Multi-state Model Analysis Reveals Increased Risk of Severe COVID-19 Outcomes in Hospitalized Patients with Diabetes
Résumé fourni par la source
Diabetes mellitus is a well -established risk factor for severe outcomes in COVID -19.However, the detailed clinical trajectory of hospitalized COVID -19 patients with diabetes, and the dynamic progression of illness, have not been fully characterized.This study explores the clinical trajectories of COVID -19 inpatients with and without diabetes using advanced modeling techniques, including multistate models, to provide a novel visualization of disease progression.We conducted a retrospective cohort study at Tokyo Medical University(Tokyo, Japan) among 413 COVID -19 patients hospitalized between 2020 and2021, of whom 123 had pre -existing diabetes.Patient data on demographics, comorbidities, laboratory tests, including baseline estimated glomerular filtration rate (eGFR), treatments, and outcomes were collected.We compared outcomes between patients with and without diabetes using multivariable logistic regression models, adjusting for potential confounders.We employed a continuous -time multistate Markov model to analyze transitions between clinical states over time and generated stacked probability plots to visualize the evolving probability of being in each state.Patients with diabetes tended to be older and had more comorbidities, higher inflammatory markers, and lower baseline renal function at admission.Compared to those without diabetes, they showed elevated rates of ICU admission, intubation, CHDF, ARDS, AKI, and mortality.These patterns persisted after adjusting for covariates, though the differences in renal and respiratory outcomes were diminished when baseline eGFR was included in the model.The multistate model indicated more rapid and frequent transitions to critical illness among patients with diabetes, with a greater cumulative probability of ICU care or death by day 14.The hazard of ICU transfer remained higher in the diabetes group even after full adjustment (HR ~2.4).This study provided a novel depiction of COVID -19 progression in patients with diabetes using a multistate modeling approach.Our findings emphasized the importance of close monitoring in the early phase of illness and suggested that impaired renal function may partially explain the increased vulnerability in this population.
Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.
Contrôle bibliographique ouvert
Institutions déclarées
Une affiliation ne permet pas de déduire la nationalité d’un auteur.