Timed intercourse versus intra-uterine insemination with or without ovarian hyperstimulation for subfertility in men.

backgroundAlthough intra-uterine insemination (IUI) is widely used, however its effectiveness remains a matter of debate. Although IUI is less invasive and expensive than IVF or GIFT, it should only be applied if the probability of conception is improved significantly as compared to the natural chance of conceiving. To increase the number of available oocytes at the site of fertilization, controlled ovarian hyperstimulation (COH) can be applied in conjunction with IUI. Uncontrolled studies suggest a beneficial effect of COH in combination with IUI, also when a male factor is present. To be able to draw firm conclusions whether IUI and/or COH improve the probability of conception, several comparisons should be performed in randomized controlled trials (RCTs).

objectivesTo determine for male subfertility whether intrauterine insemination (IUI) improves the probability of conception compared with timed intercourse and whether the addition of controlled ovarian hyperstimulation influences the results. SEARCH STRATEGY: 1. The specialist database of the Cochrane Menstrual Disorders and Subfertility Group. 2. Medline search. 3. Embase search. 4. DDFU search. 5. BIOSIS search. 6. SCIsearch. 7. Manual searching of references mentioned in the obtained studies. 8. Personal communication and write letters to experts (14) in the field. 9. Abstracts of The American Society for Reproductive Medicine and European Society for Human Reproduction and Embryology Meetings. When important information is lacking from the original publications the authors will be contacted. SELECTION CRITERIA: Randomized controlled trials only. DATA COLLECTION AND ANALYSIS: Independently by the first 2 authors: 1. Trial design characteristics. 2. Baseline characteristics of participants. 3. Types of intervention. 4. Outcomes where pregnancy is the outcome of main interest. Number of multiple pregnancies and number of cycles with ovarian hyperstimulation syndrome (OHSS) are secondary outcomes. Analysis of agreement between the two observers was determined for the following items: inclusion or exclusion of a trial, method of randomization, definition of male subfertility, design of the trial, number of pregnancies and completed cycles. Sensitivity analysis is performed. MAIN

resultsSeventeen trials fulfilled the selection criteria for this review and were included. Four trials are pending. Crude agreement concerning inclusion or exclusion of trials occurred for 41 of 43 (95%) trials reviewed (kappa 0.90). The included trials comprised 3,662 completed cycles. In natural cycles intrauterine insemination (IUI) significantly improved the probability of conception compared with timed intercourse (TI) (combined odds ratio with 95% confidence intervals: 2.43, 1.54 - 3.83). In cycles with controlled ovarian hyperstimulation (COH) IUI significantly improved the probability of conception also compared with TI (combined odds ratio with 95% confidence intervals: 2.14, 1.30 - 3.51). Despite clinical heterogeneity, these results are based on strong evidence. Intrauterine insemination in cycles with COH improved the probability of conception compared with IUI in natural cycles but significance was not reached (combined odds ratio with 95% confidence intervals: 1.79, 0.98 - 3.25). Comparing IUI in COH-cycles with TI in natural cycles the first treatment modality significantly improved the probability of conception (combined odds ratio with 95% confidence intervals: 6.23, 2.35 - 16.52). REVIEWER'S

conclusionsIntra-uterine insemination offers couples with male subfertility benefit over timed intercourse, both in natural cycles and in cycles with COH. In the case of a severe semen defect (with more than 1 million motile sperm after semen preparation and no triple sperm defect) IUI in natural cycles should be the treatment of first choice. The value of COH need to be further investigated in RCTs. Mild ovarian hyperstimulation with gonadotrophins is advised in cases with less sever