Design: Systematic review and meta-analysis

Setting: Databases evaluation (PUBMED, SCOPUS and EMBASE), with inclusion criteria articles with symptomatic or asymptomatic patients with amenorrhea for more than 12 months, without tamoxifen use, with transvaginal ultrasound evaluation of endometrial thickness and its correlation with the anatomopathological findings of endometrial neoplasm. Methodological quality assessed using the Score of Newcastle – Ottawa tool.

The meta-analysis was performed by dividing the groups into symptomatic and asymptomatic patients and the analysis of sensitivity, specificity, positive predictive value, negative predictive value and endometrial thickness cut-off points of endometrial atrophy findings, benign findings or malignant findings was performed.

Patients or Participants: Eligibility criteria were clinical trials and observational studies that contemplated transvaginal ultrasound assessment of endometrial thickness in asymptomatic menopausal patients or postmenopausal vaginal bleeding and correlation with benign or malignant findings identified in invasive procedures of the uterine cavity

Interventions: Not applicable.

Measurements and Main Results: Thirty one studies eligible for analysis were obtained. Twenty one studies with symptomatic patients, four with asymptomatic patients, three with both and three with patients using hormonal therapy. The results indicate that the endometrial thickness in symptomatic menopausal patients with greater evidence of correlation with neoplastic findings is 5mm, with the sensitivity of 92,9%, specificity 57,5%, positive predictive value 43,0% and negative predictive value 96,0%. No value of endometrial thickness was conclusive for asymptomatic patients or for those using hormone replacement therapy.

Conclusion: Further studies are needed to identify an ideal cut-off point for endometrial thickness on transvaginal ultrasound examination to validate endometrial investigation looking for intrauterine neoplasia. In the case of menopausal patients with vaginal bleeding, an endometrial thickness of 5 mm appears to be adequate in identifying patients eligible for invasive uterine investigation.