Project R-10225

PRIBE: Personalised Reference Intervals for integrated Biomarkers through nEw statistical methods and unique longitudinal data (Research)

Clinicians use reference intervals (RI) to interpret various test outcomes of patients, be it clinical parameters (e.g. weight), predictive or prognostic biomarkers (e.g. cholesterol level). Whether test outcomes lie within or outside an established RI determines diagnostic and treatment decisions so it is important that RIs are calculated in a way that is correct, interpretable and applicable to the increasing spectrum of biomarkers used in clinical practice (e.g. proteins or metabolites measured with new high-throughput technologies). The current methods to estimate RIs are problematic because they 1) are based on the population distribution of parameter values rather than at individual level or their combination; 2) can only be used in a cross-sectional setting whereas longitudinal follow-up of patients is needed to evaluate treatment response; 3) are too sensitive to outliers or unrealistic assumptions and 4) not well applicable to modern omics data. In the PRIBE project we aim to include subject-specific longitudinal information about the biomarker in order to advance the methodological framework to calculate individual RIs. Using quantile regression, Bayesian and Empirical Bayes methods and a unique longitudinal (omics) dataset, we will provide methodological and software solutions that will allow researchers and clinicians to better understand the variability of biomarker data and use the individual RIs for follow-up of patients in a preventive medicine approach.

16 October 2019 - 15 October 2023