Cluster analysis integrating age and body temperature for mortality in patients with sepsis: a multicenter retrospective study

Abstract

It is not clear whether mortality is associated with body temperature (BT) in older sepsis patients. This study aimed to evaluate the mortality rates in sepsis patients according to age and BT and identify the risk factors for mortality. We investigated the clusters using a machine learning method based on a combination of age and BT, and identified the mortality rates according to these clusters. This retrospective multicenter study was conducted at five hospitals in Korea. Data of sepsis patients aged ≥ 18 years who were admitted to the intensive care unit between January 1, 2011 and April 30, 2021 were collected. BT was divided into three groups (hypothermia < 36 °C, normothermia 36‒38 °C, and hyperthermia > 38 °C), and age groups were divided using a 75-year age threshold. Kaplan‒Meier analysis was performed to assess the cumulative mortality over 90 days. A K-means clustering algorithm using age and BT was used to characterize phenotypes. During the study period, 15,574 sepsis patients were enrolled. Overall, 90-day mortality was 20.5%. Kaplan‒Meier survival analyses demonstrated that 90-day mortality rates were 27.4%, 19.6%, and 11.9% in the hypothermia, normothermia, and hyperthermia groups, respectively, in those ≥ 75 years old (Log-rank p < 0.001). Cluster analysis demonstrated three groups: Cluster A (relatively older age and lower BT), Cluster B (relatively younger age and wide range of BT), and Cluster C (relatively higher BT than Cluster A). Kaplan‒Meier curve analysis showed that the 90-day mortality rates of Cluster A was significantly higher than those of Clusters B and C (24.2%, 17.1%, and 17.0%, respectively; Log-rank p < 0.001). The 90-day mortality rate correlated inversely with BT groups among sepsis patients in either age group (< 75 and ≥ 75 years). Clustering analysis revealed that the mortality rate was higher in the cluster of patients with relatively older age and lower BT.

Research frontiers and trends in the application of artificial intelligence to sepsis: A bibliometric analysis

“…As shown in the keyword cluster analysis in Figure 9A , Vosviewer divides keywords into three clusters, including “risk factors” in addition to the subject words “machine learning” and “sepsis.” Moon Seong Baek et al analyzed through machine learning that relatively older age and lower body temperature were risk factors for death in sepsis patients ( 51 ). Traditional machine learning algorithms such as Neural network, Bayes and Random forest had been used in early diagnosis, precise treatment and prognosis assessment of sepsis, these training models had shown excellent performance ( 52 – 54 ).…”

A new approach to predicting mortality in dialysis patients using sociodemographic features based on artificial intelligence

“…Identifying and characterizing high-risk clusters in a heterogeneous ICU population [13] MCA, Hierarchical clustering and K-Means Identifying lupus patient profiles [14] K-Medoids clustering and Kaplan-Meier survival analysis with Cox regression Effects on urticaria remission [15] Consensus clustering analysis, Kaplan-Meier curves and Cox proportional hazard models Identifying novel chronic kidney disease subgroups that best represent the data pattern [16] K-Means, Kaplan-Meier and log-rank test Mortality in patients with sepsis [17] K-means clustering and Bootstrapping Phenotyping of very old patients on admission to ICU [18] Unsupervised consensus clustering Identifying distinct phenotypes of patients with acute kidney injury requiring renal replacement therapy…”

“…In [10], the authors use the hierarchical clustering algorithm to obtain two subpopulations of interest where the Kaplan-Meier method is used to estimate survival curves, and the logrank test is used to test if statistical differences exist between the survival distributions of the two final clusters. Other studies where authors apply similar approaches are [11,14,16], among others. In [13], the authors use MCA with hierarchical clustering, K-Means and partitioning around medoids to identify profiles of lupus patients regarding treatment preferences.…”

Association between first 24-h mean body temperature and mortality in patients with diastolic heart failure in intensive care unit: A retrospective cohort study

“…Body temperature (BT) is a routinely measured vital sign that has been used to evaluate outcomes in patients with various diseases, such as sepsis ( 1 ), trauma ( 2 ), and multiorgan failure ( 3 ). BT abnormalities, including hypothermia and hyperthermia, are the most common symptoms in critically ill patients and the rationale for patient assessment and management ( 4 , 5 ).…”