Background By 2022, COVID-19 vaccine uptake had plateaued. We tested whether unvaccinated individuals in 2022 were wilfully avoiding vaccine information, or if
Overall, reviewers expressed concern about the limitations of the trial and whether the stated conclusions followed from the results and study design.
RR:C19 Evidence Scale rating by reviewer:
Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.
This is an important study and authors are to be commended on generating evidence in understanding vaccine hesitancy. Authors make valid points that few studies provide evidence on how to specifically convince the most vaccine resistant to reconsider their position and that fewer intervention studies have been tested for addressing this, and that are scalable. Having said this, it is somewhat surprising with this much effort, that the group of authors did not have vaccine specific communication theory guide their efforts. One example might be the 3/5 now 7C model of vaccine readiness (Cornelia Betsch et al) with its original model (3C model, 2012 by WHO SAGE group) acknowledging confidence, complacency, and constraints to explain vaccine hesitancy across vaccines and countries.
As a scholar who designs vaccine promotion video interventions, I am very interested and invested in this type of research. There were not too strong of difference between the treatments, but they appeared to each show an improvement over control video on all 4 outcomes (intent, efficacy, side effect concern and trust) with some exceptions.
Clarification for Table 2 would be helpful on whether the adjusted p-values (adjusted for multiple hypothesis testing with Romano & Wolf 2016 test) were significant, which was was not clear.
The operationalization of information seeking was creative - comparing videos optional (V1) with videos required (V2). There is some literature on if you provide users "choice" or "options" when they navigate a health kiosk, dashboard, or choice of videos that their evaluation increases (in a positive manner). This stems from user interface literature I believe. The number of videos a user is willing to voluntary watch in a competitive message environment is a big question I have encountered in my research now in the last 10 years. It's a challenge compressing and distilling the amount of health/vaccine information into short videos.
The authors creatively operationalized and measured info seeking with the number of additional videos watched and time watching the videos. However, for Figure 2 (Panel B), the crux of the authors’ claims seems to rest on this result? That slightly larger % of subjects from V1 watched 3 videos - was this even significant? The evidence of Figure 2, Panel C is clear but not clear for Panel B. Please elaborate in the explanation for Table 3. Also, the authors should elaborate in the manuscript (not supplement) on the differential attrition and its implications for interpretation.
This study does have considerable limitations some of which are mentioned in limitations section like not collecting vaccine uptake. The majority white sample (73%) should be mentioned as a limitation. It's not clear to me that the "unvaccinated" which represent a heterogeneous group and continuum only reflect the very resistant as authors made claims early in the paper. The magnitude of findings/effects were modest at best.
The authors should correct the typo on page 17 (discussion) where authors write "a possible reason for this discrepancy is that subject in "this" study were encouraged to watch an "important video" with details while our study...." They both refer to "our study" but I think the authors meant for the first part to refer to literature study?
Secondly, in the discussion section - the implications seem to be quite a leap (and without vaccine communication theory) from the results that providing enough detail about "mRNA technology" (vaccine confidence) will suffice when we have learned from vaccine com theory (like 7C model) that likely a combination of low confidence coupled with high complacency and perhaps low access or other reasons may be at play. Multiple aspects may need to be addressed in messaging. The authors and literature/discussion also ignore the literature that contextualizing vaccine messaging in narratives has proven to result in increased message acceptance over informational vaccine messaging.
It would be helpful in discussion section to understand how "heterogeneous" the "unvaccinated" group is in your study. I would possibly remove from intro "interventions focusing on the most vaccine hesitant" since the "unvaccinated" group in this study is well unvaccinated but there is no characterization that this group represents "the most vaccine hesitant". for added value, would also remove 'how to convince the most vaccine resistant" since this study generates modest evidence that unvaccinated are still willing to listen to vaccine information. But we don't know whether these are "the most vaccine resistant".