The IVF toolkit keeps expanding, but the supporting evidence does not expand evenly. Some tests and interventions are well anchored in outcomes, others remain plausible but unproven, and a third group can look beneficial only when judged by narrow endpoints. A practical way to read the field is to treat IVF as a pathway rather than a single event: stimulation, retrieval, fertilization, embryo culture, optional biopsy and testing, optional cryopreservation, transfer strategy, and follow-up transfers. In that pathway framing, it becomes clear why evidence can feel contradictory and why the same intervention can look “effective” in one analysis and neutral in another. Medscape’s recent overview leans into this variability and reinforces a point that is sometimes underemphasized in the add-on era: age remains the most reliable predictor of IVF success, and many additional tests do not outperform it as a global predictor. 

Issue 1: The endpoint problem (per transfer vs per started cycle)

Many claims in IVF are supported by intermediate outcomes that do not necessarily translate into higher cumulative live birth per started cycle or per retrieval. Implantation rate, clinical pregnancy per transfer, euploidy rate, and miscarriage rate can all be meaningful, but they are vulnerable to pathway effects. Any intervention that changes how many patients reach transfer, how many embryos remain available for transfer, how long it takes to reach transfer, or how much cost and burden accumulates can shift per-transfer statistics without improving, and sometimes while worsening, cumulative outcomes. This is why intention-to-treat analyses and cumulative live birth remain the load-bearing endpoints when deciding whether a test or add-on should be routine, selective, or research-only.

Issue 2: Analytic validity is not clinical utility

Reproductive medicine has become extremely good at measuring signals, categorizing embryos, and generating scores. That technical progress is real. The scientific gap is that a measurable signal is not automatically a clinically actionable signal. Clinical utility requires that the test changes management and that the management change improves outcomes that matter. This distinction is central to the current debate around several popular tools: the assays can be repeatable and biologically plausible, yet still fail to improve live birth, shorten time-to-pregnancy, or reduce burden when deployed in routine care.

Issue 3: Ovarian reserve testing works best when treated as a response tool, not a fertility forecast

AMH and antral follicle count are useful for predicting ovarian response and helping guide stimulation planning, expected oocyte yield, and risk counseling. The misstep is turning them into broad predictions about natural fertility, egg quality, or time to pregnancy in people without infertility. ACOG’s committee opinion states that using serum AMH levels for fertility counseling in women without a diagnosis of infertility is not currently supported by high-quality data.  In other words, ovarian reserve markers can be valuable within their validated use-case, but they do not function as general-purpose fertility prediction tools and should not be framed that way to patients or clinicians.

Issue 4: PGT-A shows how embryo selection can improve a slice of the pathway without improving the whole pathway

PGT-A is one of the most common places where endpoint choice and pathway effects collide. The promise is efficiency: avoid transfers with low implantation potential and reduce miscarriage by prioritizing euploid embryos. The evidence, however, remains mixed when evaluated at the level of patient-important outcomes across diverse populations and clinic workflows. ASRM’s 2024 committee opinion states that the value of PGT-A as a routine screening test for all IVF patients has not been demonstrated.  That framing does not claim “no one benefits,” but it does set a high bar for routine use and highlights the need to match PGT-A to specific indications, clear counseling goals, and trial-quality evidence using cumulative endpoints.

Issue 5: Endometrial receptivity testing has not earned a routine role in randomized evidence

Endometrial receptivity testing is appealing because it offers a clean narrative: identify an individual window of implantation and time transfer accordingly. In a large randomized clinical trial in patients undergoing single euploid frozen embryo transfer, timing guided by endometrial receptivity testing did not improve live birth compared with standard timing, and the authors conclude the findings do not support routine use to guide timing.  This is a good example of why personalization claims need to clear a higher evidentiary threshold than mechanistic plausibility, especially when the test adds biopsy burden, cost, and cycle complexity.

Issue 6: Add-ons are increasingly treated as an evidence governance issue

A growing proportion of IVF innovation arrives as “add-ons,” often priced as optional enhancements and adopted unevenly across clinics. The scientific and ethical problem is not innovation itself, but routine adoption without robust, patient-relevant outcome data. ESHRE’s add-ons guidance and its good practice recommendations explicitly focus on safety, efficacy, and transparent counseling about whether an add-on is likely to improve live birth.  The HFEA has operationalized a related goal with its evidence ratings for treatment add-ons, designed to make evidence quality legible and to discourage routine use when benefit is unclear or unsupported.  Even outside the add-on label, this approach helps bring the field back to a consistent standard: routine care should be reserved for interventions with credible evidence of meaningful benefit.

Issue 7: Emerging genomics changes the risk profile because precision can outpace validity

The next wave of embryo selection includes polygenic embryo screening and other high-dimensional scoring approaches. These differ from earlier add-ons because they can look quantitatively precise while carrying unresolved issues around predictive accuracy, ancestry portability, calibration, and clinical meaning at the individual embryo level. ASRM’s Ethics and Practice Committees concluded in December 2025 that polygenic embryo screening is not ready for clinical use, citing lack of proven clinical utility and significant scientific and ethical concerns.  This is likely a preview of how the field will increasingly evaluate new tools: not just whether they produce a number, but whether that number is stable, generalizable, and tied to outcomes that matter.

Issue 8: Standardization pressure is rising, driven by evidence gaps and practice variation

Clinical variation is often a marker of uncertainty, not personalization. NICE is actively updating fertility guidance, with an expected publication date of March 19, 2026, reflecting ongoing efforts to reduce practice variation and scrutinize unproven interventions.  Across regions and regulators, the shared direction is a higher expectation that routine interventions should demonstrate benefit on outcomes like live birth and time-to-pregnancy, not only on intermediate metrics or mechanistic plausibility.

Taken together, these issues point to a simple scientific posture that is increasingly useful in IVF. Treat new tests and add-ons as hypotheses until they demonstrate clinical utility across the pathway. Prefer evidence anchored to intention-to-treat analyses and cumulative live birth. Be explicit about what a test predicts well and what it does not. And keep age, diagnosis, and pathway design as the baseline comparators, because many tools are competing against a surprisingly strong predictor and a surprisingly complex sequence.

Sources

• Medscape. December 05, 2025
 IVF Tests and Treatments Are Backed by Varying Evidence

• American Society for Reproductive Medicine. 
The use of preimplantation genetic testing for aneuploidy: a committee opinion 

• American Society for Reproductive Medicine. 8 Dec 2025. 
ASRM Ethics and Practice Committees Release New Report Concluding Polygenic Embryo Screening Is Not Ready for Clinical Use 

• ESHRE. 
Add-ons 

• Human Reproduction. 
Good practice recommendations on add-ons in reproductive medicine 

• Human Fertilisation and Embryology Authority. 
Treatment add-ons with limited evidence 

• Fertility and Sterility. 
Live birth after transfer of a single euploid vitrified-warmed blastocyst according to standard timing versus timing as recommended by endometrial receptivity analysis

• NICE. 19 Mar 2026 (expected publication).
Fertility problems: assessment and treatment – update 1 and 2

https://www.medscape.com/viewarticle/ivf-tests-and-treatments-are-backed-varying-evidence-2025a1000y2w