Interventions for unexplained infertility: a systematic review and meta-analysis

Review question
Researchers in Cochrane reviewed the evidence on the effectiveness and safety of ovarian stimulation (OS), intrauterine insemination (IUI), OS-IUI, and in vitro fertilisation (IVF) with or without intracytoplasmic sperm injection (ICSI) versus expectant management in couples with unexplained infertility.

Background
Treatment options for unexplained infertility include expectant management as well as active treatments such as ovarian stimulation (OS), intrauterine insemination (IUI), OS-IUI,  and in vitro fertilisation (IVF) with or without intracytoplasmic sperm injection (ICSI). Network meta-analysis synthesises evidence of direct and indirect comparisons of interventions and enables researchers to simultaneously assess the effectiveness of more than two interventions for the same condition, so that clinicians can use the evidence to offer the best treatment. Therefore, we compared all these different treatment options by using network meta-analysis, to better inform clinical decision-making.

Study characteristics
We found 27 randomised controlled trials comparing these treatments with each other in a total of 4349 couples with unexplained infertility. The evidence is current to September 2018.

Key results
Evidence of differences in live birth between expectant management and the other four treatments (OS, IUI, OS-IUI, and IVF/ICSI) was insufficient. If the chance of live birth following expectant management is assumed to be 17%, the chance following OS, IUI, OS-IUI, and IVF would be 9% to 28%, 11% to 33%, 15% to 37%, and 14% to 47%, respectively. Compared to expectant management/IUI, OS may increase the chances of multiple pregnancy, and OS-IUI probably increases the chances of multiple pregnancy. Evidence showing differences between IVF/ICSI and expectant management for multiple pregnancy was insufficient. If the chance of multiple pregnancy following expectant management/IUI is assumed to be 1%, the chance following OS, OS-IUI, and IVF/ICSI would be 1% to 5%, 1% to 5%, and 0% to 6%, respectively.

Certainty of the evidence
The certainty of evidence overall was low to moderate. The main limitations were imprecision (not enough couples have been studied) and heterogeneity (couples in existing studies had different clinical characteristics).

Authors' conclusions: 

There is insufficient evidence of differences in live birth between expectant management and the other four interventions (OS, IUI, OS-IUI, and IVF/ICSI). Compared to expectant management/IUI, OS may increase the odds of multiple pregnancy, and OS-IUI probably increases the odds of multiple pregnancy. Evidence on differences between IVF/ICSI and expectant management for multiple pregnancy is insufficient, as is evidence of a difference for moderate or severe OHSS between IVF/ICSI and OS-IUI.

Read the full abstract...
Background: 

Clinical management for unexplained infertility includes expectant management as well as active treatments, including ovarian stimulation (OS), intrauterine insemination (IUI), OS-IUI,  and in vitro fertilisation (IVF) with or without intracytoplasmic sperm injection (ICSI).

Existing systematic reviews have conducted head-to-head comparisons of these interventions using pairwise meta-analyses. As this approach allows only the comparison of two interventions at a time and is contingent on the availability of appropriate primary evaluative studies, it is difficult to identify the best intervention in terms of effectiveness and safety. Network meta-analysis compares multiple treatments simultaneously by using both direct and indirect evidence and provides a hierarchy of these treatments, which can potentially better inform clinical decision-making.

Objectives: 

To evaluate the effectiveness and safety of different approaches to clinical management (expectant management, OS, IUI, OS-IUI, and IVF/ICSI) in couples with unexplained infertility.

Search strategy: 

We performed a systematic review and network meta-analysis of relevant randomised controlled trials (RCTs). We searched electronic databases including the Cochrane Gynaecology and Fertility Group Specialised Register of Controlled Trials, the Cochrane Central Register of Studies Online, MEDLINE, Embase, PsycINFO and CINAHL, up to 6 September 2018, as well as reference lists, to identify eligible studies. We also searched trial registers for ongoing trials.

Selection criteria: 

We included RCTs comparing at least two of the following clinical management options in couples with unexplained infertility: expectant management, OS, IUI, OS-IUI, and IVF (or combined with ICSI).

Data collection and analysis: 

Two review authors independently screened titles and abstracts identified by the search strategy. We obtained the full texts of potentially eligible studies to assess eligibility and extracted data using standardised forms. The primary effectiveness outcome was a composite of cumulative live birth or ongoing pregnancy, and the primary safety outcome was multiple pregnancy. We performed a network meta-analysis within a random-effects multi-variate meta-analysis model. We presented treatment effects by using odds ratios (ORs) and 95% confidence intervals (CIs). For the network meta-analysis, we used Confidence in Network Meta-analysis (CINeMA) to evaluate the overall certainty of evidence.

Main results: 

We included 27 RCTs (4349 couples) in this systematic review and 24 RCTs (3983 couples) in a subsequent network meta-analysis. Overall, the certainty of evidence was low to moderate: the main limitations were imprecision and/or heterogeneity.

Ten RCTs including 2725 couples reported on live birth. Evidence of differences between OS, IUI, OS-IUI, or IVF/ICSI versus expectant management was insufficient (OR 1.01, 95% CI 0.51 to 1.98; low-certainty evidence; OR 1.21, 95% CI 0.61 to 2.43; low-certainty evidence; OR 1.61, 95% CI 0.88 to 2.94; low-certainty evidence; OR 1.88, 95 CI 0.81 to 4.38; low-certainty evidence). This suggests that if the chance of live birth following expectant management is assumed to be 17%, the chance following OS, IUI, OS-IUI, and IVF would be 9% to 28%, 11% to 33%, 15% to 37%, and 14% to 47%, respectively. When only including couples with poor prognosis of natural conception (3 trials, 725 couples) we found OS-IUI and IVF/ICSI increased live birth rate compared to expectant management (OR 4.48, 95% CI 2.00 to 10.1; moderate-certainty evidence; OR 4.99, 95 CI 2.07 to 12.04; moderate-certainty evidence), while there was insufficient evidence of a difference between IVF/ICSI and OS-IUI (OR 1.11, 95% CI 0.78 to 1.60; low-certainty evidence).

Eleven RCTs including 2564 couples reported on multiple pregnancy. Compared to expectant management/IUI, OS (OR 3.07, 95% CI 1.00 to 9.41; low-certainty evidence) and OS-IUI (OR 3.34 95% CI 1.09 to 10.29; moderate-certainty evidence) increased the odds of multiple pregnancy, and there was insufficient evidence of a difference between IVF/ICSI and expectant management/IUI (OR 2.66, 95% CI 0.68 to 10.43; low-certainty evidence). These findings suggest that if the chance of multiple pregnancy following expectant management or IUI is assumed to be 0.6%, the chance following OS, OS-IUI, and IVF/ICSI would be 0.6% to 5.0%, 0.6% to 5.4%, and 0.4% to 5.5%, respectively.

Trial results show insufficient evidence of a difference between IVF/ICSI and OS-IUI for moderate/severe ovarian hyperstimulation syndrome (OHSS) (OR 2.50, 95% CI 0.92 to 6.76; 5 studies; 985 women; moderate-certainty evidence). This suggests that if the chance of moderate/severe OHSS following OS-IUI is assumed to be 1.1%, the chance following IVF/ICSI would be between 1.0% and 7.2%.