arXiv subjects

Alexandra Kautzky-Willer

Publications and source records attributed to Alexandra Kautzky-Willer.

6 recordsLinked to original sources

arXiv2026

Multi-center Medical Data Mining with FL-Net - A One-stop Shop for Federated Learning

Federated learning enables collaborative training without sharing patient-level data, but most studies remain simulations. Based on five requirements derived from the literature, we analyzed 14 FL frameworks and found that none fully satisfied these requirements. We present FL-Net, a novel federated clinical research framework to fulfill all requirements. It integrates modular data harmonization, data discovery, disclosure control, securely built versioned FL-Net-Tools and containerized federated workflow execution into a persistent network. It enables the re-use of harmonized data and workflows across studies. FL-Net's end-to-end capabilities were evaluated through harmonization, cross-study patient discovery across MIMIC and US-130, and reproducible, audited federated workflows with up to 50 concurrent clients. FL-Net is being developed within the dAIbetes and Microb-AI-ome EU projects and will cover over 800,000 patients across 10 hospitals in 9 countries covering longitudinal and single point in time data, FL-Net provides a practical foundation for interoperable, reproducible, and privacy-preserving multicenter clinical research.

cs.LG↗

arXiv2025

Projecting Multimorbidity and Mortality under Demographic Change and Preventive Interventions

As populations age, the rise of multimorbidity poses a significant healthcare challenge. However, our ability to quantitatively forecast the progression of multimorbidity remains limited. Leveraging a nationwide dataset comprising approximately 45 million hospital stays spanning 17 years in Austria, we develop a new compartmental model for chronic disease trajectories across 132 distinct multimorbidity patterns (compartments). Each compartment represents a distinct constellation of co-occurring chronic conditions, with transitions modeled as age- and sex-dependent probabilities. We use the compartmental disease trajectory model (CDTM) to simulate disease trajectories to 2030, estimating the frequency of all empirically observed co-occurrence patterns among more than 100 diagnosis groups. We demonstrate the model's utility in identifying high-impact prevention targets. A 5% reduction in new cases of hypertensive disease (I10--I15) leads to a 0.57 (SD 0.06)% reduction in all-cause mortality over a 15-year period, and a 0.57 (SD 0.07)% reduction in mortality for malignant neoplasms (C00--C97). We also evaluate long-term impacts of SARS-CoV-2 sequelae, projecting earlier and more frequent hospitalizations across a range of diagnoses. Our fully data-driven modelling approach identifies leverage points for proactive preparation by physicians and policymakers to reduce the overall disease burden in the population, emphasizing patient-centered healthcare planning in aging societies.

physics.soc-ph↗

arXiv2023

Unraveling cradle-to-grave disease trajectories from multilayer comorbidity networks

We aim to comprehensively identify typical life-spanning trajectories and critical events that impact patients' hospital utilization and mortality. We use a unique dataset containing 44 million records of almost all inpatient stays from 2003 to 2014 in Austria to investigate disease trajectories. We develop a new, multilayer disease network approach to quantitatively analyse how cooccurrences of two or more diagnoses form and evolve over the life course of patients. Nodes represent diagnoses in age groups of ten years; each age group makes up a layer of the comorbidity multilayer network. Inter-layer links encode a significant correlation between diagnoses (p $<$ 0.001, relative risk $>$ 1.5), while intra-layers links encode correlations between diagnoses across different age groups. We use an unsupervised clustering algorithm for detecting typical disease trajectories as overlapping clusters in the multilayer comorbidity network. We identify critical events in a patient's career as points where initially overlapping trajectories start to diverge towards different states. We identified 1,260 distinct disease trajectories (618 for females, 642 for males) that on average contain 9 (IQR 2-6) different diagnoses that cover over up to 70 years (mean 23 years). We found 70 pairs of diverging trajectories that share some diagnoses at younger ages but develop into markedly different groups of diagnoses at older ages. The disease trajectory framework can help us to identify critical events as specific combinations of risk factors that put patients at high risk for different diagnoses decades later. Our findings enable a data-driven integration of personalized life-course perspectives into clinical decision-making.

physics.med-ph↗

arXiv2022

Stress-testing the Resilience of the Austrian Healthcare System Using Agent-Based Simulation

Patients do not access physicians at random but rather via naturally emerging networks of patient flows between them. As retirements, mass quarantines and absence due to sickness during pandemics, or other shocks thin out these networks, the system might be pushed closer to a tipping point where it loses its ability to deliver care to the population. Here we propose a data-driven framework to quantify the regional resilience to such shocks of primary and secondary care in Austria via an agent-based model. For each region and medical specialty we construct detailed patient-sharing networks from administrative data and stress-test these networks by removing increasing numbers of physicians from the system. This allows us to measure regional resilience indicators describing how many physicians can be removed from a certain area before individual patients won't be treated anymore. We find that such tipping points do indeed exist and that regions and medical specialties differ substantially in their resilience. These systemic differences can be related to indicators for individual physicians by quantifying how much their hypothetical removal would stress the system (risk score) or how much of the stress from the removal of other physicians they would be able to absorb (benefit score). Our stress-testing framework could enable health authorities to rapidly identify bottlenecks in access to care as well as to inspect these naturally emerging physician networks and how potential absences would impact them.

physics.soc-ph↗

arXiv2019

Identification of gatekeeper diseases on the way to cardiovascular mortality

Multimorbidity, the co-occurrence of two or more chronic diseases such as diabetes, obesity or cardiovascular diseases in one patient, is a frequent phenomenon. To make care more efficient, it is of relevance to understand how different diseases condition each other over the life time of a patient. However, most of our current knowledge on such patient careers is either confined to narrow time spans or specific (sets of) diseases. Here, we present a population-wide analysis of long-term patient trajectories by clustering them according to their disease history observed over 17 years. When patients acquire new diseases, their cluster assignment might change. A health trajectory can then be described by a temporal sequence of disease clusters. From the transitions between clusters we construct an age-dependent multilayer network of disease clusters. Random walks on this multilayer network provide a more precise model for the time evolution of multimorbid health states when compared to models that cluster patients based on single diseases. Our results can be used to identify decisive events that potentially determine the future disease trajectory of a patient. We find that for elderly patients the cluster network consists of regions of low, medium and high in-hospital mortality. Diagnoses of diabetes and hypertension are found to strongly increase the likelihood for patients to subsequently move into the high-mortality region later in life.

physics.med-ph↗

arXiv2013

Quantifying age- and gender-related diabetes comorbidity risks using nation-wide big claims data

Currently emerging "big data" techniques are reshaping medical science into a data science. Medical claims data allow assessing an entire nation's health state in a quantitative way, in particular with regard to the occurrences and consequences of chronic and pandemic diseases like diabetes. We develop a quantitative, statistical approach to test for associations between the incidence of type 1 or type 2 diabetes and any possible other disease as provided by the ICD10 diagnosis codes using a complete set of Austrian inpatient data. With a new co-occurrence analysis the relative risks for each possible comorbidity are studied as a function of patient age and gender, a temporal analysis investigates whether the onset of diabetes typically precedes or follows the onset of the other disease. The samples is always of maximal size, i.e. contains all patients with that comorbidity within the country. The present study is an equivalent of almost 40,000 studies, all with maximum patient number available. Out of more than thousand possible associations, 123 comorbid diseases for type 1 or type 2 diabetes are identified at high significance levels. Well known diabetic comorbidities are recovered, such as retinopathies, hypertension, chronic kidney diseases, etc. This validates the method. Additionally, a number of comorbidities are identified which have only been recognized to a lesser extent, for example epilepsy, sepsis, or mental disorders. The temporal evolution, age, and gender-dependence of these comorbidities are discussed. The new statistical-network methodology developed here can be readily applied to other chronic diseases.

stat.AP↗