Wa! Ningeningma arakba akina da! (Oh! Now I know, that’s it!)

We previously reported persistently high rates of cannabis use in three Indigenous communities in Arnhem Land in the Northern Territory; in our longitudinal studies,1-3 72% of males and 23% of females aged 13–36 years were current users at baseline (2001).2 We also found that the prevalence of symptoms of anxiety and cannabis dependence increased with more frequent cannabis use, and we documented a heavy burden on community finances and health services.2,4

These reports informed changes in policies that featured policing strategies targeted at cannabis supply and associated problems in remote NT communities generally.5 However, the Indigenous communities we studied were not engaged in or aware of these wider strategic shifts. Indigenous researchers became alarmed at respondents’ reports of cannabis-related harms during interviews in 2005–2006 and expressed a desire to disseminate the research findings and describe their insights to the respondents, their families and the wider community. They envisaged that, through such a feedback process, their communities would become better informed about cannabis use and its consequences, and so would be able to make more informed choices about cannabis. Here, we report the approach we developed to providing feedback on research, the processes involved, and the implications.

There is widespread endorsement for disseminating research results back to study communities,6,7 and for the importance of correcting power imbalances in research involving vulnerable populations such as Indigenous Australians.8 In the past 30 years, approaches to conveying research results to the Indigenous groups studied have progressed from no feedback (pre-1970s) to findings being used as an impetus for change (mid 1990s).9 Diverse methods for doing this have been described for a wide range of audiences.6,10 However, few studies provide specific practical guidelines, especially where language and cultural differences compound the difficulties faced. One NT study used locally understood concepts of “land, body and spirit” to disseminate adult mortality data.11 Another survey, of Aboriginal health workers in the NT and South Australia, identified preferences for pictorial representations of survey information.12 Pictorial representations of program outcomes were also used to convey findings about infant birthweight in three Aboriginal communities in the NT.13 However, there is a lack of detailed examination of the processes used to communicate epidemiological data in remote Indigenous Australia.

The three study communities in Arnhem Land have been described in detail elsewhere.2 A single Indigenous language is spoken in these communities, and cultural concepts are generally intact. English is a second language; English language skills vary greatly, as does literacy in younger people.14 Our continuing studies of cannabis use3 are collaborative efforts between non-Indigenous and local Indigenous researchers. Commitment by Indigenous researchers to address cannabis-related harms in their communities since the late 1990s has been pivotal to achieving these research outcomes.

Over 3 months in 2006, workshops were held with the Indigenous and non-Indigenous researchers. These aimed to develop a shared understanding about the levels of cannabis use and to design feedback resources suitable for community-wide dissemination. Indigenous researchers translated the research findings into their own cultural concepts and language.

Challenges emerged in translating concepts of numbers, quantities and the ages of participants. The local numbering system has specific terms only for numbers 1 to 10, so we used a denominator of 10 individuals to pictorially and verbally describe prevalence of cannabis use. We also learned that adjectives used to describe quantities of countable things (eg, a lot of houses) were always different from those used to describe quantities of uncountable things (eg, a lot of wind), and that the terms were not interchangeable. Appropriate words were identified to qualitatively describe levels of cannabis use.

To represent the age groups of the population sample, the Indigenous researchers chose to use locally recognised descriptors of the life stages for males and females. These descriptors are not fixed according to calendar age, and the definition of each may vary from one individual to another depending on cultural considerations and individual characteristics. A local Indigenous artist was commissioned to draw relevant images. The Indigenous researchers chose to depict cannabis users as faded figures, as they considered users to be weakened by their drug use (Box 1).

In addition to these pictorial presentations of prevalence data, the mental health harms, financial impacts and information about cannabis and NT law (provided by the local police) were described in plain English and the local language. Phrasings were translated into the local language and then back to English, and concepts were re-explained by the Indigenous researchers to K S K L after consultation across the study communities. Three resources — a book, a poster and a DVD (in the local language with English subtitles) — were developed and endorsed by community leaders for dissemination to community residents and local service providers.

Dissemination of the resources began in May 2007. Initial responses to the materials were gauged from semistructured interviews with 30 Indigenous and eight non-Indigenous participants, interviewed either individually or in groups. The main questions were about attitudes towards the materials and their appropriateness for local Indigenous people. Interviews of 15–60 minutes were conducted opportunistically across the three study communities with community members, health centre personnel, linguists, representatives of governing Indigenous organisations, police, and staff of correctional services, the aged care service and schools. Interviews with Indigenous participants were conducted by M J J and K S K L, using plain English and the local language. Most participants commented positively about the locally drawn pictures used to describe prevalence of cannabis use. Many also remarked about the importance of providing communities with this kind of information using “our ways of describing things”. Negative comments were few. Suggestions for improvements were offered, such as adding more local language words to describe cannabis use, and more clearly differentiating between the local language and English (Box 2).

Rather than providing literal translations, our efforts focused on identifying common concepts, to widen the community understanding of our studies of cannabis use. Early indications are that comprehension of the research findings was considerably enhanced among Indigenous researchers and community members. The approach also appears to be flexible enough to convey information effectively to people of different ages and with different levels of English comprehension and reading ability — a positive first step in improving community-wide literacy about frequent cannabis use and related harms, including mental health impacts.15 As explained by one community leader:

Indigenous researchers’ capacity was strengthened. They took on the challenging task of seeking community review of the feedback resources and disseminating the resources to all study communities and local service providers. They were delegated by community leaders to present their work at a national drug and alcohol conference.16 Their enhanced understanding of prevalence of cannabis use and its consequences in their communities enabled them to secure funding for a project to assist a closely affiliated community that was also experiencing high levels of cannabis use.

We are unable to comment on the transferability of the approach to other communities where cultural and language concepts may be different. Time is needed to assess the uptake and utility of these resources across the study communities. However, early reports are promising:

Building community understanding and momentum for change through a community-feedback process is important for research and health promotion efforts, whether these are in a remote Indigenous community or an urban multicultural setting. We have shown that it is possible to convey health information using this simple and strategically important approach. Some key factors made this possible. Sound relationships between the Indigenous and non-Indigenous researchers, the study communities and the service providers created a basis of trust on which to conduct the research. The role of the Indigenous researchers was pivotal. Their participation combined pragmatic, moral, interventionist and epistemological rationales for involving Indigenous people in research, consistent with best practice.17 Their capacity for comprehensive community liaison, considered guidance and willingness to share their ways of understanding the research stimulated participation from other community members. They also continually challenged the non-Indigenous researchers to seek their own insights and to consider alternative approaches that would enable their own communities to better understand the research conducted in these disadvantaged and vulnerable groups.

1 Presentation of prevalence estimates of cannabis use among 262 people aged 13–36 years at baseline (2001)

Glossary: Amarda = cannabis. Warnumamalya = Aboriginal people. Warnungkwarba = men (20–36 years). Wurridarringka = women (20–36 years). Wurranjarrngalyilya = male teenagers (13–19 years). Wurradidiyara = female teenagers (13–19 years).