Keywords
blood transfusion - blood typing - hip fractures
Introduction
Hip fractures in the elderly are a public health problem worldwide. As well as being
associated with high morbidity and mortality rates, they have a significant financial
impact on health systems.[1]
Various factors that influence the evolution of these patients are studied, including
the importance of multidisciplinary care,[2] the time elapsed between hospitalization and surgery,[3] the problem of previous use of anticoagulants,[4] the types of implants used[5] etc. Among these issues, the management of the use of blood products to treat pre-
and postoperative anemia is relevant. Blood transfusion is part of the therapeutic
arsenal in the clinical care of elderly individuals with hip fractures. Still, it
is a procedure associated with risks such as increased rates of postoperative infection[6] and increases in the cost of treatment.
The criteria to indicate transfusion of packed red blood cells in elderly patients
undergoing surgical treatment for hip fractures are still poorly defined. There are
liberal protocols that indicate transfusion when hemoglobin levels fall below 10 g/dL,
and restrictive protocols that do not indicate transfusion unless the hemoglobin level
falls below 8 g/dL or the patient has symptoms related to anemia. There is evidence
that liberal protocols do not present advantages in terms of mortality, functional
recovery, and complications compared to restrictive protocols, indicating that the
restrictive protocol may be considered the better option.[7]
Two other critical questions are in which situations and how many units of packed
red blood cells should be requested for patients undergoing surgery. Requesting a
reserve of packed red blood cells involves the performance of three procedures by
the blood bank: ABO and Rh typing, irregular antibody screening (IAS), and crossmatch
tests between donor and recipient samples.[8]
[9] Therefore, each blood bag reserve requested represents a financial impact and generates
a workload for the blood bank.
Not all patients undergoing surgical treatment for hip fractures need to receive blood
component transfusions; some studies mention an incidence of around 30% to 40%.[7]
[10]
[11]
[12] Therefore, knowing the predictive factors regarding the need for transfusion of
packed red blood cells can guide the request for blood before surgery. For patients
at low risk of transfusion during and after surgery, requesting only the typing without
requesting a reserve would represent a significant advance in managing scarce resources
such as blood products.
The present study aimed to identify predictors of the need for transfusion during
and after surgery to treat hip fractures in the elderly and to evaluate a protocol
to guide the request for blood reserves for surgery.
Materials and and Methods
Materials and and Methods
Ethical procedures
The institutional Ethics in Research Committee approved this project under CAAE number
70915423.1.0000.0099.
Design and Study Population
For data collection, this retrospective cohort was developed using the electronic
medical records of patients older than 60 years of age who underwent surgical treatment
for proximal femoral fractures at our institution between January 1st and December
31st, 2021. Our institution's Hematology Institute provided the information regarding
request and use. Patients with high-energy trauma fractures and those undergoing percutaneous
fixation with cannulated screws were excluded.
Data collection
Data on sex, age, preoperative hemoglobin level, diagnosis, and type of surgery were
tested for correlation with blood transfusion.
The transfusion index (TI) and the crossmatch test:transfusion ratio (C:T) were calculated.[13]
[14] The TI measures the average number of blood units used per procedure, and values
lower than 0.5 indicate that the routine request for blood reserves is not necessary
for the procedure. The C:T is the ratio between the number of blood units requested
and the number of units transfused, and it measures the efficiency of blood requests.
Values above 2.5 indicate excessive requests.[13]
Based on these criteria, Khan et al.[14] proposed a protocol to request blood for the surgical treatment of hip fractures
([Table 1]). We included the data obtained in our sample in this protocol to verify its applicability
in our casuistry by evaluating the results obtained for the C:T, the number of transfusions
carried out without requesting a reserve, and the IT for each hemoglobin level interval
in the protocol.
Table 1
|
Hemoglobin level (g/dL)
|
Blood requests
|
|
≥ 11
|
Typing
|
|
9.0–10.9
|
Reservation of 1 unit
|
|
8.0–8.9
|
Reservation of 3 units
|
|
7.0–7.9
|
Reservation of 4 units
|
|
≤ 6.9
|
Clinical assessment
|
Statistical Analysis
For the statistical analysis, we used the PASW Statistics for Windows (SPSS Inc.,
Chicago, IL, United States), version 18.0, with a 95% confidence interval and 5% significance
level (p < 0.05) for the tests applied. The Shapiro-Wilk test was used to verify the normality
of the data, and the paired Student t-test was used to compare the pre- and postoperative hemoglobin values in the two
fracture groups (pertrochanteric and femoral neck).
Results
Data was collected from 172 medical records of patients who met the study requirements
([Table 1]). Our sample consisted of 118 female and 54 male patients, with a mean age of 80.2
years. The diagnosis was of pertrochanteric fracture in 100 cases and of femoral neck
fracture in 72.
The mean preoperative hemoglobin level was of 11.44 g/dL; postoperatively. it was
of 9.9 g/dL. The preoperative hemoglobin levels were of 11.0 g/dL and 11.7 g/dL respectively
for pertrochanteric and femoral neck fractures. In the classification of the American
Society of Anesthesiologists (ASA), 3 patients were classified as ASA 1, 128, as ASA
2, 38, as ASA 3, and 3, as ASA 4. After surgery, the average was of 8.9 for pertrochanteric
and 9.2 for femoral neck fractures. Although a reduction in the patients' hemoglobin
was observed postoperatively (paired t-test; p = 0.0001), there were no statistical differences between the fracture groups (unpaired
t-test; p = 0.17, [Fig. 1]).
Fig. 1 Hemoglobin variation (g/dL) in the different surgical groups.
Of those diagnosed with a pertrochanteric fracture, 35 (35%) received a transfusion,
and of those diagnosed with a femoral neck fracture, 29 (27.8%) received a transfusion
of one or more units of packed red blood cells. This difference was not statistically
significant (Chi-squared test; p = 0.32; [Table 2]).
Table 2
|
Surgical groups
|
Transfusion
|
|
|
No
|
Yes
|
|
Pertrochanteric fracture
|
65
|
35
|
100
|
|
65.0%
|
35.0%
|
100.0%
|
|
Femoral neck fracture
|
52
|
20
|
72
|
|
72.2%
|
27.8%
|
100.0%
|
|
Total
|
117
|
55
|
172
|
|
68.0%
|
32.0%
|
100.0%
|
A reserve of 328 units of packed red blood cells was requested for 167 surgeries,
representing 97% of the cases, and 112 units were transfused in 55 patients, or 31.9%
of the cases. The IT was calculated at 0.65. The C:T was of 2.9 ([Table 3]).
Table 3
|
Hemoglobin level
(g/dL)
|
n
|
Number of units transfused
|
Number of units reserved
|
Number of units reserved according to the protocol
|
|
≥11
|
111
|
34
|
219
|
0
|
|
9–10.9
|
49
|
55
|
88
|
49
|
|
8–8.9
|
7
|
12
|
12
|
14
|
|
7– 7.9
|
3
|
7
|
5
|
9
|
|
≤6.9
|
2
|
4
|
4
|
0
|
|
Total
|
172
|
112
|
328
|
72
|
There was a correlation between the anesthetic risk classification (ASA) and the number
of transfused bags, according to the Pearson correlation (r = 0.23; p = 0.003; [Fig. 2]). There was also an inverse correlation between the preoperative hemoglobin level
and the number of bags transfused (Pearson correlation: r = - 0.43; p = 0.001; [Fig. 3]).
Fig. 2 Correlation between preoperative hemoglobin levels (g/dL) and the number of bags
transfused.
Fig. 3 Correlation between anesthetic risk classification (ASA) and the number of bags transfused.
Discussion
In line with other studies, the ASA classification and the preoperative hemoglobin
level were predictive factors of blood transfusion, since the groups were matched
regarding anthropometric characteristics.
The average hemoglobin level on admission was of 11.44 g/dL, and 61 patients (32.4%)
had less than 11 g/dL on admission, which characterized a large proportion of patients
who had already been admitted with anemia. We must consider the possibility that this
is related to the fact that most of the patients are elderly individuals living in
low-income families, and that factors related to the food security of these patients
may be present. The average drop in hemoglobin after surgery was of 1.54 g/dL; there
was no difference in hemoglobin results between patients with a pertrochanteric fracture
and those with a femoral neck fracture.
The observed TI of 0.65 indicates that the surgical treatment of these fractures requires
requesting a reserve of packed red blood cells. Still, the C:T of 2.9 indicates that
the practice of routinely requesting blood reserves for all patients who will undergo
hip fracture surgery has proved inefficient, and that it is necessary to adopt some
criteria for the best use of these resources.
Khan et al.[14] proposed a protocol to request blood for the surgical treatment of hip fractures
([Table 1]). In a simulation in which we used these criteria for our sample ([Table 2]), we obtained a C:T of 0.64, which would indicate efficiency in the reserve request.
We also observed that if this protocol had been used, analyzing only individuals with
a hemoglobin level greater than or equal to 11, we would have obtained a TI of 0.30.
According to these same authors,[14] surgeries with a TI lower than 0.5 do not require a request for a packed red blood
cell reserve, thus reducing the fear of not requesting a blood reserve for this group
of patients.
The limitation of the present study is that it is a retrospective cohort study from
a single hospital. Its primary clinical implication is to demonstrate that requesting
packed red blood cells indiscriminately for all patients with hip fractures is unjustified,
and that adopting the proposed protocol can safely save resources.
In addition to the financial impact, the request for blood reserves without criteria
creates an operational burden. The crossmatch test and IAS take up around 40 minutes
of the blood bank's time.
Conclusion
We found a correlation between the ASA classification and transfusion. Applying a
proposed protocol to decide between requesting a reserve and just typing for these
patients proved suitable for our casuistry.
