Study Proposes Statistical Definition for Low Fertilization in IVF and ICSI Cycles

A new study introduces a standardized statistical framework to define and assess low fertilization outcomes in assisted reproductive technology (ART), offering evidence that such outcomes may reflect a distinct clinical condition rather than random variation.

The research aimed to determine whether reduced fertilization rates observed in some IVF and ICSI cycles signify a pathological entity or fall within the bounds of expected statistical variability. To explore this, the investigators developed a statistical model grounded in two core assumptions: that all oocytes from a given cycle share the same probability of fertilization, based on benchmarks from the Istanbul Consensus, and that no underlying clinical factor is influencing the result. Under these assumptions, fertilization rates should follow a binomial distribution.

The authors defined "LOW FERTILIZATION" as an event whose probability falls below 5% under the null hypothesis, assuming the fertilization potential of the oocytes is normal. Using this framework, they generated a reference table indicating the maximum number of fertilized oocytes per total retrieved that would meet the threshold for low fertilization. They also provided R code to allow independent researchers to apply or adapt the model.

The model was then tested against real-world data from 2,653 women undergoing ART. LOW FERTILIZATION occurred in 118 cases, or 4.5% of cycles (95% CI: 3.7–5.3%), including 3.6% of conventional IVF cycles and 5.2% of ICSI cycles. In cycles using ICSI, severe male factor infertility was significantly associated with low fertilization, with an adjusted odds ratio of 2.60 (95% CI: 1.37–4.73).

The study also assessed recurrence risk. Among couples experiencing low fertilization in a first cycle, the adjusted odds of recurrence in a second cycle were 3.70 (95% CI: 1.40–8.71). For couples who underwent ICSI in both cycles, the recurrence odds increased to 7.26 (95% CI: 2.48–19.93).

These findings support the concept that low fertilization is not solely a statistical occurrence but can indicate an underlying clinical issue in a subset of patients. The proposed metric may serve as a tool for identifying at-risk cases and guiding further investigation into causes and interventions.