Meet the new member of the IVF patient care team
Data scientists are the newest members of a digitally-enabled IVF Patient Care Team.
The need for IVF continues to grow dramatically around the world. Experts on the economics of fertility estimate that there could be 800 million pregnancies requiring medical intervention over the remainder of the century.
A few social determinates are driving this increased demand for medically-assisted reproduction (MAR). People are delaying childbearing longer than ever, which impacts every person on the spectrum of gender’s fertility potential. There are new options, like egg freezing. New laws, like LGBTQ marriage legalization. New technologies, like PGT, that allow even fertile individuals to use assisted reproductive technologies to choose the exact time of pregnancy, sex of the embryo transferred, or halt the continuation of genetic disorders in their family tree.
With the increase in MAR comes a pressing need to improve current processes. This can be accomplished by leveraging the huge amount of data generated in IVF clinics.
Integrating AI technology into clinical decision making brings with it a new set of requirements and regulations. As regulations unfold, it is imperative that regulators collaborate with industry and clinical teams to determine the guardrails.
The Organization for Economic Cooperation and Development (OECD) has defined a set of principles for the use of AI systems including transparency, explainability and testing for fairness, bias, robustness, and security. In Europe, the European Commission has published Ethics Guidelines for Trustworthy AI.
In the US, the White House recently unveiled its Blueprint for an AI Bill of Rights which defines guidelines for the design, use, and deployment of artificial intelligence (AI)-based tools across industries, including healthcare. The blueprint outlines five core principles: safe and effective systems, algorithmic discrimination protections, data privacy, notice and explanation, and human alternatives, consideration, and fallback.
Individual healthcare organizations, such as IVF clinics, generally lack the infrastructure to measure the consequences of AI algorithms once they are deployed. Honestly, they often lack basic tools beyond spreadsheets to monitor and measure clinical and laboratory KPIs. While data entered “by hand” are subject to errors (duplications, transpositions, mistyping), stricter oversight is necessary for digital automation tools.
One critical issue with the use of AI algorithms is bias. The stakes are high in healthcare and there is concern that AI can introduce bias of race, gender or other variables into the data analysis. As part of the IVF patient care team, data scientists need to be able to explain how the model is making a specific decision and prove that it is not making biased decisions that negatively affect patient care.
AI explainability and transparency are crucial.
As part of the team, data scientists will not only play a critical role in the clinical validation of AI but also to monitor algorithms in action in the fertility clinic. There is tension between the urgent need to deploy new technologies to improve the success of IVF and scale serve more people, and the need for lengthy and expensive, prospective clinical “RCT”s, randomized controlled clinical trials. Industry players should welcome working closely with organizations that already monitor clinical data (for example, SART, CDC, FDA, HFEA) for post-market surveillance of algorithms.
AI is everywhere in our lives. As algorithms and automation are integrated into fertility care, there is a need to remain focused on making IVF more accessible, affordable and successful for all hopeful families.
To receive information on how AI is transforming IVF patient care, please drop us a note.