Three IVF doctors (reproductive endocrinologists), writing "as leaders of the Society for Assisted Reproductive Technology (SART)," say US IVF results hold steady at about 32% if one kind of cycle is left out: the cycles where embryos were frozen for later and none was transferred (Hill et al. 2026). Each of those cycles was still a full round of stimulation drugs and an egg retrieval for a woman. The same letter says SART's records labeled those cycles correctly only about 1 time in 10. The letter declares "no financial conflicts of interest," while CMS Open Payments lists $21,672.39 to its first author from 2023 to 2025, $20,796.01 of it from Ferring (Open Payments profile).
-
01Leaving out frozen-embryo cycles raises the rate
The SART leaders get to 32% by leaving some women's cycles out of the count. In an IVF cycle, a woman takes stimulation drugs, a doctor retrieves her eggs, and the lab makes embryos. The CDC counts a started cycle as "an ovarian stimulation cycle initiated with the intent to retrieve oocytes" (Gleicher et al. 2026). In a banking cycle, the embryos are frozen for later, and none is transferred in that cycle.
Gleicher and colleagues divided each year's US live births by every IVF cycle started that year. On CDC data, that rate fell from 29.1% in 2012 to 22.2% in 2021. Over the same years, banking cycles rose from 18,585 to 167,689 (Gleicher et al. 2026).
The SART leaders take the banking cycles out of the cycles they divide by. "When embryo banking cycles are excluded," they write, "ART cycle efficiency in the USA has been stable over the decade at around 32%" (Hill et al. 2026). The letter does not show the calculation.
Each dropped cycle was still a full round of stimulation drugs and an egg retrieval for a woman. In 2021, 167,689 of the 413,776 cycles started were banking cycles. Taking them out of the count makes the rate higher. What those women went through stays the same.
-
02SART's records logged most frozen-embryo cycles as failures
The letter says the national records got most banking cycles wrong. In 2014 to 2017 SART data, some cycles ended with embryos frozen and no transfer within a year. The registry recorded only about 1 in 10 of them as banking. In the letter's words, banking "was reported in roughly 10% of IVF cycles where embryo banking actually occurred." The rest went into the national figures as failed attempts (Hill et al. 2026).
The letter explains the gap through two definitions. The registry's label rested on what the letter calls "the precycle intention," the plan stated before the cycle began. The study the letter cites, a 2024 abstract by McCulloh and colleagues on which Hill is a co-author, labeled banking by what actually happened: embryos frozen, with no transfer attempted within a year (Hill et al. 2026).
The result is a national record that does not say what happened in most of these cycles. RRM Academy's ART Reporting Completeness (ARC) project measures this kind of gap: what each IVF registry publishes and what it leaves out. The SART leaders' letter states the gap itself: "In the CDC tables and the historic SART tables, embryo banking is grossly underreported compared to what actually happened." SART's 2021 national report also prints no figure for ovarian hyperstimulation syndrome (OHSS), ectopic pregnancy, procedure complications or maternal death (SART 2021 national report).
Gleicher's team agreed in reply that historical CDC and SART reporting structures "incompletely account for delayed transfers and prolonged embryo storage." Their reply closes by noting that the three SART leaders "conclude their critique of our paper by acknowledging important deficiencies in the CDC and SART databases" (Gleicher et al. reply).
-
03Her example: three retrievals, one baby so far
The letter's own example is one woman, three egg retrievals and one baby so far: "a 38-year-old patient wants three children and performs three IVF cycles to get six embryos. She then transfers the top embryo, which happens to be from Cycle 3, and has a live birth." The letter objects that the CDC tables count Cycles 1 and 2 as negative while their embryos sit in storage (Hill et al. 2026).
For the woman, the count is three rounds of stimulation drugs, three egg retrievals and one baby. Her other embryos remain frozen, and later transfers may add births. Gleicher's team wrote that "Patients experience those cycles as real treatment burden, cost, delay, and risk" (Gleicher et al. reply).
A per-patient count follows each woman from her first stimulation across all her cycles, and SART already prints one. In its 2021 national report, among women aged 38 to 40 using their own eggs, 42.6% of new patients had a live birth. The same report prints 26.0% live births per intended egg retrieval, counting all transfers within a year (SART 2021 national report). For women 38 to 40, the CDC's report for the same year lists an average of 4.3 intended retrievals per live-birth delivery, counting retrievals started in 2020 (CDC 2021 national report).
-
04Open Payments lists Ferring payments to Hill
The letter states: "The authors have no financial conflicts of interest." CMS Open Payments lists 32 records for Micah Hill from 2023 through 2025, totaling $21,672.39. Ferring accounts for $20,796.01 of it (Open Payments profile).
The records include $13,399.62 in Ferring Pharmaceuticals AS travel and lodging in Rome, Italy, dated September 23, 2025. They also include a $3,290 payment from Ferring Pharmaceuticals Inc. for speaking or faculty services on February 13, 2025, with Menopur listed as the product. Nine records name Menopur, Ferring's IVF stimulation drug.
The letter's references were last accessed in February 2026, so all 32 records fall within the 36 months before it was written. RRM Academy found no Open Payments records for Jennifer Kawwass. Open Payments lists $127 in meals for Jennifer Eaton.
Each banking cycle the letter defends is a stimulation cycle, and Menopur is a stimulation drug.
-
05Restorative care counts the couple
Restorative reproductive medicine (RRM) starts with the question IVF skips: why conception has not happened, in both partners. Many RRM cohort studies report outcomes per couple, followed through care. Of the IVF figures, the per-patient count is the closest match, and SART publishes it.
The letter argues over which cycles to leave out of the count. The woman in its example still went through three stimulations for one birth. For couples, the useful IVF number is the per-patient one, with the harms printed beside it.
What the paper establishes
The letter shows that three SART leaders read US IVF results as stable when cycles with frozen, untransferred embryos are left out of the count. The letter also says banking was "grossly underreported" in the CDC and historic SART tables, with most such cycles in 2014 to 2017 SART data counted as failures. The letter reports no new data. Its own example remains three stimulations and one baby so far, and its first author's Ferring payments sit beside a declaration of no financial conflicts.
05gq02987 
