Significant health disparities exist between Indigenous and non-Indigenous Australians resulting in a 10–11 year average reduction in life expectancy for an Indigenous child born between 2005 and 2007 [1]. Liver disease is the third largest contributor (11%) to this gap in life expectancy with chronic hepatitis B (CHB) contributing significantly, in the form of liver cirrhosis and hepatocellular carcinoma (HCC).

CHB is endemic in the Indigenous communities of the Northern Territory (NT) of Australia with prevalence rates estimated to be between 0.8% [for children born in the universal vaccine era (1988 onwards)] and14.2% (for adults born pre universal vaccination) [2–8], this is compared to 1% in Australia as a whole [9]. Despite the availability of effective, government subsidised treatments only an estimated 5% [10] of all people living with CHB in Australia are receiving appropriate management for their infection. This disparity in rates of Hepatitis B and low uptake of treatment is also seen in other Indigenous populations across the world [11, 12].

The barriers to people accessing care for CHB are multifactorial but among Indigenous Australians, include gaps in knowledge, low health literacy and challenges in accessing the appropriate care [13]. Both a recent situational analysis [14] and a qualitative study [15] in the Torres Strait region of Australia identified low levels of knowledge about CHB both in health care providers and Indigenous Australian patients with CHB. Christie et al. [16] have explored views of health literacy in the particular cultural context of remote Indigenous communities in the NT, as well as carrying out a scoping study looking at ways to improve health literacy in this region. Christie (2010) suggests that “effective health literacy is largely to do with effective communication” (p.40). Based on their research, they argue that building on an individual’s existing knowledge using a culturally appropriate approach (i.e. a relevant respectful partnership which is mindful of language, worldview, existing knowledge and beliefs) to achieve a shared understanding of the issue at hand is more beneficial than attempts by health practitioners to simply ‘transfer’ biomedical knowledge to their patients [17]. Although many health promotion or information resources exist for hepatitis B [18], all the above [13–17] studies as well as the Australian National Hepatitis B strategy [19] highlight the lack of culturally appropriate resources, in particular visual and multimedia resources, available to facilitate shared understandings of Hepatitis B and strengthen health literacy.

In the context of the NT Indigenous population, English is usually a second (or even third or fourth) language; therefore, achieving effective cross cultural or “culturally safe” communication can be challenging, as has been extensively documented in health care settings over the last decade [20–22]. Miscommunication between health providers and patients has been reported to be pervasive, however using interpreters and translators is perceived to be only part of the solution [20]. Different worldviews and knowledge systems that exist among Indigenous Australians, including alternative concepts of physiology, pathology and disease causation also contribute [23]. An often misinformed assumption by health providers of shared understandings [20], along with the absence of opportunities and resources to construct a body of shared understanding perpetuate this miscommunication. Two specific factors, culture and worldview, are increasingly acknowledged as important antecedents contributing to health literacy [24–27]. There are a myriad of different definitions of culture; when referring to culture in this paper we use the broad definition of the culture of a society as “… the totality of its shared beliefs, norms, values, rituals, language, history, knowledge and social character” [28].

Participatory projects working with Indigenous communities in the design and development of health education resources have been successful in improving health literacy and participation in healthcare in other disease areas [29, 30].

The aim of this research project was to explore the knowledge, perceptions and experiences of remote dwelling Indigenous adults and their health care providers relating to hepatitis B infection. We also aimed to gauge interest among Indigenous participants in further knowledge of this disease and gain perspectives on how and in what format people could best acquire the information they needed. This was the first stage of a wider participatory action research (PAR) project with the intention of using the results as the evidence base to inform development of a culturally appropriate Hepatitis B educational resource.

This project was undertaken in northern Australia between July 2012 and December 2013. It was based at the health clinic of a remote community in Arnhem Land, 521 km northeast of Darwin (the capital of the NT). This community has a population of 2,124 with an average age of 24 years; 89% are Indigenous Australians and only 9.5% of the population speak English as their first language. There is an average of 4.2 people per available bedroom and 78% of households are considered to be overcrowded. There are three general stores, a school, a library, a health clinic as well as a police station and a community church.

The overall project design was based on PAR principles; specifically, ongoing consultation, reflection and discussion with the community throughout each iterative cycle. JD (a female non-Indigenous researcher and clinician with experience in working in a cross cultural environment) and SB (a female Indigenous researcher and health worker in the remote community) worked alongside each other in constructing the interview schedule, recruitment, data collection, analysis and interpretation. This paper reports the results of the first part of this project which was the first formal PAR cycle and provides the evidence base for the development of a culturally appropriate educational tool for hepatitis B, the second phase of the project (details not presented here). However prior to this project informal discussions regarding the issues facing the community with respect to the burden of disease produced by Hepatitis B, the lack of community understanding and the difficulties health workers have in explaining Hepatitis B to community members had been discussed within clinic meetings and with the visiting liver clinic service. The impetus for the project came from the community clinic. Their enthusiasm for the project led to the development of a collaborative research partnership between the community clinic, the Royal Darwin hospital liver clinic and Menzies School of Health Research and establishment of the formal PAR process.

Ethical approval for the study was obtained from the Human Research Ethics Committee of the Northern Territory Department of Health and Menzies School of Health Research (HREC) as well as Miwatj Health Aboriginal Corporation (an Aboriginal-controlled health service representing communities across East Arnhem Land) and Charles Darwin University.

JD and SB recruited participants into the study and carried out the interviews; both had received specific training in interview techniques prior to the commencement of the project. All patients were given the option of an accredited interpreter in their first language if this was not English both for the process of obtaining written informed consent and the interview itself.

A mixture of purposive (non-probability sampling in which the researchers suggest who to approach to be included in the study based on them possessing certain characteristics [31]) and network (using existing participants to suggest other people to approach [32]) sampling was used to recruit individuals from a range of different backgrounds with a proportionate mix of gender, age and hepatitis B status. The majority of participants were recruited through the community clinic and the hospital liver clinic; however some individuals were recruited through the social and professional networks of the research team.

Interviews were carried out in numerous settings ranging from the community clinic, our hospital clinic, our research institution, individuals’ homes and gardens, under trees and at an international conference (8th Australasian Viral Hepatitis Conference, Auckland, September 2012). Interviews were audio recorded and ranged in duration from 20 to 45 minutes. Information collected in Yolŋu matha was translated into English in real time by the accredited interpreter and meaning and understanding clarified by SB (bilingual researcher present at all interviews carried out in Yolŋu matha) as part of the recording. Transcription of the interviews was in English.

An audio diary of the real time experience and reflections on the interviews was kept by JD and SB and included in the data analysis. All participants were offered an AUD$30 electricity voucher in recognition of their time and effort in contributing to the study.

In the process of exploring patients’ and providers’ knowledge, experiences and perceptions of HBV, data emerged on the potential impact of low levels of health literacy on healthcare interactions and therefore future health outcomes as well as the pathways through which this may occur. As such, we have used Paasche-Orlow & Wolf’s model [25] Figure 1 of the pathways that exist between low levels of health literacy and poor health outcomes as an organising model for our data analysis.

Data analysis was carried out by JD and SB, with input from VJ and JSD. It commenced with the first interview and was continuous throughout the project. Data immersion consisted of carrying out the interviews, reading the transcripts and listening to the audio recordings multiple times on multiple occasions dispersed over time. Sections of text were organised into codes based both on the categories covered in the interview schedules and also inductively as the text was digested and understood. Codes were also reflected upon with reference to the Passche-Orlow & Wolf model in particular with regard to the similarities and differences in using this model in this particular cultural context (Yolŋu people) for this particular disease (Hepatitis B). Concurrently and inductively the codes were organised into broader categories and themes. On multiple occasions clarification was sought regarding the cultural context of specific terms and ideas from SB. SB returned to individual participants to verbally clarify findings on a number of occasions however transcripts were not routinely returned to participants for checking. Data were organised and managed in NVivo 10 (QSR International Pty Ltd, Victoria, Australia).

JD, SB, SS & JSD are all clinical care providers as well as researchers and acutely aware of the ethical implications of this within this project particularly for those individuals interviewed who were Hepatitis B patients. Careful explanation of the fact that the research project and an individual’s clinical care are completely separate and mutually exclusive was undertaken with the Hepatitis B patient group in particular. Care was taken to conduct the interviews completely separately in both time and location from any clinical care so as to maintain this separation.

We adhered to the RATS guidelines in reporting this project.

Thirty two semi-structured interviews were carried out between July and September 2012. Participants consisted of clinic patients with hepatitis B (11), other community members (9) and key informants (12). Twenty-four (75%) were Indigenous people. Median age of participants was 45 years (IQR 35–55) and 18 (56%) were female. Highest level of education attained was junior school for one individual (3%), secondary school for 23 (72%) and tertiary education for 8 (25%). All participants had the opportunity to use an interpreter; 17 interviews were carried out using a Yolŋu matha interpreter (the principal Indigenous language spoken in the community). The remainder were carried out in English.

A culturally appropriate education resource: what we need…

When discussing educational resources, non-Indigenous key informants reported that an analogy with hepatitis B using a local animal (e.g. a crocodile or snake) to represent how the virus can lie dormant in the liver and then suddenly attack resulting in serious health consequences would be culturally appropriate. By contrast, Indigenous participants generally preferred more medical imagery requesting to see a real human-like figure with a real liver, and a story based in a culturally appropriate setting. One participant remarked that the majority of local animals are hunted as food by community members, so it would be counterproductive to use them to explain a human sickness - people would then think they could get the disease from the animal.

A strong desire to understand the detail about Hepatitis B was recurrently expressed but the need for contextual translation done in a culturally appropriate way was stressed. In general, Indigenous participants reported a preference for an electronic format with an emphasis on interactive pictures and less text. If text is utilised, it was clear from participants that it must be in Yolŋu matha and spoken as well as written.

There was a recurrent specific request for a separate “women’s business” section to speak about the issues specifically related to pregnancy.Figure 2 summarises the important aspects from the results which have been taken forward into the process of developing a culturally appropriate tool to aid in the development of effective treatment partnerships for Indigenous patients with CHB.In light of our results we have adapted Paasche-Orlow & Wolf’s model Figure 3 to highlight how the relationships between health literacy and poor health outcomes may operate for Indigenous Australians with respect to hepatitis B.

This study documents low levels of biomedical knowledge about hepatitis B which appear to be influenced by a multitude of factors including culture, gender, competing health priorities and a lack of shared understanding. Pessimistic almost fatalistic perceptions of the disease predominated across all groups of individuals interviewed. In terms of experiences the major theme identified was communication particularly the importance of having information available in an individual’s first language to aid in effective cross cultural communication. Indigenous individual’s repeatedly expressed a desire for increased knowledge and insight into the ability of this knowledge to reduce disempowerment and improve Hepatitis B specific health literacy. Ideas as to how to best enable this to happen included using visual aids, electronic formats, simple language and the absolute requirement for information to be available in Yolŋu matha.

Knowledge and beliefs are important patient factors in the patient-provider interaction component of Paasche-Orlow & Wolf’s model linking low levels of health literacy and poor health outcomes. A lack of biomedical knowledge about hepatitis B was identified in Indigenous individuals across all groups interviewed. This is consistent with data from Indigenous individuals in the Torres Strait [15] as well as non-Indigenous Australians from culturally and linguistically diverse backgrounds [13]. Lack of knowledge and erroneous beliefs about hepatitis B, as well as contributing to low levels of health literacy, may lead to a reduced ability or willingness to participate in decision making about management plans. This in turn may influence adherence with the plan and subsequent necessary self-care activities. Multiple factors affecting the provider side of the patient-provider interaction were also identified. Communication skills to allow shared understandings to be developed as well as insight into how best to achieve this are crucial in our context, where there are multiple competing priorities; however lack of these skills is identified in our results as an ongoing barrier to achieving shared understandings. In the context of Australian Indigenous peoples where English is not the first language and culture and worldview are very different we would suggest that the patient-provider interaction not only significantly contributes to health literacy but is a pre-requisite to allowing access & utilisation of care and self-care to occur and so ultimately influencing health outcomes (Figure 3).

As well as the patient-provider factors described above, extrinsic factors such as support technologies, health education and resources are identified as key factors to allow optimisation of self-care. The wider project that this research is part of was initiated due to a lack of culturally appropriate resources about CHB for use in clinical practice. Our data identified a real desire for more knowledge and understanding around CHB for all in the community to motivate and empower people living with CHB and community members, which in turn should increase self-management in relation to CHB. Our results identify a clear ambition by community members and people living with CHB towards ‘critical health literacy’ as defined by Nutbeam et al. [33] as the tertiary level of health literacy encompassing not only communication of information and development of personal skills but also personal and community empowerment.

There is now increasing experience with the use of innovative, interactive, internet, mobile phone and tablet-based resources to improve health literacy in other settings [34, 35]. In the context of Indigenous Australia, several groups have produced apps in the area of mental health [36] but robust evaluation of their value is still awaited. In northern Australia, Christie’s research group has proposed a tablet-based, easily transportable, touch pad body resource, which does not contain any embedded health messages, but rather focuses on aspects of a healthy body. Their vision is that this could be used as the foundation for a further discussion about the impact of chronic diseases on the body and how treatments act to return the body to a healthy state [16]. The evidence derived from this project that will be taken forward to phase 2 of the PAR process and used to guide the development of a culturally appropriate educational tool about Hepatitis B is summarised in Figure 2.

Effective communication is not only central to improving health literacy [37], it is a crucial element in achieving culturally safe healthcare, which in essence can be defined as “shared respect, shared meaning, shared knowledge and experience of learning together”[38]. More recently, research suggesting that some Indigenous patients believe that health care workers deliberately withhold information from them highlights the extreme lack of trust that can develop as a consequence of ineffective communication [21]. As communication transcends all aspects of health literacy, hence “culturally safe communication” at both a system and individual level is clearly integral to its improvement. Culturally safe communication has also been suggested as being important in reducing ethnic and racial disparities in healthcare [39]. Specifically in the Australian Aboriginal context, involvement of the local community in developing and implementing health education programmes, so they are culturally safe, has been shown to directly influence their effectiveness [29, 30, 40] and attention to worldview and language are argued to be integral to achieving improvements in health education [41]. It is therefore disappointing that more than a decade after the publication of Cass et al’s [20] paper documenting the pervasive nature of miscommunication between Indigenous people and their health care professionals, our results show the major barrier to achieving critical health literacy is still poor cross-cultural communication. Consistent with the view of Vass et al. [41] who suggest “the health literacy of Indigenous Australians can be improved by promoting the oral use of the peoples’ first language in the health sphere” Indigenous participants anticipate they will better understand and be able to process and act on information given to them in their own language.

Our results also provide further insight into the complexity of achieving effective and culturally safe communication in this setting, when, for example, the lack of a shared understanding of one word – “silent” – which is used so commonly in clinical practice with hepatitis B patients can lead to such significant misunderstanding. We have also highlighted the potential for miscommunication to be perpetuated in health settings when communities are not adequately consulted about health education and health promotion resources. The well-meaning but mistaken beliefs among non-Indigenous key informants in this study about the appropriateness of using animal analogies when discussing how hepatitis B affects the liver or the mistaken belief that the lack of a direct translation of a word prohibits meaningful translation of key messages, are two examples from our data.

The negative perceptions and fear of hepatitis B as a disease may originate from the low levels of health literacy documented and contribute to stigma and potential non-disclosure of diagnosis as well as having implications for individual clinical care and the success of public health interventions. This pessimism may have been confounded by the lack of shared understanding and different health beliefs about causation in the context of provider-patient interactions. Additionally, the non-Indigenous key informants in this study perceived that there are multiple logistical barriers and competing priorities to providing effective and appropriate long term care for people living with CHB and felt overwhelmed by the task. This negativity is likely to adversely influence an individual’s access and utilisation of care and so contribute to the relationships between limited health literacy with inequitable health outcomes as per Paasche-Orlow & Wolf’s model.Our study is limited by the fact it only included one community and because of multiple previous education and research projects in this community in the discipline of infectious disease, it is likely that this community has higher health literacy that most regarding infectious diseases specifically. Cultural practices, traditions and world view may be totally different to other Australian Indigenous peoples; however, our findings about the importance of communications and shared understandings are likely to transcend region and apply to all Indigenous Australians. This view is supported by the similarities between our findings, and those of Preston-Thomas et al., who investigated HBV knowledge in a completely different group of Indigenous Australians – Torres Strait Islanders. Although not directly translatable to other cultures, it is likely that the modified factors highlighted in Figure 3 will be of greater importance to those people living with CHB from culturally and linguistically diverse backgrounds, particularly if they are receiving care in a country where the language of health care is not their own first language.

Although low levels of biomedical knowledge about CHB are clearly a significant barrier and an important influence on health literacy our findings resonate more clearly with Christie et al’s [16] definition of health literacy. In this context, what is really critical to improving health literacy is developing a shared understanding between patients and providers, which hinges on effective communication. If we can use the insight we have gained from this study and work with the people who provided it to develop an educational tool grounded in their culture, in their first language and make it easily accessible, that would be a first step to improving health literacy about CHB. Qualitative research using a participatory approach holds promise of breaking cross-cultural barriers in health communication and health care. We acknowledge that there will also need to be appropriate implementation and evaluation of the resulting resource to ensure its success.

Biomedical knowledge about Hepatitis B is low in this Indigenous community in the Northern Territory, experiences and perceptions about CHB are in general negative and at times nihilistic. However there is a strong desire for increased knowledge and evidence of increased understandings with contextual translation of information. Patient provider interactions leading to the development of shared understandings between Indigenous people living with CHB and the health care professionals looking after them are the foundation for improving health literacy and so health care outcomes related to CHB. Language and using a culturally appropriate worldview are crucially important in developing an educational resource to aid in developing treatment partnerships for Indigenous patients with CHB. Maintaining a participatory approach to development should help to reduce disempowerment and overcome some of the barriers to its implementation and success.

^aSorcery as a cause of disease is a commonly held belief in Indigenous communities in Arnhem Land particularly where a death is sudden, unexplained or happens to someone who is seen outwardly as healthy. It can be a form of retribution or punishment but is not always viewed in this way.