotein and other virulence factors.24,25 Against this argu-ent is the lower throat carriage rate of GGS in the

OR

IGIN

AL

AR

TIC

LE

258 McDonald et al. Heart, Lung and CirculationPharyngitis and pyoderma in a Central Australian Aboriginal community 2007;16:254–259

Central Australian community as compared to the Top End(Table 1).

Perhaps the most striking finding is the difference in theincidence of pyoderma. This is despite the prevalence ofpyoderma in the Top End study being much lower thanpreviously reported in that region26 and recently docu-mented in other Top End communities.27 The lower ratesare probably because of previously successful healthy skinprograms in each of the Top End study communities. TheCentral Australian community has a swimming pool, a fac-tor associated with reduce rates of skin infection in tworemote Western Australian Aboriginal communities.11

However, two of the three Top End communities used forcomparison also have swimming pools.

In the Top End, there was a correlation between pyo-derma and household crowding.10 The median numberof people per bedroom in the two largest Top End com-munities is more than three times that found in theCentral Australian community. There is a paradox herebecause the prevalence rates of RHD in the two regionsare very similar. A correlation between crowding, pharyn-gitis and ARF has been shown in military settings28 andin urban Baltimore,29 but has not been demonstratedso comprehensively elsewhere, including in tropical andtraditional communities.30 A relationship between house-hold crowding and pyoderma has also been reported inother studies.31,32 In the Top End communities, there was

of ARF/RHD and low rates of pyoderma, GAS pharyngitisis common, and the reverse is true in communities withhigh rates of pyoderma.

The risk of RHD in remote Aboriginal communitiesis much the same across the NT. However, in the Cen-tral Australian community, the clinical emphasis remainswith pharyngitis, and treatment will continue to be basedon recommendations in the CARPA Standard TreatmentManual. In the Top End, it remains to be seen whetherhealthy skin programs currently in progress will reducethe incidence of pyoderma in children of remote com-munities only to witness the emergence of symptomaticGAS pharyngitis, and whether that has an impact on thereported rates of ARF.

Major limitations of this study include the small num-ber of consultations in the Central Australian communityand missing data because of restricted community vis-its. The absolute numbers of symptomatic pharyngitis areso low as to make comparisons tenuous at best. In thisregard, the findings demonstrate the difficulties of con-duction community-based epidemiological research inthese settings. The restrictions on community visits wereunavoidable and follow-up of individuals was particu-larly problematic because of the highly mobile populationwith many people travelling to Alice Springs and othercommunities on a regular basis.3 Variability in collectionand transport of throat and skin swabs is also a concern

15

no correlation between BHS throat carriage and crowding,and there were too few cases of symptomatic pharyn-gitis and ARF to make any conclusions. If crowding isless critical to the incidence of GAS pharyngitis than ofGAS pyoderma and GAS pyoderma does contribute to thepathogenesis of ARF in the Top End, this may go some waytowards explaining the observed differences. Otherwise,the inconsistency may just reflect the paucity of data.

The regional difference may also be related to climate.A similar reciprocal relationship has also been reported inIndia. In the hot southern Indian state of Tamil Nadu, Brah-madathan and others identified high rates of childhoodpyoderma (prevalence rate of 6.9%) and relatively low inci-dence of GAS pharyngitis (9.4 episodes per 100 per years)33

whilst in a cooler northern Indian region (Chandigarh), theincidence of symptomatic pharyngitis in children was 350per 100 person years and the prevalence of pyoderma was1.9%.34

We have hypothesised that recurrent streptococcalpyoderma plays a role in the pathogenesis of ARF in Abo-riginal communities of the Top End of the NT.8 RecurrentGAS pyoderma may invoke an immune response that, tosome extent, protects against GAS throat infection. Pha-ryngeal GCS and GGS may also modulate GAS throatcolonisation and infection. One aspect of the hypothesisis that GAS pyoderma contributes to the immune prim-ing required for the exaggerated immunological responsethat leads to the clinical features of ARF35 in much thesame way as priming from GAS pharyngitis, but perhapsless efficiently than GAS pharyngitis. As to whether theGAS infection ultimately resulting in an episode of ARFcan be other than pharyngeal remains speculative. Nev-ertheless, it appears that in communities with high rates

in remote communities. Furthermore, the study com-munity is better endowed with community facilities andinfrastructure than many others in the region and the epi-demiological picture may not be representative. This canonly be determined through further surveillance studies.

As such, the results of this study cannot be regarded asconclusive. Nevertheless they do provide further informa-tion for hypothesis generation. Should it be establishedthat there are real differences in the epidemiology ofpharyngitis and pyoderma across communities in Centraland Northern Australia, then the findings could trans-late to different long-term approaches to prevention ofARF/RHD.

Acknowledgements

We wish to thank the families in the community, the com-munity council, the Aboriginal research officers and healthcentre staff for their participation. Norma Benger andJames Taylor made substantial contributions to recruit-ment, data collection and community feedback of studyresults. We appreciate the generosity of Western Diag-nostic Pathology and Medvet Science in Alice Springs forallowing us to use their laboratories. The study was sup-ported by a grant from the National Heart Foundation ofAustralia (G 03D 1065).

References

1. Australian Bureau of Statistics. The health and welfare ofAustralia’s Aboriginal and Torres Strait Inlander peoples-4704.0.Canberra: Commonwealth Government; 2005.

OR

IGIN

AL

AR

TIC

LEHeart, Lung and Circulation McDonald et al. 259

2007;16:254–259 Pharyngitis and pyoderma in a Central Australian Aboriginal community

2. Ewald DP, Hall GV, Franks CC. An evaluation of a SAFE-styletrachoma control program in Central Australia. Med J Aust2003;178:65–8.

3. Warchivker I, Tjapangati T, Wakerman J. The turmoil of abo-riginal enumeration: mobility and service population analysisin a central Australian community. Aust N Z J Public Health2000;24:444–9.

4. Brown A, Purton L, Schaeffer G, Wheaton G, White A. Cen-tral Australian Rheumatic Heart Disease Control Program: areport to the Commonwealth, November 2002. NT Dis ControlBull 2003;10:1–8.

5. Kelly A. Top End Rheumatic Heart Disease Program: a reportto the Commonwealth, February–November 2002. NT Dis Con-trol Bull 2003;10:9–11.

6. Fisher DA, Carapetis JR, Marks PJ, Currie BJ. Acute rheumaticfever (ARF) in non-aboriginal patients in the Northern Terri-tory. Aust N Z J Med 1998;28:63–4.

7. Carapetis JR, Steer AC, Mulholland EK, Weber M. The globalburden of group A streptococcal diseases. Lancet Infect Dis2005;5:685–94.

8. Carapetis J, Currie B. Group A streptococcus, pyoderma, andacute rheumatic fever. Lancet 1996;347:1271–2.

9. McDonald M, Currie B, Carapetis J. Acute rheumatic fever:a chink in the chain that links the heart to the throat? LancetInfect Dis 2004;4:235–40.

10. McDonald M, Towers RJ, Andrews R, Benger N, Currie BJ,Carapetis JR. Low rates of streptococcal pharyngitis and highrates of pyoderma in communities where rheumatic fever ishyperendemic. Clin Infect Dis 2006;43:683–9.

11. Lehmann D, Tennant MT, Silva DT, McAullay D, Lannigan

ondary prevention programme in an Australian Aboriginalcommunity. ANZ J Public Health 2005;29:521–5.

20. Hoffmann S. Incidence and management of sore throat ingeneral practice. Scand J Prim Health Care 1986;4:S143–50.

21. Kvaerner KJ, Nafstad P, Jaakkola JJ. Upper respiratory mor-bidity in preschool children: a cross-sectional study. ArchOtolaryngol Head Neck Surg 2000;126:1201–6.

22. Danchin MH, Rogers S, Selvaraj G, Kelpie L, Rankin P, VorichR, Howson M, Carlin JB, Curtis N, Nolan TM, Carapetis JR.The burden of group A streptococcal pharyngitis in Mel-bourne families. Indian J Med Res 2004;119:S144–7.

23. Williams GS. Group C and G streptococci infections: emerg-ing challenges. Clin Lab Sci 2003;16:209–13.

24. Davies MR, Tran TN, McMillan DJ, Gardiner DL, Currie BJ,Sriprakash KS. Inter-species genetic movement may blur theepidemiology of streptococcal diseases in endemic regions.Microbes Infect 2005;7:1128–38.

25. Currie B. Group A streptococcal infections of the skin: molec-ular advances but limited therapeutic progress. Curr OpinInfect Dis 2006;19:132–8.

26. Currie BJ, Carapetis JR. Skin infections and infestations inAboriginal communities in northern Australia. Australas J Der-matol 2000;41:139–43.

27. Bailie RS, Runcie MJ. Household infrastructure in aboriginalcommunities and the implications for health improvement.Med J Aust 2001;175:363–6.

28. Wannamaker L. The epidemiology of streptococcal infec-tions. In: McCarty M, editor. Streptococcal infections. New York:Columbia University Press; 1954. p. 157–75.

29. Gordis L, Lilienfeld A, Rodriguez R. Studies in the epidemiol-

F, Coates H, Stanley FJ. Benefits of swimming pools in tworemote Aboriginal communities in Western Australia: inter-vention study. Br Med J 2003;327:415–9.

12. Wannamaker L. The chain that links the heart to the throat.Circulation 1973;48:9–18.

13. CARPA. Standard treatment manual. 4th ed. Alice Springs,Northern Territory: Central Australian Rural PractitionersAssociation; 2003.

14. National Health and Medical Research Council. Values andethics: guidelines for ethical conduct in Aboriginal and TorresStrait Islander health research. Canberra; 2003.

15. McDonald M, Towers R, Fagan P, McKinnon M, Benger M,Andrews R, Currie BJ, Carapetis J. Recovering streptococcifrom the throat in remote tropical communities: a prac-tical alternative to direct plating. J Clin Microbiol 2006;44:547–51.

16. Del Mar CB, Glasziou PP, Spinks AB. Antibiotics for sorethroat. Cochrane Database Syst Rev 2004. CD000023.

17. Schobben X. 2004 Indigenous Community Housing SurveyCommunity Report for Santa Teresa: a report to the LtyentyeApurte Community Government Council. Darwin: Environmen-tal Health, Northern Territory Department of Health andCommunity Services; 2004.

18. Brennan R, Patel M. Acute rheumatic fever and rheumaticheart disease in a rural central Australian Aboriginal com-munity. Med J Aust 1990;153:335–9.

19. Eissa S, Rosemary L, Binns P, Garstone G, McDonald M.Assessment of a register-based rheumatic heart disease sec-

ogy and preventability of rheumatic fever II: socio-economicfactors and the incidence of acute attacks. J Chron Dis1969;21:655–66.

30. Steer AC, Carapetis JR, Nolan TM, Shann F. Systematic reviewof rheumatic heart disease prevalence in children in develop-ing countries: the role of environmental factors. J Paediatr ChildHealth 2002;38:229–34.

31. Bailie RS, Stevens M, McDonald E, Halpin S, Brewster D,Robinison G, Guthridge S. Skin infection, housing and socialcircumstances in children living in remote Indigenous com-munities: testing and methodological approaches. BMC PublicHealth 2005;5:128.

32. Gibbs S. Skin disease and socioeconomic conditions in ruralAfrica: Tanzania. Int J Dermatol 1996;35:633–9.

33. Brahmadathan KN, Rajkumari BR, Jose JM, Abraham V,Joseph A. Molecular epidemiology of group A streptococcalinfections in south Indian children. In: Proceedings of the XVIthLancefield International Symposium on Streptococci and Strepto-coccal Diseases. 2005.

34. Kumar R, Chakraborti A, Vohra H, Bandhyopadhyay S, SagarV, Sharma V, Dhaliwal RS, Sharma M, Shah B, Ganguly NK.Epidemiology of group A streptococci in a North Indian com-munity. In: Proceedings of the XVIth Lancefield InternationalSymposium on Streptococci and Streptococcal Diseases. 2005.

35. Guilherme L, Fae KC, Oshiro SE, Tanaka AC, PomerantzeffPM, Kalil J. Rheumatic fever: how S. pyogenes-primed periph-eral T-cells trigger heart valve lesions. Ann N Y Acad Sci2005;1051:132–40.