Research Presented at ECRD 2026 Highlights Potential for Earlier Identification of Rare Diseases in Primary Care
Research presented at ECRD 2026 highlights how machine learning could support earlier identification of rare and underdiagnosed diseases in primary care.
Manchester, UK, 8 June 2026 - Research presented at the European Conference on Rare Diseases (ECRD 2026) has highlighted the potential for machine learning to support earlier identification of rare and underdiagnosed diseases in primary care.
The poster, "Assessing the accuracy of an automated machine learning approach for the detection of rare and underdiagnosed diseases in primary care", was presented by Sara Elgott, Founder of OpalMedica. The work was conducted in collaboration with Dr David McMinn, Dr Kate Scoffings, Dr Youssef Hassan and Dr Safwaan Adam.
Rare diseases affect an estimated 1 in 17 people, yet many patients experience years of uncertainty before receiving an accurate diagnosis. Delayed diagnosis can result in repeated healthcare visits, unnecessary investigations, delayed treatment and significant impact on quality of life.
The study evaluated machine learning models designed to identify features suggestive of Cushing's syndrome and primary hyperaldosteronism from free-text symptom descriptions. The models were tested using patient journey narratives from individuals living with these conditions alongside control cases.
The findings demonstrated high levels of accuracy. Symptom extraction accuracy reached 86% for Cushing's syndrome and 90% for primary hyperaldosteronism. Disease determination achieved true positive rates of 89% and 100% respectively, while maintaining low false positive rates of 6% and 0%.
"Rare diseases often present with symptoms that are common, non-specific and spread across multiple consultations, making recognition extremely challenging in routine clinical practice," said Sara Elgott, Founder of OpalMedica. "These findings suggest that machine learning could help support clinicians by identifying patterns that may warrant further investigation, with the ultimate aim of reducing diagnostic delay."
The research represents an early proof-of-concept and demonstrates the potential utility of machine learning-based clinical decision support tools within primary care settings. Such tools are intended to support clinical judgement by highlighting patterns that may otherwise go unnoticed, rather than replacing clinician decision-making.
Further validation studies are planned in collaboration with NHS partners in Greater Manchester, UK, including the use of anonymised real-world primary care data. While the next phase is focused on NHS evaluation, the longer-term ambition is to create scalable tools that can support earlier identification of rare and underdiagnosed diseases internationally.
OpalMedica welcomes discussions with healthcare organisations, research institutions, patient organisations and industry partners interested in collaborating on future validation studies, implementation research and initiatives aimed at reducing diagnostic delay in rare diseases.
About OpalMedica
OpalMedica is a UK-based medical technology company focused on reducing diagnostic delay in rare and underdiagnosed diseases. Through its patient insight platform, My Rare Journey, and its patent-pending Clinical Flags decision support technology, OpalMedica aims to help healthcare professionals identify patterns that may support earlier investigation and referral.