Short Message Service Reminder Nudge for Parents and Influenza Vaccination Uptake in Children and Adolescents With Special Risk Medical Conditions

This randomized clinical trial investigates the effectiveness of a parent short message service (SMS) reminder nudge to improve influenza vaccination uptake in children and adolescents with special risk medical conditions.

Any deviations from the planned analyses in this SAP will be documented with reasons in a post-analysis version of the SAP. Any post-hoc analyses not described in this SAP will also be clearly identified.

Background
Children and adolescents with special risk medical conditions (SRMC) are at significantly higher risk of influenza-associated hospitalisation and death. Immunisation is the most effective strategy available to prevent influenza and its complications. Despite a comprehensive National Immunisation Program, the uptake of influenza vaccination in children and adolescents with SRMC is inadequate.

Objectives
To determine if text message reminders sent to parents of children/adolescents with SRMC increases uptake of influenza immunisation.

Study design
This is a two-arm parallel group randomised controlled trial. The primary outcome is the proportion of children/adolescents who have received at least one dose of flu vaccination between randomisation and the end of the study period (30 September 2021).

Randomisation
Parents of children/adolescents will be randomised to the standard care or intervention arm in a 1:1 ratio. The randomisation schedule will be prepared by an independent statistician and use randomly permuted blocks, stratified by age group (<5, 5-14, >14 -<18 years).

Sample size
A sample size of 540 parents of children/adolescents with SRMC (270 per group) is required to have 80% power to detect a 30% relative increase in the percentage of children vaccinated from 40% in the standard care arm to 52% in the intervention arm (two-tailed alpha = 0.05).
Previous studies have shown a 30% to 70% relative increase following other immunisation interventions. A 30% relative increase in the percentage vaccinated would be considered clinically meaningful.

Framework
All comparisons will be undertaken assuming a standard superiority hypothesis testing framework.

Statistical interim analyses and stopping guidance
No interim analyses are planned.

Timing of final analysis and unblinding
The database will be locked for analysis once data collection and cleaning are complete and the SAP is approved. Following database lock, unblinded treatment groups will be made available to the trial statistician and analysis of the listed outcomes performed.

Confidence intervals and p values
For each outcome variable, a 95% confidence interval will be reported to express uncertainty about the estimated treatment effect. The statistical significance of the estimated treatment effect will be assessed at the 0.05 level using a two-sided comparative test. In summarising the effects of the intervention, multiple hypothesis tests will be performed due to multiple outcomes. No multiplicity adjustment will be made for the number of secondary analyses, as these are of less importance than the analysis of the primary outcome. In the absence of a formal procedure for controlling the type-I error rate, less emphasis will be placed on the results of secondary analyses.

Protocol deviations
Frequencies and percentages will be presented by randomised group for the following protocol deviations: ineligible parent/child pairs randomised; randomised in the wrong stratum; randomised to the wrong group; withdrawal from study; loss to follow-up.

Analysis population
For all outcomes, the planned analyses will be performed using an intention to treat (ITT) approach. The ITT analysis population will include all parent/child pairs, analysed as randomised, except where their exclusion criteria preclude the child receiving the influenza vaccine during follow-up. Randomised children who have a medical condition contraindicated with the flu vaccine, or have already received a flu vaccine in 2021, will be excluded from the ITT analysis. Such exclusions are not expected to bias treatment group comparisons since ineligible participants will be identified consistently and objectively using information available for all participants (e.g., previous receipt of the 2021 influenza vaccine) without knowledge of the assigned treatment group.
Descriptive statistics will also be reported using the ITT analysis population. For the primary outcome only, a sensitivity analysis will be performed in only eligible patients.

Screening data, eligibility, recruitment, and withdrawal/follow-up
Information will be presented by treatment group (where appropriate) on the number of parent/child pairs: • Screened • Eligible to participate • Ineligible to participate (by reason) • Randomised

• Withdrawal
• With primary outcome data available

Baseline characteristics
A descriptive comparison of the randomised groups will be conducted on the baseline characteristics in Table 1. Frequencies and percentages will be reported for each variable. The clinical importance of any observed imbalances will be noted.
Residential location and Socio-Economic Indexes for Areas (SEIFA) will be obtained by linking postcode to ABS postal areas from the following datasets:

Overall analysis approach
The intent is to compare the proportion of participants who received a flu vaccine between randomisation and 30 September 2021 between the standard and intervention groups. The proportion will be compared between groups using logistic regression, with the effect of treatment described as an odds ratio with 95% confidence interval. Logistic regression will also be used for the secondary outcome of flu vaccination received in the optimal period (1 April 2021 -30 June 2021). This secondary analysis will be restricted to parent/child pairs randomised on or before 15 June 2021, to allow at least two weeks to access a flu vaccination following randomisation/intervention.
Secondary analysis comparing the time between randomisation and vaccination between standard and intervention groups will be performed using a Cox proportional hazards model, as well as compared descriptively with a plot of Kaplan-Meier survivorship estimates. The proportional hazards assumption will be checked by fitting a smoothing spline to the Schoenfeld residuals from the Cox model for each predictor and testing for a significant (p<0.05) relationship between residuals and failure time.

Covariate adjustment
Analyses will adjust for child age (<5, 5-14, >14 -<18) since this variable was used to stratify the randomisation. The adjustment variable will be treated as a categorical fixed effect. Where any participants have been randomised in the wrong age stratum, their true age category will be adjusted for, rather than the randomisation age category.

Subgroup analyses
Pre-specified subgroup analyses will be performed for the primary outcome to assess heterogeneity of the intervention in relation to: 1) Age group 2) Residential location (metropolitan or regional) 3) Paediatric subspecialty (diabetes/endocrine, respiratory medicine/cardiology, neurology, rheumatology clinic/gastroenterology) As the study is not powered for subgroup analysis, this analysis is considered exploratory.
Subgroup analysis will be performed by fitting a model that includes an interaction term between the treatment group and each variable. The models for residential location and paediatric subspecialty will also be adjusted for the stratification variable, age group. The estimates of treatment effect in specified subgroups and 95% confidence intervals from each model will be presented, as well as an overall test for interaction. If there is no evidence of interaction (p>0.05), any differences between subgroups should be regarded as due to chance.

Methods for addressing outlying values
Outliers will be queried during data collection and the statistical analysis. Unless confirmed as a data entry error, outliers will not be excluded from any analyses.

Methods for handling missing data
Missing data will be summarised descriptively by treatment group for all baseline characteristics, outcome variables and covariates for adjustment. Analyses will incorporate all available outcome data under the assumption that data are missing at random conditional on randomised group and covariates for adjustment.

Harms
No statistical analysis is planned to assess harms, as Flutext-4U is a behavioural intervention and the risk of the intervention is comparable to standard care.

Exploratory analyses
Results from parental acceptability survey will be summarised descriptively, with frequencies and percentages reported.

Analysis Software
All analyses will be performed using R version 4.1.1 (R Core Team, 2021) or later.