Design, setting and sampling
We conducted a cross-sectional survey of patients in General Practices in the London Borough of Brent, the most ethnically diverse and one of the most deprived boroughs in the United Kingdom [18]. The sample frame was all General Practices that had a television in the waiting area and showed the Life Channel. The Life Channel consists of six health promotion television advertisements (dental health, smoking cessation, chlamydia, contraception, HIV/hepatitis and influenza), a full cycle of which is shown over a ten minute period interspersed with commercial advertising. The dental health advert, which lasted approximately 80 s and was targeted to adults, stated that everyone should visit a dentist at least every 2 years, and gave details of how to contact both NHS and emergency dentists. Of all 68 General Practices in Brent, 49 (75.1 %) showed the Life Channel. We contacted these 49 General Practices, inviting them to participate in an anonymous, interviewer-administered survey. All General Practices agreed to participate.
The study took place between February and March 2013 (five weeks). Eighteen trained facilitators visited each General Practice at least once, the time of which occurred according to facilitator suitability and not at random. After taking informed consent they sampled all patients in the waiting room; for newly-entered patients they waited at least five minutes before approach, which enabled some viewing of the Life Channel. No staff were sampled and due to the exploratory nature of this study no sample size was calculated (although we conducted a post-hoc power analysis). Patients were able to complete either a paper copy or digital copy of the survey (the latter via a Smartphone). Each survey took approximately five minutes to complete.
Questionnaire and measures
The survey was a self-developed, pre-piloted but non-validated tool which was originally piloted in Brent to 754 patients in 16 General Practices in 2010. The low completion rate then led to a reduction in the number of questions and a refinement in question wording. The current survey used in this study contained 28 items which covered two main categories. Sixteen questions asked about the Life Channel, including its noticeability, acceptability, and whether it had changed knowledge and intentions about access dental and smoking cessation services. It did not ask any questions about change in knowledge and intentions about the other four health topics seen on the Life Channel (chlamydia, contraception, HIV/hepatitis, and influenza). Twelve questions gathered socio-demographic and health data. The full survey is available upon request.
This study had three main binary outcome measures. The first outcome measure was those who selected “I know how to find an NHS dentist in Brent because of seeing the Life Channel today” after being asked “Which one of the following statements about finding an NHS dentist is most accurate for you?” (other options: “I don’t know how to find an NHS dentist in Brent”, “I know how to find an NHS dentist in Brent because of something else”, and “None of the above”). The second outcome measure was those who selected “I know how to contact a dentist in an emergency from seeing the Life Channel today” after being asked “Which one of the following statements about finding an emergency is most accurate for you?” (other options: “I don’t know how to contact a dentist in an emergency”, “I know how to contact a dentist in an emergency from another source”, and “None of the above”). The final outcome measure was those who selected “I’m planning to contact a dentist because of seeing the Life Channel today” after being asked “Which one of the following statements is most accurate for you?” (other options: “I’m not planning to see a dentist at the moment”, “I’m planning to contact a dentist for another reason”, and “None of the above”). According to the Theory of Planned Behaviour, intention to commit to a health behaviour is regarded as one of the most powerful indicators of behaviour change [19].
Predictor variables included age (29 or under/30 to 39/40 to 49/50 to 59/60 or over), gender, ethnicity (White/South Asian/Black/Mixed or other), education (primary/secondary/college/university/other), the presence of a chronic disease (ascertained by the question “Are your day-to-day activities limited because of a health problem or disability which has lasted, or is expected to last, at least 12 months (including problems related to old age)?), smoking status (non-smoker/ex-smoker/smoker) and whether a dentist had been visited in the last 2 years (yes/no).
Statistical analysis
Observations with missing data were not removed prior to analysis. The denominator for the first two outcome measures excluded those who knew how to contact an NHS/emergency dentist from another source as this study was only interested in evaluating the service for those who did not already know how to contact an NHS or emergency dentist. Similarly, the denominator for the third outcome measure excluded those who were already intending to contact a dentist for another reason.
We conducted a simple bivariate analysis by cross-tabulating outcome measures by predictor variables. We conducted three forced multi-level logistic regression models to examine which predictors were independently associated with our three outcome measures. These were adjusted for all covariates in addition to the waiting times of each patient, and whether patients could hear the Life Channel. A two-level random-intercept model was used to account for clustering effects around General Practices (Stata command: xtmelogit). We reported adjusted odds ratios (AORs) with a 95 % confidence interval (CI) and took an alpha probability value of 0.05 as statistically significant. We measured the observed power of each model in a post-hoc analysis. Finally, we ran descriptive statistics on other evaluative aspects of the Life Channel (e.g., noticeability, acceptability). All statistical analyses were performed using Stata 12 (StataCorp).