How should cumulative live birth rates be calculated and reported?

[Tier B — no verified reporting standard in hand] The critical issue is the denominator, and it is the single most common way IVF success figures mislead. Per embryo transfer excludes cycles that produced no transferable embryo, and so systematically overstates the chance for a patient starting treatment. Per oocyte retrieval includes cycles with no transfer but excludes cancelled stimulations. Per cycle started is the most conservative and the most honest from the patient's standpoint. Cumulative across one stimulation, counting the fresh transfer plus all subsequent frozen transfers from that retrieval, is the figure most useful for counselling — provided the denominator is stated. Note that ESHRE (2026) uses cumulative live birth rate per started cycle and live birth rate per started cycle as its critical efficacy outcomes. Two further sources of non-comparability: Follow-up duration. Cumulative figures rise as more frozen embryos are used, so a clinic reporting at 12 months and one reporting until embryos are exhausted are not comparable. Case mix. Age distribution and diagnosis distribution differ between clinics, so unadjusted comparison between units is not meaningful. The Vienna Consensus (2017) declined to adopt live birth rate as a laboratory performance indicator at all, on the grounds that too many other variables affect it — a useful caution when LBR is used to judge laboratory quality. Reviewer: if your unit publishes success rates, the denominator, follow-up period and age banding should be stated alongside every figure. This answer does not cite a reporting standard because I could not verify one.