What is the IVF success rate?
The cycles analyzed and how we verified the outcomes
We have followed closely the outcomes of almost 2000 treatment cycles from recent years in the Polyclinic and ART Center. Fresh non-donor IVF cycles (with or without ICSI) accounted for 56%, Frozen embryo transfers (FET) for 36% and Egg donations cycles for 8%.
All patients having treatment in our center were contacted 3 weeks after the ET to obtain their pregnancy test results. Many of the pregnant patients were followed up in our antenatal care (ANC) clinic, so pregnancy outcomes were obtained from their charts. Pregnant patients who were followed up in another clinic or lived out of town or the country were contacted periodically to obtain the relevant information. Pregnancies which manifested only elevations of the pregnancy hormone (hCG) and could not be visualized by transvaginal ultrasound (TVUS) were termed “chemical pregnancies”. Pregnancies in which one or more pregnancy sacs were observed by TVUS at 3-4 weeks after ET were termed “clinical pregnancies”. When calculating pregnancy success rates only clinical pregnancies were included. Clinical pregnancies which terminated in the first few months were classified as “miscarriages” and those which resulted in the delivery of a live baby (or babies) as “live births". Live Births were further divided into those which only one baby was born (“singleton live birth”) and those which involved the births of 2 or three babies (“multiple live birth”). Multiple live births usually involved twins or triplets, as we did not encounter a higher number of babies born.
In the average fresh IVF cycle we retrieved 13 eggs, 11 of which appeared under the microscope as mature eggs (ready for fertilization). Fertilization occurred in 75% of the mature eggs and normal embryos developed in more than 90% of the fertilized eggs. In the average cycle, we transferred 2.2 embryos and we froze about 5 embryos for later transfer (FET). In the average frozen embryo transfer (FET cycle) we thawed 4 embryos with more than 85% surviving the freeze-thaw procedure, but only 2.8 were finally transferred.
In vitro fertilization and fresh embryo transfer (IVF & ET)
Of all ovarian stimulation cycles (started) cycles, 94% underwent egg-retrieval. In the remaining 6% the main reason was failure to stimulate an adequate number of mature ovarian follicles.
Of all egg-retrieval cycles, almost 90% had ET and in 12% ET was cancelled, mostly because of ovarian hyperstimulation or patient convenience.
Of all embryo transfers more than 35% had a positive pregnancy test but 31% had clinical pregnancies observed by TVUS.
Of all clinical pregnancies 1.5% were not in the uterine cavity (ectopic), 20% were miscarriages and the remaining 78% resulted in live births.
Of all live births only 20% were of more than one baby (“multiple live births”), the great majority of which were twins.
Frozen embryo transfers (FET) cycles
Of all cycles where embryos were thawed in more than 97% embryos were transferred (ET) and in less than 3% ET was cancelled because the embryos did not survive the freeze-thaw procedure.
Of all ET cycles 31% had a positive pregnancy test and 26% had clinical pregnancies documented by TVUS.
Of all clinical pregnancies 2% were outside the uterine cavity (ectopic), 16% were miscarriages and 82% resulted in live births.
Of all live births 21% were of more than one baby (“multiple live births”), the great majority of which were twins.
Egg donation (ED) cycles
Of all egg donation cycles where either fresh or frozen embryos were transferred 49% had a positive pregnancy test and 42% were clinical pregnancies.
Of all clinical pregnancies 3% were ectopic 15% were miscarriages and 82% resulted in live births.
Of all live births 25% were of more than one baby.
Since any clinic success rates largely depend on its patients characteristics and populations cannot be equalized between different IVF centers, the comparison of success rates across clinics is inappropriate. However, comparison to treatment results of large registries which contain results from hundreds of clinics may be more appropriate as they are less susceptible to patient selection and the large number of cycles better reflects the “general population”. We have therefore chosen to present our success rates alongside with the most recent published results of the European Society of Human Reproduction and Embryology (ESHRE). Europe leads the world in ART, initiating approximately 71 % of all reported ART cycles (not including Asia). The ESHRE registry contains outcome data of close to half a million treatment cycles per year, collected systematically from 33 European countries. The results of the ESHRE registry are provided here only as “reference data” for our IVF success rates.
As previously stated the most meaningful outcome in terms of IVF success is the chance (or likelihood) of delivery or live birth. The following describes this data for IVF, FET and ED in cycles where embryos were transferred and provides some of the factors which can affect the success rate. It should be noted that ESHRE reports “deliveries” which may include also stillbirths, whereas our data refer to live births.
% Age >= 40 years Average no. Twins
IVF FET ED (% of patients) embryos transferred (% of deliveries)
Our Center 24 21 33 16.9% 2.6 21%
ESHRE 24 14 29 14.4% 2.0 20%
In conclusion:
Compared to the ESHRE reference data our Centers patients are slightly older, more embryos were transferred and the IVF delivery rates and multiple pregnancy rates were similar. However, FET and ED resulted in higher delivery rates per ET.
Definition
The first question is how we define “success” of an ART (IVF) treatment? While some may think that achieving a pregnancy is a successful outcome, it may be argued that some of these pregnancies may not be so “successful”. Thus, a “chemical pregnancy” (characterized only by a temporary rise in the pregnancy hormone), a “clinical pregnancy” (observed on ultrasound) which terminates in spontaneous miscarriage, or an ectopic pregnancy (not in the uterine cavity)- are definitely unsuccessful outcomes. There is no doubt that delivery (childbirth) is a better outcome measure of an IVF treatment, but when it is complicated by fetal death it may not represent “success”. Unfortunately not all ART clinics and registries report their results as the live birth rate for the cycle (% of couples whose treatment resulted in the delivery of a live baby). However, even this statistic is not ideal as some deliveries are very premature (after 6-7 months of pregnancy) with very small babies, which hardly survive (or not) and require long hospitalization in the NICU (Neonatal Intensive Care Unit). While such “early” deliveries rarely occur in singleton pregnancies they represent a “real risk” in multiple pregnancies with increasing number of fetuses (usually triplets and more). Since IVF treatments are associated with a higher chance of multiple pregnancies it is appropriate to report the rate of singleton live birth or the % of multiple pregnancies among the live births.
Factors which affect the success rate
There are multiple factors which affect the chances of achieving a pregnancy after an ART (IVF) treatment, as well as the “success rate” of the clinic. Among these the important factors are the age of the woman and her ovarian reserve, the duration of infertility and outcomes of previous infertility treatments, the male factor, the causes for infertility and the presence of several abnormalities together. In addition success rates depend on the response to the ovarian stimulation, the quality of the embryos in the culture and the number of embryos transferred to the uterus. While transferring more embryos will increase the success rate, the rate of multiple pregnancies is also increased.
Nature too, is ineffective and can result in an unsuccessful outcome
If you plan on an infertility treatment (or any other), you have to realize that the outcome is not always desired, it may be unfavorable and it cannot be predicted or guaranteed. You may find it comforting that “this is nature”. This is especially true when it comes to getting pregnant, carrying pregnancy, delivering and raising children. It is a long, curving road with some obstacles and hurdles. Just remember that under the most natural conditions, without any interventions, the chances of getting pregnant in any particular menstrual cycle for a fertile couple are only 20-30%. One of the reasons for this apparently low rate is early embryonic wastage. If sperm is available at the time of ovulation, we will find relatively high rates of fertilization and early embryonic development. However, a large proportion of these embryos does not develop appropriately or fail to implant in the uterus and this is even before the expected menstrual period. So by nature, many “lives” are lost only few days after they started (and we are not aware of it). Also during IVF, like in the natural process, only a minority of the cultured embryos will implant in the uterus. But even if the embryo implants successfully and pregnancy is “recognized” (missed menstrual period), another 15-20% of these pregnancies will not progress and terminate as miscarriage (spontaneous abortion) in the first few months. While losses may also occur beyond the first few months of pregnancy, these fetal losses (called “stillbirths) are relatively rare. Finally, you may find it reassuring a large percent of the embryos who fail to implant or develop is chromosomally or structurally abnormal. So indeed, “nature can be ineffective”.
The nature of statistics (from sampled populations to the individual case)
As previously noted every treatment, procedure or intervention can have several outcomes, favorable or unfavorable. Our best approach to this problem is to use statistics (rates or %) to express the chances (or likelihood) for a desired, or undesired, outcome. However, this is only an estimate and it is very limited. The reason is that we can derive “statistics” for a group of patients with certain common characteristics (=population), but it is impossible to accurately determine the chances in a particular case. Thus every person or a couple has a UNIQUE set of circumstances that will determine his own likelihood of achieving a successful pregnancy, delivery or an unfavorable outcome.
Differences in the populations for which success rates are reported (the denominators)
A major issue in assessing success rates is what kind of population (or group of patients) was used to derive these rates, or in other words: what is the denominator? The denominator in this calculation can be the number of couples entering treatment (ovarian stimulation), or those going to egg retrieval (ER), or those having embryos replaced into the uterus (ET). Since not all couples who start treatment reach the embryo transfer (ET) stage, the success rates are different for each group and are highest for those undergoing ET. While it is acceptable to present rates only for one population/group, make sure that you interpret these rates in the same context. This means that if the success rates are expressed as livebirths per ET, these are the chances for a livebirth in couples who had ET. The same applies to couples who undergo frozen embryo transfer (FET): rates may be expressed for those who initiated the hormonal preparation, those who had embryo thawed or those who had ET.
Selecting patients and strategies to improve success rates (or how do we look better?)
The misguided focus on "success rates" has created strong incentives, economic and otherwise, for ART centers to maximize IVF statistics by “patient selection”. We can improve a clinic’s success rates by encouraging “ideal” patients to undergo treatment and discouraging “difficult” patients from having treatment. Thus by excluding or wait-listing individuals who have failed in other programs, are over 38 years old, have low ovarian reserve, have prolonged unexplained infertility, or are low responders (i.e., “difficult patients”), certain centers can improve their statistics. Similarly, higher success rates can be achieved by facilitating or strongly advocating IVF treatment to “ideal patients”, who are relatively young (less than 35), with single factor or borderline infertility, who had prior successful pregnancies and/or IVF treatments, normal semen quality, etc,. This type of “selection” is usually invisible and not apparent in published or advertised statistics. So, the key to good statistics is to have more “ideal” patients than “difficult” patients enter the program.
There are no geniuses or magicians
While there are many talented individuals in the IVF field, there are no individuals, neither physicians nor embryologists, with supernatural powers. Medicine has become very standardized and we are all following the same protocols and guidelines. Although the quality of health care and laboratory techniques are important, differences in clinics success rates are more often due to the given mix of patients who present for treatment.
The bottom line
Success rates are merely estimates and provide little meaningful guidance for an individual couple hoping to achieve pregnancy with IVF. The difference between a program with good statistics and one with less favorable pregnancy rates is more often due to the given mix of “ideal” and “difficult” patients who present for treatment. While experts in the field and statisticians have acknowledged this, the public does not generally understand this and places increasing emphasis on the statistics from individual centers. Don’t forget that clinics with the best statistics do not necessarily provide the best health care.
ART or IVF “is a numbers game”!
IVF simply increases the efficiency of human reproduction, which is often not very efficient naturally. By providing several embryos we “compress” many months of "natural" attempts into one treatment cycle.
We have followed closely the outcomes of almost 2000 treatment cycles from recent years in the Polyclinic and ART Center. Fresh non-donor IVF cycles (with or without ICSI) accounted for 56%, Frozen embryo transfers (FET) for 36% and Egg donations cycles for 8%.
All patients having treatment in our center were contacted 3 weeks after the ET to obtain their pregnancy test results. Many of the pregnant patients were followed up in our antenatal care (ANC) clinic, so pregnancy outcomes were obtained from their charts. Pregnant patients who were followed up in another clinic or lived out of town or the country were contacted periodically to obtain the relevant information. Pregnancies which manifested only elevations of the pregnancy hormone (hCG) and could not be visualized by transvaginal ultrasound (TVUS) were termed “chemical pregnancies”. Pregnancies in which one or more pregnancy sacs were observed by TVUS at 3-4 weeks after ET were termed “clinical pregnancies”. When calculating pregnancy success rates only clinical pregnancies were included. Clinical pregnancies which terminated in the first few months were classified as “miscarriages” and those which resulted in the delivery of a live baby (or babies) as “live births". Live Births were further divided into those which only one baby was born (“singleton live birth”) and those which involved the births of 2 or three babies (“multiple live birth”). Multiple live births usually involved twins or triplets, as we did not encounter a higher number of babies born.
In the average fresh IVF cycle we retrieved 13 eggs, 11 of which appeared under the microscope as mature eggs (ready for fertilization). Fertilization occurred in 75% of the mature eggs and normal embryos developed in more than 90% of the fertilized eggs. In the average cycle, we transferred 2.2 embryos and we froze about 5 embryos for later transfer (FET). In the average frozen embryo transfer (FET cycle) we thawed 4 embryos with more than 85% surviving the freeze-thaw procedure, but only 2.8 were finally transferred.
In vitro fertilization and fresh embryo transfer (IVF & ET)
Of all ovarian stimulation cycles (started) cycles, 94% underwent egg-retrieval. In the remaining 6% the main reason was failure to stimulate an adequate number of mature ovarian follicles.
Of all egg-retrieval cycles, almost 90% had ET and in 12% ET was cancelled, mostly because of ovarian hyperstimulation or patient convenience.
Of all embryo transfers more than 35% had a positive pregnancy test but 31% had clinical pregnancies observed by TVUS.
Of all clinical pregnancies 1.5% were not in the uterine cavity (ectopic), 20% were miscarriages and the remaining 78% resulted in live births.
Of all live births only 20% were of more than one baby (“multiple live births”), the great majority of which were twins.
Frozen embryo transfers (FET) cycles
Of all cycles where embryos were thawed in more than 97% embryos were transferred (ET) and in less than 3% ET was cancelled because the embryos did not survive the freeze-thaw procedure.
Of all ET cycles 31% had a positive pregnancy test and 26% had clinical pregnancies documented by TVUS.
Of all clinical pregnancies 2% were outside the uterine cavity (ectopic), 16% were miscarriages and 82% resulted in live births.
Of all live births 21% were of more than one baby (“multiple live births”), the great majority of which were twins.
Egg donation (ED) cycles
Of all egg donation cycles where either fresh or frozen embryos were transferred 49% had a positive pregnancy test and 42% were clinical pregnancies.
Of all clinical pregnancies 3% were ectopic 15% were miscarriages and 82% resulted in live births.
Of all live births 25% were of more than one baby.
Since any clinic success rates largely depend on its patients characteristics and populations cannot be equalized between different IVF centers, the comparison of success rates across clinics is inappropriate. However, comparison to treatment results of large registries which contain results from hundreds of clinics may be more appropriate as they are less susceptible to patient selection and the large number of cycles better reflects the “general population”. We have therefore chosen to present our success rates alongside with the most recent published results of the European Society of Human Reproduction and Embryology (ESHRE). Europe leads the world in ART, initiating approximately 71 % of all reported ART cycles (not including Asia). The ESHRE registry contains outcome data of close to half a million treatment cycles per year, collected systematically from 33 European countries. The results of the ESHRE registry are provided here only as “reference data” for our IVF success rates.
As previously stated the most meaningful outcome in terms of IVF success is the chance (or likelihood) of delivery or live birth. The following describes this data for IVF, FET and ED in cycles where embryos were transferred and provides some of the factors which can affect the success rate. It should be noted that ESHRE reports “deliveries” which may include also stillbirths, whereas our data refer to live births.
% Age >= 40 years Average no. Twins
IVF FET ED (% of patients) embryos transferred (% of deliveries)
Our Center 24 21 33 16.9% 2.6 21%
ESHRE 24 14 29 14.4% 2.0 20%
In conclusion:
Compared to the ESHRE reference data our Centers patients are slightly older, more embryos were transferred and the IVF delivery rates and multiple pregnancy rates were similar. However, FET and ED resulted in higher delivery rates per ET.
Definition
The first question is how we define “success” of an ART (IVF) treatment? While some may think that achieving a pregnancy is a successful outcome, it may be argued that some of these pregnancies may not be so “successful”. Thus, a “chemical pregnancy” (characterized only by a temporary rise in the pregnancy hormone), a “clinical pregnancy” (observed on ultrasound) which terminates in spontaneous miscarriage, or an ectopic pregnancy (not in the uterine cavity)- are definitely unsuccessful outcomes. There is no doubt that delivery (childbirth) is a better outcome measure of an IVF treatment, but when it is complicated by fetal death it may not represent “success”. Unfortunately not all ART clinics and registries report their results as the live birth rate for the cycle (% of couples whose treatment resulted in the delivery of a live baby). However, even this statistic is not ideal as some deliveries are very premature (after 6-7 months of pregnancy) with very small babies, which hardly survive (or not) and require long hospitalization in the NICU (Neonatal Intensive Care Unit). While such “early” deliveries rarely occur in singleton pregnancies they represent a “real risk” in multiple pregnancies with increasing number of fetuses (usually triplets and more). Since IVF treatments are associated with a higher chance of multiple pregnancies it is appropriate to report the rate of singleton live birth or the % of multiple pregnancies among the live births.
Factors which affect the success rate
There are multiple factors which affect the chances of achieving a pregnancy after an ART (IVF) treatment, as well as the “success rate” of the clinic. Among these the important factors are the age of the woman and her ovarian reserve, the duration of infertility and outcomes of previous infertility treatments, the male factor, the causes for infertility and the presence of several abnormalities together. In addition success rates depend on the response to the ovarian stimulation, the quality of the embryos in the culture and the number of embryos transferred to the uterus. While transferring more embryos will increase the success rate, the rate of multiple pregnancies is also increased.
Nature too, is ineffective and can result in an unsuccessful outcome
If you plan on an infertility treatment (or any other), you have to realize that the outcome is not always desired, it may be unfavorable and it cannot be predicted or guaranteed. You may find it comforting that “this is nature”. This is especially true when it comes to getting pregnant, carrying pregnancy, delivering and raising children. It is a long, curving road with some obstacles and hurdles. Just remember that under the most natural conditions, without any interventions, the chances of getting pregnant in any particular menstrual cycle for a fertile couple are only 20-30%. One of the reasons for this apparently low rate is early embryonic wastage. If sperm is available at the time of ovulation, we will find relatively high rates of fertilization and early embryonic development. However, a large proportion of these embryos does not develop appropriately or fail to implant in the uterus and this is even before the expected menstrual period. So by nature, many “lives” are lost only few days after they started (and we are not aware of it). Also during IVF, like in the natural process, only a minority of the cultured embryos will implant in the uterus. But even if the embryo implants successfully and pregnancy is “recognized” (missed menstrual period), another 15-20% of these pregnancies will not progress and terminate as miscarriage (spontaneous abortion) in the first few months. While losses may also occur beyond the first few months of pregnancy, these fetal losses (called “stillbirths) are relatively rare. Finally, you may find it reassuring a large percent of the embryos who fail to implant or develop is chromosomally or structurally abnormal. So indeed, “nature can be ineffective”.
The nature of statistics (from sampled populations to the individual case)
As previously noted every treatment, procedure or intervention can have several outcomes, favorable or unfavorable. Our best approach to this problem is to use statistics (rates or %) to express the chances (or likelihood) for a desired, or undesired, outcome. However, this is only an estimate and it is very limited. The reason is that we can derive “statistics” for a group of patients with certain common characteristics (=population), but it is impossible to accurately determine the chances in a particular case. Thus every person or a couple has a UNIQUE set of circumstances that will determine his own likelihood of achieving a successful pregnancy, delivery or an unfavorable outcome.
Differences in the populations for which success rates are reported (the denominators)
A major issue in assessing success rates is what kind of population (or group of patients) was used to derive these rates, or in other words: what is the denominator? The denominator in this calculation can be the number of couples entering treatment (ovarian stimulation), or those going to egg retrieval (ER), or those having embryos replaced into the uterus (ET). Since not all couples who start treatment reach the embryo transfer (ET) stage, the success rates are different for each group and are highest for those undergoing ET. While it is acceptable to present rates only for one population/group, make sure that you interpret these rates in the same context. This means that if the success rates are expressed as livebirths per ET, these are the chances for a livebirth in couples who had ET. The same applies to couples who undergo frozen embryo transfer (FET): rates may be expressed for those who initiated the hormonal preparation, those who had embryo thawed or those who had ET.
Selecting patients and strategies to improve success rates (or how do we look better?)
The misguided focus on "success rates" has created strong incentives, economic and otherwise, for ART centers to maximize IVF statistics by “patient selection”. We can improve a clinic’s success rates by encouraging “ideal” patients to undergo treatment and discouraging “difficult” patients from having treatment. Thus by excluding or wait-listing individuals who have failed in other programs, are over 38 years old, have low ovarian reserve, have prolonged unexplained infertility, or are low responders (i.e., “difficult patients”), certain centers can improve their statistics. Similarly, higher success rates can be achieved by facilitating or strongly advocating IVF treatment to “ideal patients”, who are relatively young (less than 35), with single factor or borderline infertility, who had prior successful pregnancies and/or IVF treatments, normal semen quality, etc,. This type of “selection” is usually invisible and not apparent in published or advertised statistics. So, the key to good statistics is to have more “ideal” patients than “difficult” patients enter the program.
There are no geniuses or magicians
While there are many talented individuals in the IVF field, there are no individuals, neither physicians nor embryologists, with supernatural powers. Medicine has become very standardized and we are all following the same protocols and guidelines. Although the quality of health care and laboratory techniques are important, differences in clinics success rates are more often due to the given mix of patients who present for treatment.
The bottom line
Success rates are merely estimates and provide little meaningful guidance for an individual couple hoping to achieve pregnancy with IVF. The difference between a program with good statistics and one with less favorable pregnancy rates is more often due to the given mix of “ideal” and “difficult” patients who present for treatment. While experts in the field and statisticians have acknowledged this, the public does not generally understand this and places increasing emphasis on the statistics from individual centers. Don’t forget that clinics with the best statistics do not necessarily provide the best health care.
ART or IVF “is a numbers game”!
IVF simply increases the efficiency of human reproduction, which is often not very efficient naturally. By providing several embryos we “compress” many months of "natural" attempts into one treatment cycle.