Scoping review of active surveillance practices for vaccine safety: the importance of identification of diverse evidence in an iterative process

Article type
Authors
Cai T1, Liu LL1, Yao XY1, Liu ZK1, Yang Y2, Meng RG2, Zhan SY1
1School of Public Health, Peking University
2Center for Data Science in Health and Medicine, Peking University
Abstract
Background: active surveillance for vaccine safety (ASVS) is emerging worldwide to make up the defects of traditional passive surveillance. ASVS has been performed in routine surveillance systems, dedicated projects or specific studies through diverse practices. Scoping reviews are appropriate for understanding concepts or methods in a broad field, as their fundamental methodology framework indicates an iterative process for identification of comprehensive evidence. However, this principle is not fully implemented in most scoping reviews due to the workload and time costs.

Objectives: to identify diverse ASVS practices from literature and clarify their methods and features; to demonstrate the importance of searching iteratively and screening different types of evidence in scoping reviews.

Methods: we searched MEDLINE, Embase, and Web of Science for English literature published up to 31 December 2018, which described ASVS practices, including routine systems, dedicated projects, and methods used in specific studies. Firstly, we selected all literature that mentioned any potential ASVS practices, including those without detailed descriptions. We conducted reference mining of the uninformative reports for relevant citations, and performed web searching of named systems/programs for grey literature and webpages. Based on all material, we then confirmed the final eligible ASVS practices, and included only the informative literature for these practices. We charted predefined data items from the literature using a standardized form. We clarified mechanisms of the identified ASVS systems/programs through Framework Synthesis, and summarized the methods of other specific studies narratively.

Results: initially, we identified 145 items of published literature - involving 73 potential ASVS practices - from database searching. We selected 67 further supplementary publications from reference mining of 37 uninformative reports, and located 45 documents/webpages from web searching (Figure 1). Finally, we included 156 publications describing 46 eligible ASVS practices, among which were 89 publications supporting 35 practices identified from reference mining and web searching. We classified the 46 practices, which included six systems, 12 projects (Table 1), and 28 study methods, into three types of ASVS according to their features: data linkage, active follow-up, and active case ascertainment.

Conclusions: iterative searching and screening in scoping reviews can improve the completeness of evidence and provide supplementary information. Current ASVS practices present as three types with different advantages and limitations.

Patient or healthcare consumer involvement: we initiated this project considering the public voice for improving vaccine safety surveillance, and designed the review based on the needs from the China FDA and CDC.