PURPOSE: Although decreases in Hemoglobin (Hgb) and Hematocrit (Hct) correlate with volume of fluid infused, it correlated poorly with changes in PCWP. However PCWP increased by > 6 mmHg after fluid resuscitation in 20 of 21 patients with septic shock with a decreased Hgb of more than 1 gm/dl. We sought to determine the effect of hemoglobin (Hgb) on outcome and if decreases in Hgb correlated with improved survival during resuscitation of patients in severe sepsis and septic shock.
METHODS: Clinical data and outcome variables were extracted from the medical records of 162 adult patients with severe sepsis admitted to the intensive care unit at the University of Oklahoma, Health Sciences Center. Informed consent was obtained. Patients with bleeding or hemolysis were excluded. A multivariate logistic regression was constructed with mortality as the dependent variable. Data are presented as mean ± SEM.
RESULTS: Death occurred in 72 patients and was associated with a higher APACHE II score (28.8±0.9 vs. 23.1±0.8; p<0.05) and a higher MODS score (9.2±0.4 vs. 7.3±0.3; p<0.05). Mortality was not affected by the absolute Hgb level (10.7±2.1 vs. 10.4±1.9; p=0.36). A Hgb level >10 gm/dl was not associated with a better outcome (Table 1). The decrease in Hgb was more pronounced in the survival group at 24 hours (1.0±0.1 vs. 0.5±0.1; p:0.03) and at 48 hours (1.4±0.2 vs. 0.7±0.2; p<0.03). A decrease in Hgb >1gm/dl at 24 hours was associated with a lower mortality (33 vs. 48%; p<0.05), and a shorter duration on mechanical ventilation (MV) (5.5±1.2 vs. 8.6±1.1 vent-days; p:0.03). In the multivariate analysis, a higher survival was associated with a decrease of hgb > 1.0 at 24 hours (OR=1.9; 95 CI: 1.1–3.6) and >1.5 at 48 hours (OR: 2.2; 95 CI: 1.2–4.5).
CONCLUSION: Decreases in hemoglobin during fluid infusion reflects better resuscitation and correlate with improved survival. Targeting a Hgb >10g/dl is unnecessary.
CLINICAL IMPLICATIONS: Decreases in hemoglobin can serve as a marker of the adequacy of fluid resuscitation in patients with septic shock.
DISCLOSURE: Mohammad-Ali El-Harakeh, None.