PROMOtion of VAccination in Canada (PromovaC Study)

OBJECTIVES

Main objective: To evaluate the impact of the motivational interviewing (MI) intervention on parents' intention to vaccinate their infant and their vaccine hesitancy (VH) score in a Pan-Canadian context. In order to explore differences in the intervention's impact according to cultural and organizational factors, all analysis will also be conducted at the level of each center.

Specific objectives:

1. To assess parents' intention to vaccinate their infant before the intervention and immediately after the intervention, globally and at each center level.
2. To assess parents' VH scores before the intervention, immediately after the intervention and when their infant is 3 months of age, globally and at each center level.
3. To assess parents' knowledge, attitudes and beliefs (KAB) about vaccination before the intervention, immediately after the intervention and when their infant is 3 months of age:

• Descriptive analysis according to a composite model inspired from the Health Belief Model (HBM), the Theory of Planned Behavior, and the VH measures (based on a validated questionnaire);
• Research of the determinants of vaccination intention.

1. To compare vaccine coverage (VC) at 3 months of age of infants whose parents has received (intervention group) or not (control group) the intervention in the maternity ward during the recruitment period.

STUDY DESIGN

This study is a multicentre randomized controlled trial (RCT) with single unit of randomization that will be conducted in maternity wards of 5 provinces: Quebec (QC), British Columbia (BC), Ontario (ON), and Nova Scotia (NS)/ Prince Edward Island (PEI). In each center, parents will be randomized to receive the MI intervention (intervention group) or to receive an information flyer on childhood vaccines (control group). The flyer that will be given to mothers of the control group will be the one that is usually given to parents at the maternity ward or at the first home visit after birth during the first week. Thus, flyers will be different according to the participating regions in this study. Randomization will be stratified by center and will be made in blocks of 4, 6 and 8 participants.

STUDY POPULATION

Study participants

The study population will include 1,600 mothers of infants born in the participating maternity wards during the recruitment period.

Study centers

The study will be conducted in 6 maternity wards located in 4 different centers:

1. Centre hospitalier universitaire de Sherbrooke (CHUS), Sherbrooke, QC;
2. B.C. Women's Hospital \& Health Centre, Vancouver, BC and Surrey Memorial Hospital, Surrey, BC;
3. North York General Hospital, Toronto, ON;
4. IWK Health Centre, Halifax, NS and Queen Elizabeth Hospital, Charlottetown, PE.

The choice of these 6 maternity wards is the best compromise between the study feasibility and the diversity of the Canadian population. More than 2,500 deliveries per year occur in each maternity ward (except for Queen Elizabeth Hospital with approximately 1200 annual births), which ensures the feasibility of the study.

CONDUCT OF STUDY

Enrollment

Each subject must participate in the informed consent process and sign and date the informed consent form before any procedures specified in this protocol are performed. Mothers attending the participant maternity wards will be entirely free to participate in the study. Benefits of the study will be the chance for the parents to discuss childhood immunizations in a non-confronting and non-judgemental climate and the improvement of their knowledge. Participation in the study will not involve risk, besides the risk of breach of confidentiality for which precautions will be taken.

Regulatory considerations

The study will be registered with ClinicalTrials.gov.

Recruitment

Mothers will be approached to participate during the postpartum stay at the maternity ward. The MI intervention will be delivered by a research assistant (RA) (or a research nurse) that will have first received a standardized training session on the content and techniques of MI. Since the postpartum stay may last less than 24 hours in some provinces, RAs will recruit mothers in a chronological way according to the birth time of the mother to ensure to meet the highest number of mothers as possible.

The recruitment will be conducted during 4 to 6 months. In the Quebec RCT, 25 participants were recruited in average each week, so it is feasible to recruit 400 participants within 4 to 6 months.

Study plan

After the mothers agree to participate, they will have to complete an initial questionnaire (Q1) to assess their intention to vaccinate, their VH scores and their KAB about vaccination, as well as their socio-demographics characteristics. RAs will collect this questionnaire and proceed to the randomization of the participating mother to receive the MI intervention or to receive a flyer on childhood immunizations. Depending on the randomization result, RAs will perform the MI intervention with the parents or only give the information flyer. RAs will then distribute a similar questionnaire (Q2) to parents of both groups to assess their intention to vaccinate, their VH scores and their KAB about vaccination, as well as their satisfaction of the intervention received (MI or flyer). Before leaving the maternity ward, parents will have to return their completed questionnaire to the RA or in a designated place. When the child is 3 months of age, an online survey (Q3) will be sent to participants by email to assess their VH score and their KAB as well as the child's vaccine status.

Standardization of the intervention

To ensure that the content of the information sessions will be standardized and similar for each center or RA, a reference document on the primary vaccination (vaccines administered before the age of 1 year) will be made according to the provincial recommendations for each center and will be used for the training of RAs. This training will be conducted over 3 days including supervision via audio recordings (2 days + 1 day). First, RAs will participate in a 2-day centralized training seminar in Montreal on the MI techniques given by a certified trainer before the beginning of the study. Subsequently, RAs will practice their MI skills during a 4 to 6-week trial period in their respective maternity ward. During this time, RAs will send an online audio recording of their MI intervention with a mother in the maternity ward (a separate informed consent form will be signed by these mothers to authorize the record and the sending of the MI intervention). The audio recordings will be securely transferred to the certified MI trainer through a web-based application named REDCap (Research Electronic Data Capture). Based on this recording, the certified MI trainer will assess their MI skills and offer them support and feedback to improve their practice. After 6 weeks of trial, the study coordinator and the certified MI trainer will visit each center to give individual feedback and coaching to each RA. They will observe and validate the MI intervention performed by the RAs and identify possible barriers to the implementation of the study according to each maternity ward. This training framework is consistent with the literature since MI training indeed requires training in 2 steps with feedback training and coaching in real situations after the initial training. Data collected during the 6-week trial period will be discarded since the RAs will be practising conducting the study. Data will only be used to monitor good clinical practices (GCPs) in each center during the audit by the study coordinator. After this training, RAs will officially start the recruitment in the maternity wards. The MI trainer will be available to answer questions and offer support to the RAs regarding their MI techniques throughout the study.

Confidentiality protection

Source documents containing names, addresses and other contact information of participants will be stored at each local site in individual folders, identified by the assigned participant identification number (PID) number. These documents are stored in locked filing cabinets in a secure research facility, with controlled access, limited to authorized staff and investigators. No personal identifiers will appear in any summary or report of the study data and will not be sent to the coordinating center.

STATISTICAL CONSIDERATIONS

Sample size

In order to identify a statistically significant amelioration of 8% in parents' intention to vaccinate, and taking account of a basic intention to vaccinate of 78%, a risk of alpha error of 0.05 and a power of 80%, a total of 400 parents by center will be included. This sample size will also allow detect a minimum of 5% reduction in VH scores in each center. The total sample of 1,600 families will allow detecting a significant amelioration of 6% in infants' VC.

Data analysis

A descriptive analysis of responses from Q1, Q2 and Q3 will be conducted. Q1 and Q2 responses as well as Q1 and Q3 responses (KAB, vaccination intention and VH scores) will be compared for mothers belonging to the intervention group in order to assess intervention's effectiveness, globally and for each participating center. Q1, Q2 and Q3 responses (KAB, vaccination intention and VH scores) will be also compared between intervention and control groups in order to demonstrate the MI intervention's effectiveness compared to the information flyer. Parents' intention to vaccinate their infant measured before the intervention will be dichotomized as "certainly" vs. "probably"; "probably not" and "certainly not" and compared to all covariates measured before the intervention. Multivariate logistic regression analysis will be conducted to identify independent determinants influencing vaccination intention. Regarding VC, it will be defined as the proportion of children having a complete vaccine status on the total of participating children. A child is considered having a complete vaccine status if he has received all vaccines and antigens recommended by his respective provincial immunization schedule at 3 months of age.

Q1 and Q2 responses (KAB, vaccination intention and VH scores) will be first compared using McNemar's test and Wilcoxon's signed rank test for paired data for mothers belonging to the intervention group. Thereafter, Q1, Q2 and Q3 responses (KAB, vaccination intention and VH scores) will be compared between intervention and control groups using the Chi-square test or the Fisher's exact test as appropriate while continuous variables will be compared using the Student's t-test or the Mann-Whitney U test as appropriate. Moreover, covariates will be compared with vaccination intention using the appropriate tests mentioned. In multivariate analysis, collinearity of the variables included in the model will be checked and the model adequacy will be assessed by the Hosmer-Lemeshow test and the deviance test. Finally, VC will be analyzed globally, comparing proportion of properly vaccinated children between intervention and control groups using the Chi-square test. The SAS software (version 9.4, SAS Institute, Inc., Cary, NC) will be used for all statistical analysis.