Describing Characteristics and Differences of Neutrophils in Sepsis, Trauma, and Control Patients in Routinely Measured Hematology Data

Abstract

Neutrophils have an important role in the immune response. These cells can be subjected to an impaired function and a shift in population depending on disease states. In sepsis, this shift is recognized and flagged by automated hematology analyzers, including the presence of band neutrophils, while these cells, although present, appear not to be detected in trauma patients. To better understand this suspected error in flagging, we set out to distinguish neutrophil populations of these two patient groups and compared these with controls. Different data-driven methods were used compared to standard algorithms used by the software of the analyzers. Using K-means clustering, we extracted neutrophils from raw hematology analyzer datafiles, and compared characteristics of these clusters between the patient groups. We observed an increased neutrophil size for both sepsis and trauma patients, but trauma patients had a smaller increase. Trauma patients also had a high proportion of cells with relatively high nuclear segmentation, which is contradictory with the presence of band neutrophils. This, in combination with the smaller size increase, might explain the inability to flag band neutrophils in trauma.

Neutrophils, Fast and Strong

“…There are differences in neutrophil characteristics in sepsis, trauma, and control patients. Thorough analysis was performed by Haitjema and colleagues, and increased neutrophil size for both sepsis and trauma patients was reported [ 20 ].…”

Swarm Thinking: Can Humans Beat Artificial Intelligence?

“…Though AI has demonstrated accuracy in making diagnoses, the judgment of experienced clinicians is necessary. For example, AI exploits data in EHRs to provide early warning for sepsis; however, patients with severe trauma with no sepsis can be flagged incorrectly because of the activation of neutrophils after tissue damage (Joosse et al, 2022). In this example, an experienced clinician would not start a sepsis protocol because it would not be indicated.…”