Method: Recruitment of women pregnant with an Aboriginal child occurred throughout South Australia from Feb 2011 to May 2012. Recruitment was largely through hospitals, Aboriginal community-controlled health organisations and 22 other key stakeholder groups. Self-reported data were collected at baseline to evaluate self-rated oral health, self-efficacy and socio-demographic, psychosocial, social cognitive and risk factors.
Result: Data from 446 women was obtained (age range 14 to 43 years); 224 were randomly allocated to the intervention group and 222 to the control group. The sample was representative of the total population of pregnant Aboriginal woman in South Australia in the recruitment period. In an unadjusted model, those with low self-efficacy had 1.58 times the prevalence of rating their oral health as ‘fair’ or ‘poor’ (95% CI 1.26–1.98). In an adjusted model, which controlled for socio-demographic, psychosocial, social cognitive and behavioural risk factors, the prevalence of fair or poor oral health among those with low self-efficacy was attenuated by 15 percent (PR 1.35, 95% CI 1.03, 1.78).
Conclusion: Recruitment strategies yielded a representative sample of the target population. Randomisation allocated approximately equal numbers of participants to intervention and control groups. Low self-efficacy persisted as a risk indicator for poor self-rated oral health after adjusting for confounding among this vulnerable population.