Univfy's AI Models Outperform National Registry in Predicting IVF Success
In a groundbreaking development for fertility treatments, a new study reveals that Univfy's predictive models for live birth rates following in vitro fertilization (IVF) are significantly more effective than those based on U.S. national registries. This study, published in the esteemed journal Nature Communications, highlights the crucial metrics that can enhance access to IVF, its financial viability, and overall clinical outcomes.
Located in the Bay Area of San Francisco, California, Univfy is a leading innovator in fertility and health AI technologies. The peer-reviewed study, titled Machine learning center-specific models show improved live birth predictions over US national registry-based model, showcases the clinical and real-world validation of Univfy's unique models designed to predict the likelihood of live births after IVF. These advanced metrics play a vital role in establishing cost-effective solutions and improving the accessibility of IVF treatments.
Specifically, the study demonstrated that Univfy's models surpassed the performance of the national registry model in several critical areas, including the F1 scorean important measure of a model's accuracy that combines precision and recalland the area under the precision-recall curve (PR AUC). These improvements are not merely academic; they signify a model's ability to minimize false positives and false negatives, essential qualities for real-world solutions that support value-based care and actuarial models.
To put the findings into perspective, the study analyzed data from 4,645 patients across six centers. Univfy's models accurately predicted that 76% of these patients had a 50% or higher chance of achieving a live birth after their first cycle of IVF. Even more striking, 23% of patients who were assigned a probability of live birth (PNV) of 50% or greater by Univfys model were classified with a lower probability by the U.S. national model. Furthermore, Univfy successfully predicted that 11% of patients had a PNV of 75% or higher, correlating with an actual live birth rate of 81%, while the national registry model failed to assign a PNV of 75% or greater to any of those patients.
This publication signifies a significant milestone in validating the scientific principles that underpin Univfys AI and machine learning platform, highlighting its potential to facilitate economically viable solutions such as value-based IVF care. The results build upon previous research that indicated a 2 to 3 times increase in IVF utilization when patients received recommendations based on Univfy's pre-IVF report. It also emphasizes the importance of employing locally validated AI-optimized solutions within real clinical contexts, covering not just facilities in the United States but also those in the UK and the EU.
Dr. Mylene Yao, CEO and co-founder of Univfy, remarked, This publication is a testament to the scientific rigor of Univfys AI/ML platform, allowing us to accurately predict the excellent IVF outcomes achieved by our collaborators and the broader IVF ecosystem, which are conventionally underappreciated. She continued, Univfy was founded to enhance patient-centered care, particularly through prognostic counseling for IVF. With our efforts to improve cost transparency and IVF success rates, we have established a platform that enables the scalable production of validated, cost-effective solutions that benefit patients, providers, and healthcare stakeholders. We are urgently committed to helping more women and couples access and afford IVF to start their families.
Key Benefits of Findings for Stakeholders:
- For Patients: Enhanced transparency regarding IVF costs and success rates helps inform fertility care and family planning decisions.
- For Healthcare Providers: Improved patient counseling and streamlined clinical workflow enhances efficiency and avoids underestimating IVF effectiveness and treatment delays.
- For Insurers and Benefits Programs: Better member experiences and transparency, supporting providers and expanding IVF coverage with predictability and cost savings.
About Univfy
Founded as a Series B company in the Bay Area of San Francisco, Univfy focuses on enhancing the success rate, access, and financial feasibility of IVF. Developed by researchers from Stanford University, Univfys proprietary AI and machine learning platform provides precise, personalized, and validated pre-treatment probabilities of having a child through IVF and other therapeutic options, facilitating informed decision-making for each patient. Univfy's platform also enables providers to offer value-based IVF pricing at scale, making IVF more affordable. The company has marketed solutions to providers in the United States, the UK, and the EU through a B2B model to assist healthcare providers with patient counseling, clinical analysis, and business analytics, including client relationship management tools. Univfy's AI/ML platform is designed to support enterprise-level health plans and benefits programs to achieve savings by enhancing patient-centered care and facilitating value-based IVF delivery. The technology and products developed by Univfy are protected by a portfolio of intellectual property rights in the U.S. and worldwide, including issued and pending patents and copyrights.
Understanding Infertility, IVF, and the IVF Market
Infertility refers to the need for medical assistance to conceive a child, affecting one in six individuals of reproductive age. This translates to over 200 million people globally, with about 7 to 10 million in the U.S., 25 million in the EU, 3 to 5 million in the UK, and 186 million in other countries. Annually, approximately 4 million IVF treatments are performed worldwide, resulting in over one million births each year. In the U.S., around 2% of babies born each year are conceived through IVF. Despite being a safe and effective procedure, several cycles might be necessary. High costs, lack of insurance coverage, and uncertainty of success remain significant barriers, leading to a dismal IVF utilization rate of about 3% in the U.S. Patients struggling with fertility often find it challenging to afford and access IVF treatments.
The above estimates are based on reports from the World Health Organization (WHO), the American Society for Reproductive Medicine (ASRM), RESOLVE: The National Infertility Association, the Centers for Disease Control and Prevention (CDC) in the U.S., the European Society for Human Reproduction and Embryology (ESHRE), and others. A comprehensive list of references is available here.
The global and U.S. IVF market has seen significant consolidation by private equity, with the IVF market size estimated at approximately $25 billion in 2023 and projected to reach around $44 billion by 2033, with a compound annual growth rate (CAGR) of 5.57% between 2024 and 2033 (Biospace, April 2024).
Media Contact:
