ALLOGENEIC BLOOD-TRANSFUSION, HOSPITAL CHARGES, AND LENGTH OF HOSPITALIZATION - A STUDY OF 487 CONSECUTIVE PATIENTS UNDERGOING COLORECTAL-CANCER RESECTION
Ec. Vamvakas et Jh. Carven, ALLOGENEIC BLOOD-TRANSFUSION, HOSPITAL CHARGES, AND LENGTH OF HOSPITALIZATION - A STUDY OF 487 CONSECUTIVE PATIENTS UNDERGOING COLORECTAL-CANCER RESECTION, Archives of pathology and laboratory medicine, 122(2), 1998, pp. 145-151
Objective.-To investigate the independent association of allogeneic bl
ood transfusion with longer hospital stays and higher hospital charges
, after adjustment for the effects of confounding factors that are rel
ated to both these outcomes and the receipt of a perioperative transfu
sion. Design and Methods.-The records of 487 consecutive patients unde
rgoing colorectal cancer resection and assigned a diagnosis-related gr
oup code between 146 and 149 were reviewed. The influence of allogenei
c transfusion in explaining the variation in the observed length of st
ay and total hospital charges was calculated after adjustment for the
effects of 20 confounding factors that related to severity of illness,
difficulty of operation, and risk of postoperative infection. Results
.-Length of stay (mean +/- SE) was 16.7 +/- 0.81 days in the transfuse
d group, compared with 10.3 +/- 0.26 days in the untransfused group (P
<.0001); hospital charges were $28101 +/- 1121 and $15978 +/- 265, res
pectively (P<.0001). After adjustment for the effects of 20 confoundin
g factors, length of stay increased by 1.3% (95% confidence interval 0
.5% to 2.1%) per unit of red blood cells and/or platelets transfused (
P<.001), and hospital charges increased by 2.0% (95% confidence interv
al, 1.4% to 2.6%) per unit (P<.001). Conclusion.-Allogeneic transfusio
n was independently associated with longer hospital stays and higher h
ospital charges. This association may be due to a relationship between
allogeneic transfusion and a higher incidence of septic complications
of surgery or may reflect the function of blood transfusion as a surr
ogate marker for severity of illness.