PGT-A and ICSI are worth the extra cost only when they match a specific fertility history. Hard evidence supports PGT-A after recurrent pregnancy loss, a prior aneuploid embryo, or advanced maternal age, and ICSI for severe male factor or previous total fertilization failure. For younger patients in an initial IVF cycle, randomized trials show no live-birth gain from either add-on, and routine ICSI can even lower fertilization quality.
Most US fertility clinics now present a menu of add-ons before a cycle begins: genetic testing of embryos, intracytoplasmic sperm injection, time-lapse imaging, endometrial receptivity arrays, immune protocols. The menu looks uniform across clinics. The evidence behind each item is not, and the deciding variable is rarely the clinic's reputation. It is the patient's own reproductive history, written down in the chart: how many losses, how many failed transfers, what the semen analysis actually showed. Reading that history correctly is where the money either buys something real or buys reassurance.
Which fertility histories give PGT-A hard evidence, and which ones don't?
PGT-A (preimplantation genetic testing for aneuploidy) costs roughly $1,500 to $5,000 per cycle in the US, depending on the embryology lab and the number of embryos biopsied, with exact pricing published by each clinic. That money is well spent in three situations, and the American Society for Reproductive Medicine (ASRM) said so explicitly in its 2023 committee opinion on PGT-A, published in Fertility and Sterility.
The three histories with genuine supporting evidence
Recurrent pregnancy loss. ASRM's 2020 guideline defines recurrent pregnancy loss as two or more clinical pregnancy losses. When embryos from these patients are tested, aneuploidy rates run higher than age-matched averages, and a chromosomally abnormal embryo that reaches the uterus produces exactly the kind of early loss these patients have already endured. Selecting euploid embryos for transfer interrupts that pattern. Among all the proposed uses of PGT-A, this one carries the strongest clinical logic and the most consistent observational support.
A prior aneuploid embryo or affected pregnancy. A previous trisomic miscarriage, or a PGT-A report from an earlier cycle showing aneuploid embryos, shifts the probability estimate for the next cycle. Testing has a defined target to look for.
Advanced maternal age. The proportion of aneuploid eggs rises steeply through the late thirties. Here the evidence is real but narrower than most marketing suggests, and it is worth reading the actual trial numbers.
A randomized trial published in the New England Journal of Medicine in 2021 (Yan and colleagues, the ESTHEM trial) compared PGT-A against standard morphologic embryo selection and found no significant difference in live-birth rates per treatment cycle. An earlier multicenter randomized trial led by Munn?? and colleagues, also in the New England Journal of Medicine in 2019, found that PGT-A lowered the miscarriage rate modestly in women aged 24 to 37, while cumulative ongoing pregnancy rates across all embryos from one retrieval stayed essentially the same. Put plainly: in these trials PGT-A reduced bad outcomes per pregnancy without producing more babies per egg retrieval.
Why an initial cycle at 28 can end with fewer usable embryos after PGT-A
This is the most underestimated point in the entire add-on conversation. PGT-A does not improve embryos. It sorts them. In a young patient producing a healthy number of embryos, sorting adds little, because the natural selection process of implantation is already efficient at that age. What testing adds is a new set of risks that operate in only one direction.
First, the error rate. A biopsy removes five to ten cells from the trophectoderm, the outer layer that later becomes the placenta. The embryo's inner cell mass, which becomes the fetus, is not sampled. Technical limitations of next-generation sequencing, plus the sampling mismatch, produce false positives in a few percent of cases. For a patient with three embryos, one false positive can eliminate a third of her transferable inventory, and that embryo will never be proven wrong because it sits frozen, labeled abnormal, or discarded.
Second, the arithmetic of fewer embryos. A 28-year-old with two embryos faces a different equation than a 38-year-old with eight. Testing two embryos costs the same per-embryo fee as testing eight, consumes both embryos' time in the lab, and offers a much smaller pool from which to find a euploid result. If one of the two is called abnormal incorrectly, the patient has halved her chances on the strength of a test whose errors cluster exactly at the margins.
Third, the freeze-all requirement. PGT-A requires vitrifying every embryo and waiting for results, which converts the cycle into a frozen transfer plan by default. That is workable for most patients, but it adds a month of time, additional storage and thaw fees, and it removes the fresh transfer option from the table entirely.
The UK regulator, the Human Fertilisation and Embryology Authority (HFEA), maintains a public traffic-light rating of IVF add-ons based on evidence strength. PGT-A sits at amber, meaning the evidence of benefit is unclear for general use. Patients can verify that rating themselves on the HFEA website, and it aligns with the ASRM 2023 position that routine PGT-A for all IVF patients is not supported.
A practical rule: if you are under 35, in your initial cycle, with no pregnancy losses and no prior abnormal embryo, ask the clinic what specific problem PGT-A would solve in your case. If the answer is general reassurance, the $2,000 to $5,000 buys information you will likely never act on.
Why can one mosaic embryo be "transferable" at one lab and discarded at another?
Even patients who accept the evidence for PGT-A rarely learn about its messiest output: the mosaic result. Understanding this single issue changes how you should interrogate any clinic before paying for testing.
An embryo sometimes carries two genetically distinct cell lines, one normal and one with a chromosomal gain or loss, arising from a mitotic error after fertilization. When the biopsy samples five to ten cells from the trophectoderm, next-generation sequencing reads the mix of DNA and reports an intermediate copy-number signal, typically in the 20% to 80% abnormality range. That signal is real. Its interpretation is where laboratories diverge.
Each embryology lab sets its own reporting thresholds. One lab reports mosaicism detected at 20% abnormal cells and above. Another uses a 30% floor. A third discards anything at 40% or higher as unreportable. Some labs separate whole-chromosome mosaicism from segmental mosaicism (a deleted or duplicated piece of a chromosome) and handle them differently; others lump them. Some platform-specific differences matter too, because the analytical pipelines behind different commercial testing platforms apply different smoothing and cutoff algorithms to the same raw sequencing data.
The consequence is concrete. An embryo with a 45% monosomy 16 signal may be reported by one lab as a low-level mosaic eligible for transfer, with published series behind that policy. The identical biopsy data interpreted under a stricter lab's rules may produce a "high-level mosaic, not recommended" report, or no report at all. Patients have paid for a second round of testing, or traveled to a different clinic, over an embryo whose biology never changed. The Preimplantation Genetic Diagnosis International Society (PGDIS) addressed this directly in its 2019 position statement and its 2022 update, which recommend that labs report mosaic results transparently, distinguish low-level from high-level mosaics, and prioritize low-level mosaics for transfer after euploid embryos are exhausted. Not every lab has adopted the full framework.
What to ask the lab before you pay for PGT-A
These questions are answerable in writing, and a clinic that handles the conversation well is telling you something about its overall practice:
- What percentage cutoff defines a mosaic result in your lab, and what percentage defines "high-level" versus "low-level"?
- Do you report segmental mosaicism separately from whole-chromosome mosaicism, and do you transfer embryos with segmental-only findings?
- What is your transfer priority policy for mosaics? Standard practice places them after euploid embryos, but some labs refuse high-level mosaics outright.
- Do you discard embryos called mosaic, or do you freeze them so the patient can seek a second interpretation?
- What are your published outcomes for mosaic transfers, and what prenatal diagnostic follow-up (chorionic villus sampling or amniocentesis) do you require or recommend?
The last point deserves emphasis. Published series show that mosaic embryos, especially low-level ones, still produce healthy live births at meaningful rates, generally below euploid transfers but far from zero. Many such pregnancies resolve to normal fetuses, likely because the abnormal cell line is concentrated in the placenta or gets outcompeted. Because of the residual uncertainty, professional guidance consistently recommends prenatal diagnostic testing after any mosaic transfer. A clinic that transfers mosaics without mentioning amniocentesis is skipping a step that the evidence base considers standard.
How did ICSI shift from a rescue tool to a default setting?
Intracytoplasmic sperm injection was developed for a narrow, desperate problem. Since the first ICSI birth in 1992, the technique has allowed men with severely abnormal semen parameters, or sperm obtained surgically, to achieve fertilization that conventional IVF could not. A single sperm is selected under the microscope and injected directly into the egg, bypassing every natural barrier. For that population, ICSI is transformative and non-optional.
Walk into many US clinics today and ICSI will be proposed for couples with entirely normal semen analyses. The add-on typically runs $1,000 to $2,500 per cycle. The ASRM Practice Committee addressed this trend in its 2024 committee opinion on ICSI for non-male-factor infertility, concluding that the evidence does not support routine ICSI when male factor is absent, and that fertilization rates with conventional IVF are comparable in these patients. Some randomized data even show conventional insemination producing equal or better fertilization when sperm quality is normal, since ICSI injects one sperm chosen by an embryologist in place of thousands of sperm competing naturally.
The commercial logic behind the default
Why does the practice persist? The incentive structure explains it better than any medical rationale. Total fertilization failure, where conventional insemination yields no viable embryos, is a catastrophic outcome for a patient who has paid $12,000 to $20,000 for a retrieval cycle. ICSI reduces that risk to near zero. From the clinic's perspective, universal ICSI is insurance. From the patient's perspective, the patient is paying the premium on that insurance, every cycle, regardless of whether the risk applies to her.
Standardization reinforces the pattern. A lab that performs ICSI on every case runs one protocol, trains embryologists on one workflow, and never has to explain a fertilization failure. The billing system completes the loop: ICSI appears as a line item, revenue appears with it, and no one is required to justify the indication in the chart.
There is also a long-term data gap that patients should weigh. ICSI bypasses the natural sperm selection process, and for couples with severe male factor, the underlying cause of the sperm abnormality may itself carry genetic implications. Published follow-up studies of ICSI-conceived children are generally reassuring so far, but those cohorts are still young, and much of the observed risk elevation in existing studies concentrates in severe male-factor populations, where the infertility itself may be the driver. Extending the technique to couples with no male factor adds exposure without a demonstrated benefit to offset it. The HFEA rates ICSI for non-male-factor infertility at amber on its add-on list, consistent with the ASRM position.
When ICSI genuinely changes outcomes
The legitimate indications are specific and checkable against your own chart:
- Severe male factor: markedly low sperm concentration, poor motility, or severe teratozoospermia on a repeated semen analysis, confirmed across at least two samples.
- Surgically retrieved sperm: sperm obtained from the epididymis or testicle cannot fertilize conventionally at reliable rates.
- Prior fertilization failure: a previous conventional IVF cycle with zero or near-zero normal fertilization (few or no two-pronuclear zygotes) is the classic rescue indication.
- Vitrified oocytes: freezing alters the egg's outer membrane, and ICSI improves fertilization consistency with previously frozen eggs.
- Preimplantation genetic testing for monogenic disease: ICSI reduces sperm DNA contamination on the biopsy, protecting the accuracy of the genetic result.
For borderline male factor, a middle path exists and is worth requesting. Some clinics offer split insemination, dividing the retrieved eggs between conventional IVF and ICSI, or hold back a cohort of eggs for next-day rescue ICSI if conventional fertilization underperforms. Rescue ICSI outcomes run below planned ICSI, so this approach suits patients with adequate egg numbers. A clinic that refuses to discuss split insemination while charging for universal ICSI is making a business decision, and patients deserve to hear it framed that way.
How do major US clinics handle add-on indications, and what can you verify yourself?
Clinic choice matters less for whether add-ons exist than for how they are prescribed. Large networks differ in their defaults, and all of the following publish cycle-level outcome data through the Society for Assisted Reproductive Technology (SART) and the CDC's annual ART surveillance reports, both of which are searchable by patients at no cost.
Shady Grove Fertility operates more than 40 locations across the US, anchored in Maryland. Its scale means standardized protocols and published per-clinic SART data, which lets patients compare fertilization and live-birth outcomes at the specific location they would attend rather than the network average. When evaluating add-ons there, ask whether PGT-A and ICSI are recommended case by case or bundled into standard cycle pricing.
CCRM Fertility, headquartered in Colorado with a national network, has built its model around laboratory-heavy care, including extensive embryo testing infrastructure. Patients with recurrent loss or advanced age who have already decided on PGT-A may find the deep lab integration useful. The same question applies: request the lab's written mosaic reporting policy before paying for testing, since network scale does not guarantee threshold transparency.
RMA (Reproductive Medicine Associates), with locations including New York, has published extensively on embryo genetics and was among the networks that helped generate the early PGT-A trial data discussed above. Its physicians are therefore well positioned to explain both what the trials showed and what they did not, including the flat cumulative live-birth findings. Ask directly how the clinic counsels patients under 35 who request PGT-A without an indication.
Across all three, the verification steps are identical and free: pull the SART report for your specific location, read the clinic's add-on price sheet line by line, and request written answers on mosaic thresholds and ICSI defaults. Clinics differ in how willingly they put these policies in writing, and that willingness is itself a usable signal.
What should you settle before signing the add-on consent form?
Three decisions cover most of the financial and clinical exposure. First, match each add-on to a line in your medical history. PGT-A needs recurrent loss, a prior aneuploid result, or advanced age to justify its cost; ICSI needs a confirmed male factor, prior fertilization failure, frozen eggs, or monogenic testing. An add-on with no corresponding line item is a purchase, not a treatment.
Second, if you proceed with PGT-A, obtain the lab's mosaic policy in writing before the biopsy, including cutoffs, segmental handling, and whether mosaic embryos are frozen or discarded. This single document determines whether a borderline embryo remains recoverable if you later want a second opinion.
Third, on ICSI, ask what happens with conventional insemination if fertilization fails, and whether split insemination is available. For patients with normal semen analyses, the honest framing is that ICSI protects the clinic's workflow and the cycle's worst-case outcome, at your expense, with no evidence of improved live birth and a long-term offspring dataset that remains incomplete. For patients with severe male factor, the same technique is the reason the cycle can succeed at all. The difference between those two sentences is where the extra cost is either worth paying or worth declining.
