Systematic review: fetal death reporting and risk in Zika-affected pregnancies
Susannah Hopkins Leisher1, Arin A Balalian1, Hanna Reinebrant2
1Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY, USA.
Objectives:
Zika virus is linked to several adverse pregnancy outcomes. We assessed whether Zika infection during pregnancy is associated with increased risk of foetal death (miscarriage, stillbirth, abortion) and whether there is incomplete reporting of such deaths.
Methods:
We searched PubMed, Embase, CINAHL, Web of Science and LILACS for studies reporting Zika-affected completed pregnancies (ending in foetal death or live birth), excluding studies whose aim required live birth. Studies 'allowed' foetal death if their populations were defined to encompass both live births and foetal deaths, regardless of whether deaths were actually found. Two authors independently extracted data and assessed study quality. Foetal death absolute and relative risks in Zika-affected vs. unaffected pregnancies were calculated.
Results:
We found 108 reports including 24 699 completed, Zika-affected pregnancies. The median absolute risk in 37 studies of completed, Zika-affected pregnancies was 6.3% (IQR 3.2%, 10.6%) for foetal death and 5.9% (IQR 0%, 29.1%) for non-fatal adverse outcomes (e.g. microcephaly). More studies allowed non-fatal adverse outcomes (95%) than foetal death (58%). Of studies which allowed them, 94% found at least one foetal death. In 37% of reports, it was unknown whether foetal deaths were allowed. Only one study had sufficient data to estimate a foetal death relative risk (11.05, 95% CI 3.43, 35.55).
Conclusions:
Evidence was insufficient to determine whether foetal death risk is higher in Zika-affected pregnancies, but suggests quality of foetal death reporting should be improved, including stating whether foetal deaths were found, how many, and at what gestational ages, or justifying their exclusion.
