A percentage needs a complete definition
A fertility-treatment success rate combines an outcome with a group of people or treatment events. Without both parts, two percentages can look comparable while answering different questions. A clinic-wide figure also needs a reporting period and enough detail about the patients included.
For instance, a positive blood pregnancy test occurs earlier than an ultrasound-confirmed clinical pregnancy. A live-birth delivery is a later outcome. A rate using the earliest endpoint will not describe the same result as a rate using the latest endpoint.
Start with the outcome
| Endpoint | What has been observed | What it does not yet show |
|---|---|---|
| Positive pregnancy test | A biochemical signal of pregnancy, usually measured through hCG. | An ultrasound-confirmed pregnancy or a live birth. |
| Clinical pregnancy | A pregnancy meeting the study or clinic’s clinical/ultrasound definition. | That the pregnancy will continue to birth. |
| Live-birth delivery | A delivery resulting in at least one live-born baby. | An individual patient’s chance without considering the population and denominator. |
| Cumulative live-birth outcome | Birth outcomes across a defined series of treatment attempts or transfers. | A universal result unless the starting point, follow-up and included attempts are specified. |
Scroll the table sideways to see all columns.
The same results can produce three different percentages
Consider an illustrative group of 100 treatment cycles started. Suppose 80 reach egg collection, 60 reach one embryo transfer, and 24 of those cycles result in a live-birth delivery. These invented numbers demonstrate the arithmetic; they are not Tang IVF Clinic results.
| Denominator | Calculation | Rate |
|---|---|---|
| Cycles started | 24 ÷ 100 | 24% |
| Egg collections | 24 ÷ 80 | 30% |
| Embryo transfers | 24 ÷ 60 | 40% |
Why the highest-looking number is not the whole picture
A per-transfer rate describes people who reached transfer. It leaves out cycles that stopped earlier or produced no embryo for transfer. That can make the number look more favourable than a rate measured from treatment start. Neither calculation is automatically wrong, but the denominator must be visible.
This is particularly relevant when comparing testing or embryo-selection approaches. A transfer group may become more selected while the chance of reaching transfer changes. The overall treatment experience includes both the chance of having an embryo to transfer and the outcome of that transfer.
Cumulative outcomes need a time window
One egg collection can lead to a fresh transfer and later frozen transfers. A cumulative figure may combine the outcomes from those embryos, or it may refer to several complete collection cycles. Those are different starting groups. The definition should identify the included attempts and how long patients were followed.
A cumulative rate cannot be obtained by simply adding individual transfer percentages. The same person may contribute more than one attempt, and people stop or continue treatment for different reasons. Published methods should explain how incomplete follow-up was handled.
Compare like with like
A small group can produce a striking percentage that changes substantially with just a few additional outcomes. Population differences can also matter more than a small difference between two headline rates. Statistics are most helpful when they explain the relevant group and the limits of comparison.
- Age at egg collection, and whether the eggs are the patient’s own or donated.
- Outcome definition, denominator and reporting years.
- Whether cancelled cycles and cycles with no transfer are included.
- Number of people or cycles in the group and the uncertainty around the estimate.
- Whether results are single-transfer, single-collection or cumulative outcomes.
