×
IVF Information

News Center

Current position: Blessed Future > IVF Information >

IVF Cycle in the United States: How Many Rounds Patients Typically Try at Clinics Like Boston IVF and Shady Grove Fertility Before Success

Release time: 2026-11-17 Blessed Future

Most patients at large U.S. clinics such as Boston IVF and Shady Grove Fertility do not succeed in a single IVF cycle. National data collected by SART and the CDC show that cumulative live birth rates keep rising through a second and third attempt, especially for women under 35 using their own eggs. The practical question is how many rounds the statistics actually justify, and when the marginal gain of one more cycle stops paying for itself.

How many IVF cycles does it really take? Reading the cumulative curve instead of the single-cycle number

The number most patients carry in their head comes from the wrong chart. Fertility clinic websites and the CDC's annual ART Fertility Clinic Success Rates Report both publish per-cycle outcomes, broken into age bands: under 35, 35 to 37, 38 to 40, 41 to 42, and over 42. A woman of 33 might see a per-transfer live birth rate in the neighborhood of 40 to 50 percent at strong programs, and she reads that as her chance of having a baby. It is her chance per attempt. The question "how many rounds do patients try" only becomes answerable when those per-cycle numbers are compounded across cycles, which is exactly what the Society for Assisted Reproductive Technology (SART) does in its National Summary data and in its online IVF Success Estimator, a free tool that projects cumulative live birth probability over up to three cycles based on a patient's age, height, weight, and prior cycle history.

Compounding changes the picture substantially. If a single cycle carries, say, a 40 percent chance, three independent cycles would carry roughly 78 percent, and even after discounting for the reality that later cycles tend to produce fewer eggs and lower-quality embryos, published cumulative curves for younger patients typically land well above any single-cycle figure. For patients in the 38 to 40 band, the compounding effect matters even more, because their per-cycle numbers look discouraging in isolation while their three-cycle cumulative numbers remain clinically meaningful. This is the single most underused piece of information at the first consultation, and patients who ask for it directly, using the SART estimator as a starting point and their clinic's own age-banded data as a correction, walk into cycle planning with a fundamentally different mental model than patients who anchor on one number from a website table.

Large-volume networks such as Boston IVF and Shady Grove Fertility sit on a genuine informational advantage here. A program that runs thousands of cycles a year has internal, recent, age-stratified outcome data, and it can in principle show a patient where she falls on a cumulative curve rather than on a single-cycle bar. Whether it does so proactively varies by clinic and by physician, so the burden falls on the patient to ask. A concrete request works better than a vague one: ask the physician to estimate your cumulative live birth probability over three cycles given your AMH, antral follicle count, and any prior response, and ask what fraction of patients in your age band at that clinic ultimately succeed within three attempts. The answer may be an honest range rather than a precise figure, and any range should be checked against the SART and CDC published data for that clinic, which are updated annually and are the only externally verifiable reference point.

Why the first failed cycle is a baseline, not a verdict

A failed first cycle feels like an ending, and the emotional weight of that feeling causes real statistical errors. The first cycle is the only cycle for which the clinic has no personal data about you. Everything before it, the AMH level, the antral follicle count, the FSH value, was a prediction. The first cycle is the first measurement. It tells you how your ovaries actually responded to a specific dose of stimulation medication, how many mature eggs were retrieved, how many fertilized normally, how many reached the blastocyst stage by day five, and, if genetic testing was performed, how many were chromosomally normal. Those five numbers form a baseline that is worth more than any population statistic, because the second cycle can now be designed against observed behavior instead of predicted behavior.

Consider how the math works with an illustrative example. Suppose a 34-year-old patient's first cycle yields eight eggs, five blastocysts, and, with testing, two chromosomally normal embryos that fail to implant. Her per-transfer chance is not the generic 40 to 50 percent for her age band; it is now anchored to a known euploid blastocyst production rate. If her euploid rate per blastocyst was 40 percent and her clinic's euploid transfer success runs in the 50 to 60 percent range for her age, the cumulative question becomes arithmetic: how many euploid embryos does she need in storage before the probability that at least one succeeds crosses her personal threshold? Patients who run this calculation, even roughly, with their physician stop experiencing each failed transfer as a coin flip and start seeing it as one draw from a distribution they now understand. These figures are illustrative only; the actual inputs must come from your own cycle report and your clinic's published per-age-band outcomes.

The boundary condition matters as much as the baseline. A first cycle with a strong response but failed implantation points in a different direction than a first cycle with a weak response and no blastocysts at all. The first pattern keeps the cumulative curve largely intact, because embryo production was adequate and the problem sits downstream. The second pattern bends the curve downward, and it should trigger the protocol review described in the next section rather than an immediate repeat of the same stimulation. Patients who treat any failed cycle as proof that the whole endeavor is failing conflate two very different situations, and they make the decision to continue or stop with worse information than they actually have.

What should change between cycle 1 and cycle 2? The protocol rewrite window

Between the first and second cycle lies the highest-leverage decision window of the entire treatment arc, and it is frequently squandered because patients arrive at it exhausted. Three variables dominate the rewrite: the stimulation dose, the protocol type, and whether preimplantation genetic testing for aneuploidy (PGT-A) enters the plan. Each addresses a different failure mode identified in the first cycle's data.

Dose adjustment responds to observed ovarian behavior. A patient who produced few mature eggs on a standard gonadotropin dose may be escalated, within safety limits set by the clinic to control ovarian hyperstimulation risk. A patient who produced many eggs but a low mature-egg rate may have her dose timing or trigger medication changed, since premature ovulation triggers and incomplete maturation show up in the retrieval report as a gap between eggs collected and eggs that are actually mature. These adjustments are routine, but they are not automatic; you need to ask your physician to walk through the retrieval report line by line and state which parameter was furthest from target.

Protocol switching is a larger structural change. An antagonist (short) protocol and a long agonist protocol suppress the patient's own hormonal cycle in different ways, and patients who recruited a poorly synchronized cohort of follicles on one protocol sometimes recruit more uniformly on the other. The decision rests on the pattern in the first cycle: widely spread follicle sizes and a low fertilization rate are the classic signals that prompt a switch. Neither protocol is universally superior; large clinics rotate between them routinely, and the choice is empirical, driven by what the first cycle revealed.

Adding PGT-A is the most consequential and the most debated change. Testing every blastocyst for chromosome count does not improve the quality of any individual embryo; it changes the information available at transfer. Without testing, each transfer carries the combined uncertainty of embryo quality and chromosomal status, and failed transfers remain ambiguous. With testing, transfers proceed only with euploid embryos, and repeated failure of euploid embryos becomes a distinct, clinically meaningful signal that points away from embryo selection and toward the uterine environment or other factors. For patients over 35, where the aneuploid rate per embryo rises steeply, this reframing often changes the cumulative calculation significantly, because it converts unexplained failures into explained ones. The tradeoff is added cost per cycle and the possibility that a cycle produces no testable blastocysts at all. The American Society for Reproductive Medicine (ASRM) has published committee opinions on the appropriate and inappropriate uses of PGT-A, and those documents are worth reading before the decision, because they lay out where the evidence is strong and where it remains contested.

The reason this window gets missed is timing. Protocol decisions are usually made at a post-cycle review appointment scheduled two to six weeks after a failed cycle, which is precisely when motivation and analytical clarity are at their lowest. Patients either accept a repeat of the identical plan out of fatigue, or they quit before the review happens at all. One concrete practice changes this: prepare a written list of questions before the review appointment, covering the four numbers from your baseline (mature eggs, fertilization rate, blastocyst rate, euploid rate if tested) and asking explicitly which single change to the plan has the largest expected effect on those numbers. A clinic that cannot or will not answer that question in specific terms is giving you a reason to seek a second opinion before committing to round two.

Where is the dropout cliff, and why do patients leave while the odds still favor them?

Real-world treatment data show a steep attrition pattern. A widely cited 2013 meta-analysis by Gameiro and colleagues in Human Reproduction Update examined why patients discontinue fertility treatment and found that the burden of treatment, financial cost, and a perceived poor prognosis ranked among the dominant reasons, with a substantial share of patients exiting after one to three unsuccessful cycles. The pattern holds across systems with very different payment structures, which suggests the driver is cumulative strain rather than any single healthcare policy. In the United States, where a single IVF cycle commonly runs in the range of $15,000 to $25,000 before medication costs of several thousand dollars more, the financial slope of the cliff is steeper still, and patients frequently face the second-cycle decision with five figures already spent and no live birth to show for it.

The uncomfortable observation is that many of these exits happen while the cumulative odds still favor continuing. A patient whose first cycle produced blastocysts but no pregnancy, and whose age-band cumulative curve still projects a solidly favorable three-cycle probability, is statistically in a very different position from a patient whose first cycle produced nothing usable. Both feel identical from the inside. Both hear the same sentence in their own head: it did not work. The difference is that no one at the clinic translated the first patient's position into a number she could weigh against the cost of another round. The marginal benefit of cycle two, expressed as the change in cumulative live birth probability from attempting it, is a computable quantity, and it is almost never computed for the patient in plain terms.

Fixing this is a matter of asking for the calculation directly. At the review appointment, request three numbers: your estimated cumulative probability of a live birth if you stop now, if you complete one more cycle, and if you complete two more. The gap between the first and second numbers is the marginal gain you are buying. Weighing that gain against the quoted cycle cost, your remaining financial capacity, and the emotional load you can document from your own experience over the past months turns a decision that usually gets made in despair into one that gets made on a spreadsheet. It does not remove the pain of a failed cycle. It does prevent the specific error of quitting at the point where the expected value of continuing was still positive, which is the most common and most regretted exit point reported by patients who later return to treatment.

When should you stop? Three trigger conditions that turn quitting into a decision

Stopping treatment carries a stigma of surrender, and that framing causes patients to continue past the point where data say the expected gain has collapsed, and to quit before it has. The alternative is to define stopping criteria in advance, the way a clinical trial defines its endpoints. Three conditions carry the strongest signal, and each has boundary conditions worth stating precisely.

The first trigger is recurrent production of chromosomally abnormal blastocysts across multiple stimulation cycles. A single cycle with all aneuploid results is bad luck within the range of probability, particularly over age 38. Two or three consecutive cycles in which every tested embryo is aneuploid, especially when blastocyst counts were adequate, shift the diagnosis toward an egg-quality ceiling that additional stimulation cycles are unlikely to move. The boundary condition is age: the same pattern at 41 points toward a biological limit that a change of clinic or protocol rarely alters, while at 31 it may justify one more protocol variation before any conclusion is drawn, because young patients occasionally produce an aneuploid-dominant cohort for reasons that resolve on retry.

The second trigger is a measurable collapse in ovarian reserve or response. When antral follicle counts and AMH values drop sharply between cycles, or when the number of retrieved eggs falls cycle over cycle despite escalated doses, the input side of the pipeline is shrinking. Each subsequent cycle then costs the same money and physical burden while purchasing progressively less raw material. A decline between consecutive cycles is normal; a steep decline across two consecutive cycles, documented in your own lab reports, is the point at which the cumulative curve flattens and the marginal gain of another round approaches zero. This is a trend judgment, not a single-lab-value judgment, and it should be confirmed with a repeat measurement before acting on it.

The third trigger is three completed stimulation cycles with no usable embryo to transfer, meaning no blastocyst at all, or no euploid blastocyst if testing was in use. At that point the accumulated evidence says the bottleneck is upstream of transfer, and the realistic options narrow. This is the natural moment for a formal second opinion at a different program, ideally one with a different laboratory, before any final decision. A fresh embryology lab and a fresh protocol review occasionally find correctable problems, and even when they do not, the second opinion converts an open-ended "keep trying" into a documented, evidence-based endpoint that patients report being able to live with.

None of these triggers obligates anyone to stop. They mark the points where continuing becomes a decision that needs an affirmative reason, rather than the default. Patients who cross a trigger and choose to continue anyway, for reasons of finances, values, or simply wanting no regrets, make a legitimate choice, but they make it knowingly, which is the entire point of defining the triggers in advance.

How do Boston IVF and Shady Grove Fertility fit into this picture?

The clinics named in this article's title are useful examples of the large-network model, because their scale shapes what a patient can realistically ask of them. Boston IVF operates as a New England network with multiple locations and a high annual cycle volume, which means its physicians see the full range of protocol outcomes repeatedly and its internal data on age-banded cumulative results is current. Shady Grove Fertility, headquartered in Maryland with more than 40 locations nationwide, runs one of the highest-volume operations in the country and maintains published outcome reporting plus in-house financial counselors, which matters directly to the dropout-cliff problem, since payment planning before cycle two is one of the few interventions that demonstrably reduces mid-treatment exits. CCRM Fertility, a national network headquartered in Colorado, is known for its emphasis on embryology laboratory infrastructure and on testing-heavy treatment plans, which makes it a common destination for the second-opinion step described above. All published success figures for these programs should be read directly from their SART and CDC reports rather than from marketing materials, and figures change year to year.

The practical implication of choosing a high-volume network is access to standardized cycle review processes and, in many cases, shared decision-making tools that smaller clinics lack. The implication of the network's scale for the questions in this article is straightforward: ask for cumulative projections, ask for the marginal-gain calculation before each subsequent cycle, and ask what fraction of patients in your exact situation proceed to a second and third cycle. A large program has answers to these questions in its own data. Whether it volunteers them is a different matter, and the willingness of the team to answer them in numbers, rather than in reassurance, is one of the more reliable signals of how the rest of the relationship will go.

Which questions should you bring to your next appointment?

  • Based on my age and my first-cycle numbers, what is my estimated cumulative live birth probability over three cycles, and how does it compare with the SART estimator output for my profile?
  • Of the four baseline numbers from my last cycle, mature eggs, fertilization rate, blastocyst rate, and euploid rate, which one was furthest from its target, and which single protocol change addresses it?
  • What is the marginal gain, in percentage points of cumulative probability, if I complete one more cycle? And two more?
  • If this next cycle fails, what specific result would trigger a protocol switch, and what result would trigger a conversation about stopping?
  • What does the full per-cycle cost look like including medications, and are there multi-cycle or refund-style programs available at this clinic that change the financial math?

How many rounds patients try before success is, in the end, a number that can be estimated, revised after every cycle, and compared against cost and capacity. Patients who track their own cumulative curve, demand the marginal-gain calculation before each additional round, and define their stopping triggers in advance end up in one of two places: a live birth, or a documented, evidence-based endpoint they chose deliberately. Both are better outcomes than the default path, which is drifting out of treatment at cycle two or three without ever seeing the math that was running underneath the whole time.

Contact Us