The same patients, two success rates, one of them three times the other
This is the most important single idea in this part.
The CDC collects results from every reporting US fertility clinic and publishes them three ways. For women over 40 using their own eggs, in the 2022 reporting year, the same women produced these figures:
Both numbers are true. They describe the same women. The difference is that the second one deletes everyone whose cycle was cancelled or who never had an embryo to transfer.
CDC also publishes the statistic that explains the gap: the average number of transfers per intended egg retrieval is 1.0 at under 35, and 0.3 at over 40. Roughly seven in ten intended retrievals in women over 40 never reach a transfer at all.
| Own eggs, all US clinics, 2022 | Under 35 | 35 to 37 | 38 to 40 | Over 40 |
|---|---|---|---|---|
| Live birth per intended retrieval | 49.7% | 36.3% | 23.1% | 7.7% |
| Live birth per transfer | 49.1% | 46.0% | 41.2% | 25.4% |
| Transfers per intended retrieval | 1.0 | 0.8 | 0.6 | 0.3 |
| Intended retrievals per live birth | 2.0 | 2.8 | 4.3 | 12.9 |
Under 35 the two figures are almost identical, so the denominator barely matters. Over 40 it changes the answer by more than a factor of three. Advertising almost always uses the per-transfer figure.
Two more definitions worth knowing. These are cumulative rates, counting transfers that occur within a year of a retrieval. And a live-birth delivery is counted once, so a twin delivery counts as one, not two.