Association between cell-free hemoglobin, acetaminophen, and mortality in patients with sepsis: an observational study

Abstract

Objective: To determine the association of circulating cell-free hemoglobin with poor clinical outcomes in patients with sepsis and to characterize the potential protective effects of acetaminophen, an inhibitor of hemoprotein-mediated oxidation.

Design: Retrospective observational study.

Patients: A total of 391 critically ill patients with sepsis in multiple ICUs in an academic tertiary care hospital.

Interventions: None.

Measurements and main results: Nonsurvivors had significantly higher plasma cell-free hemoglobin concentrations (median 20mg/dL, interquartile range 10-40) measured on enrollment compared to survivors (10mg/dL, interquartile range 10-30, p = 0.002). After controlling for potential confounders, patients with higher cell-free hemoglobin concentrations were significantly more likely to die in the hospital (odds ratio 1.078, 95% confidence interval 1.012-1.149, p = 0.02). In addition, receiving acetaminophen in the setting of increased cell-free hemoglobin was independently associated with a protective effect against death (odds ratio 0.48, 95% confidence interval 0.25-0.91, p = 0.026) and lower plasma concentrations of the lipid peroxidation product F2-isoprostanes (18.5 pg/mL, interquartile range 9-22.2) compared to no acetaminophen (42 pg/mL, interquartile range 29.7-86, p = 0.009).

Conclusions: In critically ill patients with sepsis, elevated concentrations of circulating cell-free hemoglobin are independently associated with an increased risk of death. Acetaminophen may exert a protective effect by reducing cell-free hemoglobin-induced oxidative injury.

Figure 1. Plasma concentrations of cell-free hemoglobin and in-hospital mortality

Survivors had significantly lower cell-free hemoglobin (10 mg/dl) than non-survivors (20 mg/dl)(*p = 0.002). Values are medians (middle bold horizontal line), interquartile ranges (IQR: box borders), and 10^th to 90^th percentiles (whiskers).

Figure 2. In-Hospital Mortality based on increasing concentrations of Plasma Cell-free hemoglobin

There was an increasing mortality trend amongst patients with higher plasma cell-free hemoglobin concentrations (Pearson Χ² for a trend, p = 0.004).

Figure 3. F₂-isoprostanes and acetaminophen exposure

Patients exposed to acetaminophen in the 48 hours prior to their blood draw had significantly lower plasma F₂-isoprostane concentrations (18.5 pg/ml) compared to those who received no acetaminophen (42 pg/ml, *p = 0.009). Values are medians (middle bold horizontal line), interquartile ranges (IQR: box borders), and 10^th to 90^th percentiles (whiskers).