Design: systematic review and multivariate meta-analysis of observational studies
Population: 156 studies of pregnant women tested for genital mycoplasmas, literature to February 2026
Primary outcome: preterm birth, low birthweight and small-for-gestational-age birth
Effect: cervicovaginal Ureaplasma parvum and preterm birth OR 1.63 (95% CI 1.36-1.96); Mycoplasma hominis and preterm birth adjusted OR 1.75 (1.21-2.53)

A systematic review screened 11,957 records and pooled 156 studies on Mycoplasma genitalium, Mycoplasma hominis and Ureaplasma species in pregnancy. The question it set out to answer was whether detection in cervicovaginal samples - rather than in amniotic fluid or placenta, where the link is better established - carries the same signal.

It does. In sensitivity analyses restricted to cervicovaginal sampling, Ureaplasma parvum was associated with preterm birth (odds ratio 1.63, 95% CI 1.36-1.96), with low birthweight (OR 1.56, 1.33-1.83) and with small-for-gestational-age birth (OR 1.47, 1.19-1.80). Mycoplasma hominis was associated with preterm birth (adjusted OR 1.75, 1.21-2.53) and low birthweight (OR 1.81, 1.51-2.16). First-trimester detection of Ureaplasma carried a stronger association than second-trimester positivity (P=0.044).

The authors then do something worth noting: they argue against acting on their own finding. All the included data are observational, residual confounding is unresolved, and they read cervicovaginal detection as a marker of the vaginal microbial environment rather than as an isolated pathogen. That conclusion matters more than the odds ratios, because the practical temptation - swab, find Ureaplasma, prescribe a macrolide - has never been shown to improve any outcome, and would add antibiotic exposure in the first trimester to a population already over-prescribed.

Do not add genital mycoplasma screening to routine antenatal work-up on the strength of this
Where a swab has already been sent and returns Ureaplasma, resist treating an asymptomatic result
Read a positive result as a marker of vaginal dysbiosis, not as an infection needing eradication
Reserve testing for symptomatic infection or investigation of established preterm labour
Note that first-trimester positivity carried the stronger association, if you are interpreting a result already in the notes

Why it matters

The finding that would tempt a clinician to screen is the same finding the authors say does not justify screening.

The statistics, in plain English

Odds ratios of 1.5 to 1.8 pooled from observational studies sit in the range where unmeasured confounding can plausibly account for the whole effect: socioeconomic position, smoking and prior preterm birth all predict both the organism and the outcome. A meta-analysis of 156 studies has a great deal of statistical power and none of the design strength of a single randomised trial - pooling observational data makes the estimate more precise, not less confounded.