Reducing smoking among Australia's Indigenous people has been identified as a government priority in its efforts to "Close the Gap" between Indigenous and non-Indigenous life expectancy [1]. Approximately half the adult Indigenous population smokes, with similar rates for men and women [2]. Identified drivers of smoking among Indigenous Australians include a history of colonisation and dispossession, socio-economic disadvantage and marginalisation, acceptability and normalisation of smoking within Aboriginal social networks and the role of tobacco in social exchange [3–7].

Addressing tobacco smoking during pregnancy could bring significant health gains. Studies with pregnant Australian Indigenous women report smoking prevalence rates between 50% and 67% [8–14], approximately three times that in the non-Indigenous population [14]. Smoking during pregnancy is associated with increased risk of maternal and infant adverse outcomes. For the mother, these include higher rates of placental abruption, placenta praevia, premature labour and premature rupture of membranes [15, 16]. For the baby, adverse outcomes include low birth weight, preterm birth, intra-uterine growth retardation, perinatal death and Sudden Infant Death Syndrome [15, 16]. Examination of population-level data confirms these adverse outcomes among Aboriginal women [10].

Pregnancy is considered a "teachable moment"-a time when women are more motivated to modify their behaviour than at other times [17]. Numerous studies in the general population have demonstrated high spontaneous quit rates, with additional women reducing the amount smoked [17–21]. There is also some evidence to support pregnancy as a "teachable moment" among Aboriginal women. A qualitative study with pregnant Aboriginal women from Western Australia found that a few women quit smoking when they became pregnant, but the majority preferred to try to reduce the number of cigarettes smoked as quitting was too difficult and the benefits of smoking outweighed those of quitting [6]. A recent survey of pregnant Aboriginal women in New South Wales (NSW) and the Northern Territory (NT) found that 21% of women smoking at the beginning of their pregnancy reported quitting, while a further 46% reported reducing their smoking [22]. However, smoking during pregnancy among Aboriginal women remains common, perinatal data indicate low quit rates [13], and audits have revealed gaps in provision of advice [23], suggesting that more could be done to capitalise on this opportunity.

Antenatal care providers see pregnant women multiple times throughout their pregnancy, with national data indicating that 77% of Indigenous women attended for at least five antenatal visits [24]. There is strong evidence that provision of smoking cessation counselling in the antenatal period is effective [25]. In other populations, studies have confirmed that women consider provision of smoking cessation support within the antenatal clinic setting to be appropriate [26–28], that information on smoking is best provided by health professionals [29], and that this advice is an important factor in helping women quit smoking [30]. However, the approach and manner of this advice is important [31].

Providers caring for pregnant Aboriginal women are thus potentially important change agents for addressing the harms associated with smoking. Assessment of client smoking is an essential first step in providing tailored cessation advice and support. Exploring the knowledge and views of providers on the value and effectiveness of addressing smoking with pregnant Aboriginal women, and factors associated with provision of care will assist in determining the best approaches to optimising their effectiveness. To date, there have been no studies published on this topic.

This paper explores perceptions of health care providers who provide care to pregnant Aboriginal women regarding:

a. Their estimate of the prevalence of smoking among Aboriginal women in their community;

b. Optimal and actual assessment of smoking status;

c. Their knowledge and attitudes to providing smoking cessation advice; and

d. The factors associated with self reported assessment of smoking among pregnant Aboriginal women.

A cross-sectional survey was undertaken with staff providing antenatal care from two jurisdictions: those working in remote medical services in the NT and those providing care through the Aboriginal Maternal and Infant Health Strategy (AMIHS) in NSW. AMIHS teams are comprised of community midwives and Aboriginal Health Workers (AHWs) working together to provide outreach antenatal care for Aboriginal women in multiple sites across NSW. While both professionals work together, the midwives usually focus on the clinical aspects of antenatal care, and the AHWs address health education, community development and social needs. In the NT, staffing of remote clinics varies, with most clinics employing midwives or nurses to provide clinical aspects of antenatal care, and AHWs supporting clinical care and providing health education. Aboriginal Health Workers are a specific category of Australian health professional, working in both government health services and Aboriginal Community Controlled Health Organisations. They are usually members of the local community and work to help bridge the cultural gap between Aboriginal people and the western medical system.

A community reference group (CRG) of Aboriginal women and service providers from rural NSW was formed to guide the study and ensure it was conducted in a culturally secure manner [32]. The CRG provided input into the content and wording of the questionnaire, interpretation of findings and endorsed the reports and papers from the study.

In this study we found that better smoking cessation knowledge, a positive attitude towards providing cessation advice and being employed by a government health service were significantly associated with higher rates of self-reported assessment of smoking status while providing antenatal care to pregnant Aboriginal women. There is some indication that being a non-smoker, and disagreeing with statements expressing concern that women would be pushed away from antenatal care or about having inadequate skills were also associated with higher rates of assessment, but these relationships were not statistically significant at the 5% level.

To our knowledge, this is the first study to specifically explore the knowledge and attitudes regarding smoking among service providers caring for pregnant Aboriginal women in Australia. The study was undertaken in two jurisdictions among antenatal care providers in remote, regional and urban Australian settings. The number of eligible providers was less than anticipated and the consequent small sample may have limited our ability to identify other significant associations. However, the response rate was good and the sample is likely to represent this group of service providers reasonably well.

Consistent with studies among other antenatal care providers [28, 42, 43], the majority of respondents considered that assessing smoking status of all women was integral to good antenatal care and a professional and service responsibility. The majority also indicated that they do ask all women about their smoking. However, over one fifth reported not always asking all women, indicating a missed opportunity for addressing a major preventable risk factor for adverse birth outcomes. While there may be over-estimation of rates of assessment due to social desirability bias, the positive attitudes expressed by the majority of respondents are an asset with potential to be enhanced by training and skills development.

In general, knowledge of risks associated with smoking was high, particularly in relation to birth outcomes and infant illness, but not for childhood health problems. An earlier study of Australian directors of antenatal clinics identified poor specific knowledge of risks [33]. Others have reported that providers did not consider smoking a serious risk to infant health [35, 44]. The uncertainty regarding risks of ongoing problems in childhood indicates gaps in knowledge and lost opportunities for conveying the true burden of antenatal smoking. As provider knowledge of risk is essential for conveying clear messages regarding risk to women, education of providers regarding the specific risks associated with smoking is essential.

Knowledge of smoking cessation was poor and inversely associated with level of assessment. Only half the respondents recognised that complete and sudden cessation was an effective quitting method; three quarters incorrectly indicated that gradual reduction was effective; and one third incorrectly indicated that NRT shouldn't be used in pregnancy despite national guidelines stating that NRT can be used in pregnancy [41]. Other studies have also identified a preference among antenatal providers for advising reduction rather than complete cessation [33–36, 42, 45]. Smoking cessation interventions are poorly covered in nursing curricula [46] and in training for Aboriginal Health Workers [47], which may explain the low level of knowledge and that approximately one fifth felt they didn't have the skills to provide advice. Perceived skill level is associated with provision of tobacco interventions [37, 48], and lack of skills have repeatedly been identified as a barrier to smoking cessation counselling by practitioners [31, 49–51]. In our study, both knowledge scores and perceptions of skills were related to level of smoking assessment, suggesting that provision of culturally appropriate, pregnancy-specific training and resources would increase confidence and skills and consequently assessment and management of antenatal smoking.

Although only a small proportion of respondents agreed that giving advice was not worth it, this perception was strongly associated with level of reported assessment, suggesting that pessimism regarding the impact of advice may contribute to non-assessment of smoking status. Pessimism about the effectiveness of interventions has been identified as a barrier to providing cessation counselling in other Australian antenatal settings [37, 42, 51] and internationally [38, 43, 44, 52, 53]. The poor knowledge of smoking cessation identified is likely to contribute to low efficacy of any advice provided, further contributing to a perception that advising cessation is futile. Within the context of providing care to women with multiple complex care needs with constrained resources [5, 6, 23], providers who anticipate low success rates may prioritise other activities which are easier to implement or have greater chance of success.

Provider smoking has been shown to be negatively associated with smoking cessation counselling [48, 54] and smoking is perceived to be a barrier to providing cessation counselling among Aboriginal Health Workers [6, 39, 55–57]. Several studies have suggested assisting AHWs to quit in order to increase their comfort in providing cessation support [55, 56]. Although only 19 (15%) of the respondents smoked, provision of smoking cessation support to those who do is likely to be beneficial for the individuals, enhance their willingness to provide cessation supports, and give them personal experience in quitting.

One quarter of the respondents indicated concern that providing advice might push women away and this was associated with lower smoking assessment, although this was not significant at the 5% level in the multivariable analysis. In a Western Australian study of smoking during pregnancy, AHWs expressed discomfort about raising smoking as they wished to maintain positive relationships with women [6]. Similar concerns have been identified in other studies with AHWs, although they report feeling more comfortable discussing smoking with pregnant smokers than with other smokers [39, 56, 57]. In other antenatal settings, a perception that clients are not interested or do not expect advice, has been identified as a barrier [37, 52], and midwives have expressed concern about potentially damaging their relationship with women if they address their smoking [28, 31, 50]. By contrast, women consider provision of smoking cessation advice within antenatal care to be acceptable [26], and state that it doesn't affect their relationships with their midwives [27]. However, the manner of providing care is important, and should not be authoritarian [28, 50, 54, 58, 59]. Greater community and provider understanding of the real risks of smoking and benefits of cessation may increase community support and help providers feel more comfortable addressing smoking.

Limitations to this research should be considered in interpreting the findings. In addition to the small sample size and potential social desirability bias mentioned above, the cross-sectional nature of the survey prevents assessment of causality in the relationship between assessment and the knowledge and attitudinal variables. Factors other than the knowledge and attitudinal variables included in our study may be determining respondents smoking cessation activities, and the reported attitudes may then reflect a rationalisation on the part of respondents to justify their behaviour. Trials of interventions that aim to address knowledge and attitudes would be beneficial in assessing this. A further limitation is that the study did not assess the amount or type of advice that the clinicians provide to women.

This study has identified factors constraining the provision of evidence-based antenatal care in relation to tobacco use, but has also found strengths on which to build. The majority of providers recognised that smoking increased the risk of adverse outcomes, considered that giving cessation advice was important, and believed that providing advice was their responsibility. The majority reported assessing all women for smoking and saw it as part of optimal antenatal care.

The poor knowledge of providers regarding smoking cessation reinforces the call from others for development of culturally appropriate training and resources for providers caring for Aboriginal peoples [5, 47, 55, 60], including those specific to pregnancy [6, 61]. Addressing knowledge of risks and smoking cessation counselling among antenatal providers is a priority and should improve both confidence and ability, and increase the frequency and effectiveness of counselling. Programs designed to support pregnant women to quit smoking need to address the many drivers of smoking, including high levels of stress and disadvantage, and social norms of smoking [4–7]. Additionally, the health system must provide supports to providers through appropriate policy and resourcing, to enable them to address this issue. Recent government initiatives in Indigenous smoking are likely to raise recognition of the importance of addressing smoking at every opportunity and should be accompanied by broader efforts to address Indigenous disadvantage.