Federated Search Engine Architecture For Donor Tracking in ART Clinics

Abstract

Clinics and sperm banks offering Assisted Reproductive Technology (ART) make extensive use of donor sperm and oocytes; however, information related to donors is often dispersed among these banks. Such repeated use of the same donor across clinics results in significant births from the same donor. It also raises critical ethical, legal, and public health concerns, such as unintended genetic relationships among offspring, and the inability to perform long-term monitoring. While centralized donor databases have been proposed earlier as a solution, they raise substantial concerns regarding data privacy, regulatory adherence, and acceptance by participating organizations. This paper proposes a federated search engine framework to support secure donor tracking across ART clinics without centralizing the storage of sensitive records. The proposed framework deploys decentralized query processing, with privacy-enhancing cryptographic identifiers, and aggregates non-identifiable donor indicators to enable controlled information sharing. Formal mathematical formulations help describe donor identity abstraction, federated query resolution, and detection of potential consanguinity risks. Performance evaluations conducted on synthetically generated, multi-clinic datasets indicate that the architecture is scalable, compliant with privacy constraints, and effective in mitigating donor overuse. Overall, the proposed framework is a practical and ethically responsible framework for donor tracking within distributed ART environments.