The recently released preliminary 2015 SART clinic outcomes statistics for US clinics have a lot of people in a panic. Perhaps they were most shocking when comparing donor egg cycles, and that’s the focus of this post. Reputable clinics in some cases seemingly have alarmingly lower live birth results per fresh cycle reported, and it was hard at first glance to decipher exactly why. I had some theories, and after doing some digging and talking to clinics, I think I’ve figured out a few things that might be helpful for patients in understanding what they are reporting.
The purpose of this is to help patients more effectively know what they are looking at when viewing the SART statistics as a whole and for any specific clinic, so that they can have a more meaningful discussion with their doctor about their own case. SART is very careful to warn people about how they should NOT use this data, but the practical reality is that patients will compare clinics and do initial narrowing of choices based on this data, so despite what they say, I feel that it’s important for patients to understand the relevance of the data reported and factors that may skew perceptions as a result. Here’s the important SART disclaimer:
“Accurate and complete reporting of ART success rates is complicated. Clinics may have differences in patient selection, treatment approaches, and cycle reporting practices which may inflate or lower pregnancy rates relative to another clinic. This report is best understood in consult with your physician. Use of the data in the report for comparing clinics, ranking clinics, making insurance coverage decisions, discouraging patients from seeking care at a given clinic, or for any other commercial purposes is strictly prohibited. By accessing this report, you agree not to use the data in the report for any of the prohibited purposes, and you further agree to abide by the SART Website Terms and Conditions of Use.”
First, my perspective is that in this new reporting format, SART is doing patients a disservice. While their intentions with the report format updates were good, the practical ramifications of this new reporting format are that patients are even more confused and in some cases alarmed, and they can’t figure out why the results appear so drastically different than what is being reported by the CDC for the same clinics in the same time period. As a result, we might even see patients relying more heavily on the CDC data simply because they can understand it better.
What I noticed with the 2015 reporting is that SART is now prominently displaying the live birth rate per cycle start rather than per transfer. One could argue that the per transfer statistic is what patients have come to expect to see and is more consistent with what we see reported in other countries, although most of those are reporting clinical pregnancy rates per transfer, which is really more reflective of the time that the RE is in charge of the patient. And while in some cases the per cycle start results might be helpful, the outcome reporting is further impacted by the fact that many clinics are doing many more CCS/freeze all cycles and that is further skewing the results, because those cancelled cycles get reported in the fresh “cycle starts”.
Let me give an example – let’s say clinic A does 260 donor egg cycles, 60 of them are intended for fresh transfer and 200 are intended to be PGS/freeze all cycles. Let’s say that of those 260 cycles, 5% of those cancelled prior to retrieval (3 of the fresh cycles and 10 of the CCS/freeze all cycles) and the rest went to transfer. And let’s suppose that they have a 75% live birth rate per TRANSFER. Let’s say that clinic B does only fresh cycles and the exact same number of fresh cycles as clinic A – 60 cycles with a 5% cancellation are. Here’s how the results reporting would vary comparing live birth per transfer vs. live birth per cycle start for each of those clinics despite having the exact same number of cases/live births for those fresh cycles:
Fresh donor egg starts = 70 (60 fresh + the 10 cancelled cycles that were intended for CCS/freezing that end up getting categorized here since they cancelled)
Fresh donor transfers = 57
Number of live births = 43
Live births per transfer = 75%
Live birth per cycle start = 61.43%
Fresh donor egg starts= 60
Fresh donor egg transfers = 57
Number of live births = 43
Live births per transfer = 75%
Live births per cycle start = 71.67%
The more CCS/freeze cycles that get cancelled, the more those live birth per start numbers get skewed. Practically speaking – both of those clinics did the exact same number of fresh cycles, exact same number of transfers, and had the exact same live births per transfer – but SART would show one as 71% and the other as 61%. Most patients would consider that a big difference! Yet, the real cycle start and live birth cases were exactly the same in both clinics in my example.
As a patient, should I care about how many cycles are cancelled? Yes, absolutely, but that information is NOT a priority over the live birth per transfer information and should not have replaced it. It also should not be skewed by nuances in reporting cancelled cycles. A non-clinician should be able to easily see how many cycles were started, cancelled, and went to transfer, but typically the patient’s primary interest is in what their chances are for success for each transfer (ideally information to compare eSET results to multiple embryo results).
So, what happened to the results of all of the CCS/freeze all cycles from Clinic A? Well, those are now going to be reported the year that they are transferred, with all of the other donor egg frozen embryo transfers. So, the CCS tested embryos are lumped in with the donor egg embryos that were frozen 5 years ago and with the ones that were left after higher quality embryos were already used. We’d expect a clinic doing a large percentage of CCS cycles to have better than average frozen embryo transfer live birth rate per transfer (and per cycle) and in several clinics that I randomly checked, the reported live birth rates for thawed embryos were higher than their fresh donor egg cycle rates. So those might give someone a false sense of the likely outcome if they are comparing their older stored embryos to the CCS tested thawed embryo results.
Another percentage that I think is misleading is the SART expression of singleton, twin, triplet or more live births. These are being reported as a percentage of cycle starts – which is not terribly helpful and perhaps even somewhat misleading.
Let’s go back to our example clinics:
Clinic A is reported as 70 fresh donor egg starts; Clinic B is reported as 60 fresh donor egg starts
Both had fresh donor transfers = 57
Both had number of live births = 43
For both, of the 43 live births, 30 were singleton, 12 were twins; 1 was triplet or more.
Because SART reports this based on cycle starts, SART will report as follows:
Clinic A: 42.86% singleton, 17.14% twins and 1.4% were triplet or more
Clinic B: 50% singleton, 20% twins and 1.67% were triplet or more
In reality, of the pregnancies that delivered, 69.8% were singleton, 28% were twins and 2.3% were triplet or more – so over 30% were multiple pregnancies! Suddenly it appears that maybe the clinics should be working on improving their singleton pregnancy rates, whereas in the SART reporting, it seems that they’re doing pretty good by having close to 20% or fewer multiple gestation pregnancies.
Would that reinterpretation of the multiple gestation pregnancy rate potentially impact an intended parent’s choice of transferring one or two embryos? Maybe. It’s certainly worth exploring more. Most intended parents would think harder about the risk of multiple gestation pregnancy rate at over 30% vs. 17%. However, I trust that the ethical doctors are having conversations to properly inform patients of the risks of transferring more than one embryo at a time.
So, what do you think about the new changes to the SART reporting? You can read more on the 2015 preliminary results page about how they are classifying cases. In the section at the top titled “understand this report” be sure to choose “read more” so that you get the full explanation of the SART changes. Also, at the bottom of that page, they offer you the option to give them feedback about the reporting, I encourage you to do so – whether you love the new reporting or not, it’s helpful for them to hear from actual patients who are using the data!
Again, the information that I’m sharing is simply intended to help patients more effectively know what they are looking at when viewing the SART statistics as a whole and for any specific clinic. The goal is to be able to have a more meaningful discussion with the doctor specific to an individual case. Nothing can take the place of having an informed conversation with the doctor!