If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Trauma and complex cardiac surgery are associated with a high risk of bleeding complications. The difference in costs between patients who require bleeding control measures and those who do not is poorly understood. Our goal was to assess the cost of care and outcomes for patients in these settings.
Methods
Patients >18 years of age, who were discharged between January 2010 and December 2012, were retrospectively identified in the Premier Hospital Database based on International Classification of Disease, Ninth Revision codes. These patients were categorized as having received blood products (“bleeding patients”) or not (“nonbleeding patients”). Patients with costs and length of stay (LOS) of zero were excluded. Differences in treatment costs and outcomes were assessed using univariate analysis and multivariate modeling.
Findings
Bleeding trauma patients (n = 8800) had a 150% higher total cost of care (P < 0.001; 146% after excluding costs of agents used for bleeding control, P < 0.001), an 81.3% longer hospital LOS (P < 0.001), and a 65.2% longer intensive care unit (ICU) LOS (P < 0.001) than nonbleeding patients (n = 53,727). Bleeding complex cardiac surgery patients (n = 82,832) had a 133.2% higher total cost of care (P < 0.001; 128.7% after excluding costs of agents used for bleeding control, P < 0.001), a 155.6% longer hospital LOS (P < 0.001), and an 89.3% longer ICU LOS (P < 0.001) than nonbleeding patients (n = 380,902).
Implications
Trauma and cardiac surgery patients who experienced bleeding and received allogeneic blood product transfusions had significantly worse outcomes, including longer LOS, greater inpatient mortality, and higher costs of care (even when excluding costs of agents used for bleeding control) than those who did not.
The true cost of allogeneic blood products, such as fresh frozen plasma (FFP), red blood cells (RBCs), and cryoprecipitate, is an issue that is poorly understood and not well documented in the literature.
It is a complex matter, because costs have to take into account not only the direct acquisition cost of the product, but also the associated indirect costs, such as blood collection, testing, distribution, and storage.
Furthermore, the total cost should also take into account all costs associated with hospital stay and the treatment of any adverse events and poor outcomes associated with blood product transfusion. Blood products have been associated with a number of adverse events, including transfusion-associated circulatory overload (TACO), transfusion-related acute lung injury (TRALI), pathogen transmission, allergic reactions, and even increased mortality,
which significantly increases the cost of their treatment; it has been estimated that hospital stay accounts for >65% of all transfusion-related costs.
In addition, the cost of blood products themselves is rising. This can be attributed to a number of factors, including increased demand, tighter restrictions on blood donor eligibility, and increasing laboratory testing and treatment of blood to keep it free from infectious diseases and bacteria.
For example, the introduction of pathogen reduction technology, such as treatment of plasma with amotosalen or riboflavin, can more than triple the cost of a unit of blood.
The clinical settings of complex cardiac surgery and trauma are known to be associated with high rates of transfusion. Up to 15% of the US blood supply is transfused in the setting of cardiac surgery,
Society of Thoracic Surgeons Blood Conservation Guideline Task Force 2011 update to the Society of Thoracic Surgeons and the Society of Cardiovascular Anesthesiologists blood conservation clinical practice guidelines.
which results in a high transfusion rate in this setting. Hemorrhage is associated with poor outcomes, such as increased morbidity, mortality, and even death,
which leads to an increased use of resources by these patients during hospitalization.
To determine the cost implications of allogeneic blood product transfusion in trauma and complex cardiac surgery, we assessed the cost of care and outcomes for patients in these 2 clinical settings based on bleeding status (ie, bleeding patients vs nonbleeding patients).
Patients and Methods
Trauma and Cardiac Surgery
In this retrospective analysis, patients >18 years old discharged between January 1, 2010 and December 31, 2012 were identified in the US Premier Hospital Database based on select International Classification Disease, Ninth Revision (ICD-9) procedure codes within the trauma or complex cardiac surgical population definitions (Appendix). The Premier Hospital Database is a Health Insurance Portability and Accountability Act compliant dataset and is considered exempt from approval from an institutional review board based upon the Code of Federal Regulation, §46.101b(4), from the United States Department of Health and Human Services.
Complex cardiac patients included those who underwent operations of the valves, septa, or vessels of the heart, as well as those with incision or excision of the aorta or other thoracic vessels. Trauma patients included those with a diagnosis code that indicated internal injuries to the heart, lung, or other specified internal organs. A listing of the codes used to identify patients is presented in the Appendix. Patients were then categorized as bleeding or nonbleeding, based on specific ICD-9 codes and blood product use (patients who received at least 1 U of blood products [RBCs, FFP, cryoprecipitate, or platelets] were characterized as bleeding). Use of the following agents for bleeding control was extracted from the database: RBCs, FFP, platelets, cryoprecipitate, recombinant activated factor VII, and hetastarch. Patients who received blood products before the day of surgery or patients with costs and a hospital length of stay (LOS) of zero were excluded. The following information relating to patient demographic characteristics was extracted from the database: age, gender, race, ethnicity, comorbidities and comorbidity score, and hospital costs. Products and services that might have influenced the use of blood products in bleeding patients, including fibrin sealants, tranexamic acid, protamine, and the use of an intraoperative cell saver, were identified and used as effect modifiers in multivariate analyses. All costs were based on hospital-reported costs for procedures, medications, other therapies, and laboratory costs taken directly from chargemaster-detailed use.
The primary outcome was hospital cost per patient (both including and excluding costs of agents used for bleeding control); secondary outcomes included hospital LOS, intensive care unit (ICU) admission and LOS, inpatient mortality, hospital readmission, and adverse events (including TACO, TRALI, thromboembolic events, pulmonary edema, and acute respiratory distress).
Statistical Analysis
Differences in outcomes between bleeding and nonbleeding patients were assessed using both univariate analysis and multivariate regression models. Statistical significance for univariate comparisons was assessed using χ2 tests for categorical variables and Student’s t tests for continuous variables. Multivariate analysis was used to calculate odds ratios for inpatient mortality, 30-day readmission for bleeding, and ICU admission; confidence limits for these parameters were generated using the Wald test. A fully specified model was developed that included all covariates that demonstrated significant differences in univariate comparisons from patient demographic and hospital characteristic variables, as well as the Charlson Comorbidity Index score and ICU admission. All models used backward elimination with a P value of < 0.1 to determine variable inclusion in the final model.
Results
Trauma
A total of 62,527 trauma patients were identified for inclusion in the study; of these, 8800 received at least 1 U of blood products (bleeding patients), and 53,727 did not (nonbleeding patients). Baseline patient characteristics are shown in Table I.
Table IBaseline characteristics of trauma and complex cardiac surgery patients
In the bleeding patients group, 56.2% (n = 4942) of patients received RBCs (mean [SD] 4.8 [6.7] U), 44.3% (n = 3893) received FFP (mean 6.4 [10.1] U), 2.7% (n = 240) received cryoprecipitate (mean 9.6 [2.9] U), and 13.5% (n = 1180) received platelets (mean 2.9 [3.9] U). Mean costs for these blood products per transfused patient were $1112 for RBCs, $578 for FFP, $1463 for cryoprecipitate, and $1486 for platelets (Table II).
Table IIAgents used for bleeding control in trauma and complex cardiac surgery patients
Trauma
Complex Cardiac Surgery
Characteristics
Nonbleeding Patients (n = 53,727)
Bleeding Patients (n = 8800)
Nonbleeding Patients (n = 380,902)
Bleeding Patients (n = 82,832)
n (%)
Mean (SD) Median Cost per Recipient ($)
n (%)
Mean (SD) Median Dose per Recipient (U)
Mean (SD) Median Cost per Recipient ($)
P Value
n (%)
Mean (SD) Median Cost per Recipient ($)
n (%)
Mean (SD) Median Dose per Recipient (U)
Mean (SD) Median Cost per Recipient ($)
P Value
PRBCs
—
—
4942 (56.2)
4.81 (6.65) 3.0
1112 (2001) 582
—
—
42,691 (51.5)
4.51 (4.90) 3.0
1034 (1450) 627
FFP
—
—
3893 (44.3)
6.43 (10.08) 4.0
578 (1183) 252
—
—
30,118 (36.4)
4.22 (5.45) 0.0
323 (591) 192
Platelets
—
—
1180 (13.5)
2.89 (3.91) 2.0
1486 (2427) 742
—
—
22,914 (27.7)
2.95 (4.77) 0.0
1281 (1583) 830
Cryoprecipitate
—
—
240 (2.7)
9.64 (12.72) 7.0
1463 (2295) 759
—
—
4892 (5.9)
5.53 (6.81) 5.0
968 (1481) 563
rFVIIa
—
—
190 (2.2)
5.77 (6.49) 3.5
7549 (8352) 4895
—
—
1551 (1.9)
4.91 (6.65) 3.0
7349 (14,691) 4453
Hetastarch
—
—
1264 (14.4)
1.67 (1.50) 1.0
51 (77) 29
—
—
25,148 (30.4)
1.74 (1.47) 1.0
45 (54) 30
Fibrin sealants
128 (0.2)
97 (1.1)
—
<0.001
2320 (0.6)
2474 (3.0)
—
<0.001
600 (430) 513
897 (1066) 676
0.011
581 (446) 507
669 (601) 505
<0.001
Tranexamic acid
39 (0.1)
52 (0.6)
—
<0.001
1472 (0.4)
2187 (2.6)
—
<0.001
168 (202) 111
255 (421) 147
0.200
452 (318) 531
193 (254) 67
<0.001
Protamine
420 (0.8)
341 (3.9)
—
<0.001
58,470 (15.4)
55,575 (67.1)
—
<0.001
23 (25) 14
30 (38) 20
0.002
36 (35) 27
38 (41) 27
<0.001
Intraoperative cell saver
375 (0.7)
506 (5.8)
—
<0.001 0.462
24,377 (6.4)
26,928 (32.5)
—
<0.001
385 (637) 150
422 (849) 152
495 (803) 228
415 (653) 215
<0.001
FFP = fresh frozen plasma; PRBCs = packed red blood cells; rFVIIa = recombinant activated factor VII.
Multivariate models adjusting for patient demographic characteristics demonstrated that, when the costs of agents used for bleeding control were included, bleeding patients had a 150% higher total cost of care ($46,033 vs $18,434; P < 0.001) than nonbleeding patients (Figure 1; Table III). When the costs of agents used for bleeding control were excluded, bleeding patients had a 146% higher total cost of care ($45,375 vs $18,441; P < 0.001).
Figure 1Comparison of total hospital costs for bleeding and nonbleeding patients in trauma and complex cardiac surgery.
Bleeding patients had a 81.3% longer LOS (13.6 days vs 7.5 days; P < 0.001). They were also 3.53 times as likely to be admitted to the ICU (P < 0.001), 3.75 times as likely to be readmitted for bleeding within 30 days (P < 0.001), and 4.09 times as likely to die in hospital (P < 0.001). The mortality rate differed significantly between the 2 groups (10.72% bleeding patients vs 3.14% nonbleeding patients; P < 0.001). Compared with nonbleeding patients, bleeding patients had higher rates of thromboembolic events (7.84% vs 6.70%; P < 0.001), TACO (2.00% vs 0.76%; P < 0.001), and TRALI (0.13% vs 0.01%; P < 0.001). The incidence of acute respiratory distress did not differ between the 2 groups (0.31% vs 0.33%; P = 0.730), and no cases of pulmonary edema were observed in either group.
Cardiac Surgery
A total of 463,734 patients (82,832 bleeding; 380,902 nonbleeding) were included. Baseline demographic patient characteristics are shown in Table I. Bleeding patients were older (>65 years: 62.3% vs 49.8%; P < 0.001), sicker (Charlson Comorbidity Index >3: 22.5% vs 15.4%; P < 0.001), and had higher rates of cerebrovascular disease (10.2% vs 4.3%; P < 0.001) and peripheral vascular disease (18.2% vs 9.4%; P < 0.001) than nonbleeding patients.
In the bleeding patients group, 51.5% (n = 42,691) of patients received RBCs (mean [SD] 4.5 [4.9] U), 36.4% (n = 30,118) received FFP (mean 4.2 [5.5] U), 5.9% (n = 4892) received cryoprecipitate (mean 5.5 [6.8] U), and 27.7% (n = 22,914) received platelets (mean 3.0 [4.8] U). Mean costs for these blood products per transfused patient were $1034 for RBCs, $323 for FFP, $968 for cryoprecipitate, and $1281 for platelets (Table II).
Multivariate models that adjusted for patient demographic characteristics demonstrated that, in bleeding patients, the total cost of hospital stay was 133.2% higher when the costs of agents used for bleeding control were included ($50,344 vs $21,590; P < 0.001) and 128.7% higher when these costs were excluded ($49,378 vs $21,590; P < 0.001) than in nonbleeding patients (Figure 1; Table III).
Bleeding patients were 4.54 times as likely to be admitted to the ICU, 3.40 times as likely to die, and 4.59 times as likely to be readmitted for bleeding within 30 days (all P < 0.001). They also had a 49.1% longer hospital LOS (P < 0.001) than nonbleeding patients. Mortality was significantly higher in bleeding patients than in nonbleeding patients (5.30% vs 1.43%; P < 0.001). Bleeding patients also experienced higher rates of a number of adverse events, including TACO (6.38% vs 1.80%; P < 0.001), TRALI (0.07% vs 0.01%; P < 0.001), and acute respiratory distress (0.36% vs 0.22%; P < 0.001). In contrast, more nonbleeding patients experienced thromboembolic events than bleeding patients (46.12% vs 29.92%; P < 0.001).
Discussion
Our study showed that, in the settings of both trauma and complex cardiac surgery, bleeding patients had significantly worse outcomes than nonbleeding patients. Bleeding patients had longer hospital and ICU LOS, greater inpatient mortality, and higher total costs of care, even when the costs of agents used for bleeding control were excluded. In addition, bleeding patients were more likely to be readmitted to the hospital than nonbleeding patients.
The true total cost of allogeneic blood products is difficult to define, because it must take into account all aspects of the blood collection, production, and administration process. A 2007 study by Shander et al.
suggested that there were 9 cost areas that must be taken into account, ranging from the costs incurred by blood donors to the cost of organizing and maintaining nationwide and/or international hemovigilance programs. The authors estimated that, taking into account only 3 of the 9 described cost elements (cost of producing blood components for transfusion, transfusion logistics, and the treatment of transfusion-transmitted disease and associated adverse outcomes), the real cost of blood per unit in 2007 was a minimum of $1400.
However, the costs are difficult to compare across studies, because each of these studies was performed in a different clinical setting, and varying methods were used to calculate the total.
A large proportion of the total cost of blood products can be attributed to costs incurred post-transfusion, such as the treatment of adverse events and hospital LOS. On average, bleeding patients in the present study stayed in hospital for 4 days longer than nonbleeding patients, an increase in LOS of 38.9%. Of the patients who were admitted to the ICU, bleeding patients stayed an average of 2 days longer than nonbleeding patients. The impact of transfusion on hospital LOS was shown previously, with 1 UK study estimating that hospital LOS accounted for 66% of total costs.
Further analysis of these data showed that the majority of this increased cost could be attributed to increased hospital LOS. Another economic study of patients who underwent colorectal cancer resection showed that total hospital charges of transfused patients were significantly higher than those of nontransfused patients ($28,101 vs $15,978; P < 0.0001), and that charges were shown to increase by 2.0% per unit of RBCs and/or platelets transfused after adjustment for confounding factors.
Allogeneic blood transfusion, hospital charges, and length of hospitalization: a study of 487 consecutive patients undergoing colorectal cancer resection.
In addition, hospital LOS was significantly longer in the transfused group (16.7 days vs 10.3 days; P < 0.0001), and was shown to increase by 1.3% for every unit of RBCs and/or platelets transfused.
A number of previous studies showed that the transfusion of allogeneic blood products is associated with a number of adverse events, including TACO, TRALI, allergic reactions, and the transmission of pathogens.
In the present study, bleeding cardiac surgery and trauma patients who received transfusions experienced significantly higher rates of TACO and TRALI than those who did not receive tranfusions. Studies showed that avoidance of blood product transfusion reduced the risk of complications,
One recent study showed that the implementation of a blood product conservation initiative in the setting of cardiac surgery reduced postoperative morbidity and mortality, reduced health care costs, and improved resource utilization.
Furthermore, the implementation of hemostatic therapy algorithms based on point-of-care (POC)-guided coagulation management was shown to reduce blood product transfusion in different clinical settings.
In the setting of cardiac surgery, one such study compared a POC-guided algorithm with conventional therapy (whereby hemostatic therapy was driven by laboratory coagulation testing).
Patients randomized to the POC algorithm group received fewer FFP and platelet transfusions, had shorter ICU LOS, experienced fewer adverse events, and had a lower mortality rate than patients in the conventional therapy group.
Therefore, we hypothesize that the use of patient blood management strategies may improve clinical outcomes by focusing not only on the reduction of blood product transfusion, but also on minimizing blood loss.
Preoperatively, this consists of identifying patients at risk of bleeding as soon as possible, and ensuring any necessary treatment is promptly administered. Intraoperatively, the use of meticulous hemostasis, surgical, and anesthetic techniques can all help to minimize blood loss. For example, maintaining physiologic body temperature during surgery can reduce blood loss and transfusion requirements.
Influence of body core temperature on blood loss and transfusion requirements during off-pump coronary artery bypass grafting: a comparison of 3 warming systems.
and when used in combination with a transfusion algorithm, POC-guided coagulation therapy in the setting of cardiac surgery has been shown to reduce both transfusion requirements and costs.
Our study had limitations similar to other observational studies. ICD-9 codes were used to determine patient inclusion; identification of comorbidities and coding practices might differ among hospitals. Similarly, data regarding use and costs relies on hospitals reporting from their chargemaster records. Finally, our results only considered the variables identified, and the unobserved differences, including differences in the types of surgeries or trauma types, might affect the outcomes considered.
Our study showed that, in the settings of trauma and complex cardiac surgery, bleeding patients who subsequently receive allogeneic blood product transfusions have poorer outcomes, which results in higher costs and longer hospital LOS. Minimizing blood loss or prompt resolution of bleeding with targeted therapy is likely to reduce additional risks and costs incurred by patients with trauma or who are undergoing complex cardiac surgery. These concerns are important both to care providers and to hospital administrators, whose cost perspectives we took. These professionals both worry, to varying degrees, about quality of care and the effects on hospital finances. Procedure-related outcomes generating additional costs and LOS without reimbursement adjustments bode poorly for the finances of the entire provider institution. Further, the results of this study provide break-even estimates for administrators and clinicians to use when evaluating the costs of therapies designed to avoid surgical bleeding.
Conflicts of Interest
Art Zbrozek is an employee of CSL Behring. Glenn Magree is an employee of Premier Research Services, which received funding from CSL Behring to conduct this study. The authors have indicated that they have no other conflicts of interest regarding the content of this article.
Acknowledgments
Both authors designed this study and prepared the article, and approved the final article. Medical writing assistance for this study was provided by Meridian HealthComms, funded by CSL Behring. The input of Joanna Whyte, while at CSL Behring Global Health Economics, is acknowledged.
Allogeneic blood transfusion, hospital charges, and length of hospitalization: a study of 487 consecutive patients undergoing colorectal cancer resection.
Influence of body core temperature on blood loss and transfusion requirements during off-pump coronary artery bypass grafting: a comparison of 3 warming systems.
Occurring in bleeding versus nonbleeding patients.
00861-9&id=gr1.jpg){kind=link}