The study is a randomized controlled trial to test an oral health promotion intervention aimed at new-born children and their parents/carers in metropolitan Perth (population 1.7 million) and the regional cities of Bunbury (population 32,000) and Busselton (population 31,000) in Western Australia. It is a collaboration between Child and Adolescent Community Health and Dental Health Services (DHS), both Department of Health state-wide agencies delivering services for children.
Study participants will be recruited from all child/community health clinics in metropolitan Perth (fluoridated; 0.8 mg/L fluoride) and Bunbury/Busselton (non-fluoridated; < 0.2 mg/L fluoride). Children/parent dyads attending the child health clinics from these locations will be invited to participate, and the randomisation procedure will be based on individual randomisation stratified on the basis of residential location of Perth or Bunbury/Busselton. All new-born WA children are referred to their local child health clinic and receive a home visit from the child health nurse within 6 weeks of birth and are seen at the local child health clinics at regular intervals for health assessments throughout the early years. Parents of a new-born child presenting to selected child health clinics will be invited to participate and provided with information about the study including an oral health promotional package, a consent form and a self-complete questionnaire by the child health nurse. Upon receipt of consent to participate, consenting parent/caregiver and child dyads will be randomly allocated to either an intervention or control group by a central study coordinator using a computer generated random permuted blocks (block sizes; 4, 6, 8, 10, 12). The blocks will ensure that at most, an imbalance between test and control groups will be 6 participants at the conclusion of participant recruitment. Blinding of parent participants and counsellors will not be possible.
The self-completed questionnaire will collect socio-demographic (age, sex, education level, family income, Aboriginality) and psychosocial information (oral health-related self-efficacy, knowledge about appropriate baby bottle use and children’s oral hygiene and belief in oral fatalism, instrumental social support and parental stress) and parental perceptions of treatment need and oral health care behaviours. The psychosocial measures have been used in other studies and have had their psychometric properties validated [38
The psychosocial measures are shown in List of Psychocosial Measures. Oral health self-efficacy comprised nine items scored on a Likert scale from 1 (not at all confident) to 4 (very confident); higher scores indicate greater self-efficacy. Oral health knowledge on appropriate use of baby bottle and children’s oral hygiene comprise 10 items scored on a Likert scale from 1 (strongly agree) to 5 (strongly disagree); higher scores indicate better knowledge. Oral health fatalism comprised one item, “Most children eventually develop dental cavities”, scored 1 (strongly agree) to 5 (strongly disagree). The parenting stress instrument comprised six items scored on a Likert scale, 1 (never) to 5 (almost always); higher scores indicate higher levels of stress. The social support instrument comprised four items scored “yes” or “no”. Mean scores from sum of the item scores will be used for the oral health self-efficacy, oral health knowledge and parenting stress scales while the oral health fatalism score will be converted to a binary variable reflecting parent/caregiver agreement with the statement and each item of the instrumental social support instrument will be
presented as frequency scores and used as a binary variable.List of Psychological measures
Oral Health Self-efficacy; scale, 1
not at all confident, 4
How confident do you feel that you are able to make sure that your child’s teeth are brushed before bedtime when you are:
1. Under a lot of stress?
4. Feeling like you do not have time?
6. Worrying about other things in your life?
7. Bothered by your child crying?
8. Bothered because your child doesn’t stay still when you want him or her to brush?
9. Told by your child that he/she does not feel like brushing right now?
Knowledge of bottle use; scale, 1
strongly agree, 5
1. Putting a baby to bed with a bottle helps the child to be better fed.
2. Putting a baby to bed with a bottle helps the child sleep better.
3. Putting a baby to bed with a bottle helps the child to gain weight and grow.
4. There is nothing wrong with putting the baby to bed with a bottle.
Knowledge of children’s oral hygiene; scale, 1
strongly agree, 5
1. Cavities in baby teeth don’t matter since they fall out anyway.
2. Keeping baby teeth clean is not very important; after all, they fall out.
3. There is not much I can do to stop my child from developing dental cavities.
4. There is not much I can do to help my child have healthy teeth.
5. Children don’t need to brush everyday until they get their permanent teeth.
6. Children don’t really need their own toothbrush until all their teeth come in.
Oral health fatalism; scale 1
strongly disagree, 5
1. Most children eventually develop dental cavities.
Parental stress; scale 1
1. How often do you feel that you have too little time to spend by yourself?
2. How often do you wish you did not have so many responsibilities?
3. How often would you say your child (or children) gets on your nerves?
4. How often do you feel that your child/ren is/are making too many demands on you?
5. How often do you feel that being a parent is much more work than pleasure?
6. How often do you feel tired, worn out, or exhausted from raising a family?
Instrumental support; scale “yes”, “no”
If you need to, is there someone you can count on to……
1. ……Run errands for you?
2. ……Lend you money?
3. ……Watch over your child/ren?
4. ……Lend you their car or give you a lift?
We will also seek information relating to the mother’s health during the last trimester of pregnancy and the child’s health in the peri-natal period at base-line. The weight and height of each child and primary caregiver will be measured at 24- and 36- month follow-up visits. Parents will recall 24-hour and usual dietary information for each child using a self-administered short food and beverage frequency questionnaire at the 12- and 36-month follow-ups. The 12- and 36-month questionnaire will also collect information relating to oral health care behaviours for the child including toothbrushing and use of fluoride toothpastes, fluid consumption, and use of professional oral health care services.
Dietary related information collected via a 24 hour recall questionnaire will be coded for cariogenic potential of the foods and drinks using the approach outlined by Lee and Messer (2010) [42
] and the approach outlined by Bennet et al. (2009) for obesogenic potential of the foods and drinks [43
The intervention group will be provided with oral health promotion delivered by a trained dental assistant (oral health counsellor, OHC) using the MI/AG protocol. The training of the counsellors involved a workshop comprising two days of didactic presentations, role-playing exercises and a practice counselling session with a parent–child dyad at the DHS’ School Dental Therapy training facility. An evaluation of the training program was undertaken and consisted of a MI pre- and post-test knowledge–base questionnaire, a Helpful Response Questionnaire and a trainer evaluation questionnaire. The knowledge and trainer evaluation questionnaires were based on those developed by Yale University School of Medicine [44
]. The Helpful Response Questionnaire utilised was developed by Miller et al. [45
At this stage, preliminary analyses have been undertaken of the OHCs’ knowledge of MI and trainer evaluation. The questionnaire to assess the OHCs' knowledge of the principles and practices of MI was administered as a pre–test (immediately before training) and post–test (immediately after training). It consisted of ten multiple-choice questions. OHCs scored one point for each correct answer.
From a maximum possible score of 10, the OHCs’ knowledge pre–test mean score was 4.3 (SD
1.49) and post–test score was 6.4 (SD
1.43). Comparison of mean scores using a paired two–sample t–test revealed a statistically significant difference (p=0.01).
It was concluded that the two-day training workshop increased OHCs’ knowledge of MI principles and practices. Full details of the evaluation of the training workshop for OHCs will be presented elsewhere (Manuscript in preparation).
Monitoring, feedback and improvement of the OHCs’ competence in the use of MI is on-going through the analysis of audio recordings.
Counsellors will be further assisted and supported by the researchers listening to selected interview sessions to monitor the fidelity of the MI/AG approach. Three, one-day follow-up meetings with the counsellors was also undertaken to further develop and refine the MI/AG counselling approach. Throughout the research period the fidelity of the MI intervention will be monitored with audio recordings, which will later be analysed using the Motivational Interviewing Treatment Integrity (MITI) instrument developed by Moyers et al. (2005) [46
]. The MITI is a behavioural coding system to measure competence in the use of MI and enable the provision of feedback about MI practice. The MITI consists of two components: the global scores and the behaviour counts. Using the global score system, the coder assigns a number along a seven-point Likert scale to rate the dimensions of interviewer empathy and MI spirit. The behaviour count requires the coder to tally instances of interviewer behaviours relating to giving information, questioning, reflection, MI adherent and non-adherent skills. Global scores and behaviour counts are then compared with recommended thresholds in the MITI coding manual (Moyers et al. 2003) [47
]. This approach enables OHCs to be provided with structured, formal feedback about ways to improve MI practice.
The OHCs employed in this intervention are fully qualified and experienced Dental Therapists and Dental Clinic Assistants. They receive education in basic oral health messages as part of their studies and through regular in-service training that their employer DHS conducts. In addition, comprehensive background notes on early childhood oral health promotion and AG were provided to the OHCs.
The control group will receive the standard care delivered through the universal “lift the lip” program (available throughout WA since 2011). The “lift the lip” protocol requires a child who presents to a child/community health clinic for a review at 8, 18 and 36 months of age to be screened for signs of dental decay by a child/community health nurse. An inspection of the oral cavity is undertaken, usually by lifting the upper lip and inspecting the upper anterior teeth for signs of decay. If signs of decay are detected, the parent is offered a referral to a dental practitioner, either to a private practitioner, or to a government general dental clinic if eligible for government supported dental care. All children, test and control, will continue to be provided with the standard lift the lip screening throughout the study period.
The OHC will undertake the first counselling session (home visit, expected to last up to 30 minutes) within 4 weeks after consent with two additional counselling sessions, negotiated between the OHC and the study participant as to face-to-face or phone contact, in the first 6 months (expected to last 15–30 minutes). The protocol for MI/AG will be adapted from Weinstein et al. and Plutzer & Spencer [30
The MI/AG approach will follow a structured approach of rapport establishment and identification of oral health and nutritional needs using empathic reflective listening; presentation of menu of options and information with permission; discussion of options and elicitation of “change talk”; elicitation of importance and confidence in behavioural change; the development of behaviour change plan; and a schedule of follow-up. All discussions will be undertaken in a collaborative, person centred approach central to the MI “spirit”. Whilst the menu will present a range of options with potential to affect early childhood dental decay, the discussion will be limited to one or two topics of primary concern to the study participant, in keeping with the MI approach. Negative responses during the counselling session will be accepted without rebuttal and issues left for exploration at a subsequent session. AG will follow the anticipated milestones of teething and transition from wholly breast- or formula-fed to solid foods. Again, topics for discussion will arise in conversation with the study participant with guidance from the counsellor and the counsellor will provide information only when requested to do so. The information will incorporate appropriate dietary and nutritional guidance, following the national guidelines, to assist in the development of healthful behaviours, with the counsellors linking oral and general health impacts of these behaviours [48
]. Review of audio recordings of counselling sessions will be undertaken to monitor protocol compliance.
Sample size and power
Sample size of approximately 514 children in each arm of the trial is required to detect a 35% reduction in dental decay experience [based on decay experience of children within DHS; dmft 1.26 (SD 2.63), test dmft 0.8] with a probability of 0.8 (power 80%) and rejecting the likelihood of no difference at probability 0.05 (α
]. This is a conservative sample allowance to detect a difference, since an Australian intervention study on early childhood decay found 79% reduction in decay incidence among 20 month-old children [23
]. Allowing for drop-off and loss to follow-up over the following 3 years, an initial sample of approximately 750 in each group of the trial would be required (an earlier study of caries prevention within SDS found attrition rate of 9% per annum over two years [50
]). The estimated sample size will have a power of 94% to detect a 35% reduction in the prevalence of early childhood decay (dmft
0; from 0.25 to 0.16) between the groups (α
0.05). The estimated sample size will also permit detecting a 33% difference in prevalence of overweight or obese children between the two groups (from 20% to 14%) with 80% power and α
Ethical approval for the study has been provided by the University of Western Australia Human Research Ethics Committee (RA/4/1/4469), the Princess Margaret Hospital for Children Ethics Committee (Reg. No. 1900/EP), and the WA Country Health Services Ethics Committee (ref 2011:05). Figure shows the study sample and pathway for the study participants in the test and control arms. All child participants will have the same clinical and other information collected again at approximately 5 years of age. Sample maintenance strategies, including card mail-outs on the child’s birthday, change of address notification and regular updating of the study will be undertaken.
The primary outcome measures are the incidence of dental decay in primary teeth and the prevalence of obesity. Children will be examined at the dental therapy clinics of the DHS of Western Australia for the incidence of dental decay by two blinded, calibrated examiners at 24 and 36 months of age, using the criteria specified by Workshop on Reporting Early Childhood Caries [51
] and at 60 months using the criteria of the WHO [52
Examiner training will involve examination of early childhood decay under the guidance of a specialist paediatric dentist. Further examiner calibration will be undertaken by examination of 10 children consenting to participate in an oral screening program of DHS. During field examinations, approximately 5% of study participants will be re-examined to test for intra- and inter-examiner reliability. Child height/length and weight will be measured and Body Mass Index calculated (BMI; weight/height2
) and BMI z score determined at 24 and 36 months, using the age-specific WHO growth standards [53
]. Number needed to treat as an effect measure of the program will be estimated [54
]. Examiner training will be repeated for each cycle of the clinical examinations to maintain reliability over the course of the study.
Secondary outcome measures will be changes in knowledge, attitude, behaviour and self-efficacy of parents towards the oral health of their child and differences in nutritional and dietary patterns between groups. Long-term outcomes will be the incidence of dental decay and referral for care under general anaesthesia among the groups when the child is 5 years of age.
The research will adhere to guidelines within the Australian Code for the Responsible Conduct of Research [55
]. All collected data will be transcribed into electronic format and stored on a password-protected computer accessible only by the researchers. A separate electronic file will be created, excluding identification data for data analysis.
Oral health outcomes will be defined by counts of decayed, missing and filled teeth/surfaces and include non-cavitated lesions. Predictor variables will be on the basis of group allocation to intervention or control group and the posited psychosocial mediating variables will be defined using the approach adopted by Finlayson et al. [39
Appropriate parametric and non-parametric statistics will be used to test the aims of the study. Differences in outcomes between the groups will be tested on an intention to treat and per protocol basis. Univariate analyses using t-tests for continuous variables (oral health knowledge, behaviour, attitudes, self-efficacy and parental stress of parents/caregivers), and Chi square analyses for categorical variables will be undertaken to examine the baseline factors and their association with group assignment. Univariate analyses will also be used to examine the association between the primary outcomes of dental decay incidence and prevalence, and early childhood obesity prevalence, with group allocation.
Multivariate analysis will be used to control for potential confounding factors to examine the effects of treatment strategies, and will use linear regression for continuous outcome variables and test for mediator effects using the approach outlined by MacKinnon and Luecken (2011) [56
], logistic regression for binary outcomes (dental decay prevalence, obesity prevalence) and Poisson regression to estimate rates of disease occurrence. Site-specific measures of caries incidence within an individual will use GEE analysis to control for correlation within an individual. Effect size will be estimated through calculation of incidence density ratios for dental decay and prevalence ratios for obesity. Effect measures will also be presented through calculation of number needed to treat and its associated 95% confidence interval. The statistically significant level will be at α