Background: There is limited comprehensive evaluation of the methodology and reporting quality of observational studies of vaccine safety.
Methods: Databases including Medline, Embase, Web of Science, Scopus, and Chinese databases were searched from inception to 31 May 2021. All observational studies regarding vaccine safety using an SCCS design were selected. Information regarding methodological elements were extracted. In addition, reporting quality was assessed using the REporting of studies Conducted using Observational Routinely collected health Data statement for PharmacoEpidemiology (RECORD-PE).
Results: : Of the 105 studies identified, administrative databases were the main data source for vaccination records and adverse events following immunization (AEFI). Twenty-eight articles (27%) used multiple designs to verify the association, and the results obtained with the SCCS design were robust. The top three AEFI studied were intussusception, Guillain-Barré syndrome, and convulsions. Only 21 studies (20%) reported the approach for case validation by chart review. The healthy vaccinee effect was considered by 51 studies (49%), with 16 of them (31%) using extended SCCS models to alleviate this effect. Overall, the reporting quality of included studies could be improved.
Conclusions: Administrative databases were the main data source for vaccination records and adverse events following immunization. Case validation, the validity of assumptions for standard SCCS, and quality of reporting should be given more importance in future research projects.
Keywords: RECORD-PE; Self-controlled case-series design; reporting quality; safety; systematic review; vaccine.