Relationship Between White Blood Cell Count and Bacteremia Using Interval Likelihood Ratios in Hospitalized Patients

Area of Science:

• Medical Diagnostics
• Infectious Disease Research
• Hematology

Background:

• Elevated white blood cell (WBC) and neutrophil counts are common in patients with bacteremia.
• However, these markers alone are insufficient predictors of bacteremia.
• There is a lack of data on the continuous relationship between WBC response and bacteremia.

Purpose of the Study:

• To characterize the relationship between WBC count, neutrophil percentage, and absolute neutrophil count (ANC) and bacteremia.
• To utilize interval likelihood ratios (ILRs) for a more precise estimation of bacteremia probability.
• To compare the discriminative power of ILRs against traditional threshold approaches.

Main Methods:

• Retrospective cohort study of 25,776 hospitalized patients from 2017-2018.
• Included non-surgical inpatients with complete blood count (CBC) with differential and blood culture.
• Analyzed WBC count, ANC, and neutrophil percentage as predictors of bacteremia using ILRs and AUROC.

Main Results:

• Neutrophil percentage demonstrated the highest discriminative ability (AUROC 0.74) for bacteremia.
• Bacteremia probability increased exponentially with neutrophil percentages from 80% to 100%.
• A threshold approach using 70% neutrophils underestimated bacteremia risk at higher levels.

Conclusions:

• Interval likelihood ratios (ILRs) offer a more discriminating method for estimating bacteremia probability compared to single thresholds.
• The magnitude of WBC and neutrophil abnormalities is crucial for predicting bacteremia risk.
• Physicians should consider the full spectrum of values, not just dichotomized results, for accurate risk assessment.