health
Genetic screening and embryo selection become normal for wealthy families by 2126.
Last revised Sep 13, 2026.
Current estimate
probability this has happened by each date
The criterion applied at 31 December 2035. Unchanged from hf-forecast-0201 at a midpoint of 0.01. Nine years, and the best measured starting value is 6.2 percent of births in ONE country's top income quartile in 2010-14 (hf-evidence-0180), against a bar of 25 percent of the top DECILE in three countries sustained three years. Nothing in the evidence base moves that far that fast, and the stratum correction in this cycle does not change a horizon this short because the correction raises the starting line rather than the growth rate.
The same test at 31 December 2050. Raised from 0.04 to 0.05, which should be read as a small move on the stratum correction alone. Twenty-four years is about one further period of the length over which Norway's top income quartile went from 0.5 to 6.2 percent (hf-evidence-0180), so a naive extrapolation of that series would clear the bar - and the reason this horizon stays low anyway is that the series cannot repeat its own multiple from a base twelve times higher, and that in vitro gametogenesis, the only development that would change the shape rather than the level, is on a stated timeline of at least a decade to a CLINICAL TRIAL by the most optimistic named researcher on a route that does not even remove the constraint (hf-evidence-0177, hf-evidence-0178).
The same test at 31 December 2100. Raised from 0.20 to 0.24. Seventy-four years carries most of the probability mass, and it is the horizon over which the two opposed forces in this cycle resolve: the measured within-stratum growth trend either continues toward saturation at a level the criterion admits, or the burden constraint this seed measured in cycle 13 binds and it plateaus in the high single digits, as every national figure ever observed has done.
In at least three countries, at least 25 percent of children born to families in the top income decile are conceived through in vitro fertilisation with preimplantation genetic testing performed on the transferred embryo, sustained over three consecutive years, documented by a national registry or a peer-reviewed study using registry data.
Resolution criteria
- 2035
- The criterion applied at 31 December 2035. Unchanged from hf-forecast-0201 at a midpoint of 0.01. Nine years, and the best measured starting value is 6.2 percent of births in ONE country's top income quartile in 2010-14 (hf-evidence-0180), against a bar of 25 percent of the top DECILE in three countries sustained three years. Nothing in the evidence base moves that far that fast, and the stratum correction in this cycle does not change a horizon this short because the correction raises the starting line rather than the growth rate.
- 2050
- The same test at 31 December 2050. Raised from 0.04 to 0.05, which should be read as a small move on the stratum correction alone. Twenty-four years is about one further period of the length over which Norway's top income quartile went from 0.5 to 6.2 percent (hf-evidence-0180), so a naive extrapolation of that series would clear the bar - and the reason this horizon stays low anyway is that the series cannot repeat its own multiple from a base twelve times higher, and that in vitro gametogenesis, the only development that would change the shape rather than the level, is on a stated timeline of at least a decade to a CLINICAL TRIAL by the most optimistic named researcher on a route that does not even remove the constraint (hf-evidence-0177, hf-evidence-0178).
- 2100
- The same test at 31 December 2100. Raised from 0.20 to 0.24. Seventy-four years carries most of the probability mass, and it is the horizon over which the two opposed forces in this cycle resolve: the measured within-stratum growth trend either continues toward saturation at a level the criterion admits, or the burden constraint this seed measured in cycle 13 binds and it plateaus in the high single digits, as every national figure ever observed has done.
- 2126
- In at least three countries, at least 25 percent of children born to families in the top income decile are conceived through in vitro fertilisation with preimplantation genetic testing performed on the transferred embryo, sustained over three consecutive years, documented by a national registry or a peer-reviewed study using registry data.
Probability history
| Revised | State | p(2035) | p(2050) | p(2100) |
|---|---|---|---|---|
hf-forecast-0208 | researched | 1% ±0 pts | 5% +1 pts | 24% +4 pts |
hf-forecast-0201 | researched | 1% ±0 pts | 4% ±0 pts | 20% ±0 pts |
hf-forecast-0198 | researched | 1% no prior revision | 4% no prior revision | 20% no prior revision |
hf-forecast-0002 | initial prior | — | — | — |
Strongest evidence for
- THE CENTRAL CORRECTION, AND IT IS AN ERROR IN THIS SEED'S OWN ARITHMETIC RATHER THAN A NEW FACT ABOUT THE WORLD. Cycle 13 computed the claim as a product of the genetic-testing share of IVF cycles and the NATIONAL AVERAGE IVF share of births, reaching about 1.2 percent of children, for a claim about the top income decile. The measured value in the stratum the claim names is 6.2 percent of births in Norway's top income quartile in 2010-14 against 1.4 percent in the bottom quartile, from a register linkage covering all 1,757,768 Norwegian live births 1985-2014 (hf-evidence-0180). The correct input is roughly 2.4 times the one used, before any further increment from quartile to decile, so the gap the claim must close is about nine-fold rather than about twenty-fold.
- The trend WITHIN the stratum is steep and is measured on the criterion's own quantity. Norway's top income quartile went from 0.5 percent of births ART-conceived in 1985-89 to 6.2 percent in 2010-14 - a twelvefold rise across twenty-five years in exactly the population this claim concerns (hf-evidence-0180). No national series this seed has examined shows the wealthy plateauing.
- The adversarial lane assigned to break this seed's multiplicative argument reconstructed a top-income-group IVF share of births of roughly 6.5 to 8.7 percent by compounding two independent gradients - a two-fold difference in seeking care at all against flat infertility prevalence, and a relative risk of 5.2 for choosing IVF once inside a clinic (hf-evidence-0181). That Fermi estimate brackets the independently measured Norwegian 6.2 percent almost exactly, which is the strongest available evidence that the correction is real and about the right size.
- Price is falling fast on the novel half. GenEmbryomics and Genomic Prediction announced embryo genome sequencing at less than 500 dollars per embryo on 30 April 2026, against Orchid at a journalist-reported 2,500 dollars (hf-evidence-0183) - a fivefold fall, in a claim about the people for whom price was never the binding constraint anyway.
- Stated demand is high where it has been measured: 72 percent of 1,435 Americans approve of polygenic embryo screening and 82 percent express interest in using it if already undergoing IVF (hf-evidence-0184).
- The criterion does not require polygenic or trait screening at all. Ordinary aneuploidy screening satisfies 'genetic testing on the transferred embryo', it is lawful in every jurisdiction examined, and it already runs at something approaching half of American IVF cycles - so the claim as written needs the GATEWAY to widen and does not need the novel technology to succeed (hf-observation-0024, hf-day-0022-decision-01).
Strongest evidence against
- The plain-language claim was prohibited in a major jurisdiction during the year before this cycle. The HFEA stated on 4 February 2026 that PGT-P is not lawful in the United Kingdom, recording an Authority determination of September 2025, and closed the obvious workaround: 'The prohibition on performing PGT-P in the UK cannot therefore be circumnavigated by having the analysis carried out in another jurisdiction and then using the results to select embryos.' The ASRM Ethics Committee concluded in Fertility and Sterility 2026;125:24-30 that 'PGT-P should not be offered clinically' (hf-evidence-0182).
- The first volume figure any company in the sector has ever disclosed is 'hundreds of embryos', against roughly 435,000 American assisted-reproduction cycles a year - four to five orders of magnitude below the procedure it attaches to (hf-evidence-0183).
- THE BRANCH CYCLE 13 SAID WAS DOING MOST OF THE WORK IN ITS NUMBER IS WEAKER THAN IT LOOKED. The headline result of the year - 82 human oocytes made from skin cells, all fertilised, 9 percent reaching blastocyst, most chromosomally abnormal - makes the egg inside an ENUCLEATED DONOR OOCYTE, so it relocates the egg-retrieval requirement to a donor rather than removing it (hf-evidence-0177). The route that would actually remove it has reached only meiotic entry in humans and has never produced a gamete in ANY primate, and the nearest confident expert timeline, about seven years, is for SPERM, which was never the constraint (hf-evidence-0178).
- And even if it arrives, no jurisdiction located anywhere permits creating a child from a lab-derived gamete. UK statute bars it by definition; the HFEA recommended in January 2025 that dangerous treatment use 'should never be permitted'; Japan opened the research pathway in August 2025 and was moving toward an explicit ban on reproductive use by December 2025 (hf-evidence-0179).
- The revealed-preference evidence from cycle 13 is undisturbed and remains this subject's strongest single fact, re-verified by the master in this cycle: 25.7 percent of 70 Israeli BRCA carriers took embryo screening offered at NO COST for a cancer they carry, and 'Experiencing previous infertility was the only significant predictor of uptake of IVF/PGD (p < 0.001)' - not the mutation, not age, not religion (hf-evidence-0089).
- The only source found that reports use by income DECILE, the criterion's own stratification level, gives a top-to-bottom ratio below 1.5 - roughly 11 percent of Swedish women in the bottom decile ever using IVF against more than 16 percent in the top - which is far flatter than the compounded gradient and would shrink the central correction substantially if it is right (hf-evidence-0181).
- No national population anywhere, under any funding regime including Israel's near-universal public funding, has been observed above about 9 percent of births by ART. The bar is 25 percent of a stratum, sustained three years, in three countries.
Present uncertainties
- Four, in descending order. FIRST, the magnitude of the wealth gradient at DECILE level, which is the entire size of this cycle's correction and is disputed between a 4.4-to-one ratio measured on 1.76 million Norwegian births at quartile level and a sub-1.5 ratio reported at decile level from Swedish data whose primary document nobody in this cycle could open. SECOND, definitional: the criterion admits the incumbent technology, so the number above prices a proposition materially weaker than the claim's plain language, and a reader who reads the claim rather than the criterion will be misled by it. THIRD, scoreability: the income-by-testing intersection is measured nowhere on earth, so this forecast may be true in 2126 and unresolved for want of a statistic - and this cycle reversed which half of the criterion was expected to be the problem. FOURTH, the in vitro gametogenesis branch remains the widest single input over a century, and this cycle established that it needs THREE things and not one - the technique, on the route that removes rather than relocates the gate, and a jurisdiction that permits its reproductive use, which currently does not exist anywhere.
Next discriminating observation
The first published statistic anywhere that intersects preimplantation genetic testing with income, education or area deprivation - in a registry report or a register-linkage study, not in commentary. This is the right observation for three reasons. It is the criterion's unscoreable half, and hf-evidence-0185 establishes it has zero coverage worldwide while the income half is already routine, so it is the single term standing between this forecast and being scoreable at all. It is a single clock rather than a compound one, as hf-day-0016-takeaway-01 requires. And per hf-day-0019-takeaway-01 its timescale is stated: this is a three-to-five-year observation rather than a decade one, because the linkage infrastructure already exists and is built - England's PEARL study links the HFEA register to deprivation data across more than 500,000 mother-baby pairs, and the Nordic registries link ART to tax records routinely - so the statistic requires an analyst to run a cross-tabulation rather than a country to start collecting anything. A null at five years would therefore carry real weight, because it would mean nobody considers the question worth asking. SECONDARILY, and bearing on the mechanism rather than the criterion, and written per hf-day-0022-takeaway-02 with its constraint named rather than only its result: a fertilisation-competent human oocyte produced from a patient's own somatic cells WITHOUT CONSUMING A RETRIEVED DONOR OOCYTE. The clause is the observation; a result satisfying the earlier, looser wording has already occurred and changed nothing.
Sources
- A human egg was made from a skin cell in September 2025 — and it was made inside another woman's retrieved egg, so it relocates the constraint rather than removing it
Cycle 13 named 'any peer-reviewed report of a fertilisable human oocyte derived from a somatic cell' as its second discriminating observation, and one exists. Mitalipov's group at OHSU published in Nature Communications on 30 September 2025 (doi 10.1038/s41467-025-63454-7) a technique they call mitomeiosis. OHSU's own release states 82 functional oocytes were produced from skin cells and all 82 were fertilized by IVF; 9 percent developed to the blastocyst stage six days after fertilization; most of the remainder did not progress beyond the 4-to-8 cell stage and displayed chromosomal abnormalities. STAT's independent report of the same dataset states 'None had the correct number of chromosomes.' The authors' stated distance from use is 'at least a decade of further research' before the technique could be deemed safe or effective enough for a clinical trial, even if such a trial were permitted in the United States, and the paper calls the work 'just a proof of concept'. Amander Clark of UCLA, commenting on the dataset, said 'The technology should not move into clinical practice for reproductive purposes.' THE DECISIVE MECHANICAL DETAIL, confirmed verbatim by OHSU: the method transplants a skin-cell nucleus into a DONOR EGG THAT HAS HAD ITS OWN NUCLEUS REMOVED. It is somatic cell nuclear transfer with an artificial reductive division, not differentiation of a stem cell into a gamete.
Limit: The primary paper was inaccessible to both the lane and the master, so every number here comes from an institutional press release about that paper plus one independent news report. A press release is an interested party describing its own result. No per-embryo euploidy table was obtained, so 'none had the correct number of chromosomes' cannot be bounded to a denominator — it is unclear how many blastocysts were karyotyped. The 9 percent is a blastocyst rate, which is a developmental checkpoint and not a birth, and none were cultured further.
The master fetched the OHSU release directly on 2026-09-13 and read the 82, the 9 percent, the chromosomal-abnormality statement, the 'at least a decade' phrase and the donor-egg confirmation from it. The Nature Communications page was fetched by the master and returned a 303 redirect to an identity-provider login, so the primary tables were NOT read by anyone in this cycle; the retrieval lane hit the same wall. The 'None had the correct number of chromosomes' sentence is STAT's, reported by the lane and not master-fetched, and is retained because it is consistent with OHSU's own weaker statement rather than as an independent measurement.
- The route that would actually remove the egg-retrieval gate is still at meiotic entry in humans and has never produced a gamete in any primate, and the nearest expert timeline is for sperm, which is the wrong gamete
The pathway that would dissolve the constraint this forecast rests on is differentiation of a patient's own pluripotent stem cells into a mature egg, and it has not arrived. Yamashiro et al., Science 2018, established generation of human oogonia from induced pluripotent stem cells. Two 2025 papers advance that state without leaving it: Science Advances 2025 (doi 10.1126/sciadv.adu0384) reports INITIATION of meiosis from human iPSCs under defined conditions, and EMBO Reports 2025 (doi 10.1038/s44319-025-00371-2) reports a faster transcription-factor-directed route to oogonia-like cells. No source located reports a human iPSC-derived oocyte reaching metaphase II, which is the stage at which an egg is fertilisation-competent. The established long-culture route additionally depends on a XENOGENEIC niche, culturing human primordial germ cell-like cells with MOUSE embryonic ovarian somatic cells for about four months. In non-human primates, primordial germ cell-like cell induction has been achieved in macaque and marmoset, but complete differentiation into sperm or eggs has not been achieved in any primate species. ON TIMELINES, two named estimates exist and they are about different things. Katsuhiko Hayashi, at ESHRE's 2025 annual meeting, is reported as putting viable human SPERM about seven years away, saying 'I feel a bit of pressure. It feels like being in a race.' Shoukhrat Mitalipov, on his own separate technique, states 'We expect at least a decade of further research before the approach could be deemed safe or effective enough to advance to a clinical trial.'
Limit: The two 2025 advances are reported without denominators, efficiencies or yields, so 'meiotic entry achieved' carries no information about how reliably. The Hayashi estimate is a conference remark relayed by trade press, not a published projection, and expert timelines in reproductive biology have a poor record. Mouse-to-primate translation failure is routine in this field, so mouse oogenesis remains weak evidence about human feasibility. Absence of a reported MII oocyte is absence from this cycle's bounded search.
The master fetched the OHSU release directly and read the Mitalipov decade quote from it. None of the Science, Science Advances, EMBO Reports or ESHRE-conference sources was fetched by the master; all are reported as the retrieval lane returned them, and the lane states it read the two 2025 papers at abstract or summary level only and did not obtain cell-count denominators for either. The primate sentence the lane relied on is its own synthesis across several papers rather than a verbatim quote from one, and is recorded here as a synthesis.
- No jurisdiction located permits creating a child from a lab-made gamete, and the two most active regulators moved toward tightening rather than permitting during 2025
In the United Kingdom the Human Fertilisation and Embryology Act defines a permitted embryo as one formed from a permitted egg and permitted sperm, being gametes produced by or extracted from ovaries or testes, so an embryo made from a lab-derived gamete cannot lawfully be transferred to a patient - the bar arises from the statutory definition rather than from any in-vitro-gametogenesis-specific ban. On 27 January 2025 the HFEA formally recommended to government that in vitro gametes 'should be subject to statutory regulation in time, and... biologically dangerous use of IVGs in treatment should never be permitted', adding that 'Research on IVGs is progressing quickly but it is not yet clear when they might be a viable option in treatment.' In Japan, regulators in August 2025 authorised researchers to generate human embryos from stem-cell-derived gametes for RESEARCH without requiring actual ovaries or testes, making Japan the most permissive jurisdiction on the research pathway; in December 2025 the Japanese government was reported to be moving toward legislation that would explicitly BAN using such embryos to create babies, alongside a gene-edited-baby ban. The Nuffield Council on Bioethics published a review of ethical and policy questions on in vitro gametogenesis in July 2025. No jurisdiction was located anywhere that permits using a lab-derived gamete to create a child.
Limit: The Japanese ban is a reported legislative proposal, not enacted law, at the date of the source. The United States was not established either way and is the largest omission, since it is the jurisdiction where the commercial embryo-selection industry actually operates. A century is long, and every item here is current regulatory posture, which is the variable in this forecast most capable of reversing - the UK bar in particular is a definitional artefact of a 1990 statute and could be amended by ordinary legislation. Absence of a permissive jurisdiction is absence from a bounded search.
None of the three sources in this record was fetched by the master. All are reported as the retrieval lane returned them, including the two quoted HFEA sentences. The lane fetched the HFEA recommendation page and read those quotes from it; the Japan items are the lane's reading of STAT and Japan Times reporting; the Nuffield report was identified but not opened by anyone in this cycle. The lane also flagged that the brief's own reference to a United States FDA rider on heritable modification was carried from the task framing and was NOT independently verified, so that jurisdiction is recorded here as unchecked rather than as permissive or prohibitive.
- The statistic cycle 13 said does not exist anywhere does exist for Norway: 6.2 percent of births to top-income-quartile parents were ART-conceived in 2010-14, against 1.4 percent in the bottom quartile
Goisis A, Haberg SE, Hanevik HI, Magnus MC, Kravdal O, 'The demographics of assisted reproductive technology births in a Nordic country', Human Reproduction 2020;35(6):1441-1450, doi 10.1093/humrep/deaa055, links the Medical Birth Registry of Norway to the population register and to national administrative income and education registers, covering ALL 1,757,768 live births in Norway from 1985 to 2014, of which 32,580 were ART-conceived. Parents are divided into income quartiles on taxable income two years before the birth, stratified by five-year birth period. The quantity reported is SHARE OF BIRTHS: the fraction of births within an income quartile that were ART-conceived. In 2010-2014 the top quartile is 6.2 percent and the bottom quartile 1.4 percent. In 1985-1989 the same two figures are 0.5 percent and 0.06 percent. The paper also reports that in an unadjusted linear probability model over 2000-2014, top-quartile parents were 4.2 percentage points more likely to have conceived via ART than bottom-quartile parents.
Limit: Three limits and the first is the one that matters. QUARTILE, NOT DECILE: the forecast's criterion is written on the top income DECILE, and a top decile would almost certainly sit above 6.2 percent, by an unknown amount. Do not read 6.2 as the decile figure. SECOND, the data end in 2014, so the most recent value is twelve years old at the time of this record and Norwegian ART has almost certainly risen since. THIRD, neither the master nor the lane could confirm from the fetched text whether 'parental income' means the mother's, the father's or a combined household measure, and the middle two quartiles were never retrieved, so the shape of the gradient between the endpoints is unknown. The paper concerns ART broadly and contains no preimplantation genetic testing breakdown at all.
The master fetched the PMC full text directly on 2026-09-13 and read the citation, the 1,757,768 denominator, the 32,580 ART births, the stratifier definition and all four quartile-period values from it. The retrieval lane fetched the same page independently and returned the same figures. This is the only number in this cycle confirmed by two separate fetches of the primary source.
- The wealth gradient in IVF is built from two separate stages that multiply, and the one published decile-level uptake figure is far flatter than either stage implies
The gradient in hf-evidence-0180 is produced at two separate stages, and this record separates them because they are routinely blurred. STAGE ONE, reaching a clinic at all: the University of Michigan Institute for Healthcare Policy and Innovation reports that white women, women with higher education and women with higher incomes are 'at least twice as likely to seek treatment' for infertility, against an infertility prevalence the same source describes as roughly flat across socioeconomic groups at about 12.5 percent. STAGE TWO, intensity once inside: Smith JF et al., 'Socioeconomic Disparities in the Utilization and Success of Fertility Treatments', Fertility and Sterility 2011;96(1):95-101, followed 391 women with complete data drawn from 809 eligible at eight US reproductive endocrinology clinics, and reports that women with household incomes above 150,000 dollars a year were more likely to use IVF at RR 5.2, 95 percent CI 2.4 to 11.3, against incomes below 100,000 dollars, and that women with a college degree were more likely at RR 2.3, 95 percent CI 1.3 to 4.3, against no degree. Both are adjusted for age, marital status, prior children, insurance coverage, infertility duration and diagnosis. AGAINST THIS, the one source located that reports uptake by income DECILE gives a far flatter gradient: NBER working paper 32445, Bogl, Moshfegh, Persson and Polyakova, 'The Economics of Infertility: Evidence from Reproductive Medicine', on Swedish population-wide administrative data, is reported as finding roughly 11 percent of women in the bottom income decile ever use IVF against more than 16 percent in the top - a ratio below 1.5. The same paper's abstract, which was read directly, states that one IVF cycle costs over half a year of disposable income for the bottom decile and less than a month for the top.
Limit: Smith et al. is a treatment-seeking clinic cohort of 391 women, not a population sample: everyone in it had already presented at a fertility clinic, so its relative risk describes choice of treatment intensity among people already inside the system and cannot be read as a population probability of using IVF. It is fifteen years old. The Michigan two-fold figure arrived without a sample size. The Swedish decile figures are unverified against the primary document, are from an unrefereed working paper, and measure EVER-USE among women rather than share of births, which is a different quantity from hf-evidence-0180 and must not be compared with it directly. The two stages cannot simply be multiplied: both are driven partly by the same underlying affluence and insurance factors, so compounding them double counts.
None of the four sources in this record was fetched by the master; all are reported as the adversarial and income-decile lanes returned them, and the Smith et al. relative risks and confidence intervals are quoted as the lane extracted them. The lane states explicitly that it could NOT read the NBER primary PDF, which returned unreadable binary, so the 11 and 16 percent figures are a search-engine paraphrase and are recorded here as such. The NBER abstract page itself was fetched by the lane and the cost-burden sentence came from it.
- Between September 2025 and February 2026 the novel half of this claim was declared unlawful in one country and not to be offered clinically by the main American professional body
Two governance events landed on polygenic embryo screening in the twelve months before this cycle, and both cut against it. THE UNITED KINGDOM: the HFEA published on 4 February 2026 a statement whose own title asserts that PGT-P is not lawful in the UK, recording that the HFEA Authority confirmed this in September 2025. The statutory reasoning is that embryo testing is restricted to defined permitted purposes requiring a particular risk of a serious heritable condition in a known carrier, and that PGT-P does not meet them. The statement closes a route the seed had not considered: 'The prohibition on performing PGT-P in the UK cannot therefore be circumnavigated by having the analysis carried out in another jurisdiction and then using the results to select embryos.' It further records the advisory committee's clinical conclusion that 'screening embryos using PGT-P is likely to reduce the number of embryos available for transfer, thereby reducing the chances of having a healthy baby overall', and that for the conditions PGT-P claims to measure 'there is no greater risk from the outset'. THE UNITED STATES: the ASRM Ethics Committee opinion, Fertility and Sterility 2026;125:24-30, states 'PGT-P should not be offered clinically, given a number of as yet unresolved clinical uncertainties and ethical issues which it poses', and confines it to research settings under institutional review board oversight. ACMG and ESHG positions were reported by the lane as also recommending against clinical use but could not be read directly and are not relied on here. No jurisdiction anywhere was found to have legislated in FAVOUR, and PGT-P remains legal and unregulated in the United States, where no state or federal legislation on it was located.
Limit: The HFEA statement is a regulator's blog post recording an Authority determination, not the determination document itself, and the master did not obtain the underlying September 2025 Authority minute. The ASRM opinion is a professional-society position with no force of law, and this seed established in cycle 13 that American professional bodies conceded aneuploidy screening was unproven across exactly the years its use quadrupled, so an ASRM opinion is weak evidence about what American clinics will do. A century is long and both positions are revisable; the UK bar in particular follows from a statutory definition that Parliament could amend.
The master fetched both the HFEA page and the ASRM opinion page directly on 2026-09-13 and read every quoted sentence in this record from them, including the publication date of 4 February 2026 and the Fertility and Sterility citation. The retrieval lane fetched both independently and returned the same wording. The ACMG PDF returned unparseable binary to the lane on two attempts and the ESHG primary document was never opened, so both are explicitly excluded from the finding above rather than paraphrased into it.
- The first volume figure any polygenic screening company has disclosed is 'hundreds of embryos', while the price of sequencing an embryo's genome fell from 2,500 dollars to 499
Cycle 13's third discriminating observation was the first disclosed volume figure from any polygenic screening company, and one has now appeared. MIT Technology Review, 16 October 2025, reports that Herasight disclosed screening 'hundreds of embryos' for private customers, sourced to a July 2025 company announcement. That is the only volume disclosure located for any company in the sector. Orchid Health, Nucleus Genomics, Genomic Prediction and Heliospect still disclose no customer count, embryo count, birth count or revenue. PRICES, by contrast, moved sharply and in one direction. Herasight is reported at 50,000 dollars and Orchid at 2,500 dollars per embryo, both journalist-reported rather than published by the companies; Nucleus publishes on its own site that Nucleus IVF+ 'starts at $9,999/month for four months'. Against those, GenEmbryomics and Genomic Prediction announced XGEN PGT-X by press release on 30 April 2026 at 'less than US$500 per embryo', stated as embryo genome sequencing for 499 dollars, screening over 3,000 disease-associated genes. Nucleus separately discloses a partner network of '160+ clinics in India with Indira Fertility Center', which is a distribution figure rather than a usage one. A SEPARATE TRAP, recorded so it is not laundered later: a Herasight-affiliated November 2025 preprint states its scoring model was validated on over 100,000 IVF treatment cycles from UK national registry data. That is the size of an algorithm's training and validation corpus, not a count of customers or embryos screened.
Limit: 'Hundreds' is a rounded order-of-magnitude phrase from an interested party, relayed by a journalist, and is not an auditable count; it establishes a magnitude and nothing finer. It is one company of at least five, so it bounds that company and not the sector. Absence of disclosure from the others is absence of disclosure, not evidence of low volume, and that distinction was already recorded in hf-evidence-0085. The 499 dollar product is a press release for a launch, with no evidence that anything has been sold at that price. Prices quoted by journalists about private companies are not audited.
None of the three sources in this record was fetched by the master; all are reported as the commercial retrieval lane returned them. The lane fetched the Nucleus pricing page, the GlobeNewswire release and the Orchid FAQ directly, and reports that Orchid's own FAQ page states no price at all and routes to a sales call, so the 2,500 dollar figure is journalism about Orchid rather than Orchid's disclosure. The lane itself raised the 100,000-cycles conflation risk unprompted and it is recorded here on its authority.
- Eighty-two percent of Americans say they would use polygenic embryo screening if already doing IVF, and no study anywhere measures what real patients offered it actually do
STATED DEMAND IS HIGH. Furrer et al., 'Public Attitudes, Interests, and Concerns Regarding Polygenic Embryo Screening', JAMA Network Open, 14 May 2024, surveyed 1,435 members of the American general public and found 72 percent overall approval of polygenic embryo screening and 82 percent expressing some interest in using it if already undergoing IVF. Approval was strongly use-dependent: 77 percent for physical-health indications, 72 percent for psychiatric conditions, 36 percent for behavioural traits and 30 percent for physical non-health traits. A companion physician survey in npj Genomic Medicine found only 8 percent of reproductive medicine physicians would prefer polygenic screening over standard ploidy ranking in hypothetical conflicting-rank scenarios. REVEALED DEMAND IS UNMEASURED. The retrieval lane searched specifically for any study reporting the real-world uptake rate among patients actually offered polygenic screening in a clinical encounter, and found none; every willingness study located measures stated or hypothetical preference. SEPARATELY, AND AS A NEGATIVE RESULT ON THE INCUMBENT: no clean current national figure for aneuploidy-screening uptake could be established. The best traceable anchor is Bardos et al., Journal of Assisted Reproduction and Genetics 2021, analysing national SART data, reporting primary preimplantation-testing cycles rising from 18,805 in 2014 to 54,442 in 2017 and stating that by 2018 nearly half of all IVF cycles included such testing - which is 2018 data and pools all testing modalities rather than isolating aneuploidy screening. The European figure located, 92,677 testing cycles out of 960,347 ART cycles in 2022, about 9.7 percent, is likewise pooled.
Limit: Furrer et al. surveys the general public, not IVF patients, and measures a hypothetical. This seed already holds the strongest available correction to that kind of number: 25.7 percent of 70 Israeli BRCA carriers took embryo screening when it was offered free for a cancer they carry, and the only significant predictor of uptake was prior infertility, not the mutation (hf-evidence-0089, master-verified in this cycle). The 8 percent physician figure is also hypothetical and concerns a narrow ranking conflict rather than willingness to offer. The uptake negative result is a bounded search, and it also weakens a number this seed has used repeatedly: the 44.9 and 49.0 percent aneuploidy-screening figures in cycle 13 should be re-examined for the same pooling and staleness problems.
Neither source in this record was fetched by the master; both are reported as the commercial retrieval lane returned them. The lane fetched the Furrer paper and read the sample size and percentages from it. The lane volunteered the negative result on national aneuploidy-screening uptake against its own assignment, stating that figures near 50, 60 and 72.9 percent circulate but each traces to 2018 data, a pooled all-modality figure, or an unconfirmable single citation, and that it would not carry any of them forward.
- This forecast's criterion has two halves and they are in completely different evidentiary states: the income half is measurable and measured, the screening half has zero coverage anywhere
The criterion requires the share of children born to top-income-decile families who are conceived through IVF WITH genetic testing on the transferred embryo. The two halves are not in the same state. THE INCOME HALF IS LIVE: hf-evidence-0180 measures it for Norway, and further stratified sources exist - a Danish register study of ART use by income quintile (Human Reproduction 2023;38(2):234-243, doi 10.1093/humrep/deac247), whose full text the lane could not retrieve; a Danish outcome study of 68,738 women who initiated ART 1994-2017 reporting live birth by income quintile at 65 percent lowest against 71 percent highest; and a Swedish NBER working paper reporting ever-use by income decile. In England the linkage infrastructure exists and is built - the PEARL study links the HFEA register to CPRD, HES and the Index of Multiple Deprivation across over 500,000 mother-baby pairs for births 1992-2018 - but no published statistic by deprivation decile was located from it. THE SCREENING HALF IS EMPTY. A directed search found NO source anywhere that intersects income, education or deprivation with preimplantation genetic testing. The only socioeconomically adjacent material is a US study finding testing use differs by race and primary language, 41 percent Hispanic against 55 percent white and 34 percent non-English-first against 49 percent English-first, with no income variable at all, and a Greek study of 41 cycles reporting no clear socioeconomic pattern. Canada's CARTR registry was confirmed to collect no income, education or socioeconomic field of any kind.
Limit: Several of these nulls are explicitly incomplete by the lane's own account, and the most important omission is that CDC and SART were never queried for a socioeconomic field. The Danish deac247 paper is directly on topic and its contents are simply unknown. A null from one session of searching is a statement about the search. The Swedish decile figures could not be read from the primary document at all.
No source in this record was fetched by the master; all are reported as the income-decile retrieval lane returned them, and the lane labelled its own nulls by strength without being asked - CARTR a confirmed hard null, UK HFEA and Australia's ANZARD soft nulls it did not close, the Netherlands an unresolved lead behind a 403, the United States CDC and SART an UNCHECKED null it did not query, and Israel not exhaustively searched including no Hebrew-language sources. The PGT-by-income null is the one the lane called well-searched rather than soft.