Complex Intervention to Promote Human Papillomavirus (HPV) Vaccine Uptake in School Settings: A Cluster-Randomized Trial

"Changes to school-based vaccination requires consideration of the organisational model and multiple levels of influence."

In Australia, adolescents are primarily vaccinated against the human papillomavirus (HPV) on designated "vaccination days" on school grounds after obtaining parental or guardian consent. Logistical and organisational factors, including communication with parents and students, can influence the successful implementation of school-based vaccination. To investigate the impact of addressing these issues collectively, this group of researchers undertook a community-based cluster-randomised controlled trial, known as HPV.edu, in 40 Australian schools.

The study was undertaken in 40 high schools (clusters in the cluster RCT) in Western Australia and South Australia between 2013 and 2015 with with 6,967 adolescents aged 12-13 years. Nineteen control schools conducted the HPV vaccination programme as per their usual practice, following standard procedures described in state guidelines. In contrast, the 21 intervention schools received:

1. Adolescent in-class education and vaccination-day guidelines that were designed to improve vaccination literacy, psychosocial outcomes, and the vaccination experience (e.g., instructions about optimal vaccination clinic room set up to minimise student anxiety and maximise privacy);
2. A decisional support tool booklet that was designed to promote shared parent-adolescent decision-making; and
3. A logistical component that included, for example, consent form return strategies (e.g., daily reminders to students, email/letters to parents, phone calls, and school newsletter notices) and instructions for conducting in-school catch-up of missed doses.

The main outcome was school vaccine uptake. Secondary outcomes included consent forms returned and mean time to vaccinate 50 students.

There was no difference between intervention and control (3-dose mean 75.7% and 78.9%, respectively). Following adjustment for baseline covariates, absolute differences in coverage in favour of the intervention group were: dose 1, 0.8% (95% confidence interval (CI), -1.4,3.0); dose 2, 0.2% (95% CI, -2.7, 3.1); dose 3, 0.5% (95% CI, -2.6, 3.7). The percentage of returned consent forms in intervention schools (91.4%) was higher than in control schools (difference: 6%, 95% CI, 1.4, 10.7). There was a shorter mean time to vaccinate 50 students at dose 3. The difference for dose 3 was 110 minutes (95% CI, 42, 177); for dose 2, 90 minutes (95% CI, -15, 196); and dose 1, 28 minutes (95% CI, -71, 127).

Logs revealed the inconsistent implementation of logistical strategies. Inadequate resourcing for logistical strategies and advisory board reluctance toward strategies with potential financial implications impacted the implementation of logistical components. Another possible explanation for the lack of effect is a high baseline first-dose mean uptake. The 15% of adolescents who did not receive the first dose in this study may require an individually tailored approach.

Further details about the intervention were revealed in interviews with 11 immunisation nurses conducted across jurisdictions:

• Immunisation nurses reported various barriers to vaccination regarding consent forms, such as incorrect/incomplete consent forms - with common consent form errors including parents not marking correct boxes for all vaccines or having a person without legal guardianship of the adolescent complete the form. Facilitators included: a 2-week time-limited turnaround for consent form return; a good relationship with the school coordinator; "conscientious" school personnel; phone calls to parents; and immunisation guidelines outlining roles and procedures.
• Immunisation nurses reported verious following barriers to vaccine catch-up, such as adolescent vaccine refusal despite parental consent. Among the facilitators of vaccine catch-up: Processes were optimised where schools partnered with immunisation nurses to deliver the programme.

Although the complex intervention had no impact on HPV vaccine uptake, the researchers suggest that the adolescent component improved vaccination literacy, which in turn improved self-efficacy and reduced fear and anxiety. These changes may have impacted student wellbeing, behaviour, and compliance on vaccination day. Also, implementation of the logistical study guideline strategies in intervention schools, combined with the adolescent intervention, was the likely reason for the reduced time to vaccinate adolescents. The guidelines were intended to provide an environment whereby student anxiety was minimised, and vaccination processes were streamlined.

In conclusion, the researchers point to several issues impacting consent form returns that may undermine the success of school-based HPV vaccination programmes like this one. Clearly defined guidance on the steps necessary to overcome consent-related barriers could assist. For example: consent forms could be delivered electronically to parents in addition to hard copy; consent form design could be inclusive of people with low literacy and non-English speaking background; resourcing for obtaining verbal consent from parents could be allocated before vaccination day; and the school-based programme could consider verbal consent for young people after competency assessment.