This pilot study is the first assessment of skin health for urban-living Aboriginal children in Australia. We found it was feasible and informs subsequent larger studies underway. Aboriginal Elder co-design; involvement of Aboriginal researchers; and the research-service model were critical elements of success in the pilot. Historically, non-Aboriginal researchers have studied Aboriginal children and families. In this model, both Aboriginal and non-Aboriginal researchers walk together with Aboriginal children and families to discover answers to questions that are prioritised together.
Co-design is a “philosophical approach and evolving set of methodologies for involving people in the design of the services, strategies, environments, policies and processes that impact them.’39 When practiced well, co-design in Indigenous health research can lead to outcomes that can be used to advocate for policy and service delivery changes, helping achieve equity.39, 40 Co-design is recommended in research with Indigenous peoples in recognition of the history and wisdom of Aboriginal-peoples, the longest, continuous culture in the world dating back more than 65,000 years.41 Working together from a strengths-based approach to co-create Aboriginal led systems and services supports self-determination and overcomes inequities that are a result of colonisation and racism.42 The Koolungar Moorditj Healthy Skin project was developed through extensive consultation and cultural guidance from Noongar Elder co-researchers, and in partnership with DYHSAC; collaborating on all aspects of project development from needs assessment to content development, pilot testing and through to research dissemination.43
The involvement of Aboriginal investigators in all roles of the research team was crucial to the success of the pilot project. In particular, the involvement of a trusted AHP embedded in DYHSAC was key to recruiting Aboriginal families, achieving a sample size of 84 participants and surpassing the estimated pilot size of 50. The value of a week-long screening event was evident in the observed increase in participant numbers over the course of the week peaking on the final day, which is consistent with cultural protocol and awareness. Integral to the study was establishing culturally appropriate dermatological care for urban-living Aboriginal children in a place where they feel most comfortable receiving healthcare. The ACCHO setting ensured cultural safety and facilitated the research-service model, whereby over a quarter of participants received opportunistic same-day management of their skin condition.
The results of the pilot project provide the first description of skin health and skin disease frequency in urban-living Aboriginal children recruited from the waiting room of an urban ACCHO. Recently, a systematic review has synthesised the available global literature on skin health in urban-living Indigenous children in high-income countries that share a history of colonisation, displacement and subsequent ongoing health inequities.19, 44 Our results are comparable to the findings, and add further knowledge to the burden of these skin diseases, whilst scoping out a methodology for future studies.
This pilot project also acts as a hypothesis-generating study for associations of the more common childhood skin diseases. Using internal comparison to reduce bias, a number of possible disease associations were suggested.
Infectious Skin Diseases
For BSI and scabies we found a significantly lower prevalence in this urban-living cohort (5% and 1%, respectively) compared to the median prevalence reported for remote-living Aboriginal children (44.5% and 35%, respectively).4, 5 S. aureus was the only species cultured from BSI swab specimens in our pilot, while in the remote setting impetigo is mainly driven by S. pyogenes, with S. aureus playing a secondary role.45 We found a history of iron deficiency was associated with a greater-odds of ever having BSI. To our knowledge this association has not previously been described in Aboriginal children; however, both conditions occur more frequently in children where health inequities exist, including Indigenous, refugee and immigrant children.46, 47 In the pilot cohort, low birthweight was associated with a greater-odds of ever having BSI, consistent with an earlier study of WA Aboriginal children that demonstrated low birthweight to be a risk factor for skin infection hospitalisation.48
In this pilot cohort there appeared to be a high burden of pediculosis capitis (23%) and dermatophyte infection (19%), however background prevalence in urban Australian children is unknown. We identified a point prevalence of 11% for tinea capitis, which is the most infectious and most common paediatric dermatophyte infection globally especially in low-income and resource-poor settings.49, 50 This compares with up to 20% of the general population in developing countries and more than 30% of children in some urban areas of the United States.51 T. tonsurans was the only species cultured from hair pluck specimens in our cohort. This anthropophilic species has historically been one of the most common species responsible for tinea capitis in Australia and is currently the most common cause of tinea capitis in the United States.52, 53
Clustering of skin infections (dermatophyte infection, BSI and pediculosis capitis) was suggested in this pilot study. These skin infections are among the most common childhood skin disorders globally and co-infection is frequent, particularly in settings where poor housing conditions and health infrastructure exist, leading to ongoing transmission.31, 54 A recent narrative review on systemic housing-level contributions to infectious disease transmission for Indigenous Australians found skin infections were associated with malfunctioning health hardware, including infrastructure required to wash clothes and bedding.55 Consistent with this, the absence of a working washing machine was predictive of dermatophyte infection in our study. It has been shown that domestic laundering at 60°C (as opposed to 30°C) is required to completely eliminate dermatophytes; these temperatures are unlikely to be achieved with hand-washing of clothing, inability to afford heating of water for washing clothes, or a malfunctioning washing machine.56 If this disease association is reproducible in the larger multisite observational studies, advocacy to address this modifiable factor in housing infrastructure programs may help reduce the burden of dermatophyte infection in urban-living Aboriginal children.
An interesting hypothesis-generating result in our study found bed-sharing, defined as sharing a mattress with another person/people, was also predictive of dermatophyte infection. This can be explained by the contagious nature of dermatophyte infection, with spread known to occur from close contact with an infected person or contaminated bedding/clothing. Our pilot study failed to collect information on the reasons for bed-sharing, which is crucial for interpreting this result. It is known that sleeping arrangements are strongly influenced by cultural tradition, with bed-sharing reported to be one of the parental choices most influenced by cultural practice and beliefs.57, 58 In many Aboriginal communities purposeful bed-sharing is a cultural norm; however, bed-sharing may also be reactive and the result of environmental influences including socioeconomic factors, housing arrangements, heating and economic access to beds.57, 59, 60 We hope the results of the larger multisite observational studies will provide the information needed to better translate this result, be it with advocacy for adequate housing infrastructure and/or empowering the community to consider health promotion messages that are appropriate and culturally strong.
Regarding protective factors for skin infections, the data suggest ‘sometimes’ (as compared with ‘never’) swimming in the ocean and ‘sometimes’ (as compared with ‘never’) swimming in a chlorinated pool may protect against dermatophyte infection and BSI, respectively. This latter result is supported by a systematic review investigating the health benefits of swimming pools in remote Aboriginal communities, in which the included prospective studies consistently demonstrated reduced skin sore prevalence associated with access to swimming pools.61
Non-Infectious Skin Diseases
We found the lifetime prevalence of parent-reported ‘eczema ever’ (19%) in the pilot cohort to be less than that documented in all Australian children by age 4-years (28%), and less than that recently reported in a systematic review of all urban-living Indigenous children in high-income countries (median prevalence 25%).15, 19 Current AD was dermatologist-diagnosed on examination in 15% in our cohort (0–18 years); which, after taking into account the natural history of AD and tendency to improve over time, is not dissimilar to the 20% population prevalence of clinician-diagnosed AD described in urban Australian children.14 A history of iron deficiency was predictive for current AD in the pilot, consistent with evidence indicating adequate iron status in the perinatal and infantile period may protect against AD.62
Untreated acne was detected in over half (55%) of all 10–19 year-olds in this pilot. An Australian population-based study has previously shown acne to be a common problem in school students, with prevalence ranging from 28% in 10–12 year-olds to 93% in 16–18 year-olds.63 While little is known about the prevalence of acne in Aboriginal youth, a cross-sectional analysis of data from primary care reported a lower frequency of acne presentations by Aboriginal youth; suggesting either a lower incidence of acne or lower rates of Aboriginal youth seeking medical care for acne.64 Further, among Australian studies profiling the case-mix of patients seen in various dermatology outpatient clinics, presentation for acne was uncommon in Aboriginal patients.21, 22, 65
To our knowledge, this pilot data represents the first attempt to describe sun-protection behaviour and sunburn in urban-living Aboriginal children. Using the FSP classification, which is the most commonly used strategy to assess skin sensitivity to ultraviolet radiation despite inconsistencies being described in skin of colour, we identified urban-living Aboriginal children with FSP II and III to be a subgroup with most frequent sunburns where targeted sun protection messaging may be helpful. 33, 34, 66–71 This is further supported by research showing urban-living Indigenous adults with FSP II and III are at the highest risk of developing skin cancer, representing 19 of the 22 (86%) Indigenous patients diagnosed with skin cancer between 2003 and 2017 in Sydney, Australia.72
Strengths and limitations
The strengths of this study include its co-design with Noongar Elders, the research-service model embedded in an urban ACCHO, Aboriginal researchers involved in all study procedures, capacity building of AHPs, clinical diagnosis by a dermatologist and comprehensive data collection.
Limitations of this study include selection bias; most notably recruitment from a healthcare facility which may not represent all urban-living Aboriginal families therefore impacting on generalisability. Also, self-selection, in that parents of children with a dermatological concern may be more inclined to enrol their child skewing the results towards a greater level of disease.
Questionnaire responses carry the potential for recall bias and information bias may be present with the absence of validated questions for disease diagnosis (including eczema) and the absence of validated measures of sun exposure and sun protection practices. The FSP classification system has also been criticised for potential inconsistencies in skin of colour, with some studies suggesting the questionnaire lacks reliability and functionality in this group.33, 34, 67, 68 Better understanding of this topic in urban-living Aboriginal families is needed and may be a direction for future research.
The nature of this brief screening week taking place in spring, means seasonal bias may be introduced for those identified skin conditions that show seasonal variation. Finally, the disease association results are limited by small sample sizes reflected in the wide confidence intervals. The findings of this pilot study must be considered in the context of these sources of potential bias.