How to Actually Read SART Success-Rate Data

Where SART data comes from, and what it is

SART data is the outcome record that member fertility clinics report through the SART CORS system, broken out by patient age and by cycle 1Ref 1Society for Assisted Reproductive Technology (SART) (2024).National Summary Report (SART CORS Online).Supports that SART reports IVF outcomes by patient age band and by cycle, with per-retrieval, per-new-patient, per-transfer, and cumulative groupings, and how SART defines and reports success.. It sits alongside a federal dataset: the Fertility Clinic Success Rate and Certification Act of 1992 requires every U.S. clinic performing assisted reproductive technology to report its outcomes, which the CDC compiles and publishes at both national and clinic levels 2Ref 2Centers for Disease Control and Prevention (2024).ART Success Rates.Supports that a 1992 federal law mandates ART success-rate reporting and that the CDC publishes national and clinic-level data, distinct from SART's own registry..

Both live on public sites you can open without an account. The federal data lets you look up national tables and individual clinics, and the SART site lets you pull a clinic's report or the national summary directly 2Ref 2Centers for Disease Control and Prevention (2024).ART Success Rates.Supports that a 1992 federal law mandates ART success-rate reporting and that the CDC publishes national and clinic-level data, distinct from SART's own registry.1Ref 1Society for Assisted Reproductive Technology (SART) (2024).National Summary Report (SART CORS Online).Supports that SART reports IVF outcomes by patient age band and by cycle, with per-retrieval, per-new-patient, per-transfer, and cumulative groupings, and how SART defines and reports success.. Arriving at these before a consult changes the conversation: instead of being handed a number, you can ask a clinic to explain the number the public record already shows.

Reading either one well starts with accepting that it is a structured report, not a single score. The same system carries national vs clinic data, and the two answer different questions: the national tables describe the field as a whole, while a clinic report describes one practice. Neither is a rating you can skim; both reward a reader who knows where to look, and treating the report as a document to be read in order is the whole method.

The denominators: per retrieval, per transfer, per patient

The core of a SART report is that it states outcomes on several denominators, and each answers a different question. Live-birth rate per intended egg retrieval counts from the start of a stimulation cycle; per new patient follows a person across their care; per transfer counts only cycles that reached an embryo transfer; and cumulative figures combine multiple transfers or cycles 1Ref 1Society for Assisted Reproductive Technology (SART) (2024).National Summary Report (SART CORS Online).Supports that SART reports IVF outcomes by patient age band and by cycle, with per-retrieval, per-new-patient, per-transfer, and cumulative groupings, and how SART defines and reports success..

A concrete way to hold these apart: the per-retrieval rate answers 'if I start one stimulation cycle, what is my chance of a baby from it,' while the per-transfer rate answers the narrower 'if I get as far as a transfer, what is my chance then.' The cumulative rate answers the biggest question, 'across a course of care, what is my chance overall.' A single report can show all of these for the same patients, which is why skimming for the largest percentage is the fastest way to misread it 1Ref 1Society for Assisted Reproductive Technology (SART) (2024).National Summary Report (SART CORS Online).Supports that SART reports IVF outcomes by patient age band and by cycle, with per-retrieval, per-new-patient, per-transfer, and cumulative groupings, and how SART defines and reports success..

Find the denominator label before you read the percentage, because the numbers are not interchangeable. A per-transfer rate will always be higher than a per-retrieval rate at the same clinic, since it drops the cycles that never reached transfer.

In a SART report, the word after 'per' matters more than the number itself.

Reading the age bands

SART reports outcomes in age bands because age is the strongest predictor of IVF success, and a single blended average hides that. The report separates results into brackets, younger patients and successively older ones, so the only row that truly describes you is the one matching your age at treatment 1Ref 1Society for Assisted Reproductive Technology (SART) (2024).National Summary Report (SART CORS Online).Supports that SART reports IVF outcomes by patient age band and by cycle, with per-retrieval, per-new-patient, per-transfer, and cumulative groupings, and how SART defines and reports success..

When you open your own age band, read across the row rather than fixating on one cell. A bracket will typically show the number of cycles behind the percentage, and a rate built on very few cycles is noisier and less reliable than one built on many. This is also where a clinic's caseload shows through: a practice that treats mostly younger patients will have thin data in older brackets, so its numbers for a 42-year-old may rest on a handful of cycles and should be read with that in mind 1Ref 1Society for Assisted Reproductive Technology (SART) (2024).National Summary Report (SART CORS Online).Supports that SART reports IVF outcomes by patient age band and by cycle, with per-retrieval, per-new-patient, per-transfer, and cumulative groupings, and how SART defines and reports success..

Reading age-stratified success rates, rather than the blended headline, is the single most important habit in interpreting this data. A clinic that treats many younger patients will show a strong overall number that says little about an older patient's odds, so when you look at SART age-stratified data, ignore the top-line figure and go straight to your own bracket. It is also worth checking how wide a bracket is, because a band that lumps several ages together can hide a real decline within it; the narrower the bracket, the more the number actually describes you.

Why each report is a noncumulative snapshot

A SART or CDC reporting year is a photograph of one period, not a running total you can add up. The national surveillance system defines exactly how cycles are counted and how a success is measured, whether a clinical pregnancy, a live-birth delivery, or a singleton versus a multiple, and it treats each year's national figures as noncumulative 3Ref 3Centers for Disease Control and Prevention (2024).NASS Technical Notes.Supports how cycles are counted, how success measures (clinical pregnancy, live-birth delivery, singleton versus multiple) are defined, and that national ART figures are noncumulative..

That is why you cannot combine two years, and why the most recent year describes cycles rather than a lifetime probability. When a summary presents a single reporting year, it is answering 'how did cycles started that year turn out,' not 'what are my overall odds across a course of treatment.'

This is also why the reporting lag exists and why it is a feature, not a flaw. Counting a live birth means waiting out a pregnancy and confirming a delivery, so the system trades speed for accuracy 3Ref 3Centers for Disease Control and Prevention (2024).NASS Technical Notes.Supports how cycles are counted, how success measures (clinical pregnancy, live-birth delivery, singleton versus multiple) are defined, and that national ART figures are noncumulative.. Be wary of any source quoting suspiciously current 'success rates,' because the honest numbers for the most recent cycles simply do not exist yet. Patience with the data is part of reading it well.

Cumulative outcomes: the fairest single number

Because IVF is usually a course of care, the number that best reflects real odds is the cumulative one: the chance of a live birth across several transfers or cycles, which SART reports as its own grouping 1Ref 1Society for Assisted Reproductive Technology (SART) (2024).National Summary Report (SART CORS Online).Supports that SART reports IVF outcomes by patient age band and by cycle, with per-retrieval, per-new-patient, per-transfer, and cumulative groupings, and how SART defines and reports success.. Independent research shows how much this matters. In a large UK cohort of more than 156,000 women, live-birth rates climbed from about 29.5% after one cycle to roughly 65% after six overall, and were far lower, about 12% rising to 31.5%, for women aged 40 to 42 4Ref 4Smith ADAC, Tilling K, Nelson SM, Lawlor DA (2015).Live-Birth Rate Associated With Repeat In Vitro Fertilization Treatment Cycles.Supports that cumulative IVF live-birth rates rose from about 29.5% after the first cycle to roughly 65% by six cycles overall, and were far lower (about 12% to 31.5%) for women aged 40 to 42..

cumulative live birth reached ~65% by six cycles overall, but only ~31.5% at ages 40-42 4Ref 4Smith ADAC, Tilling K, Nelson SM, Lawlor DA (2015).Live-Birth Rate Associated With Repeat In Vitro Fertilization Treatment Cycles.Supports that cumulative IVF live-birth rates rose from about 29.5% after the first cycle to roughly 65% by six cycles overall, and were far lower (about 12% to 31.5%) for women aged 40 to 42.

The cumulative view reframes what a single failed cycle means. Because odds accumulate across attempts, one negative result is a data point, not a verdict, and the steepness of the cumulative curve, and how sharply it drops with age, is what actually informs planning 4Ref 4Smith ADAC, Tilling K, Nelson SM, Lawlor DA (2015).Live-Birth Rate Associated With Repeat In Vitro Fertilization Treatment Cycles.Supports that cumulative IVF live-birth rates rose from about 29.5% after the first cycle to roughly 65% by six cycles overall, and were far lower (about 12% to 31.5%) for women aged 40 to 42..

Reading only the per-transfer headline flatters a single attempt and hides the multi-cycle reality; reading only a first-cycle figure does the reverse. The cumulative, per-patient number sits between them and is the fairest single thing to plan around.

SART or CDC: two reports of the same reality

SART and the CDC publish overlapping but distinct reports, and knowing the difference keeps you from double-counting. The CDC's dataset exists because federal law mandates it and covers reporting clinics nationally; its most recent national summary recorded 94,039 live-birth deliveries from 435,426 ART cycles, performed on 251,542 patients at 457 clinics, and is published as an online dashboard 5Ref 5Centers for Disease Control and Prevention (2024).National ART Summary (2022 data dashboard).Supports the most recent single-year national ART totals (435,426 cycles, 251,542 patients, 457 clinics, 94,039 live-birth deliveries) and that current data are published as an online dashboard.. SART's is the member society's own registry, reported through SART CORS 2Ref 2Centers for Disease Control and Prevention (2024).ART Success Rates.Supports that a 1992 federal law mandates ART success-rate reporting and that the CDC publishes national and clinic-level data, distinct from SART's own registry..

Comparing sart vs cdc data means matching the reporting year and the definitions before you compare any figure, because the two can count and label things slightly differently. In practice, if you find a clinic in one dataset but not the other, or you see two different-looking numbers, do not assume the worst; assume a definitional or timing difference and check it.

Confirm that you are comparing the same reporting year, the same denominator, and the same age band across both sources. The two reports are complementary, and reading the CDC ART success rates report and the SART report together, rather than picking whichever shows the friendlier figure, is how you see the whole picture instead of a corner of it. Reading them together also guards against a subtler mistake, which is treating a clinic's own summary slide as if it were the primary source; the report itself is the source, and any paraphrase a clinic offers should agree with it.

Preliminary versus final data years

The most recent year in any fertility report is usually preliminary, because a live birth cannot be counted until roughly nine months after the cycle, plus time to confirm the delivery. A cycle started late in a year may have no recorded outcome when the first tables publish, so early figures are revised as births are confirmed and the definition of a live-birth delivery is met 3Ref 3Centers for Disease Control and Prevention (2024).NASS Technical Notes.Supports how cycles are counted, how success measures (clinical pregnancy, live-birth delivery, singleton versus multiple) are defined, and that national ART figures are noncumulative..

When a report labels its data years, check whether you are reading preliminary or final numbers before drawing conclusions. A preliminary year can shift as data mature, so treating its percentages as settled, or comparing a preliminary year at one clinic against a final year at another, quietly compares unlike things.

There is a related trap in comparing across sources or clinics: a preliminary year at one place versus a final year at another are not the same measurement, even if both are labeled with a percentage. Before you let a comparison drive a decision, make sure both sides are final, or both preliminary, and both drawn on the same denominator. When in doubt, favor the older, finalized year over a shinier but unsettled recent one.

Your odds are not the report's headline

No published rate, national or clinic, per transfer or cumulative, is your personal probability. The CDC's IVF Success Estimator converts national data into an individual estimate using age, height, weight, and diagnosis, for ages 20 to 50, but it is still built on national averages rather than one clinic's results 6Ref 6Centers for Disease Control and Prevention (2024).IVF Success Estimator.Supports that individualized IVF live-birth estimates exist for ages 20 to 50 and are based on national averages, not any single clinic's results.. Your diagnosis, ovarian reserve, and history move your real number in ways a report cannot.

There is no single 'good' success rate in the abstract; a good rate is one measured on an honest denominator, in your age band, for patients like you. And some numbers sit outside your own tables entirely: donor-egg success rates are reported on separate denominators, so they should never be read next to your own-egg age band as if they described the same thing.

So the closing habit is simple. Take any success rate you are shown, find its denominator, its age band, and its data year, and then ask what your own diagnosis and history would do to it. A number that survives all four questions, and that describes patients genuinely like you, is worth something. A glossy headline that dissolves under any one of them was never describing you in the first place. That discipline is portable: it works on a clinic's brochure, on a national dashboard, and on a headline in an article, because every honest fertility statistic can name its denominator, its age band, and its year when asked.