Professor of OBGYN, Columbia University Medical Center
Medical Director, REI, Northwestern Medicine
Associate Director, REI, Weill Cornell Medical Center
Former President, ASRM
Director, Male Fertility, Stanford School of Medicine
Medical Director, RMA of Northern California
Founder & Medical Director, Colorado Center for Reproductive Medicine
When assessing fertility treatment success rates, we often focus on two measures:
How often a treatment leads to a live birth. The higher the live birth rate, the better the treatment
How often a pregnancy from a treatment results in a multiple birth (twins, triplets, or more). Multiple births are dangerous (as you can see below) and so the lower the multiple birth rate, the better the treatment.
Across all patients, IUI cycles have live birth rates per cycle of 5 - 15% and multiples-per-pregnancy of 8 - 30%. As we mentioned, while this “multiples-per-pregnancy” rate is high, since IUI has such modest pregnancy rates, only 0.5% - 2.5% of IUI cycles actually result in twins or triplets being born.
For context, compared with IVF, IUI delivers significantly lower live birth rates (2 - 10x lower) and potentially far-higher (8 - 30x) rates of multiple-births-per-pregnancy. However, IUI is 5 - 40x cheaper than the IVF alternative.
However, it is important to point out how IUI success rates vary depending upon the age of the woman, the drugs (if any) that she takes, and, relatedly, the number of follicles she grows.
Probably the most compelling data on the influence of female age on IUI success comes from an analysis of 17,000 patients. As you can see below, the decline is meaningful between each age bucket and tracks a similar phenomenon amongst patients who receive IVF.
To illustrate how drugs impact success, see how adding letrozole, clomid and gonadotropin each successively improve success rates in younger patients with unexplained Infertility. However, also notice how adding these drugs increases the rate of multiple births per pregnancy.
Determining which drugs offer the best reward to risk ratio is a matter of personal preference, but today letrozole and clomid are often considered to have the best profile for most patients.
Related to the above two points (age and drug selection), the number of follicles a woman has growing during her IUI clearly correlates with rates of success. The best analysis to date was performed by a Dutch team combining data from over 14 studies across 11,000 cycles.
Unexplained infertility refers to patients who had no significant abnormalities found on the diagnostic workup. In these patients, the goal of an IUI cycle is to have 2 or 3 eggs ovulated in a given cycle (rather than the typical one egg per cycle) to increase the chance of pregnancy.
Generally speaking IUI leads to a live birth rate for these patients in 7 - 10% of cycles and requires at least 3 IUIs to equate to the success rates of one IVF cycle.
There are many studies comparing success rates between clomid, letrozole, or gonadotropins for patients with unexplained infertility, but two stand out as the best and most informative. The first study was conducted at multiple sites across the country and was termed the AMIGOS trial. In this study, gonadotropins produced the highest pregnancy rate, followed by clomid, and then letrozole. However, almost one third of all pregnancies in the gonadotropin arm was either a twin or triplet gestation. This was significantly higher than the clomid or letrozole arms.
The second study by Huang et al. demonstrated nearly equivalent pregnancy rates between the three medications. Furthermore, the twin risk was not significantly elevated in any of the three groups. The key difference between these studies is that the dose of gonadotropins was higher in the AMIGOS study (150 units) than the Huang study (75 units). A higher dose often means more eggs ovulated and a greater risk of twins or more.
The take home message of these studies is that when gonadotropins are given to this patient population at a dose that doesn’t put the patient at significant risk of twins (like the 75 units in the Huang study), they are no better than oral medications (clomid and letrozole) at producing successful pregnancies. This is important because clomid and letrozole are also cheaper to purchase (hundreds vs. thousands of dollars) and easier to administer (oral rather than injectable). For this reason, many clinics have moved away from using gonadotropins in IUI cycles.
The AMIGOS study suggested that clomid provided the best balance of a high pregnancy rate with a reasonably low multiple rate among couples with unexplained infertility. However, the authors of the Huang study concluded that on balance letrozole was better. From our vantage point, given that clomid more commonly causes multiple eggs to be ovulated, it seems like the slightly better option between the two because the whole point of treatment in unexplained infertility patients is to increase the odds of delivery by increasing the number of eggs ovulated.
IUI can be a successful treatment strategy for patients with mild (not severe) male factor infertility but its rate of success is dramatically lower than what is achieved with IVF for this group.
There is major debate whether an IUI should be used for male patients with a modest (less than 10 million) Total Motile Count or TMC. TMC is an equation combining 3 parameters from the semen analysis: volume, concentration, and motility.
The data we have available is conflicting. A University of Iowa study paints a pessimistic picture and shows IUI live birth rates at 1.5% for a patients with a TMC of less than 10 million. A much larger study from Shady Grove reveals higher rates of success with three additional observations:
IUI success rates are the same so long as a man has TMCs anywhere over 10 million
IUI pregnancies are indeed possible using sperm with TMC below 10 million
Below 10 million TMC, the rate of decline for success is shallow, calling into question the wisdom of arbitrary TMC cutoffs
Unfortunately, a couple cannot look at the TMC from a semen analysis and know precisely where they fit into this rubrik. That is because the TMCs recorded above were measured after the man already had his sperm washed for IUI. A sperm wash can lower TMCs by roughly 30% and a credible rule of thumb for a general estimate would be to take the “pre-wash” TMC from a semen analysis, subtract a third, then presume that will be the “post-wash” TMC and apply it to the above charts.
However, couples should carefully record their pre and post-wash TMCs on the day of IUI. This will tell them how drastically the wash impacts their TMCs. If another IUI is necessary, they can apply a more refined adjustment (versus subtracting 30%) to predict their post-wash TMC and handicap their IUI’s odds of success.
Here we address two types of patients: those unable to have a period (called anovulatory) and those with irregular periods.
Having no period means ovulation isn’t taking place at all, so a pregnancy can’t happen because no eggs is making itself eligible to be fertilized. Similarly, having irregular periods makes achieving pregnancy difficult, because it’s hard to time intercourse properly -- if sperm and egg aren’t at the same place at the same time, there is no chance of pregnancy.
Amongst both groups (those with no period or irregular periods) there are three types of patients: those with hypothalamic dysfunction, those with PCOS, and those with no diagnosis (but who have had hypothalamic dysfunction ruled out).
Patients with hypothalamic dysfunction are not producing signals within their brains to tell the ovary to mature an egg. They are diagnosed because they have an extremely low FSH and a low LH (almost zero). Neither clomid nor letrozole will help them. For these patients, IUI must be accompanied by gonadotropin to be effective. From here on in this section, none of the data we’ll reference refers to patients with hypothalamic dysfunction.
Next we have patients who have either PCOS or no diagnosis. We classify them together because so as long they don’t have a hypothalamic dysfunction, the underlying mechanism leading to their irregular period, or lack of a period, is likely similar.
A lot of studies you’ll see below refer to women with PCOS but the findings should also apply to women with no periods, or irregular periods, regardless of whether they’ve been given a PCOS diagnosis (so long as they don’t suffer from hypothalamic dysfunction).
Amongst these two patient populations (which we’ll refer to now as PCOS), IUI with clomid and letrozole generally shows no improvement versus when these drugs are taken without IUI. When IUI is coupled with gonadotropins, success rates reach 14 - 26% per cycle, which is higher than almost any other patient class, but still pales in comparison to IVF. It’s also worth noting that PCOS patients who are taking gonadotropins have a very high risk of multiples.
When PCOS patients who don’t ovulate, or irregularly ovulate, must choose between taking clomid or letrozole, without IUI, the data typically favors letrozole.
The best study in the field enrolled 750 women to receive clomid or letrozole, followed them for 5 courses of therapy and revealed that the group receiving letrozole had higher live birth rates and fewer multiple gestations. The data is of exceptional quality, and there’s no reason to believe the conclusion doesn’t also apply to the choice of drugs if these patients proceeded on to IUI.
A closer look at the data suggest that the benefit of letrozole over clomid depended on the BMI of the participants. For patients with a BMI of less than 30 kg/m2, the cumulative live birth rate was approximately 30% for each group. However, for patients with a BMI over 30 kg/m2, twice as many patients had a live birth in the letrozole group than the clomid group.
Next, we’ll look at data that compares PCOS patients who took clomid alone or clomid accompanied by IUI (no studies use letrozole, so we are forced to use a study with clomid). Clomid alone produced a 17% live birth rate after 3 cycles, while clomid plus IUI produced a 19% live birth rate. Investigators concluded there was no statistically significant difference.
One might conclude there is thus no benefit to adding an IUI to a clomid cycle. However, the applicability of this study to all patients may be limited because it excludes women with a BMI over 30 (common for women with PCOS) and used outdated sperm washing techniques.
IUI (preferably with letrozole instead of clomid) is probably a good option for women with PCOS or irregular cycles in the event that:
There is evidence of a mild male factor
Ovulation induction medication has produced ovulation but no pregnancy after timed intercourse
If letrozole or clomid haven’t worked in causing ovulation, patients are considered “resistant” to these medications. Note that we’re talking about a failure to cause ovulation, not failure to cause pregnancy.
There are multiple strategies for causing ovulation in clomid or letrozole resistant patients. Some of these include adding medications such as dexamethasone or metformin to the treatment regimen. Another approach is changing to gonadotropin injections. Rather than tricking the brain into sending a stronger signal to the ovaries to cause follicle recruitment, gonadotropins directly stimulate the ovary to recruit multiple follicles.
The available data on how effective a gonadotropin-plus-IUI strategy is in patients with irregular, or non-existent, cycles is limited and unfortunately includes other drugs. However, these studies give us a sense for how well a patient can expect to do if she has irregular, or non-existent, cycles and is using a gonadotropin-plus-IUI approach.
The good news is that the pregnancy rate is favorable: 20 - 26% per cycle compared with the study above that showed clomid-plus IUI-was ~19% across three cycles.
The bad news is that gonadotropins produce a high multiple pregnancy rate, especially in this population. In the Palomba study, the IUI-plus-gonadotropin arm recorded over a third of pregnancies were multiple birth. By contrast, when these patients do IUI with just clomid, the percentage of pregnancies that resulted in twins was 9%.
In PCOS patients, IUI cycles that include gonadotropins can take a long time (sometimes 3 or 4 weeks). Doctors need to be careful to prescribe enough drugs to elicit a response but not so much that many follicles start growing at once. This balance is hard to find and can take time.
When one fallopian tube is blocked, IUI can still lead to live births, but the location of the blockage itself on the tube (proximal or distal) is likely a determinant of success.
As you can see below, patients with a proximal-only blockage fare better than those with a distal-only block, and both fare worse on a per-cycle basis than similar patients who receive IVF.
In 35% of cases when an HSG, the primary method for diagnosing a blocked fallopian tube, diagnoses a patient with proximal blockage, no blocage actually exists. That’s because during an HSG it’s easy to mistake a simple muscle spasm for a blockage. As a result, many of these proximal-only blockage patients actually have two fallopian tubes open and the eggs from both ovaries are able to migrate through the fallopian tubes, improving the prognosis for these patients.
Patients with one blocked tube who undergo IUI are at greater risk for an ectopic pregnancy than patients who have two open tubes. This is because most conditions that affect the health of fallopian tubes, like endometriosis or pelvic inflammatory disease, tend to impact both tubes.
A tube that is open, but has issues, is more likely to have difficulty passing the embryo into the uterus, so the embryo can become embedded in the tube. While ectopic pregnancies are typically caught early in patients seeking fertility care, they can be life threatening emergencies if the tube ruptures and begins to bleed.
IUI is a popular option when donor sperm is needed and, in the case of lesbian women or single mothers, many haven’t received any underlying “infertility” diagnosis.
IUI success rates here are typically 3 - 16% depending upon the woman’s age and are roughly one third of what IVF patients (many of whom do have a fertility issue) experience.
As you can see below, after about 3 cycles, the additional benefits of each IUI cycle becomes small (2 - 4% per cycle). This likely means that for the women who haven’t become pregnant by that point, there is likely another factor that is making it difficult to achieve pregnancy and IVF probably should be considered.
As we alluded to in the second lesson, all things being equal, IUI success rates are higher with fresh sperm than with frozen and thawed sperm. For women and lesbian couples using a “known donor” and who have the option to inseminate with fresh sperm, this may be an important fact to consider. For women obtaining sperm from a sperm bank, IUI with frozen and thawed sperm can absolutely still work.
As you can see in the data below, IUI continues to be somewhat productive for some patients after a few cycles, while in others it stops offering any benefit and patients should consider an alternative.
A fundamental question to consider is whether it makes sense to start with IUI or advance immediately to IVF. One study tried to assess this in women over the age of 38 who had unexplained infertility. One group of patients immediately started with IVF while another group underwent to 2 IUI cycles and, if they failed, continued on to IVF.
Ultimately the group that began with IVF saw success much more quickly, but after 6 cycles (either 6 IVF cycles for the first group or 2 IUI and 4 IVF cycles for the second group) there was no statistical difference in the number of women who had a live birth.
The implication here is that if you are focused on having a child as soon as possible, advancing to IVF is preferable. If you are more price sensitive and want to begin with IUI, you stand an equally as good chance so long as you are prepared and equipped to quickly move to IVF if IUI doesn’t work after a few cycles.
A next logical question is, if patients do start with IUI and do not succeed, when should they switch over to IVF? A similar study to the one above measured this by enrolling two groups of patients with unexplained infertility: those who would do 3 IUI cycles and then proceed to IVF and those who would continue on with an additional 3 IUI cycles (doing 6 in total) before undergoing IVF.
In essence, both groups enjoyed high rates of success, but the group that started IVF after only 3 failed IUI cycles were successful about 3 months earlier than the group that underwent 6 IUI cycles before transition to IVF, if necessary.
IUI success rates vary by a woman’s age, which drugs (if any) accompany the cycle, and how many follicles grow during the cycle
On average 5 - 15% of IUI cycles lead to a live birth with gonadotropin (injectable drug) cycles performing best followed by clomid or letrozole cycles
For an average patient, IUI paired with letrozole or clomid delivers the best mix of decent birth rates with limited risks of multiple gestations. But, if you don’t ovulate using clomid or letrozole, or if you have hypothalamic dysfunction, then gonadotropins are necessary
For women with PCOS, letrozole is superior to clomid. In this group, IUIs with clomid or letrozole often do no better than taking drugs alone, with a few exceptions. Success rates are higher with IUI plus gonadotropins but so is the risk of multiples. In these patients, IVF can be especially valuable to lower the risk of multiples
For women who have no periods or irregular periods who are diagnosed with hypothalamic dysfunction (their FSH and LH are almost zero), the only option for an IUI cycle is to use gonadotropins
For patients with male factor infertility, success rates vary by a man’s total motile count, with disagreement about how likely IUI with a TMC of less than 10 million are likely to work
There are two TMC measures: pre-wash and post-wash. Most men can calculate their pre-wash TMC (directly from the semen analysis), but TMC numbers go down post-wash and the numbers cited in this course above are post-wash. Generally-speaking, a man can subtract 30% from his pre-wash TMC levels to arrive at his projected post-wash TMC levels and then apply his post-wash TMC to the odds-of-success tables. However, the actual pre and post wash differential will differ from man to man. If you have done an IUI and have your real post wash numbers, use that for the odds-of-success tables
For women with one blocked fallopian tube, IUI success rates vary depending upon where the blockage exists. Women with an HSG-diagnosed “proximal” blockage do better partially because in 35% of cases the HSG diagnoses a blockage here when none exists. Risk of ectopic pregnancies with IUI are higher in patients who have one blocked tube
For women without an infertility diagnosis (like lesbians or single women), success rates vary by age and, if possible, using fresh sperm is preferable to frozen and thawed sperm. Frozen sperm is often the only available option from a sperm bank and does work
Multiples-per-pregnancy risk is higher in gonadotropin IUI cycles than in clomid or letrozole IUI cycles. While multiples-per-pregnancy rates can be high with IUI (8 - 30%), since pregnancy is so seldom achieved, only 0.5 - 2.5% of IUI cycles actually culminate in a multiple birth. Patients most susceptible are often under 35 years old or have PCOS
While IUI success rates are lower than with IVF (2 -10x), so are costs (5 - 40x) and in early IUI cycles, cost-per-live-birth can be lower than in IVF. However, often after 3 cycles each extra IUI cycle is unlikely to work and in those cycles IUI’s cost-per-live-birth become unacceptably high
Patients starting with IVF instead of IUI have a quicker time to pregnancy. But presuming patients who fail IUI are able to do IVF after only a few IUI cycles, the likelihood of eventually having a child in both groups is about the same