What IVF success rates actually measure and hide

The number you found is real. It just may not be about you.

You looked up a clinic, saw a percentage, and now you are trying to decide whether it means anything for you. That instinct is reasonable. The problem is that "IVF success rate" is not one measurement. It is a family of measurements, and the headline figure you see is usually the most flattering member of that family.

Here is what the number actually measures, what it quietly leaves out, and how to read it without being misled.

What the percentage is counting

Every success rate is a fraction: outcomes divided by a denominator. Both halves can change.

The outcome can be a positive pregnancy test, a clinical pregnancy confirmed on ultrasound, or a live birth. These are sequential steps, and each later endpoint produces a smaller number than the one before it. A live birth rate will always look lower than a clinical pregnancy rate for the same clinic and year.

The denominator can be a cycle started, an egg retrieval, or an embryo transfer. A per-transfer figure answers "what is my chance once I reach transfer?" A per-cycle figure answers "what is my chance from the start of treatment?" Per-transfer numbers exclude cancelled cycles and cycles with nothing to transfer, so they run higher. Both are legitimate. Trouble starts when a per-transfer figure is read as a per-cycle one.

SART, which collects and audits clinic-reported data, notes that a cycle involves egg retrieval and the transfer of all embryos from that cycle, and that its predictor estimates cumulative live birth rate up to three cycles — a different question from a single-cycle percentage (SART success rates).

What the headline hides

Age band. Age at egg retrieval is the major predictor of IVF outcome, and success declines with it. A clinic's headline may come from its strongest age band while you sit in a different one.

Patient mix. A clinic that treats complex diagnoses, low ovarian reserve, or prior failed cycles will report lower numbers than one with tight acceptance criteria. The lower number is not necessarily worse care.

Donor-egg cycles. In donation, embryo competence tracks donor age rather than recipient age. Folding donor-egg cycles into a headline figure changes what the number means for someone using their own eggs.

Sample size. A rate built on a handful of cycles wobbles far more than one built on hundreds. The percentage rarely shows its denominator, so you cannot see the noise.

The year. Published data trails reality. Because live births must be tracked to delivery, reporting lags well behind the treatment year, so a strong year is not the same as a current pattern (RSC Bay Area on CDC and SART data).

Why two numbers for one clinic don't match

A clinic website may advertise 60% while its official report for the same year shows closer to 40%. Neither is necessarily wrong — they are answering different questions, using different endpoints, denominators, or patient segments. When the endpoint, denominator, time window, and patient segment are not stated, the number cannot be interpreted, let alone compared (analysis of clinic website reporting).

Expect reconciliation work. Different official formats can produce figures that look inconsistent for the same clinic. That is a reporting artifact, not proof of a contradiction.

A checklist for reading any clinic's number

Before you weigh a percentage, get these answers:

• Endpoint: Is this a positive test, a clinical pregnancy, or a live birth?
• Denominator: Per cycle started, per retrieval, or per transfer?
• Age band: Which age group does this figure describe, and are you in it?
• Diagnosis group: Does it reflect patients with your diagnosis?
• Donor eggs: Are donor-egg cycles included or excluded?
• Sample size: How many cycles is the figure built on?
• Year: Which treatment year, and how old is that data?
• Cumulative view: What is the cumulative rate across two or three cycles, not just one?

If a clinic cannot answer these, treat the number as a fragment rather than a comparison.

What the rate cannot capture at all

The percentage does not describe the doctor relationship, the quality of the lab, the total cost, or how a clinic handles a failed cycle. Those factors shape your experience and your decisions, and none of them appear in the headline. A rate is one input, not a verdict.

Quick check: do you know what the number measures?

If you can answer the endpoint, denominator, age band, and sample size for a clinic's figure, you are ahead of most people comparing clinics. If you cannot, the number is not yet usable.

How the Featured Expert Can Help

Fertility X-Ray provides detailed, plain-English reports on U.S. fertility clinic success rates, turning CDC data into actionable dossiers for patients. The service sells individual clinic reports and comparative analyses for a fee, aiming to help prospective patients make informed choices. If you want help working out which question a clinic's number is actually answering, you can explore the reports at Fertility X-Ray.