A cluster randomized controlled trial with two-arms were implemented to assess the efficacy of behavior change in decisions on antibiotic use in relation to childhood infection. For confidentiality, computer generated numbers were used as codes to recruit kindergartens and parents of kindergarteners aged under 7, as clusters, to the intervention and control groups. The subjects were selected according to the inclusion criterion that they were parents of kindergarteners aged below 7. Kindergartens were selected from the same region to ensure that the demographics of the two groups were comparable. Randomization was implemented by putting the sealed codes for selection inside a paper bag. The sample was then drawn randomly by an individual who was not associated with the study to prevent selection and confounding biases.
The parents in the intervention group were asked to join an education program on antibiotic usage and a peer support group, whereas those in the control group were only given information leaflets on antibiotics from the Centre for Health Protection.
The intervention program consisted of two weekly sessions and each session lasted for 90 minutes. The participating kindergartens sent training reminders to the parents one week before each session. Interventions addressing antibiotic use were administrated in two formats: a functional session (Week 1: the basic knowledge on viral and bacterial infections) and an interactive session (Week 2: case studies on consulting behavior, management planning and experience sharing). The parents in the intervention group, together with a pharmacist and an infection control nurse, would join a Facebook Page of Antibiotic Use at the beginning of the program. This page would allow these participants to build a social network to send and receive instant online advice amongst the team members on antibiotic issues. On the online platform, they could interact with one another, thus strengthening parent-to-parent support.
To facilitate the assessment of the participating parents, survey packets containing the questionnaires and reminders to parents were sent to both the intervention and control groups from the participating kindergartens for use before the 2-week period. One participating parent of each child was asked to complete one set of questionnaires before and after the program. The participants were instructed to fill out the questionnaires only once irrespective of the number of children attending the same kindergarten. The questionnaires took 20 minutes to complete.
Approval from the Research Ethics Committee was sought before recruitment and written consents were obtained from the parents concerned. The parents were informed that withdrawal at any time would not result in any negative consequences. All data were protected with passwords. Only the researcher and her team had access to the datasets to prevent any leakage of sensitive information.
To measure peer support, the peer support outcome protocol adapted from the outcome evaluation indicators10 was used. Specific outcomes that are available in the protocol include: demographics, service use, program satisfaction, participation in a discussion group. Parental knowledge, attitude, and their social network were measured before and after the program using the Parental Perception on Antibiotics (PAPA) scale and the GSE scale to assess differences, if any, between and within the intervention and control groups.
The PAPA scale was administered to pre-test and post-test the participants’ capability in comprehending and acting on antibiotics-related information (functional: the basic skills in understanding antimicrobial drug knowledge), Cronbach’s alpha =0.78. The PAPA scale consists of 4 sections: 1) Children’s health, 2) Antibiotics and health information, 3) Experience with antibiotics and health professionals, and 4) Personal attitudes and beliefs about antibiotics.
The PAPA scale has 32 items measuring the factors influencing the overuse of antibiotics in children, especially those with upper respiratory tract infections. Parents were asked to rate on a 5-point Likert scale ranging from strongly disagree to strongly agree or from never to always on child health-related history, including the number of cold episodes and antibiotics (courses) used for the youngest child during the previous year (ranging from never to more than 6 times a year), whether any of the children in the family has ever had a serious infectious disease or a chronic disease, and items relating to factors influencing the parental use of antibiotics including knowledge and beliefs, behaviors, adherence, information seeking, awareness about antibiotics resistance, and their perception about doctors’ prescribing behavior.5 The parents in both groups were required to complete the self-reporting Generalized Self-Efficacy scale (GSE) before and after the program. GSE is a 10-item scale with a score ranging from 1 to 4 for each question. Higher scores indicate stronger parental belief in self-efficacy. The Cronbach’s alpha coefficient for the entire scale was 0.80 and the test-retest reliability coefficient was 0.69.11
Descriptive statistics was used to group the numerical and categorical data of the study. Chi-square tests were undertaken to ascertain differences in baseline characteristics across groups of categorical data. Independent t tests were run to investigate differences in outcomes from parents’ health behavior at baseline and follow-up. Mann Whitney U test was conducted to compare the difference of attitude of antibiotic use between the intervention and control groups. Correlation analysis was performed to analyze the relationship between social support and parents’ personal attitude towards antibiotic use. For the estimation of effects, 95% confidence intervals were provided. The statistical significance for all tests was set at p<0.05.