Recommendation to improve the rigor and impact of nonrandomized studies of interventions in fertility treatment research.

Objective: To provide a framework for conducting rigorous nonrandomized studies of interventions in fertility treatment research, addressing their role as complements to randomized controlled trials (RCTs) in evaluating treatment outcomes.

Design: Multidisciplinary expert consensus on best practices for nonrandomized studies of interventions, informed by advancements in novel methodologies, including causal inference.

Subjects: Patients undergoing assisted reproductive technologies (ARTs) procedures, such as ovarian stimulation, laboratory techniques, and embryo transfer.

Intervention: None.

Main Outcome Measures: Guidance on methodological rigor, transparency, and relevance in nonrandomized studies of interventions study design and analysis.

Results: Randomized controlled trials are the gold standard for determining the efficacy and safety of fertility treatment/ART interventions but can face logistical, practical, and sometimes ethical challenges. Nonrandomized studies of interventions, when conducted with high methodological rigor, complement RCTs by offering insights into real-world clinical practices and diverse patient populations. Key limitations of nonrandomized studies of interventions include susceptibility to confounding and selection bias, which require meticulous study design and advanced analytical techniques to address. Recent innovations, such as target trial emulation studies, have enhanced the validity of causal inferences based on nonrandomized studies of interventions. This article outlines 7 recommendations to improve the credibility of nonrandomized studies of interventions in ART research: clearly define research questions with precise estimands; design nonrandomized studies of interventions as emulated trials; use directed acyclic graphs to clarify causal assumptions; preregister study protocols; separate data analysis from study planning; incorporate negative controls to detect biases; and use appropriate analytical methods to account for confounding and selection bias.

Conclusion: Integrating evidence from RCTs and well-conducted nonrandomized studies of interventions enhances clinical decision making in fertility treatment research. By adhering to these recommendations, researchers can improve the quality, transparency, and impact of nonrandomized studies of interventions, ultimately fostering robust, evidence-based clinical practices in fertility treatment/ART.