(1501) Subphenotypes of Sepsis and Differences in Treatment Effects: A Cluster Analysis for Mixed Data

Introduction: Sepsis is a heterogeneous syndrome, and the effectiveness of treatment may vary depending on its subphenotype. Additionally, patients with sepsis are treated with multiple therapies besed on their condition; however, no study has investigated the effects of these multiple treatments across subphenotypes. For appropriate subphenotyping, the dataset consisted with numerical and categorical data, called “mixed” data, is a barrier to fully using clinical data in cluster analysis. In this study, we aimed to classify patients with sepsis into subphenotypes based on mixed clinical data and examined the differences in multiple treatments effectiveness by subphenotype.

Methods: This was a secondary analysis of multicenter registries that enrolled patients with sepsis admitted to intensive care units in Japan. Fifty-two variables at admission were used in the cluster analysis. We applied the state-of-the-art clustering method k-uniform manifold approximation and projection (k-UMAP), which uses UMAP for dimensionality reduction, followed by clustering using k-means and the previous clustering methods k-prototype and KAMILA. To examine differences in the effectiveness of the six treatments by subphenotype, a logistic regression model was used for propensity score-based weighted data to calculate the odds ratios for the interaction of the treatments and subphenotype in each clustering method. The primary outcome was the in-hospital mortality rate.

Results: The analysis included 1,756 patients. Based on the consensus clustering results and the similarity between different clustering methods, we selected three clustering conditions: two results of k-UMAP ([number of neighbors (nn)=100, number of clusters (k)=3], and [nn=200, k=5]) and KAMILA [k=3]. Hospital mortality, patient characteristics, and treatment effects varied by subphenotype. In the k-UMAP [nn=100, k=3], recombinant thrombomodulin was significantly effective for subphenotype 1 (Odds ratio (OR) 0.37, 95% CI 0.13-0.80), and in the k-UMAP [nn=200, k=5], antithrombin III was significantly effective for subphenotype 2 (OR 0.49, 95% CI 0.20-0.93).

Conclusions: Using cluster analysis on mixed clinical data to identify sepsis subphenotypes may reveal appropriate treatment targets that have not been proven effective.