Research by Dervic Elma, Leutner Michael, Kautzky Alexander, Johannes Sorger, Liuhuaying Yang, Kautzky-Willer Alexandra, Klimek Peter
Created by Liuhuaying Yang
Patients become increasingly multimorbid with age leading to decreased quality of life, increased need for hospitalizations, health care utilization, mortality, and care costs. However, the typical population-scale disease trajectories along which patients become multimorbid are not yet fully understood.
We use a unique dataset containing 44 million records of almost all inpatient stays from 2003 to 2014 in Austria to investigate these disease trajectories. We developed a new, multilayer disease network approach to quantitatively analyze complex connections between two or more conditions and how they evolve over the life course of patients. Nodes represent diagnoses from the different chapters (cancer, respiratory diseases, etc.) in specific age groups in intervals of ten years. Thus, nodes in the constructed network are 3-digits ICD10 codes + 10 years age group, i.e., E66-20-29 represents obesity in the twenties. Links within a layer encode the tendency of two diagnoses to co-occur. While, directed links across layers mean that a given diagnosis is more likely to occur at a specific age, given that another diagnosis was already present at an earlier age. Network visulisation is available online: https://vis.csh.ac.at/netviewer/.
We used an unsupervised community detection algorithm for detecting overlapping communities of diagnoses within and across layers of the multilayer comorbidity network to reveal the most common disease trajectories [https://iopscience.iop.org/article/10.1088/1367-2630/11/3/033015].
This visualization shows identified trajectories of female and male patients and the Y-axis approximate age groups of nodes. More details of nodes can be found by clicking on a node.