This summer, select IMO Health experts presented a Spotlight Poster on Prader-Willi Syndrome (PWS) at the 2024 ISPE Annual Meeting, held in Berlin, Germany.
ISPE, or the International Society for Pharmacoepidemiology, “is an international organization dedicated to advancing the health of the public by providing a global forum for the open exchange of scientific information and for the development of policy, education, and advocacy for the field of pharmacoepidemiology…” per its website.
The Spotlight Poster, “Building Patient Journeys for Prader-Willi Syndrome Patients: Insights from Electronic Health Records Through Natural Language Processing,” demonstrates how leveraging advanced NLP techniques to analyze electronic health records (EHRs) can improve our understanding of rare diseases, significantly enhancing patient care and streamlining operations.
RESEARCH
Building Patient Journeys for Prader-Willi Syndrome
Patients: Insights from Electronic Health Records Through
Natural Language Processing
Background and objective
Prader-Willi Syndrome is an extremely rare genetic condition that presents a wide range of symptoms, notably a constant sense of hunger. The varied nature of symptoms complicates researchers’ understanding of the disease’s trajectory, especially when critical pieces of information are buried within unstructured clinical notes.
IMO Health’s Life Science team, headed by Xiaoyan Wang, PhD, FAMIA, sought to illustrate how extracting this data could reveal patterns in symptoms and related health issues, thus increasing our understanding of the disease’s progression.
Methods and results
IMO Health researchers extracted clinical variables from both structured data and unstructured clinical notes using NLP in healthcare, capturing trends in physical and behavioral symptoms. They then analyzed the percentage of patients exhibiting specific symptoms or behavioral outcomes across different ages within the same age group and tracked symptom evolution over time. The team also explored differences in disease progression between male and female patients.
IMO Health’s NLP-powered approach yielded high accuracy in gathering relevant data. By reviewing clinical information spanning more than two decades and encompassing 433 pediatric PWS patients, IMO Health researchers managed to successfully craft comprehensive patient journeys, revealing the prevalence and long-term patterns of various symptoms, including skin-picking, scoliosis, obstructive sleep apnea, and hypopituitarism. The analysis also revealed gender differences in some ailments, such as central sleep apnea, which was more common in females, while convulsions were more frequent in males.
Conclusion
This breakthrough demonstrates the potential of AI and NLP in healthcare to transform the way patient journeys are studied, offering a faster, more comprehensive tool that researchers could leverage for other rare diseases, driving overall innovation and efficiency in rare disease management.
