Benefits of swimming pools in two remote Aboriginal communities in Western Australia: intervention study

BMJ 2003; 327 doi: (Published 21 August 2003) Cite this as: BMJ 2003;327:415
  1. Deborah Lehmann, epidemiologist (deborahl{at},
  2. Mary T Tennant, research assistant1,
  3. Desiree T Silva, paediatrician1,
  4. Daniel McAullay, research officer1,
  5. Francis Lannigan, otorhinolaryngologist2,
  6. Harvey Coates, otorhinolaryngologist2,
  7. Harvey Coates, otorhinolaryngologist2,
  8. Fiona J Stanley, director1
  1. 1Centre for Child Health Research, University of Western Australia, Telethon Institute for Child Health Research, PO Box 855, West Perth, WA 6872, Australia
  2. 2Princess Margaret Hospital for Children, Subiaco, WA 6008, Australia
  1. Correspondence to: D Lehmann


    Objective To determine the health impact of swimming pools built with the aim of improving quality of life and reducing high rates of pyoderma and otitis media.

    Design Intervention study assessing prevalence of ear disease and skin infections before and at six monthly intervals after opening of swimming pools.

    Setting Two remote Aboriginal communities in Western Australia.

    Participants 84 boys and 78 girls aged < 17 years.

    Main outcome measures Changes in prevalence and severity of pyoderma and perforation of tympanic membranes with or without otorrhoea over 18 months after opening of pools.

    Results In community A, 61 children were seen before the pool was opened, and 41, 46, and 33 children were seen at the second, third, and fourth surveys. Equivalent figures for community B were 60, 35, 39, and 45. Prevalence of pyoderma declined significantly from 62% to 18% in community A and from 70% to 20% in community B during the 18 months after the pools opened. Over the same period, prevalence of severe pyoderma fell from 30% to 15% in community A and from 48% to 0% in community B. Prevalence of perforations of the tympanic membrane fell from 32% in both communities to 13% in community A and 18% in community B. School attendance improved in community A.

    Conclusion Swimming pools in remote communities were associated with reduction in prevalence of pyoderma and tympanic membrane perforations, which could result in long term benefits through reduction in chronic disease burden and improved educational and social outcomes.


    Indigenous Australian children have very high rates of pyoderma (known locally as “skin sores”) and otitis media. Prevalences as high as 70% have been reported for pyoderma in remote Aboriginal communities.1 The major pathogen of pyoderma, group A streptococcus, is associated with chronic renal failure, which is very prevalent in Aboriginal communities.2 3 The high prevalence of pyoderma is also the most likely reason for some Aboriginal communities having the highest reported incidence of acute rheumatic fever worldwide.4

    Impaired hearing can seriously affect performance at school, subsequent employment, and social circumstances in adulthood. Between 10% and 67% of Aboriginal school age children have perforated tympanic membranes, and 14-67% have some degree of hearing loss; remote communities have the highest prevalence.5

    Some reduction in skin infections was reported after a swimming pool opened in an Aboriginal community in the Northern Territory.6 An audit of Aboriginal communities in central Australia has also found health and social benefits of pools.7 Furthermore, communities within easy access of places to swim have been reported to have lower prevalence of otitis media than those further away.8 These reports are in contrast to concerns about pool related infections in areas with low prevalence of infectious diseases.9

    After publication of these reports suggesting that Aboriginal children may benefit from access to swimming pools, the Western Australian government built 25 metre, part shaded, salt water pools in three remote Aboriginal communities. In view of possible beneficial or potentially deleterious effects on health and the lack of previous evaluation of such an intervention, we assessed the health impact of swimming pools in the first two communities to have a functioning pool, particularly with regard to skin infections and middle ear disease in children. A secondary objective was to determine whether social benefits might result from pools in these remote communities. Financial constraints precluded inclusion of the third community.



    We compared disease prevalences before and after swimming pools were opened in two isolated Aboriginal communities (total population about 180 in community A and 250 in community B) located in a semi-arid environment more than 1000 km north of Perth. Daily temperatures range from 15°C to 45°C during the summer, falling to freezing at night in winter. Health services in community A have been erratic, with one nurse or Aboriginal health worker present most of the time and a physician visiting every two to four weeks. Community B has two nurses, two doctors, and two Aboriginal health workers. The swimming pools were opened in September 2000 and are open annually from September to April.


    We examined children four times at approximately six month intervals between July 2000 and March 2002. All children aged under 17 years present in the community were eligible for inclusion in the study if informed consent was obtained. At each survey a paediatrician examined all enrolled children who could be located, and children who moved into the communities were also given the opportunity to enrol.

    We graded pyoderma according to severity (the “sore score”), on the basis of the total number and severity of sores (crusty, pussy, or dry flat lesions).10 We classified a sore score of 1-4 as non-severe and ≥ 5 as severe. We also recorded the presence and severity of scabies, abscesses, and fungal lesions. Experienced paediatricians did otoscopic examinations. If otorrhoea was present the ear was irrigated with 1:20 betadine solution to obtain a good view of the tympanic membrane.11 The paediatricians recorded the colour and position (retracted, bulging, or normal) of the tympanic membrane, as well as the presence of wax, cholesteatoma, perforations, and otorrhoea. They also noted pathology indicative of past disease (for example, healed perforations and tympanosclerosis). During the first survey two ear, nose, and throat specialists examined children with serious ear disease or to confirm a paediatrician's diagnosis.

    To identify any pool related injuries or disease outbreaks, the research team examined data held at the local clinic and obtained information from pool managers. Pool managers keep a daily record of any incidents occurring at the pools.

    As parents were concerned about truancy, the communities introduced a “no school, no pool” policy as an incentive for children to attend school. In addition to having swimming lessons at school, children attending school are given passes permitting them to use the pool after school. We collected school attendance records for the school term immediately before each of our surveys. School attendance rate was the number of days attended/number of days enrolled.

    During the fourth visit to community A, in a house to house survey, we asked family members a series of questions to document their impressions of the pool and any perceived benefits and concerns they may have had regarding the pool. Time did not permit interviews in community B.


    We entered all data into Filemaker Pro version 5 and exported them to SPSS version 10.1 for Windows. We used the χ2 test and Fisher's exact test to compare groups of interest and the χ2 for trend to investigate trends in prevalence over time. Small numbers precluded more formal statistical analysis.


    In community A we saw 83 children (48% boys) at least once (61 enrolled during the first survey), 51 (61%) twice, 36 (43%) three times, and 12 (14%) four times. During the first survey, 20% were aged under 5 years, 66% were aged 5-11, and 15% were 12-16 years old. In community B, we saw 79 children (56% boys) at least once (60 enrolled during the first survey), 48 (61%) twice, 30 (38%) three times, and 23 (29%) four times. During the first survey, 37% were aged less than 5 years, 42% were aged 5-11, and 22% were 12-16 years old.

    Skin infections

    In community A, during the winter before the pool opened (July 2000), 38 (62%) children had pyoderma; 18 (30%) had severe pyoderma. Children aged 12-16 years had significantly less pyoderma (2/9, 22%) than younger children (36/52, 69%, χ2 = 7.4, 2df, P = 0.02). Figure 1 shows a marked reduction in pyoderma during the 18 months of follow up (χ2 for trend = 24.88, 1df, P < 0.0001). The result was almost exactly the same for children aged under 12. The prevalence of severe pyoderma also fell significantly over time (in children aged under 12, χ2 for trend = 8.46, 1df, P = 0.004). During the first survey, other skin infections included abscesses in 5 (8%) children, fungal infections in 15 (25%) children, and uncomplicated scabies in 3 (5%) children. The prevalences of any skin infection at the four consecutive surveys were 64% (39), 51% (21), 43% (20), and 18% (6) (χ2 for trend = 17.32, 1df, P < 0.0001).

    Fig 1
    Fig 1

    Proportion of children with severe, non-severe, or no pyoderma during four surveys before and after installation of swimming pools in two communities (N=number of children examined)

    In community B, a scabies epidemic was occurring at the time of the first survey (August 2000); 42 (70%) of 60 children had pyoderma, and 29 (48%) had severe pyoderma. Prevalence and severity of pyoderma did not vary with age. The swimming pool was closed for six weeks before the second survey and for a total of three months during that summer owing to a structural fault.

    Prevalence of pyoderma fell significantly over time (χ2 for trend = 17.03, 1df, P < 0.0001, fig 1). A reduction in severity also occurred, with 29 (48%) children having severe pyoderma before the pool opened, 12 (31%) one year later, and none at the fourth survey. The prevalences of any skin infection at the four consecutive surveys were 78% (47), 43% (15), 69% (27), and 20% (9) (χ2 for trend = 25.16, 1df, P < 0.0001).

    We saw 29 children in community A and 25 children in community B at all of the first three surveys. No significant difference in prevalence or severity of sores existed during successive surveys between children seen three times and those seen less often.

    Middle ear disease

    Table 1 shows the otoscopic findings during the first survey for those children in whom both tympanic membranes were fully visualised (54 in each community). Only five children in community A and 12 children in community B had no evidence of past or current disease–that is, bilateral intact tympanic membranes, normal position of the tympanic membranes, and no tympanosclerosis or healed perforations. We saw no cholesteatoma throughout the study.

    Otoscopic findings during first survey before pools were opened. Values are numbers (percentages)

    View this table:

    Figure 2 shows the prevalence of perforations of the tympanic membrane and otorrhoea through tympanic membrane perforations during the four surveys. Although no reduction in prevalence of tympanic membrane perforations occurred during the first summer after the pools opened (March 2001), a gradual decline occurred thereafter to 13% (8/44) in community A (χ2 for trend = 4.32, 1df, P = 0.04) and 18% (4/31) in community B (χ2 for trend = 2.18, 1df, P = 0.14) in the fourth survey. No significant change in prevalence of otorrhoea through perforated tympanic membranes occurred in either community. Results were similar when we considered all ears that were successfully examined (total of 113, 81, 88, and 62 ears in community A and 112, 66, 76, and 91 ears in community B, during the four surveys): during the 18 months of follow up, tympanic membrane perforation rates declined from 19% (21/113) before the pools were opened to 8% (5/62) (χ2 for trend = 3.88, 1df, P = 0.05)in community A and from 23% (26/112) to 15% (14/91) (χ2 for trend = 2.5, 1df, P = 0.11) in community B. Stratification by age made no difference to our findings in either community.

    Fig 2
    Fig 2

    Prevalence of dry perforations and otorrhoea through a perforation (wet) during four surveys before and after installation of swimming pools in two communities (N=number of children with both tympanic membranes examined)

    In community A, the prevalence of tympanic membrane perforations declined more rapidly in children seen on each of the first three surveys than in those seen less often. After the second survey, the prevalence of tympanic membrane perforation in children seen regularly was 21% (6/29) compared with 64% (7/11) in those seen less frequently (Fisher's exact test, P = 0.02); equivalent figures during the third survey were 10% (3/29) and 43% (6/14) (P = 0.04). No such differences existed in community B.

    School attendance

    In community A, 50, 47, 57, and 61 children were enrolled during the school terms preceding surveys 1, 2, 3, and 4. The proportion of children with attendance rates of at least 70% rose from 42% (21) during the term before the pool opened to 51% (24), 65% (37), and 67% (41) during the terms preceding the second, third, and fourth surveys (χ2 for trend = 8.70, 1df, P = 0.003). In community B, 79, 60, 58, and 85 children were enrolled during the terms preceding the four consecutive surveys; 52% (41), 40% (24), 33% (19), and 54% (46) of children had attendance rates of at least 70% during the terms preceding consecutive surveys, significantly lower during the terms preceding the second and third surveys than during those preceding surveys 1 and 4 (χ2 = 8.32, 3df, P = 0.04).

    Pool related incidents and social impact of pools

    No disease outbreaks have been reported, and only two reports of faecal contamination of the pools were made. One pool related trauma (broken arm or mild concussion) occurred in each community.

    The 12 extended family groups interviewed in community A during the last survey were almost unanimous in their support of the swimming pool. People generally thought that children looked happier and healthier and were learning to swim. Some comments were: “Fun, enjoy, stop getting bored, health improved,” “Keeps kids out of mischief, some place for kids to go, no other pools around, it's clean,” and “Not breaking in and being a mischief at home.” One person felt that there had been no benefit: “Kids have broken into the pool, still play money hole [gamble] and use gings [slingshots].”

    Parents were pleased that their children were encouraged to attend school through the “no school, no pool” policy. We contacted the local police by telephone, and they commented that they had noted a marked decline in petty crime.


    We have found significant improvements in health after the introduction of swimming pools in two remote Aboriginal communities. Swimming in a salt water pool provides the equivalent of a nasal and ear washout and cleans the skin. Prevalence of pyoderma declined from over 60% to around 20% during the study period. In the short term this is likely to reduce the clinical workload and prescription of antibiotics; in the longer term it may help to reduce the enormous burden of chronic renal and cardiac diseases in Aboriginal communities.24

    Chronic suppurative otitis media is very uncommon in first world countries, and a prevalence over 4% in children is considered to be a massive public health problem.12 During our first survey, 9% of children in community A and 17% in community B had otorrhoea through large perforations of the tympanic membrane (well known to be chronic from a young age in these Aboriginal communities5 11). Despite intervention and published guidelines, ear health among Australian Aboriginal children has not improved for over two decades.5 11 13 In 1986 a tympanic membrane perforation rate of 31% was reported in community B, with 17% otorrhoea (S Weeks, unpublished report, 1986). These rates remained unchanged over 14 years, as reflected in our pre-pool data. Thus the reduction in tympanic membrane perforation rates in our study from 32% to 18% in community B, although not statistically significant, is the first documented successful preventive intervention for otitis media in the past 15 years. This should lead to improvements in language development and literacy.

    A limitation of this study was the poor follow up of individual children due to the high mobility of the population. Nevertheless, at each survey, virtually all enrolled children attending school had a clinical examination. Of interest was the more rapid decline in prevalence of tympanic membrane perforations in community A among resident children compared with those who were more mobile. Whether this is because more permanent residents used the pool more or because of differences in social circumstances between the more and less mobile groups is unknown. We did not see a similar finding in community B, which may be owing to closure of the pool for a total of three months during the first summer. Children in community B tend to move between the local school and a boarding school 500 km away, which in part explains the lower attendance rate we found in the middle of the study period.

    A further limitation was the lack of a control community. This was neither logistically nor ethically possible: no comparable communities were nearby, and a control community would have been appropriate only if it had been possible to offer them a pool if the intervention proved to be effective. To our knowledge no other activities were taking place during the study period that might have reduced the burden of skin infections and ear disease. However, we cannot exclude the possibility that our four very brief visits to the communities over 18 months may have had an impact on the burden of ear disease through paediatric consultation and training of health and education staff, including dissemination of best practice guidelines. Ongoing monitoring of the impact of these pools is necessary, and funding for this has been secured.

    The pools are being used for educational and recreational functions. Virtually all school age children use the pools, thus benefiting from regular exercise and instruction in swimming and resuscitation techniques, which are important to reduce rates of drowning in this population. In the Northern Territory, 93% of drowning in Aboriginal children occurs in open waterways.14 Drowning rates are higher in Aboriginal than in non-Aboriginal children in Western Australia,15 but until now supervised swimming lessons have not been available for children in remote Aboriginal communities, which is in contrast to other areas of Western Australia where access to swimming pools and lessons is taken for granted.

    What is already known on this topic

    The very high prevalence of otitis media and its long term sequelae can seriously affect quality of life of Aboriginal children and adults

    Pyoderma is endemic in many Aboriginal communities and contributes to the high prevalence of chronic renal disease and rheumatic heart disease

    Rates of drowning are high in Aboriginal communities

    What this study adds

    Introduction of swimming pools in two remote Aboriginal communities was associated with reducedprevalence of pyoderma and tympanic membrane perforations

    Swimming pools provide important social capital in disadvantaged communities

    The pools have been well managed by the Royal Life Saving Association. Water quality is regularly monitored according to government regulations. To date no outbreaks of infectious disease have occurred, but the ongoing success of the pools will depend on adequate funding to maintain high safety and environmental standards.

    In conclusion, the pools have resulted in significant health and social gains. The community's enthusiasm for the intervention and support for the study have been key to the success of the project. Programmes to improve the health of Aboriginal Australians living in isolated communities should certainly focus on better housing, sanitation, nutrition, education, and access to health care,16 but consideration should also be given to building and maintaining swimming pools. The costs involved will be a small price to pay for the reduction in severe chronic disease and improved health, educational, and social outcomes in this seriously disadvantaged segment of Australian society.

    Editorialby Hall and Sibthorpe


    We thank Peter Richmond and John Stuart, who each participated in one survey in which they did all the paediatric examinations; Sharon Weeks for audiological expertise; and Irene Nannup, Jacinta Johnston, Kate Butler, and Claire Gordon, who assisted in the field work. We thank all members of the two communities participating in the study for their support and the teachers and health staff for assistance and hospitality. Kim Hames, previously minister for housing, Aboriginal affairs and water resources, had the vision to install swimming pools in remote communities. Kerryn Sivwright established the databases, and Peta Serna and Norries Pomat assisted in database management. Anne Read and Sandra Eades provided valuable guidance in establishing the project. We thank Marilyn Lyford, Mark Middleton, and Lester Ball from the Royal Life Saving Association and Peter Smith, department of housing and works, for their continuing support and Michael Alpers and Carol Bower for their critical review of the manuscript.


    • Contributors FJS and DL initiated the project. DL, MTT, DMcA, and DTS developed the study design. MTT, DTS, DMcA, and DL conducted the field work. HC and FL provided ear, nose, and throat examination, training, and advice. MTT and DL did the data analysis. All authors contributed to writing the paper. DL is the guarantor

    • Funding Western Australian department of housing and works and Western Australia Health Promotion Foundation (Health-way). The guarantor accepts full responsibility for the conduct of the study, had access to the data, and controlled the decision to publish.

    • Competing interests None declared.

    • Ethical approval The ethics committee of Princess Margaret Hospital for Children and the Western Australian Aboriginal health information ethics committee approved the study.


    View Abstract