Design: Systematic review and meta-analysis of pooled proportions, random-effects
Population: 44 studies; 3221 women and 748 fetuses with parvovirus B19 infection
Primary outcome: Perinatal outcomes, including intrauterine demise
Effect: Intrauterine demise 38.7% (95% CI 27.3 to 51.6) in fetal infection vs 7.3% (4.6 to 11.3) in maternal infection

Forty-four studies covering 3221 women and 748 fetuses were pooled to describe what parvovirus B19 actually does to a pregnancy. Among confirmed fetal infections, intrauterine death occurred in 38.7 per cent. Among maternal infections, where the fetus may never be infected, it was 7.3 per cent. Hydrops was seen in 58.7 per cent of infected fetuses and cardiomegaly in 39.0 per cent.

That distinction is the whole of the counselling. A woman told she has parvovirus in pregnancy hears the higher number if you give her only one. The rate that applies to her at first contact is the maternal-infection rate, and it falls further once serial middle cerebral artery Dopplers stay normal.

The 2023 to 2024 outbreak did not behave differently: no significant change in intrauterine death or ultrasound signs of anaemia against earlier waves. One finding is new rather than reassuring. Among fetuses who had prenatal MRI, 28.6 per cent had abnormal neuroimaging, which is a reason to discuss second-line imaging after intrauterine transfusion rather than to assume a transfused survivor is unaffected.

Separate maternal infection from confirmed fetal infection before quoting any risk figure.
Serial MCA peak systolic velocity is the surveillance that matters; hydrops is a late sign.
Discuss fetal MRI with the fetal medicine unit where transfusion has been needed.
Ask about occupational exposure — schools, creches, paediatric wards — when a rash illness is reported in the household.
Document the gestation at maternal infection; the fetal consequences cluster before 20 weeks.

Why it matters

The number most often quoted to a frightened woman is the one that applies to the smallest and sickest subgroup.

The statistics, in plain English

These are pooled proportions, not comparisons, so the confidence intervals describe how precisely the literature pins each rate rather than any treatment effect. The interval around intrauterine death in fetal infection is wide, 27.3 to 51.6 per cent, because the contributing series are small and referral-biased — cases reaching a fetal medicine unit are the sick end of the spectrum. The wave comparison (OR 0.84, 0.46 to 1.54) crosses 1.0 comfortably: no difference detected, and the study was not large enough to exclude a modest one.