Endometrial injury for pregnancy following sexual intercourse or intrauterine insemination

Background: Intentional endometrial injury is being proposed as a technique to improve the probability of pregnancy in women undergoing assisted reproductive technologies (ART) such as in vitro fertilisation (IVF). Endometrial injury is often performed by pipelle biopsy and is a common gynaecological procedure with established safety. However, it causes a moderate degree of discomfort/pain and requires an additional pelvic examination. The effectiveness of this procedure outside of ART, in women or couples attempting to conceive via sexual intercourse or with intrauterine insemination (IUI), remains unclear.

Objectives: To assess the effectiveness and safety of intentional endometrial injury performed in infertile women or couples attempting to conceive through sexual intercourse or intrauterine insemination (IUI).

Search methods: The Cochrane Gynaecology and Fertility Group Specialised Register, CENTRAL, MEDLINE, Embase, PsycINFO, CINAHL, LILACS, ISI Web of Knowledge, and clinical trial registries were searched from inception to 21 May 2020, as were conference abstracts and reference lists of relevant reviews and included studies.

Selection criteria: We included randomised controlled trials (RCTs) that evaluated any kind of intentional endometrial injury in women planning to undergo IUI or attempting to conceive spontaneously (with or without ovarian stimulation (OS)) compared to no intervention, a mock intervention, or intentional endometrial injury performed at a different time.

Data collection and analysis: We used standard methodological procedures recommended by Cochrane. Primary outcomes were live birth/ongoing pregnancy and pain experienced during the procedure. Due to high risk of bias associated with many of the studies, primary analyses of all review outcomes were restricted to studies at low risk of bias. Sensitivity analysis including all studies was then performed.

Main results: We included 22 RCTs (3703 women). Most of these studies included women with unexplained infertility. Intentional endometrial injury versus either no intervention or a sham procedure The primary analysis was restricted to studies at low risk of bias, which left only one study included. We are uncertain whether endometrial injury has an effect on the probability of live birth, as only one study is included in the analysis and the confidence interval is wide (risk ratio (RR) 1.11, 95% confidence interval (CI) 0.78 to 1.59; 1 RCT, 210 participants). Evidence suggests that if the chance of live birth with no intervention/a sham procedure is assumed to be 34%, then the chance with endometrial injury would be 27% to 55%. When all studies were included in the sensitivity analysis, we were uncertain whether endometrial injury improves live birth/ongoing pregnancy, as the evidence was of very low quality (RR 1.71, 95% CI 1.32 to 2.21; 8 RCTs, 1522 participants; I² = 16%). Evidence suggests that if the chance of live birth/ongoing pregnancy with no intervention/a sham procedure is assumed to be 13%, then the chance with endometrial injury would be 17% to 28%. A narrative synthesis conducted for the other primary outcome of pain during the procedure included studies measuring pain on a zero-to-ten visual analogue scale (VAS) or grading pain as mild/moderate/severe, and showed that most often mild to moderate pain was reported (6 RCTs, 911 participants; very low-quality evidence). Timing of intentional endometrial injury Four trials compared endometrial injury performed in the cycle before IUI to that performed in the same cycle as IUI. None of these studies reported the primary outcomes of live birth/ongoing pregnancy and pain during the procedure. One study compared endometrial injury in the early follicular phase (EFP; Day 2 to 4) to endometrial injury in the late follicular phase (LFP; Day 7 to 9), both in the same cycle as IUI. The primary outcome live birth/ongoing pregnancy was not reported, but the study did report the other primary outcome of pain during the procedure assessed by a zero-to-ten VAS. The average pain score was 3.67 (standard deviation (SD) 0.7) when endometrial injury was performed in the EFP and 3.84 (SD 0.96) when endometrial injury was performed in the LFP. The mean difference was -0.17, suggesting that on average, women undergoing endometrial injury in the EFP scored 0.17 points lower on the VAS as compared to women undergoing endometrial injury in the LFP (95% CI -0.48 to 0.14; 1 RCT, 110 participants; very low-quality evidence).

Authors' conclusions: Evidence is insufficient to show whether there is a difference in live birth/ongoing pregnancy between endometrial injury and no intervention/a sham procedure in women undergoing IUI or attempting to conceive via sexual intercourse. The pooled results should be interpreted with caution, as the evidence was of low to very low quality due to high risk of bias present in most included studies and an overall low level of precision. Furthermore, studies investigating the effect of timing of endometrial injury did not report the outcome live birth/ongoing pregnancy; therefore no conclusions could be drawn for this outcome. Further well-conducted RCTs that recruit large numbers of participants and minimise bias are required to confirm or refute these findings. Current evidence is insufficient to support routine use of endometrial injury in women undergoing IUI or attempting to conceive via sexual intercourse.