International IVF Success Rates: How to Interpret Registry and Global Data
Before you look at any one clinic's numbers, it helps to understand the larger datasets those numbers are judged against: the world report compiled from national registries, and national systems such as those run by the CDC in the US and the HFEA in the UK. Registry data are the best benchmark available, but they are easy to misread. This guide explains how to read them, and why a "country success rate" is rarely the comparison it looks like.
The layers of published IVF data
- Global. The International Committee for Monitoring Assisted Reproductive Technologies (ICMART) compiles a world report from data submitted by national and regional registries.
- National. Some countries publish standardised clinic-by-clinic results. The CDC publishes ART success rates by US clinic. The HFEA (2016) publishes comparable success rates for every licensed UK clinic.
- Clinic. A clinic's own figures, which may or may not follow any registry's definitions.
Each layer down is closer to your situation, but less standardised and less checked. The skill is using the upper layers to test the lower ones.
The end point also varies between sources. The ICMART report counts deliveries. The CDC tracks cycles from the intended retrieval through to a live-birth delivery. A clinic may report pregnancies instead. When you move between sources, check whether you are looking at deliveries, live births or pregnancies, and whether a twin birth is counted as one outcome or two.
One dataset, several "success rates"
The most recent ICMART world report covers treatment in 2019. It counted 3,544,683 reported cycles and 783,073 babies born (Dyer et al., Fertility and Sterility, 2025). From that one dataset, you can quote very different numbers:
| Measure (2019, worldwide) | Delivery rate |
|---|---|
| Own eggs, per egg retrieval | 22.9% |
| Own eggs, cumulative per egg retrieval (includes later frozen transfers; excludes PGT) | 38.6% |
| Own eggs, per fresh transfer | 31.3% |
| Own eggs, per frozen transfer | 31.9% |
| Per transfer after PGT | 48.7% |
| Donor eggs, per transfer (fresh and frozen) | 34.6% |
None of these figures is wrong. They answer different questions. Per retrieval counts deliveries against everyone who had eggs collected. The cumulative version also counts later frozen transfers from the same eggs. Per transfer leaves out everyone who never reached a transfer at all. The PGT figure is especially easy to misread. It is measured only among transfers of tested embryos, so it says nothing about cycles where no embryo was suitable to transfer.
The CDC (2024) uses similar layers for US data: per intended retrieval (a cycle started with the plan to collect eggs), per actual retrieval, and per transfer. Its cumulative measure combines one year's retrievals with all transfers within a year after them. Whenever you see a percentage, find its denominator before you compare it with anything.
Why country averages mislead
Practice changes faster than outcomes
The ICMART report shows how quickly treatment patterns shift. Frozen transfers made up 61.8% of own-egg transfers without PGT in 2019, up from 14.8% in 2015. "Freeze-all" cycles, where no fresh transfer is done, rose from 24.6% to 38.4% of cycles. PGT cycles grew from 2.6% to 6.6% of all reported cycles. When a country's mix shifts toward frozen transfers or PGT, its per-transfer rate can rise without any patient's overall chance of a baby changing.
Law shapes the mix
Countries differ in what clinics are allowed to do, which changes the numbers they report. In the 2024 ESHRE survey, PGT-A was not permitted in ten European countries or entities, including Denmark, France, the Netherlands, Norway and Sweden, and egg donation was banned in Germany, Switzerland and Turkey (Calhaz-Jorge et al., Human Reproduction, 2024). A country where donor eggs or PGT are common will report a different overall profile from one where they are banned, even if its clinics are no better. See PGT legal status by country for that layer.
Patients differ
The CDC notes that patient factors such as age, egg and sperm quality and cause of infertility drive results. It also points out that clinics vary in whom they accept, and that it doesn't collect data on patient selection. The HFEA makes the same point: clinics treating younger women with milder problems will show higher rates. A destination that attracts many donor-egg patients, or turns away harder cases, will look better in aggregate data.
Gaps and delays in the data
- Not everyone reports. ICMART estimates there were at least 3,855,694 cycles and more than 865,914 babies in 2019 once non-reporting clinics are counted, more than the reported totals. Coverage is incomplete, and incompleteness varies by country.
- The data are old. The 2019 world figures were published in 2025. Even national systems run behind. The CDC explains that cycles must be completed, followed up and analysed first, so its data describe treatment from more than a year before publication. Clinic staff, equipment and methods may have changed since.
- Cross-border patients fall between the cracks. The ESHRE survey authors say there are no reliable figures on cross-border reproductive care, and flag limited outcome data and poor tracking across borders. If you are treated abroad and give birth at home, the destination registry may never learn the outcome.
- Small numbers swing. The CDC warns that clinics with few cycles show more year-to-year variation. A category with one cycle can only show 0% or 100%.
A worked reading (illustrative)
Illustrative only. The clinic and its figure are hypothetical. Imagine a clinic abroad advertising a "65% success rate". Before comparing it with anything, break it down:
- Success of what? Positive test, clinical pregnancy or live birth? Only a live birth matches the registry measures above.
- Per what? If it is per transfer of PGT-tested embryos, the closest world benchmark is the 48.7% PGT per-transfer figure, not the 22.9% per-retrieval figure.
- For whom? If it covers only patients under 35, or only donor-egg recipients, the benchmark changes again.
- When? A single strong year, or an average across several?
Once each question is answered, the claim either lines up with a sensible benchmark or it doesn't. Either way, you learn more than the headline figure tells you.
How to use registry data well
Registry data can't tell you your own chances. As the CDC puts it, a clinic's rate is not a forecast for any individual patient. They are still a strong reality check:
- Match the measure. Compare a clinic's claim only with a registry figure that uses the same denominator, age group, egg source and year.
- Prefer live births. The UK Competition and Markets Authority warns that "pregnancy rates" are highly likely to be unrealistic indicators of success, and that showing only a good year or a favourable patient group can mislead (CMA, 2021, updated 2026).
- Ask which registry a clinic reports to. A clinic that submits to a national or regional registry has accepted standard definitions. One that reports to none is asking you to trust its own.
- Treat large gaps as a question. If a clinic's figure far exceeds the registry benchmark for a similar group, ask how the patients were selected and what the denominator is. Don't assume the clinic is simply better.
- Check multiple births too. ICMART reported a 17.7% multiple delivery rate after fresh own-egg transfers worldwide in 2019. A clinic's twin rate is part of its results, not a footnote.
For how to turn these checks into a destination choice, see IVF success rates abroad: what the numbers mean.
Ready to Take Next Steps?
Have questions about how this applies to your own plan? Explore Colombia's fertility clinics and resources, or schedule a telemedicine consultation to discuss your specific situation.
Frequently Asked Questions
What is the average IVF success rate worldwide?
It depends on the measure. The ICMART report for 2019 gives a delivery rate of 22.9% per egg retrieval with own eggs, rising to 38.6% per retrieval when later frozen transfers are counted cumulatively. Per-transfer rates were around 31% to 32%. None of these is a personal prediction.
Why are PGT success rates per transfer so much higher?
Because they are counted only among transfers of tested embryos. Cycles where no embryo was suitable to transfer drop out of the denominator. A per-transfer PGT rate is not the chance of a baby from starting a cycle.
Can I compare countries using registry averages?
Only cautiously. Countries differ in which treatments are legal, how many patients use donor eggs or PGT, how complete their reporting is, and which patients clinics accept. Those differences can outweigh any real difference in quality.
How old is the latest international IVF data?
Several years. The most recent ICMART world report, published in 2025, covers 2019. National systems such as the CDC's also run more than a year behind because outcomes must be followed up first.
Will my outcome abroad be counted in any registry?
Possibly not completely. The ESHRE survey authors note that cross-border treatment is poorly tracked and outcome data are limited. If you give birth at home after treatment abroad, tell your overseas clinic so its records are complete.