The 20 x 10(9)/L threshold for prophylactic platelet transfusion may be unn
ecessarily high. Few prospective studies. however, in which other trigger v
alues were tested have been published. In this study all hospitalized, thro
mbocytopenic adult hematology-oncology patients in our institution were pro
spectively evaluated daily for hemorrhage and platelet transfusion during a
one year period; no patients were excluded for bleeding or infectious prob
lems. By design, during the initial six-months: (baseline period), the prop
hylactic platelet transfusion trigger was 20 x 10(9)/L; for the second six-
months (study period) this threshold was changed to 10 x 10(9)/L. Patients
studied during the two periods did not differ significantly in age, gender.
diagnosis. blood or marrow transplant status, and duration of neutropenia.
Compliance with the thresholds was 95.6% (baseline period) and 93.5% (stud
y period). For patients with platelet counts under 20 x 10(9)/L. the mean u
se of platelet transfusions per patient per day was significantly lower in
the study period (4.47) than in the baseline period (6.48: p<0.001). Both m
ean prophylactic (1.54/patient-day) and therapeutic (2.93/patient-day) plat
elet transfusions were reduced in the study period compared with the baseli
ne period (2.16 and 4.22/patient-day, respectively). Hemorrhage was slightl
y reduced in the study period compared with the baseline period: major hemo
rrhage, 15.2% vs. 18.4% (p=0.014): minor hemorrhage. 63.6% vs. 70.1% (p<0.0
01). Thus, hemorrhage was not increased with the lower trigger level. A 10
x 10(9)/L prophylactic platelet transfusion threshold value is safe and eff
ective.