Endometrial thickness and pregnancy rates after IVF: a systematic review and meta-analysis.

BACKGROUND Thin endometrium on ultrasound in the course of ovarian hyperstimulation has been thought to be associated with poor success rates after IVF, even in the absence of prior intrauterine surgery or infection. To assess the clinical significance of endometrial thickness (EMT) for IVF outcome, we performed a systematic review and meta-analysis. METHODS The electronic databases Pubmed, Cochrane and Embase were searched up to October 2013 for articles that studied the association between EMT and IVF outcome. The articles had to be written in the English or Dutch language. Studies were included if two-by-two tables for EMT and pregnancy rates could be constructed. Study quality was scored using the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist. Summary receiver operating characteristic (sROC) curves were estimated to assess the accuracy of EMT in the prediction of pregnancy. In addition, odds ratios (ORs) with 95% confidence intervals (CIs) were calculated using a Mantel-Haenszel random effect model expressing the association between EMT and pregnancy chances. Meta-regression was performed to determine if female age and number of oocytes at retrieval interacted in the estimated effect of EMT on IVF outcome. RESULTS A total of 1170 studies was retrieved by the search. The overall quality of the 22 studies included in the review and meta-analysis was moderate. The estimated sROC curve indicated a virtually absent discriminatory capacity of EMT in the prediction of pregnancy. A thin endometrium (≤ 7 mm) was observed in only 2.4% of the reported cases (260/10 724). In these cases a trend towards lower ongoing pregnancy and live birth rates for women with EMT ≤ 7 mm was observed [OR 0.38 (95% CI 0.09-1.5)]. The probability of clinical pregnancy for an EMT ≤ 7 mm was significantly lower compared with cases with EMT > 7 mm [23.3% versus 48.1%, OR 0.42 (95% CI 0.27-0.67)]. Positive and negative predictive values for the outcome of clinical pregnancy 77 and 48%, respectively. The relationship between the number of oocytes and female age on the one hand and pregnancy on the other hand was very weak making correction for these variables unfeasible. CONCLUSIONS Current data indicate that EMT has a limited capacity to identify women who have a low chance to conceive after IVF. The frequently reported cut-off of 7 mm is related to a lower chance of pregnancy, but occurs infrequently. The use of EMT as a tool to decide on cycle cancellation, freezing of all embryos or refraining from further IVF treatment seems not to be justified based on the current meta-analysis. Further research is needed to investigate the real independent significance of EMT in IVF.