Objectives: To characterize the contemporary red cell transfusion prac
tice in the critically ill and to define clinical factors that influen
ce these practices, Design: Scenario based national survey. Study Popu
lation: Canadian critical care practitioners, Measurements and Main Re
sults: We evaluated transfusion thresholds before transfusion and the
number of red cell units ordered, under the given conditions, Of 254 C
anadian critical care physicians, 193 (76%) responded to the survey, T
he primary specialty of most respondents was internal medicine (56%),
Internal medicine respondents were in practice for an average of 8.4 /- 5.7 (SD) yrs, and worked most often in combined medical/surgical in
tensive care units. Baseline hemo globin transfusion thresholds averag
ed from 8.3 +/- 1.0 g/dL in a scenario involving a young stable trauma
victim to 9.5 +/- 1.0 g/dL for an older patient after gastrointestina
l bleeding, Transfusion thresholds differed significantly (p < .0001)
between all four separate scenarios, With the exception of congestive
heart failure (p > .05), all clinical factors (including age, Acute Ph
ysiology and Chronic Health Evaluation II score, preoperative status,
hypoxemia shock, lactic acidosis, coronary ischemia, and chronic anemi
a significantly (p < .0001) modified the transfusion thresholds, A sta
tistically significant (p < .01) difference in baseline transfusion th
resholds was noted across four major regions (with a maximum of five a
cademic centers per region) of the country, Low physician numbers in t
wo of the regions did not allow for further investigation of regional
variations, Conclusions: There is significant variation in critical ca
re trans fusion practice, with many intensivists adhering to a 10.0-g/
dL threshold, while other physicians described a much more restrictive
approach to red cell transfusion, Also, many physicians opted to admi
nister multiple units, despite published guidelines to the contrary, A
dditionally, the administration of red cells was strongly influenced b
y a number of clinical factors, many unique to intensive care unit pat
ients, There is a need for prospective studies to define optimal pract
ice in the critically ill.