ICSI does not improve reproductive outcomes in autologous ovarian response cycles with non-male factor subfertility

Abstract STUDY QUESTION Does the method of fertilisation improve reproductive outcomes in poor ovarian response (POR) cycles when compared to all other ovarian response categories in the absence of male factor subfertility? SUMMARY ANSWER ICSI does not confer any benefit in improving the clinical pregnancy or live birth (LB) outcome in autologous ovarian response cycles in the absence of male factor subfertility when compared to IVF. WHAT IS KNOWN ALREADY ICSI is associated with an improved outcome when compared to IVF in patients with severe male factor subfertility. STUDY DESIGN, SIZE, DURATION A retrospective study involving 1 376 454 ART cycles, of which 569 605 (41.4%) cycles fulfilled the inclusion and exclusion criteria for all autologous ovarian response categories: 272 433 (47.8%) IVF cycles and 297 172 (52.2%) ICSI cycles. Of these, the POR cohort represented 62 641 stimulated fresh cycles (11.0%): 33 436 (53.4%) IVF cycles and 29 205 (46.6%) ICSI cycles. PARTICIPANTS/MATERIALS, SETTING, METHOD All cycles recorded on the anonymised Human Fertilisation and Embryology Authority (HFEA) registry database between 1991 and 2016 were analysed. All fresh cycles with normal sperm parameters, performed after 1998 were included: frozen cycles, donor oocyte and sperm usage, intrauterine insemination cycles, preimplantation genetic testing (PGT) for aneuploidies (PGT-A), PGT for monogenic/single gene defects (PGT-M), PGT for chromosomal structural arrangements (PGT-SR) cycles, where the reason for stimulation was for storage and unstimulated cycles were excluded. MAIN RESULTS AND THE ROLE OF CHANCE ICSI did not confer any benefit in improving the LB outcome when compared to conventional IVF per treatment cycle (PTC), when adjusted for female age, number of previous ART treatment cycles, number of previous live births through ART, oocyte yield, stage of transfer, method of fertilisation and number of embryos transferred in the POR cohort (adjusted odds ratio [a OR] 1.03, 99.5% confidence interval [CI] 0.96–1.11, P = 0.261) and all autologous ovarian response categories (aOR 1.00, 99.5% CI 0.98–1.02, P = 0.900). The mean fertilisation rate was statistically lower for IVF treatment cycles (64.7%) when compared to ICSI treatment cycles (67.2%) in the POR cohort (mean difference −2.5%, 99.5% CI −3.3 to −1.6, P