To The Editor:
Bierbaum et al. should be commended for their extensive study
entitled "An Analysis of Blood Management in Patients Having a Total Hip
or Knee Arthroplasty" (81-A: 2-10, Jan. 1999), in which they question
the utilization of preoperative autologous blood donation in current
practice. The authors conclude that the most important predictors
of allogeneic transfusion are a low baseline hemoglobin level and
a lack of predonated units. They state that patients who predonate
more blood are expected to receive fewer allogeneic units because
of the availability of autologous units and conclude that preoperative
autologous donation decreases the risk of allogeneic transfusion.
This may not necessarily be true. Instead, patients predonating
more autologous units are expected to have higher baseline hemoglobin
levels. High baseline hemoglobin levels, regardless of the availability of
predonated units, decrease the likelihood of allogeneic transfusion.
The negative correlation between the number of autologous units
predonated and the allogeneic transfusion rate implies, but does
not prove, that predonation decreases the risk of allogeneic transfusion
in patients managed with total joint arthroplasty. Their nonrandomized,
uncontrolled, prospective study is flawed by preselection bias because anemic
patients (who have low baseline hemoglobin levels) can only be nondonors.
A low baseline hemoglobin level (less than 110 grams per liter)
disqualifies a patient from autologous predonation. In order to
justify their conclusion that autologous predonation decreases the
risk of allogeneic transfusion, all patients analyzed must be eligible
to participate in autologous predonation. I believe that anemic
patients, who are automatically excluded from predonation, are more
likely to require allogeneic transfusion and that nonanemic patients,
who are eligible to participate in predonation, customarily participate in
predonation (currently the standard of care for joint replacement
surgery). A closer analysis of their data may reveal that a high
baseline hemoglobin level predicts autologous unit predonation and
that a low hemoglobin level predicts no predonation. The more autologous
units predonated, the greater the phlebotomy-induced anemia, and
hence the higher the autologous transfusion rate (a self-fulfilling
prophecy) without necessarily significantly decreasing the allogeneic
transfusion rate. Without randomization, omitting preselection bias,
and providing standardized autologous and allogeneic transfusion guidelines,
the authors may mislead readers to interpret the negative correlation
between the number of autologous units available and the allogeneic
transfusion rate reported in their study as a cause-and-effect relationship.
In Figure 1 (on page 5 of the article), the category of patients
with a baseline hemoglobin level of 100 to 130 grams per liter is
too broad in that it that it mixes anemic and nonanemic elderly
patients as well as nondonors and donor-eligible patients together.
In elderly patients (those who are at least sixty years old), the
normal hemoglobin level ranges from 117 to 138 grams per liter in
women and from 124 to 153 grams per liter in men3.
If broken into smaller baseline hemoglobin increments, more useful
information may be found. I agree with the authors that more investigations
of blood-management strategies are warranted for joint replacement
surgery.
Dinna B. Billote, M.D.
Department of Anesthesia
Columbus Hospital
2520 North Lakeview Avenue
Chicago, Illinois 60614
E-mail address: dbillote@worldnet.att.net
B. E. Bierbaum, J. J. Callaghan, J. O. Galante, H. E.
Rubash, and R. E. Tooms reply:
We appreciate Dr. Billote's thorough review of our manuscript
and agree with several of the points presented in the letter. We
concur that the baseline hemoglobin level is an important predictor
of the need for allogeneic transfusion, a key finding of our study.
To our knowledge, ours is the first study to prospectively evaluate the
role of hemoglobin according to specific hemoglobin levels or categories.
Despite the fact that the hemoglobin level is central to blood-management
planning, we believe that our data supports the availability of autologous
units as another independent predictor of allogeneic transfusion.
As pointed out by Dr. Billote, our study was nonrandomized and noncontrolled;
however, the ordered logistic regression model enabled us to assess
the contribution of autologous units as a predictor of allogeneic
transfusion independent of other factors (for example, the baseline
hemoglobin level). While a cause-and-effect relationship has not
been established, our data is consistent with previously published
data indicating the effectiveness of autologous blood donation in
reducing the need for allogeneic transfusion2,4,5.
Dr. Billote also commented on one of our hemoglobin categories
(100 to 130 grams per liter) and suggested that a different range
could be more appropriate in the elderly. Not all of our patients
were elderly (average age, 66.6 years; range, fifteen to ninety-four
years), although we certainly agree that anemia can be more common
in the elderly and often resembles an anemia of chronic disease.
From a study-design perspective, the range of 100 to 130 grams per
liter was based on World Health Organization (WHO) guidelines1. Accordingly, some anemic patients
may be eligible for autologous blood donation provided that their
hemoglobin level is greater than 110 grams per liter. Our study
shows that there is a higher percentage of allogeneic transfusion
(or "breakthrough" transfusion) for anemic patients who are able
to donate autologous blood than for nonanemic patients.
We and Dr. Billote are in agreement that additional investigations
of blood-management strategies are warranted in joint-replacement surgery,
and we will be pleased if our work has enabled others to conduct
further research into this challenging area of patient management
in elective surgery.
Benjamin E. Bierbaum, M.D.
John J. Callaghan, M.D.
Jorge O. Galante, M.D., D.M.Sc.
Harry E. Rubash, M.D.
Robert E. Tooms, M.D.
Corresponding author: Benjamin E. Bierbaum, M.D.
New England Baptist Hospital, 830 Boylston Street
Chestnut Hill, Massachusetts 021671.