The CBC with diff in Sepsis: Do You Get Out all the Juice? by Steven P. LaRosa, M.D.

As an adult Infectious Disease physician many of the calls I receive stem from concerning labs. The most common calls are concern for sepsis in a patient with leukocytosis and “bandemia”. Leukocytosis while a reliable marker of bacteremia in the pediatric population is woefully non-specific for sepsis in adults with AUC ROC of about .500. Similarly, the presence of immature band forms in not predictive of sepsis or bacteremia. It turns out that much more can be cleaned about the likelihood of sepsis and bacteremia from other components of the WBC differential. A neutrophil to lymphocyte count ratio greater than 10 has been demonstrated in numerous studies to be almost as good as serum Procalcitonin at predicting bacteremia. Additionally an absolute lymphocyte count < 1000 is highly predictive of bacteremia. While physicians tend to notice eosinophilia as a marker of allergic reactions or fungal infection, eosinopenia with an absolute eosinophil count < 40 cells/ microliter reliably differentiates sepsis from systemic inflammation without infection. Absolute neutrophil, lymphocyte, an eosinophil counts can be lifted right off the CBC report. White cell counts are usually reported as K/ microliter so an ANC of .500 is 500 cells/microliter and an absolute eosinophil count of 0.040 is 40 cells/ microliter, etc.

One would think that red blood cell indices would hold little information when it comes to evaluation for sepsis. I wonder how many physicians actually know what RDW stands for (red cell distribution width)? The RDW represents the distribution of RBC volumes. There is a burgeoning literature of increasing RDW and sepsis as well as disease severity, evolution of organ dysfunction and mortality. The cut-off for RDW in sepsis appears to be 15%.

With platelet counts the story is biphasic. New or worsening thrombocytopenia due to sepsis-induced coagulopathy is indicative of severe disease and associated with a poor prognosis. Far more common however in infection is the presence of thrombocytosis. Young physicians often forget the value of the platelet count as acute phase reactant and that it was often called “The poor man’s sed rate”. To an ID physician thrombocytosis is often a clue to an abscess in a closed space that requires “source control”.

What is the morale of this story? When you order a CBC with diff make sure that you squeeze all the data from it that you can!