A meta-analysis of 156 studies drawn from 11,957 records examined cervicovaginal detection of genital mycoplasmas and adverse pregnancy outcomes. Cervicovaginal Ureaplasma parvum was associated with preterm birth (odds ratio 1.63, interval 1.36 to 1.96), low birthweight (1.56) and small for gestational age (1.47). Mycoplasma hominis showed adjusted odds ratios of 1.75 for preterm birth and 1.81 for low birthweight. First-trimester detection carried stronger associations than second-trimester.
The associations held in sensitivity analyses excluding amniotic fluid and placental specimens, which is the finding: this is not simply intra-amniotic infection detected by another route.
What the authors then do is the model of how to report an observational association. They state plainly that the data are observational, residual confounding is likely, and the findings do not support routine screening or treatment — and that cervicovaginal detection should be read as part of the vaginal microbial environment rather than as an isolated pathogen.
That restraint is correct. Trials of antibiotic treatment for asymptomatic genital colonisation in pregnancy have repeatedly failed to reduce preterm birth, and some have suggested harm. An association of this size, from data of this kind, is a reason to study the question, not to act on it.
- Observational throughout; residual confounding cannot be excluded
- Associations persisted excluding amniotic and placental specimens
- First-trimester detection was more strongly associated than second
- The authors explicitly do not support screening or treatment
- Previous antibiotic trials for asymptomatic colonisation have not helped
The statistics, in plain English
Odds ratios in the 1.5 to 1.8 range from observational data sit in the zone where unmeasured confounding can plausibly account for the whole effect — vaginal microbial composition travels with socioeconomic status, smoking, prior preterm birth and coexisting infections, each independently linked to the outcome. Adjustment removes only the confounders that were measured.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for obstetrics & gynaecology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free