First trimester and early second trimester transvaginal cervical length before 16 weeks among women with preterm and term birth: a systematic review and meta-analysis

Abstract

<h3>Objective</h3><p>To evaluate the difference in first and early second trimester transvaginal cervical length (TVCL) between spontaneous preterm birth (PTB) and term birth.</p><h3>Data sources</h3><p>PubMed, MEDLINE, EMBASE, and the Cochrane Library were systematically searched through August 2023.</p><h3>Study eligibility criteria</h3><p>Studies must have included: (1)TVCL measurement before the 16⁺⁰ th week of gestation and (2)TVCL measurement in a population of patients who delivered preterm and at term. Abstracts, studies with duplicated data, and those with cervical length measured by transabdominal ultrasound scan were excluded.</p><h3>Study appraisal and synthesis methods</h3><p>K.W.C. and J.L. searched for, screened, and reviewed the articles independently. The quality of the studies was assessed using the Newcastle-Ottawa Scale. Mean differences (MDs) were calculated using a random effects model and pooled through a meta-analysis.</p><h3>Results</h3><p>A total of 5727 published article were identified. Only 10 studies met the inclusion criteria which analysed 22151 pregnancies. All studies excluded iatrogenic PTB. TVCL was significantly shorter in women with spontaneous PTB compared with those who delivered at term (MD = -0.97, 95% CI = -1.65 to -0.29, p = 0.005, I² = 69%). When a linear technique was used to measure TVCL, a significantly shorter TVCL was associated with spontaneous PTB, compared with term birth (MD = -1.09, 95% CI = -1.96 to -0.21, p = 0.02, I² = 77%). A shorter TVCL measured by other techniques was also associated with spontaneous PTB at < 34-35 weeks (MD = -1.87, 95% CI = -3.04 to -0.70, p = 0.002 I² = 0%). When studies where interventions were given for ‘short’ cervix or studies with a mean TVCL ≥ 40mm were excluded, a significantly shorter TVCL among those with spontaneous PTB was noted (MD = -1.13, 95% CI = -1.89 to -0.37, p = 0.004; MD = -0.86, 95% CI = -1.67 to -0.04, p = 0.04, respectively). The optimal TVCL cut-offs were 38-39mm, yielding pooled sensitivity 0.80; specificity 0.45; positive likelihood ratio 1.16, negative likelihood ratio 0.33, diagnostic odds ratio 5.12 and area under curve 0.75.</p><h3>Conclusions</h3><p>Women with spontaneous PTB had a significantly shorter TVCL before 16 weeks of gestation than those who delivered at term. Either the linear or 2-line method is an acceptable technique to measure TVCL.</p>