Identification of Clinical Phenotypes in Septic Patients Presenting With Hypotension or Elevated Lactate

Abstract

IntroductionTargeted therapies for sepsis have failed to show benefit due to high variability among subjects. We sought to demonstrate different phenotypes of septic shock based solely on clinical features and show that these relate to outcome.MethodsA retrospective analysis was performed of a 1,023-subject cohort with early septic shock from the ProCESS trial. Twenty-three clinical variables at baseline were analyzed using hierarchical clustering, with consensus clustering used to identify and validate the ideal number of clusters in a derivation cohort of 642 subjects from 20 hospitals. Clusters were visualized using heatmaps over 0, 6, 24, and 72 h. Clinical outcomes were 14-day all-cause mortality and organ failure pattern. Cluster robustness was confirmed in a validation cohort of 381 subjects from 11 hospitals.ResultsFive phenotypes were identified, each with unique organ failure patterns that persisted in time. By enrollment criteria, all patients had shock. The two high-risk phenotypes were characterized by distinct multi-organ failure patterns and cytokine signatures, with the highest mortality group characterized most notably by liver dysfunction and coagulopathy while the other group exhibited primarily respiratory failure, neurologic dysfunction, and renal dysfunction. The moderate risk phenotype was that of respiratory failure, while low-risk phenotypes did not have a high degree of additional organ failure.ConclusionsSepsis phenotypes with distinct biochemical abnormalities may be identified by clinical characteristics alone and likely provide an opportunity for early clinical actionability and prognosis.

Derivation and Validation of Generalized Sepsis-induced Acute Respiratory Failure Phenotypes Among Critically Ill Patients: A Retrospective Study

“…The set of features analyzed is also limited; in contrast, this work considers an expanded set of clinical features for phenotyping. While many of these studies identify phenotypes in sepsis at large [22][23][24][25][26], they have not examined phenotypes that may exist speci cally within sepsis-induced ARF. Similarly, LCA-derived phenotypes of ARDS have been well-studied, but do not directly apply to ventilated patients with sepsis who do not satisfy the Berlin de nition of ARDS.…”

Emergence of a Technology-Dependent Phenotype of Pediatric Sepsis in a Large Children’s Hospital

“…This study also underscores the need to approach pediatric and adult sepsis patients differently. Our group previously identified phenotypes in adult subjects with septic shock using the Protocolized Care for Early Septic Shock (ProCESS) cohort (27). Similar to the current analysis, we focused on a sicker cohort of sepsis subjects by examining adults with a diagnosis of septic shock.…”

“…To permit comparison across ages, we converted these variables to z scores based on age and sex. Methods for calculating/ estimating SF ratio, respiratory score (a measure of mechanical ventilatory support), and VIS are described in detail in Supplemental Method SM1, http://links.lww.com/ SHK/B815 (25)(26)(27). Briefly, SF ratio, respiratory score, and VIS were means of quantifying respiratory and cardiovascular support.…”

“…Briefly, SF ratio, respiratory score, and VIS were means of quantifying respiratory and cardiovascular support. Arterial gases were prioritized for blood pH, but venous gases were converted to an arterial equivalent when arterial measurements were not available as described in Supplemental Methods SM2, http://links.lww.com/SHK/B815 (25)(26)(27). Because of low prevalence of underlying liver disease in pediatric patients, bilirubin was assumed to be normal at time zero if not available.…”

Knowledge-guided Deep Temporal Clustering for Alzheimer’s Disease Subtypes in Completed Clinical Trials

“…A pooled group of placebo-treated patients from RCTs can increase the power and diversity of AD populations. RCT databases were used to identify phenotypes or subtypes in sepsis and respiratory disease 9,10 . Considering the successful utilization of RCT data in various diseases, using multimodal data from RCT to derive the AD subtypes is promising.Temporal subtyping of AD is a data-driven, unsupervised learning task to group patients into similar disease progression patterns.…”