Strengths and limitations 
The prospective design is a strength of our study. Participants were compared with all non-participants by a retrospective registry linkage, enabling us to examine whether participants were representative of all pregnant women with COVID-19 at the index hospital during the study period. Additionally, tests for SARS-CoV-2 in Denmark are free of charge and performed at public test centres reporting the results to national databases, which reduced the number of undetected cases of COVID-19. Finally, we used a serological assay with a high diagnostic accuracy with a sensitivity of 95.3% and a specificity of 100% (13).
The study has several limitations. We included 77.8% of eligible women. However, participants and non-participants did not differ with regards to background characteristics. We had no samples from mothers or cord blood between eight and 16 days after a positive maternal pharyngeal swab, entailing the exact timing of seroconversion among pregnant women with COVID-19 unknown. Further, most vaginal swabs were kept at -20C until analysis, possibly causing a deterioration of SARS-CoV-2 during storage, decreasing our ability to detect SARS-CoV-2 in the vaginal samples. However, a previous study found that SARS-CoV-2 RNA was stable when stored at -20°C for up to 84 days (14). Lack of infection status from the participants’ partners is a limitation, since it is not fully evaluated whether SARS-CoV-2 can be transmitted through semen and cause detection of SARS-CoV-2 in the vagina (15). We measured total antibodies and did not discriminate between IgG and IgM. Finally, three participants were not tested by pharyngeal RT-PCR (one by an antigen-test and two at private test centres), thus exact details about analysis methods are lacking. However, the participants were positive for SARS-CoV-2 antibodies.