LEAVE US YOUR MESSAGE
contact us

Hi! Please leave us your message or call us at 01.800.123.456

Thank you! Your submission has been received!

Oops! Something went wrong while submitting the form

DATA

Can Science Make Personalized Predictions of First-Cycle IVF Success?

Bokyung Choi, Ph.D., Ernesto Bosch, M.D., Benjamin M. Lannon, M.D., Marie-Claude Leveille, Ph.D., Wing H. Wong, Ph.D., Arthur Leader, M.D., Antonio Pellicer, M.D., Alan S. Penzias, M.D., and Mylene W. M. Yao, M.D.

July 24, 2017
Emma Holt

I’m certain any woman going through IVF right now wishes someone could predict if their first-cycle would be a success or not. It’s a guessing game that no one wants to play. Luckily, a team of researchers from the University of Los Altos, California set out to test whether the probability of having a live birth with the first IVF cycle can be predicted for patients in diverse environments in a 2013 study.  Before we dive in though, it’s important to note why this study is so crucial. For starters, fewer than 3% of the estimated 7 million women and couples who are diagnosed with infertility have access to IVF. Obviously, there are many reasons as to why this could be such as the cost, limited insurance reimbursement, or success with other treatments. Researchers of this study predict that unclear benefits and unrealistic expectations weaken a patient’s confidence in IVF’s success rate, ultimately causing them to not even want to try. Potentially having the ability to predict success rate with IVF, personalized for each patient, is a total game changer. 

About the research

University-affiliated outpatient IVF clinics located in three different countries (Boston IVF in the U.S., IVI Valencia, and Ottawa, Canada) participated in this study to garner to most accurate results for various diverse patients. The cohort consisted of 13,076 first IVF treatment cycles using fresh autologous eggs and fresh embryo transfers. Patients underwent ovarian stimulation protocols and embryos were cultured according to each clinic’s standard protocols. (The numbers of embryos transferred was based on national and clinic guidelines as well as the patient’s needs). Each patient was then followed for at least 1 year from the start of the IVF cycles to confirm IVF and pregnancy outcomes. 

A PreIVF-Diversity (PreIVF-D) model was generated with the use of a multistep procedure. For each clinic-specific model, the researchers computed the log-likelihood and applied generalized boost models (GBM).   The researchers built PreIVF-Diversity (PreIVF-D) by blending the individual components from all three clinic-specific models to form a resulting model adjusted for the different number of cases available from each clinic. PreIVF-D was compared with an age-based control model (age model) that was generated from 10,957 cases by applying GBM to patient’s age based on age categories which are less than 35, 35-37, 38-40, and 41-42. These categories are used by the Society for Assisted Reproductive Technologies and Centers for Disease Control and Prevention. They also determined the probability of having a live birth in the first IVF cycle based on the collective profile of the patient and her male partner or the patient’s profile alone if a sperm donor was used. The authors claim predicative power can be described as the improvement in the log-likelihood of predicting the probability of having a live birth in the first IVF cycle with the PreIVF-D relative to the age model using Baseline-Diversity (Baseline-D). To put it simply, Baseline-Dis the mean probability of having a live birth in the first IVF cycle if age or any other predictor isn’t used. 

What the researchers found

The researchers reveal that 86% of cases had significantly different live birth probabilities compared with age control, and more than half had higher live birth probabilities. 42% of patients could’ve been identified by PreIVF-D to have a personalized prediction rate less than 45% whereas an age control model had no ability to differentiate them from others. Basically, the prediction of personalized live birth probabilities from diverse multiple women identify exceptional prognosis in more than half of the patients! 

While it’s unclear how many women hold back from pursuing their first IVF cycle because they don’t have faith in the process, if a small percentage of patients now feel more confident, then that improves the overall success rates and utilization of IVF. The authors of this study propose that the use of developed and validated personalized predictions in the infertility community will improve access and utilization of ART care. The desire here is to help a greater number of patience to build healthy families.

...

Also worth a read
Case Studies

Evaluating Optimal Treatment for Unexplained Infertility

This research was presented at the annual meeting of the American Society of Reproductive Medicine in Washington, D.C.

read more
Pregnancy

Using Assisted Reproduction Technology to Predict Pregnancy Outcomes

H.F. Solomon, R.M. Sneeringer, A.S. Penzias

read more
Female Fertility

Anticipating Risks of Personalized Multiple Births After IVF

Benjamin M. Lannon, M.D., Bokyung Choi, M.S., Michele R. Hacker, S.D., Laura E. Dodge, M.P.H., Beth A. Malizia, M.D., C. Brent Barrett, Ph.D., Wing H. Wong, Ph.D., Mylene W. M. Yao, M.D., Alan S. Penzias, M.D

read more
easy finder