Reply to “Caution in underrepresentation of older adults in clinical trials on COVID-19 vaccines”

Abstract

We would like to thank Chen et al. for their comments on our recently published paper in Ageing Research Reviews (Veronese et al., 2021). In this reply we would like to answer to the points raised in their letter. Regarding the first question, we decided to run our search only in PubMed/Medline and ClinicalTrials.gov since the phase II-III randomized controlled trials of SARS-CoV 2 vaccines, considering the urgency for interventions to ameliorate the pandemic situation, received a great deal of interest. Therefore, they were usually published in high impact factor journals, reported in PubMed, and their protocols were usually reported in clinicaltrials.gov. However, following the suggestion of Chen et al., we did run a similar search in Web of Science (n = 150 articles) and Embase (n = 121 articles), without finding any possible eligible paper. We also searched, as suggested, the WHO ICTRP registry, which includes also the Chinese clinical trial registry, finding 36 ongoing trials, with two eligible out of them (https://trialsearch.who.int/Trial2.aspx? TrialID=ISRCTN79815558; https://trialsearch.who.int/Trial2.aspx? TrialID=RPCE0354). However, since both trials are still undergoing, it is not possible to know which percentage of older people will be eventually enrolled, therefore there is no need to modify what was already reported in our systematic review. With regard to the concern related to the language limitation, it should be pointed out that, among the completed trials, four out of ten were performed in China, but still they were published in an English language journal (Xia et al., 2020, 2021; Zhang et al., 2021; Zhu et al., 2020). Furthermore, we added the keywords suggested in the second point of the comments to our article: in this case, limiting the search to the 01st May 2021 as done in our systematic review, we found 547 possible eligible papers, but again, no new work was eligible for our analyses. Even if we recognize that adding these specific keywords might increase sensitivity of the search, they are probably of poor specificity in finding out papers that did not use randomized clinical trial as filter, particularly considering that practically all the papers reporting phase II-III trials on vaccination were published in widely known scientific journals. We also appreciate the comments regarding risk of bias and assessment of the quality of the study, which are a mandatory step in a systematic review. However, we would like to point out that the aim of our systematic review was not to investigate the efficacy/safety of COVID-19 vaccination in older people, but to highlight that older people were usually excluded from these studies. Therefore, this evaluation was not necessary, as we did not want to evaluate the quality of the results provided by the studies, but we aimed only at verifying the inclusion of older people. Finally, we are also grateful for the comments regarding the registration of the protocol and the statistical analysis. Regarding the first point, we agree that pre-publishing the protocol is an essential step in publishing a paper, but at the same time we also believe that the protocol structure is so simple in this case (we practically only extracted data regarding age range, asking to the corresponding authors further information when needed) that pre-publishing of the protocol would have not changed anything in the final results. However, we enclose to this letter as supplementary material the protocol. Moreover, the statistical analyses were so elementary that creating a specific paragraph would not represent an added value and we believe that keeping them in the outcomes section is enough for providing the necessary information to the readers to completely understand our work. In conclusion, we would like to thank Chen et al. for their comments that allowed us to confirm the methodological strengths of our paper. Indeed, as reported in this reply, addressing the points raised did not significantly change our results indicating that our work, even if simple, is correct and the conclusions are sound.