The project was undertaken at two ACCHS located in the remote towns of Derby (Derby Aboriginal Health Service – DAHS) and Kununurra (Ord Valley Aboriginal Health Service – OVAHS) in the Kimberley region of far north Western Australia  between January 2009 and June 2012. Aboriginal researchers were trained to deliver the program. These researchers came from various backgrounds: Aboriginal Health Worker (1), other Aboriginal health service employees (4) and people without a health background (4). For most of the study project managers were based at the centralised coordinating site (Kimberley Aboriginal Medical Services Council), in Broome. Investigators and research staff met before and at least annually during the trial to discuss all procedures. The Broome-based investigators (JM, DA, CN) met regularly to discuss any problems that arose during the trial.
Study design and participants
The protocol for the BOABS Study has previously been described . Briefly this study was a prospective, randomised, trial with parallel groups. Aboriginal and Torres Strait Islander smokers (current or who had quit within two weeks of enrolling) aged 16 years or older who were residents of Derby and Kununurra and who demonstrated any desire to quit smoking or cut down on the amount of cigarettes they smoked were recruited to the study. The exclusion criteria were: unable to provide informed consent; a health condition that would prevent them from completing the trial; or unlikely to be available for follow up at 12 months.
Participants were actively and opportunistically recruited. Active recruitment was facilitated by the Aboriginal researchers through incidental encounters in the community, family and community links. Participants were passively recruited through DAHS and OVAHS routine clinic visits as well as through running dedicated chronic disease prevention and screening clinics. Clients known or found to be current smokers were offered participation in the trial by health care providers. Recruitment originally intended to last 12 months eventually concluded at 30 months due to resource constraints (see below).
Randomising and masking
As the planned intervention was labour intensive, participants were allocated to usual care or intervention groups in a 2:1 ratio. A computer generated random allocation sequence was used. Sealed envelopes containing the allocation were kept at the centralised coordinating site. Allocation occurred via telephone with envelopes being opened in sequential order for each site only when a participant had provided written consent and been enrolled in the study, eligibility confirmed, an identification number assigned, and the first questionnaire completed.
Due to the nature of the intervention the participants and research team were subsequently aware whether the participant belonged to the usual care or intervention group. Local healthcare staff may have been aware of this. The staff performing the urinary cotinine assay were blinded to the allocation.
Participants allocated to usual care received routine care relating to smoking cessation at their local primary health care service, including advice regarding quitting, pharmacotherapy, and self-initiated follow up. Participants allocated to the intervention group received usual care as above. In addition they were provided with smoking cessation counselling at face-to-face visits, which were scheduled weekly for the first four weeks, monthly to six months and two monthly to 12 months (12 sessions). The content delivered by Aboriginal researchers to participants in the intervention group included: motivational interviewing; diversions and strategies to deal with smoking triggers; action plans for preventing and dealing with short term relapses; discussion regarding the positives of smoking cessation; referral for and titration of pharmacotherapy; identification of factors driving smoking and case management to address these by linking participants with additional non-health support agencies (e.g. public housing, welfare, domestic violence and alcohol services); and strategies for smoking cessation-associated weight gain. They were also encouraged to attend a monthly smoking cessation peer support group. We planned that the Aboriginal researchers would organise weekly case conferences with the health care clinic medical officer/general practitioner to review all active participants in the intervention group.
Planned study endpoints were at six and 12 months following enrolment. The planned primary endpoint was not smoking at 12 months defined as self-reported non-smoking for at least seven days  and low urinary cotinine . Final follow-up was extended to improve completeness, with 22 and 11 follow-ups from the usual care and intervention groups, respectively, occurring more than 15 months after enrolment. Multiple questions (Are you still a smoker? Have you smoked since the last BOABS Smoking Checkup? When was your last cigarette? On an average day how many cigarettes a day do you smoke?) were used to assess self-reported quitting. Urinary cotinine was analysed using tandem mass spectrometry (Waters 2795 Autosampler with Waters Quattro Premier Mass Spectrometer, Milford, MA USA) with a cut-off to indicate not smoking of < 50 ng/mL . Secondary outcomes were the proportion of participants who: were not smoking at six months for at least seven days as determined by self-report; reported fair/poor health using a self-reported health questionnaire ; reported 20% or greater reduction in the number of cigarettes smoked each week; and various process evaluation indicators that assessed how well the intervention program was implemented according to the protocol (e.g. number of ‘face-to-face’ meetings completed). We also recorded quit attempts during the study.
The sample size estimation was based on a two-sided alpha of 0.05, power of 80%, ratio of intervention to control participants of 1:2, and an anticipated efficacy of maintaining smoking cessation at 12 months of 3% in the control group based on local program evaluation (personal communication CN) and 13% in the intervention group. Based on these assumptions the sample size required was 106 in the intervention group and 212 in the control group. To take into account those lost to follow up, the target sample size was 120 in the intervention group and 240 in the control group.
Differences in baseline characteristics, and primary and secondary outcomes were compared using χ2 tests for categorical data, t tests for continuous normally distributed data and Mann–Whitney test for continuous non-parametric data. Changes from baseline to final follow-up of nicotine dependency score ≤5 (decreased dependency, no change, increased dependency) were compared using the Cochran-Maentel-Haenzel test for linear association. Fishers exact test and relative risks (RR) were used to compare smoking cessation endpoints between the intervention and control groups. Multivariate analyses did not significantly change the results and hence have not been included. All statistical tests were two-sided and a p value less than 0.05 was taken to be statistically significant.
Systematic review and meta-analysis
Based on CONSORT recommendations of placing the results of this study into context  we carried out, post hoc, an associated systematic review of smoking cessation trials in Indigenous populations. Our selection criteria for inclusion in the meta-analysis were randomised controlled trials of interventions with validated smoking cessation in Indigenous populations. Our exclusion criteria were 1) trials comparing pharmaceutical interventions compared with placebo and 2) all studies where quitting smoking was not biochemically validated for all participants who self-reported they had quit smoking (e.g. urinary cotinine). A recent meta-analysis of four studies reported a RR of 1.43 (95% CI; 1.03-1.98) for smoking abstinence at 6–12 month follow-up in the intervention group, based on a very low quality of evidence . Additional searches of the Cochrane Tobacco Addiction Group Specialised Register of Trials (June 2013), MEDLINE (June 2013), online clinical trial databases and publication references for potential studies were also conducted. The four trials reported in the recent review did not meet the selection criteria: two were not randomised [18, 19], one assessed the efficacy of a pharmaceutical intervention (bupropion) compared with placebo , and one had smoking cessation validated in only a subset of participants who self-reported that they had quit smoking .
Only one published study  met these criteria and we combined our results with the data from this RCT. Although this study focused on pregnant women, like our study this was based on one-on-one intensive support compared with usual care. Both studies also allowed recent quitters to enrol and in line with other smoking cessation studies used 7-day point prevalence abstinence of smoking confirmed by cotinine to classify non-smokers . An estimated pooled weight average for RRs was calculated using the Mantel-Hetzel fixed-effect model, with 95% confidence intervals. This meta-analysis was not part of the original study protocol.
The trial was conducted in compliance with the study protocol , the principles of Good Clinical Practice, the NHMRC National Statement on Ethical Conduct in Human Research (2007), NHMRC Australian Code for the Responsible Conduct of Research (2007), and NHMRC Values And Ethics: Guidelines For Ethical Conduct In Aboriginal And Torres Strait Islander Health Research (2003) [23–25]. The trial received approval from The University of Western Australia Human Research Ethics Committee and the Western Australian Aboriginal Health Information and Ethics Committee, and support from the Kimberley Aboriginal Health Planning Forum Kimberley Research Subcommittee.