Objectives: We tested the hypothesis that colloid-induced hemodilution can induce hypocalcemia in the early phase of severe trauma resuscitation and tried to assess other potential causative factors of that hypocalcemia.
Design: Prospective cohort.
Setting: Level I academic trauma center.
Patients: Consecutive severe trauma patients (n = 212, mean Injury Severity Score 34) resuscitated in the prehospital phase without any blood transfusion.
Interventions: At admission, ionized calcium (corrected to an arterial pH = 7.40) was measured.
Measurements and main results: Hypocalcemia was defined as a value <1.15 mmol/L and severe hypocalcemia as a value <0.9 mmol/L. A normal ionized calcium concentration was observed in 56 (26%) patients, a mild ionized hypocalcemia (1.05 +/- 0.06 mmol/L) in 135 (64%) patients, and a severe ionized hypocalcemia (0.77 +/- 0.10 mmol/L) in 21 (10%) patients. There were significant correlations between ionized calcium concentration with the amount of infused colloid (R = .658, p < .001) and arterial pH (R = .760, p < 0.001) but not with the amount of infused crystalloid (R = .007, not significant). Despite taking into account hemodilution, arterial pH, binding of calcium to lactates, and colloids, some patients had marked differences (>15%) between calculated and observed ionized calcium, and these patients had more severe trauma and more frequently had acidosis and/or prehospital cardiac arrest. Using the TRISS methodology, survival was not significantly different from that expected in this trauma population.
Conclusion: Hypocalcemia frequently occurs on arrival at the hospital in severe trauma patients, and colloid-induced hemodilution and severe shock and/or ischemia-reperfusion appear to be important causative factors.