You are here
It Is Virtually Impossible to Interpret Fertility Clinic Success Rates
a blog by Rhonda Levy, March 20, 2014
It has never been more difficult for a patient entering an American fertility clinic to understand her likelihood of being able to have a baby with its assistance. Although the success rates of in vitro fertilization (IVF) achieved by individual fertility clinics are publicly available, the relative utility of the reporting structure has been eclipsed by new advances in the techniques of assisted reproduction. Recent changes in the approach to care by IVF programs have made it virtually impossible for the general public to interpret pregnancy outcome data.
The success rates of IVF and other assisted reproductive technology (ART) procedures have been compiled and published since 1985 by the Society for Assisted Reproductive Technology (SART), and, since 1997, by the Centers for Disease Control (CDC). These steps were taken in response to a congressional act that recognized that patients who require IVF are consumers entitled to protection. Without access to fertility clinic success rates, patients would be forced to take blind leaps of faith, making them even more vulnerable than they already are.
Nonetheless, the public availability of fertility clinic success rates has always been a double-edged sword. Competition for patients has motivated some clinics to engage in practices that put a positive spin on their success rates to attract market share, such as denying IVF to poor prognosis patients. However, the success rates of fertility clinics have never been less interpretable than in recent years, when preimplantation genetic screening (PGS) and "embryo banking" gained traction around the country. PGS refers to analyzing the cells of the embryo to ensure that the embryo has 46 chromosomes and not a different number. It was developed to address what has been one of the greatest challenges in ART, the inability of embryologists to identify normal embryos for uterine transfer in an IVF cycle. This identification is vitally important to the success of IVF, since only chromosomally normal embryos become healthy babies. Fifty percent or more of the embryos of women over 35 years of age may have an abnormal number of chromosomes.
"Embryo banking" is a process that women may engage in if they produce few eggs in one IVF treatment cycle. Often, these are women of advanced reproductive age. Doctors stimulate the ovaries with fertility drugs and retrieve the eggs over more than one menstrual cycle to enable the clinic to accumulate a more plentiful number of embryos for biopsy and PGS analysis.
Most clinics freeze the embryos while PGS analysis is being performed. If a normal embryo is identified, it is thawed and transferred into the uterus in a subsequent menstrual cycle, when some now believe it will be more likely to implant because the uterus will be in a more natural state, having recovered from the effects of the drugs used to stimulate the woman's ovaries. When circumstances permit, some clinics conduct the PGS analysis within 24 hours so that an embryo that is recognized as normal can be transferred “fresh”.
The original structure of fertility clinic success rate reports was devised long before PGS and "embryo banking" existed. In the past, fresh embryo transfers were the primary focus of IVF and frozen embryo transfers were the secondary focus. In clinics that engage in a high volume of PGS and "embryo banking" cycles, the primary focus is now reversed. This contributes in large part to the lack of transparency patients now face when attempting to interpret fertility clinic success rates.
The reporting rules allow for the exclusion of cycles from which an immediate outcome is not intended, including embryo banking cycles. As these treatments are generally pursued by more challenging patients, their exclusion as a “fresh” embryo transfer reflects positively on a clinic's success rates. Information about embryo banking cycles only appears when at least one embryo is identified as normal and transferred, without indication that multiple stimulations and transfers were involved. The transfer of a normal embryo will have high odds of resulting in the birth of a baby and this will reflect favourably on the clinic's frozen embryo transfer success rates.
Although the use of PGS is becoming more widespread with each passing day, some fertility clinics use it sparingly and transfer unscreened, fresh embryos, often because they believe that nature is the best selector of normal embryos, and they are concerned about the added cost associated with PGS. When this approach is taken, an outcome is reported for all of the clinic's cycles, including those involving older women who generally produce more abnormal than normal embryos. This reflects unfavourably on the clinic's success rates.
Given these divergent treatment patterns, patients cannot use the SART Registry or the CDC Reports to develop the insights and gain the protection they were originally intended to provide. Different types of cycles are lumped together, and patients are unable to identify the process that was engaged in for each one. Patients would benefit greatly from being able to isolate success rates for the type of cycle they are considering, but that is not an option.
SART has acknowledged that changes to the reporting structure are necessary to address the introduction of these new approaches to IVF. In October 2013, I attended the annual meeting of the American Society for Reproductive Medicine in Boston where, at 8 am one morning, it was "standing room only" when Dr. Kevin Doody, SART Registry Chair, outlined changes in the way success rates will be presented to the public beginning with treatments that occur in 2014.
I have been analyzing fertility clinic success rates since 1998, and can say with confidence that I have developed a deep understanding of their subtleties and nuances. However, as I sat in the Boston meeting, I felt overwhelmed by the changes in store and concerned that the consumer protection goal of making fertility clinic success rates publicly available will not be met if patients find the new format uninterpretable.
Under the revised format, each ovarian stimulation will be linked to the first or "primary" embryo transfer. If a normal embryo is not identified, this will be reported as a negative outcome. When possible, the outcome will be reported in the year in which the stimulations and retrievals occur. However, if the "primary" transfer is delayed to another year, the stimulations will be linked to the transfer in the year in which it takes place, provided the transfer takes place within 12 months of the cycle start.
With this approach, clinics that transfer biopsied and genetically screened embryos that have not been frozen will appear to be the most successful to the consumer. Because fresh embryos that are identified as having a normal number of chromosomes have high odds of resulting in a live birth, clinics that transfer them will be more likely to show a positive outcome than clinics that transfer only unscreened, fresh embryos.
In addition, if a clinic conducts two separate transfers of unscreened fresh embryos for the same patient and each transfer results in the birth of a single baby, the clinic will report two successes. However, if a clinic conducts two stimulations and retrievals for one patient, banks the embryos, screens them and transfers two thawed normal embryos resulting in the birth of twins, it will report only a single successful frozen transfer or, despite the birth of the same number of babies, half as many successes.
Moreover, clinics that screen embryos via PGS and determine that the patient has not produced a normal embryo are penalized under the revised format because these cycles are reported as unsuccessful. But are they really? When normal embryos are not available, it is not a failure on the part of the clinic but merely an unavoidable reality that it is impossible to change.
It remains to be seen whether the new reporting structure will motivate some clinics to practice medicine in a way that presents their success rates in the best possible light but that is not in their patients' best interests. For example, will clinics avoid PGS because their success rates will be negatively impacted when their patients do not produce normal embryos? When they use PGS will clinics transfer only fresh embryos identified as normal because this will reflect positively on their fresh embryo success rates, despite some research data that suggests that embryos are more likely to implant when frozen and transferred into the uterus in a subsequent unstimulated cycle?
This begs the question, would it have been preferable to devise a reporting structure that did not lump all types of cycles in one pot, shining a more positive light on clinics that adopt certain practice approaches as compared with those that adopt others? SART has said that under the new format patients will be able to drill down beyond the single "pot" to obtain further information. This may be so, but will patients actually take that additional step? And, if they do, will the information provided be so complex that they will have difficulty interpreting it?
The motivation behind how fertility clinics practice should always be what is in their patients' best interests, rather than their desire to present themselves in a way most likely to attract greater market share. History has shown this has not always been the case. We must hope that, going forward, what is best for patients will be the only consideration. Fertility clinics have an ethical obligation to do the right thing.