Objective: To examine the endogenous erythropoietin response in critic
ally ill children with acute anemia or acute hypoxemia. Design: A pros
pective case study of critically ill acutely anemic, and acutely hypox
emic pediatric patients compared with control groups of critically ill
nonanemic and nonhypoxemic patients and with a hemoglobin and age-mat
ched, chronically anemic patient group. Setting: Multidisciplinary, te
rtiary, pediatric intensive care unit (ICU). Patients: Critically in p
atients admitted to the pediatric ICU during an 11-month period betwee
n February 1992 and March 1993 with acute anemia (n = 21), acute hypox
emia (n = 18), or neither anemia nor hypoxemia (n = 10). Outpatients w
ith chronic anemia (n = 21) and no acute illness were also studied as
a comparison group. Interventions: None. Measurements and Main Results
: Ages were equivalent among the groups and averaged 57.4 +/- 47.2 mon
ths (range 1 to 144). Acutely hyperemic and critically ill control pat
ients had normal hemoglobin levels. Acutely anemic patients had a hemo
globin level equivalent to chronically anemic outpatients, but lower (
p < .001) hemoglobin levels than acutely hypoxemic and critically ill
control patients. The serum erythropoietin concentrations in the acute
ly anemic group were significantly lower than erythropoietin values in
the chronically anemic group (39.3 +/- 62.2 vs. 861 +/- 758 mU/mL, p
< .001) and similar to erythropoietin concentrations in the critically
ill control (13.5 +/- 10.5 mU/mL) and acutely hypoxemic (5.2 +/- 3.3
mU/mL) patient groups. Only ten of 49 critically ill patients had an e
rythropoietin concentration above normal, compared with 20 of 21 chron
ically anemic patients, whose erythropoietin concentrations were above
normal. Conclusions: The erythropoietin response to known physiologic
stimuli is blunted in critically ill children. This blunted erythropo
ietin response may result in increased transfusion requirements.