During the early stages of hospital admission, clinicians use limited information to make decisions as patient acuity evolves. We hypothesized that clustering analysis of vital signs measured within six hours of hospital admission would reveal distinct patient phenotypes with unique pathophysiological signatures and clinical outcomes. We created a longitudinal electronic health record dataset for 75,762 adult patient admissions to a tertiary care center in 2014–2016 lasting six hours or longer. Physiotypes were derived via unsupervised machine learning in a training cohort of 41,502 patients applying consensus k-means clustering to six vital signs measured within six hours of admission. Reproducibility and correlation with clinical biomarkers and outcomes were assessed in validation cohort of 17,415 patients and testing cohort of 16,845 patients. Training, validation, and testing cohorts had similar age (54–55 years) and sex (55% female), distributions. There were four distinct clusters. Physiotype A had physiologic signals consistent with early vasoplegia, hypothermia, and low-grade inflammation and favorable short-and long-term clinical outcomes despite early, severe illness. Physiotype B exhibited early tachycardia, tachypnea, and hypoxemia followed by the highest incidence of prolonged respiratory insufficiency, sepsis, acute kidney injury, and short- and long-term mortality. Physiotype C had minimal early physiological derangement and favorable clinical outcomes. Physiotype D had the greatest prevalence of chronic cardiovascular and kidney disease, presented with severely elevated blood pressure, and had good short-term outcomes but suffered increased 3-year mortality. Comparing sequential organ failure assessment (SOFA) scores across physiotypes demonstrated that clustering did not simply recapitulate previously established acuity assessments. In a heterogeneous cohort of hospitalized patients, unsupervised machine learning techniques applied to routine, early vital sign data identified physiotypes with unique disease categories and distinct clinical outcomes. This approach has the potential to augment understanding of pathophysiology by distilling thousands of disease states into a few physiological signatures.

No abstract

In the midst of an outbreak, identification of groups of individuals that represent risk for transmission of the pathogen under investigation is critical to public health efforts. Several approaches exist that utilize the evolutionary information from pathogen genomic data derived from infected individuals to distinguish these groups from the background population, comprised of primarily randomly sampled individuals with undetermined epidemiological linkage. These methods are, however, limited in their ability to characterize the dynamics of these groups, or clusters of transmission. Dynamic transmission patterns within these clusters, whether it be the result of changes at the level of the virus (e.g., infectivity) or host (e.g., vaccination implementation), are critical in strategizing public health interventions, particularly when resources are limited. Phylogenetic trees are widely used not only in the detection of transmission clusters, but the topological shape of the branches within can be useful sources of information regarding the dynamics of the represented population. We evaluate the limitation of existing tree shape statistics when dealing with smaller subtrees containing transmission clusters and offer instead a phylogeny based deep learning system (DeepDynaTree) for classification of transmission cluster. Comprehensive experiments carried out on a variety of simulated epidemic growth models indicate that this graph deep learning approach is effective in predicting cluster dynamics (balanced accuracy of 0.826 vs. 0.533 and Brier score of 0.234 vs. 0.466 in independent test set). Our deployment model in DeepDynaTree incorporates a primal-dual graph neural network principle using output from phylogenetic-based cluster identification tools (available from https://github.com/salemilab/DeepDynaTree).