Nudge and Motivational Interviewing Interventions to Improve Influenza Vaccine Uptake Among Healthcare Workers in China

A two-phase randomized controlled trial will be conducted to evaluate the effectiveness of multi-component interventions to improve vaccination uptake, targeting healthcare workers in China.

Sample size Phase 1: Cluster-Randomized Controlled Trial The sample size is determined to identify a significant difference in the primary outcome, defined as the influenza vaccination uptake among healthcare workers in community hospital, comparing the intervention and control groups. Historical surveillance data from the three participating cities indicate a baseline vaccination rate of approximately 4% without targeted interventions. The proposed intervention is expected to increase this rate to 10%. A two-sided significance level of α = 0.05 is set, adjusted to α = 0.0167 via Bonferroni correction to account for three planned pairwise comparisons. Statistical power is set at 0.80, with an assumed intraclass correlation coefficient (ICC) of 0.05 and an average cluster size of 50 participants per cluster. Based on these parameters, 23 clusters per study arm are required, resulting in a total of 69 clusters across the three arms. To account for potential attrition, the sample size will be increased to 27 clusters per arm, yielding a total of 81 clusters and an estimated 4,050 participants. All eligible healthcare workers in the participating community hospitals will be included in the study.

Phase 2: Individual Randomized Controlled Trial In phase 2, the target population comprises healthcare workers who remain unvaccinated after Phase 1, from the 54 community hospitals assigned to the intervention arms. Without further intervention, the vaccination uptake rate in this group is estimated to be approximately 3%. The motivational interviewing intervention is expected to increase this rate to 13%. Assuming a two-sided significance level of α = 0.05 and statistical power of 0.80, sample size estimation indicates a minimum of 112 participants per group is required. To ensure feasibility and maintain study power, the motivational interviewing group will aim to enroll 150 participants. Fifteen community hospitals will be randomly selected from the two intervention arms of Phase 1. Control groups are established: 1) healthcare workers from 15 community hospitals selected for the motivational interviewing intervention, excluding approximately 10 individuals per community hospital who receive the intervention, resulting in a control group of approximately 525 healthcare workers; and 2) healthcare workers from the remaining 39 community hospitals not participating in the motivational interviewing intervention, comprising approximately 1,755 individuals. Therefore, the control group in phase 2 comprises approximately 2,280 participants.

In phase 1, a stratified block randomization method will be employed to ensure balanced allocation across study groups. Stratification factors include historical influenza vaccination rates among healthcare workers and the size of the healthcare staff at each community hospital. Community hospitals with comparable levels of these factors are grouped into blocks within each city and then randomly assigned within each block to one of three groups: the standardized nudge group, personalized nudge group, or a control group.

In phase 2, a two-stage random sampling procedure will be used. Specifically, 15 community hospitals will be randomly selected from the 54 hospitals assigned to the intervention groups in the phase 1 (five hospitals per province). Within each selected hospital, 10 healthcare workers who remained unvaccinated after the first-phase intervention will be randomly sampled and enrolled in the individual-level intervention study.

Qualitative interviews and a baseline survey will be conducted before the interventions. Follow-up questionnaires will be completed two weeks after the end of each intervention phase. The primary outcome is the receipt of influenza vaccination by healthcare workers, which will be assessed two months after each intervention phase using official vaccination records.

The intention-to-treat (ITT) analysis will be used to evaluate the effectiveness of the interventions. The primary outcome is influenza vaccine uptake among healthcare workers post-intervention. Vaccine uptake will be compared between intervention and control groups, and across intervention arms, using chi-square tests. To account for clustering and adjust for potential confounders, generalized estimating equation (GEE) models will be employed to provide robust effect estimates. Timing of vaccination will be evaluated using Kaplan-Meier survival curves and log-rank tests, with Cox proportional hazards models to estimate intervention effects on time-to-vaccination. Subgroup analyses will assess heterogeneity of intervention effects across populations stratified by baseline characteristics.

Secondary outcomes will be evaluated at baseline and post-intervention. Between-group comparisons, including intervention versus control and across intervention arms, will use t-tests for continuous variables and chi-square tests for categorical variables. GEE models with repeated measures will assess intervention effects on secondary outcomes, adjusting for relevant covariates. Subgroup analyses will examine differential effects across subpopulations defined by baseline factors.