Use of fertility treatments is growing with as many as 10% of births occurring after use of assisted reproductive technologies (ART) or non-IVF fertility treatments (NIFT). These treatments often require substantial societal and individual investments of resources and time, yet there is little information on the long-term health or development outcomes for children born after treatment. Indeed, most of the existing literature centers around events occurring during birth or infancy, with limited and mixed findings about events later in life. This limited literature also reflects major methodological challenges, including (i) insufficient follow-up time to assess the development of medical conditions past adolescence and into adulthood; (ii) numerous potential confounders, including the original reason for needing treatment (infertility), and variation in access to and use of treatments, e.g., differential access by income or education; and (iii) use of analytic approaches that either do not address relevant sources of bias, or rely on ill-defined assumptions to assess treatment effects. To both address the need for more evidence and mitigate the methodological concerns, we propose to compare the risk of outcomes associated with different fertility treatments using multiple population-based, nationwide rich datasets combined with modern causal inference approaches.
Our aims are to examine the impact of fertility treatments (stratified by the type of infertility) on two types of outcomes: 1) clinical events; and 2) educational events. We will study nearly 400,000 children born after fertility treatments during 1980-2020 in four populations (in Denmark, the Netherlands, Sweden, and Massachusetts). Across these populations, we have comprehensive, individual-level information on our outcomes and potential cofounders, e.g., types of infertility, treatments, education, income, and clinic traits. We will perform both predictive and causal analyses in order to inform clinicians and potential parents as they first consider use of fertility treatments, then choose between available treatments. Our statistical methods encompass a series of approaches starting with traditional survival and repeated measures analyses, and more recently developed techniques for causal analysis with observational data, e.g., inverse-probability weighting. We also will explore alternative approaches using instrumental variables and the g-formula. In short, this study provides the best opportunity to assess long-term outcomes in children born after fertility treatments, with the longest follow-up time, most extensive set of demographic, clinical, and treatment characteristics, and use of state-of-the-art analytic approaches.
Fertility treatments are very effective in helping individuals have children, yet are costly in terms of money and time. There also is little long-term information on the health and development of the child that could guide doctors and patients when choosing fertility treatments, as such information would require comprehensive information on large populations of children and their parents, and would need to follow these children for several decades. Thus, this project combines information from multiple countries to create the largest study to date with the longest follow-up available in the world.
