Releases: zeroknowledgediscovery/ZCoR-IPF
ZCoR-IPF
Screening for Idiopathic Pulmonary Fibrosis with Co-morbid Pattern Recognition in Electronic Health Records
As a lethal fibrosing interstitial lung disease with mean survival time less than 5 years, Idiopathic pulmonary fibrosis (IPF) presents a serious health problem, lacking effective early screening tools. Non-specific presentation, combined with a limited understanding of the pathobiology of early-stage IPF, and the potential need for invasive procedures for confirmation hinders early diagnosis. We introduce a new screening tool that requires no new laboratory tests, may be universally administered in primary care settings, and does not require recognition of early symptoms by the patients or care providers. Using subtle comorbidity signatures discovered from the history of medical
encounters of individual patients, we formulate the Zero-burden Co-Morbidity Risk Score for IPF (ZCoR-IPF), which is expected to be universally and inexpensively applicable at points of care to predict the future risk of a IPF diagnosis. Our algorithm is trained on a large national insurance claims database, and consequently validated on held-back data, and two independent datasets.
ZCoR-IPF
Screening for Idiopathic Pulmonary Fibrosis with Comorbid Pattern Recognition in Electronic Health Records
As a lethal fibrosing interstitial lung disease with mean survival time less than 5 years, Idiopathic pulmonary fibrosis (IPF) presents a serious health problem, lacking effective early screening tools. Non-specific presentation, combined with a limited understanding of the pathobiology of early-stage IPF, and the potential need for invasive procedures for confirmation hinders early diagnosis. In this study we introduce a new screening tool that requires no new laboratory tests, may be universally administered in primary care settings, and does not require recognition of early symptoms by the patients or care providers. Using subtle comorbidity signatures discovered from the history of medical encounters of individual patients, we formulate the Zero-burden Co-Morbidity Risk Score for IPF (ZCoR-IPF), which is expected to be universally and inexpensively applicable at points of care to predict the future risk of a IPF diagnosis. Our algorithm is trained on a large national insurance claims database, and consequently validated on held-back data, and two independent datasets