Cost Reduction in Blood Transfusions After Implementation of Decision Protocol Based on Hemoglobin and Anemia Symptoms: A Pre-Post Analysis

Introduction

More than 100 million red blood cell (RBC) units are transfused globally each year [1]. From this number, around a quarter is transfused in the European Union member states [2], and in Poland, approximately 1.3 million units [3].

Modern medicine, especially perioperative medicine, depends on a constant supply of donor blood. RBC transfusion is a life-saving procedure in patients with massive bleeding. Recovery of patients from various diseases or procedures without donor blood would be prolonged [4].

On the other hand, blood transfusion is not without risks and can be associated with increased morbidity and mortality [5,6]. The potential risks include nosocomial infections, immunomodulation, transfusion-related cardiac overload, transfusion-related acute lung injury, thromboembolic complications [7], acute hemolytic transfusion reaction, and transfusion-transmitted bacterial infections [4,8]. According to the World Health Organization, the average incidence of severe transfusion reactions is 12.2 per 100 000 distributed blood components [9]. In Spain, more than 1500 patients receive a diagnosis of this type of adverse event annually [10]. Transfusion of RBCs can also prolong hospitalization [11], which increases the risk of hospital-acquired anemia [12].

Blood products are an expensive resource. Moreover, morbidity and prolonged hospitalization associated with RBC transfusion lead to increased financial costs [13–15]. Appropriate use of RBCs can improve patients’ outcomes and reduce the costs associated with RBC transfusion [16]. Therefore, it is essential to develop a standardized approach to determine the need for transfusion [17]. Rational use of blood components constitutes a cardinal pillar of patient blood management programs, which reduces unnecessary healthcare costs by decreasing blood transfusions [17–21]. According to the latest transfusion guidelines, hemoglobin (Hb) concentration cannot be the only criterion for transfusion [22], and modern transfusion practice moves away from this single-factorial approach [23]. In the environment of limited healthcare resources, it is important that healthcare organizations have accurate and comprehensive information on the impact of procedures designed to improve the appropriateness of RBC transfusions on costs of blood products and all transfusion-associated costs [24]. No publications are analyzing the financial impact of a 2-factorial allogeneic RBC transfusion decision protocol, which constitutes an obvious gap in knowledge.

This study’s primary aim was to assess the cost of RBC transfusions before and after implementation of an in-hospital RBC transfusion decision protocol. The secondary aim of the study was to estimate the current cost of a RBC transfusion procedure in the local healthcare system.

Material and Methods

HOSPITAL SETTING:

Our institution is a large (644 hospital beds), highly specialized third level reference academic medical center providing services in an area with a high level of social demand. Transfusions performed in 2 pediatric departments were not included in the study because the transfusion improvement procedure did not include non-adult inpatients. The dedicated Transfusion Committee is responsible for overseeing the supply and utilization of blood components in our institution and developing standard operating procedures in transfusion medicine. The local Transfusion Committee developed the current RBC transfusion decision protocol.

RBC TRANSFUSION DECISION PROTOCOL:

Our protocol was based on the most recent RBC transfusion guidelines and some additional criteria [19]. The approach to RBC transfusion decision making was 2-factorial. The first factor was Hb concentration: transfusion was generally classified as appropriate if pre-transfusion Hb concentration was <70 g/L or <80 g/L in patients with coronary artery disease, myocardial ischemia, and myelodysplastic syndrome. Due to local specificity, no other conditions were mentioned that could require higher RBC transfusion triggers, such as thalassemia or sickle cell disease. Nevertheless, if the decision was made to transfuse RBCs at a higher than recommended Hb concentration, there was an option in the computerized physician order entry to give a reason for transfusing at a higher Hb concentration. For pre-transfusion Hb concentration <60 g/L, no additional condition deemed a transfusion appropriate. The second factor was the presence of at least 1 symptom or sign of anemia, or at least 1 abnormal laboratory marker of anaerobic metabolism, despite attempts to improve the patient’s tolerance to anemia, according to the previously published algorithm [11,25]. The symptoms of anemia that were looked for were the following: impaired concentration, attention deficit, dizziness, headache, dyspnea, tachypnea, tachycardia, hypotension, ischemic changes in electrocardiogram, and angina. The anaerobic metabolism markers used in our medical center were lactate concentration and central venous oxygen saturation. All RBC transfusions performed outside the aforementioned clinical scenarios were classified as inappropriate.

CLINICAL DATA:

The starting point for searching for RBC transfusions was a local blood bank computer system (InfoMedica, Asseco Medical Solutions, Poland). All clinical data were obtained from the hospital’s electronic health records (AMMS, Asseco Medical Solutions, Poland). Clinical data included indication for transfusion, number and types of transfused RBCs, signs and symptoms of anemia before RBC transfusion, laboratory markers of anaerobic metabolism before RBC transfusion, and Hb concentration (from venous blood) up to 24 h before RBC transfusion. We calculated the transfusion rate, namely the percentage of patients who received at least 1 RBC unit. The exclusion criteria were indications for RBC transfusion of bleeding or preparation for surgery.

COST CALCULATIONS:

A single RBC transfusion consisted of a direct acquisition cost for a particular type of RBC (with or without modification) and transfusion-associated costs. We estimated that for every RBC recipient, 2 blood group tests had to be performed to get the “confirmed” blood group test result. For each RBC transfusion, 1 crossmatch test had to be performed, and 1 infusion set had to be used. If the same patient underwent transfusion more than one time, the cost of all subsequent RBC transfusions did not include the cost of blood group determination.

To calculate the cost of healthcare workers’ labor during RBC transfusion, the remuneration for the work of doctors, nurses, and laboratory diagnosticians was considered. We estimated the minimal working time necessary to perform transfusion of 1 RBC safely: 30 min for a physician, 30 min for a nurse, and 30 min for a laboratory diagnostician (10 min for each test=2 blood group tests and 1 crossmatch test). If the same patient was transfused more than one time, the cost of all subsequent RBC transfusions involved shortened labor time of laboratory diagnostician (10 min for 1 crossmatch test). To get the remuneration of healthcare workers, we used the minimal wages in the healthcare sector in Poland in 2022 and 2023, according to the Polish law [27,28]. As the minimal wage for physicians and nurses varies with the level of education, some estimations had to be made. Transfusions of RBCs in Poland are typically performed by junior doctors in specialty training programs (residents), mainly at the beginning of the specialty training. Therefore, RBC transfusion labor time costs for healthcare workers were calculated based on the following formulas:

If there was an adverse transfusion reaction, all the costs associated with its diagnosis and treatment were included in cost calculations. The costs mentioned above were summed to estimate the impact of the procedure on the total cost incurred by RBC transfusion. All costs were paid initially in a local currency (polish zloty, PLN). All costs were then converted to European currency (euro, €). The average exchange rate from the two 6-month periods in the years 2022 and 2023 given by the National Bank of Poland was used to convert PLN to €.

STATISTICAL ANALYSIS:

All statistical analyses were performed using licensed statistical software (18.0 Basic Edition, Stata, StataCorp LLC, College Station, TX, USA). Continuous variables are presented as medians and interquartile ranges, whereas categorical variables are presented as frequencies and percent. Inter-group comparisons were performed with paired t tests. P<0.05 was considered statistically significant.

Results

The average exchange rate was 4.65 PLN=€1. Based on the formulas contained in the Material and methods, calculations were made of RBC transfusion labor time costs of healthcare workers directly involved in an RBC transfusion procedure.

Healthcare worker’s RBC transfusion labor time costs in 2022 were as follows: Sp=€4.9, Sn=€4.1, Sd=€4.7. Healthcare worker’s RBC transfusion labor time costs in 2023 were as follows: Sp=€6.1, Sn=€5.6, Sd=€6.2.

The cost components of a single-unit RBC transfusion are shown in Table 1. The acquisition cost of an RBC unit depends on its type and modification. In certain patients, certain types of RBCs and modification thereof can be required. For example, washed RBCs are used when a patient has a history of transfusion allergic reaction. The universally used modification of RBC is leucoreduction, which reduces the risk of febrile non-hemolytic transfusion reaction, immunization, and transmission of leucotropic viruses. The cost of transfusion of a leucodepleted RBC (the preferred type of RBC) in the successive 6-month periods was €109.3 and €125.7, respectively. The non-direct RBC cost of transfusion in the successive periods constituted 48.2% and 52.8%, respectively, of the total cost of transfusion.

In each of the studied periods, there was 1 adverse reaction associated with transfusion. In 2022, a mild transfusion reaction occurred: febrile non-hemolytic transfusion reaction (FNHTR); and in 2023, a serious transfusion reaction occurred: transfusion-related cardiac overload. Both RBC transfusions were classified as inappropriate. In both cases, the costs associated with post-transfusion reaction assessments included the determination of AB0 and RhD antigens, as well as a comprehensive examination for the presence of immune antibodies in the recipients’ blood sample before and after transfusion, drainage segment samples, examinations on stored laboratory samples following compatibility assessments, and compatibility tests with donor RBCs. The cost of these determinations amounted to €86.7. A lymphocytotoxicity test was also performed (€23.7). A direct anti-globulin test was also performed on the patient with FNHTR (€26.2). The FNHTR treatment involved administering the patient 100 mg of hydrocortisone (€4.3), 1000 mg of paracetamol (€0.5), and 500 mL of PlasmaLyte (€0.5). Additional tests were conducted, such as complete blood count with differential (€5.4), prothrombin time (€2.1), activated partial thromboplastin time (€1.7), creatinine, and estimated glomerular filtration rate (€1.3). In a patient with transfusion-related cardiac overload, 16 mg of dexamethasone (€2.4) and 40 mg of furosemide (€1.9) was administered and anesthesiology (€32.3) and cardiology consultation with echocardiography (€36.6), arterial blood gas analysis (€6.9), catheterization (€0.3), complete blood counts with differential (€5.6), D-dimers (€8.6), C-reactive protein (€3.2), procalcitonin (€18.3), and a chest computed tomography (€64.5) was performed.

The appropriateness of RBC transfusions in the period before the introduction of the procedure was analyzed [11]. Following the implementation of the procedure, there was a significant reduction in the transfusion rate, the number of transfused RBCs, and an increase in the percentage of appropriate RBC transfusions (Table 2). A description of the impact of the new procedure on transfusion metrics was published before [29].

There was a reduction in RBC transfusion costs following the implementation of the 2-factorial RBC transfusion improvement procedure (Table 3). A detailed breakdown of the costs of appropriate and inappropriate RBC transfusions after the implementation of the procedure is presented in Table 4.

Complete elimination of inappropriate RBC transfusions within the period of January to June 2023 reduced the cost of RBC transfusions by another €25014.5 (62.9% of the cost of all transfusions). These costs were calculated in the same way as described in the methodology.

The overall length of hospitalization decreased from 3.2 to 3.1 days in the successive periods (P<0.05).

Discussion

Transfusion of RBCs is a common medical procedure performed in hospitalized patients, with an overall transfusion rate of 10% [30]. The transfusion rate depends very much on the patient population [24]: 2.5% in parturient patients [31], 19.4% in surgical patients [32], and 28% in elderly patients with a fractured hip [33]. Before implementing our RBC transfusion decision protocol, the transfusion rate was already low at 1.8%.

Apart from being common, RBC transfusion is also not neutral; quite the opposite, it is associated with potential risks and significant costs attached to it. Therefore, attempts have been made to minimize RBC transfusions [16]. Implementing our RBC transfusion decision protocol led to a 3-fold decrease in transfusion rate, from 1.8 to 0.6%, which translated to a 51.4% reduction in the number of RBC transfusions despite the increase in hospitalizations by 1570 hospital stays. Similar actions aimed at reducing the number of RBC transfusions were also undertaken in other hospitals worldwide. In Austria, Enko et al showed a 13.8% decrease in the number of RBC units transfused in the period after the implementation of the algorithm-driven anemia treatment program (2012–2017), compared with the period before the start of the program [34]. In the United States, Goel’s study in 2018 also observed a decline in blood transfusions, in which the percentage of hospitalizations with RBC transfusion decreased from 4.2% (Q4 2015) to 3.8% (Q4 2018) [35]. Also, Verdecchia et al showed a 29.9% reduction in the number of RBCs transfused between 2007 and 2015 [36]. Since 2014, a decreasing trend in the frequency of hospitalizations involving RBC transfusions has been observed in the United States, where, among all hospitalized patients since 2018, approximately 3.8% required RBC transfusion [36]. A study by Wu et al showed the impact of the implementation of a patient blood management program in an urban teaching hospital: the mean pre-transfusion Hb concentration decreased from 7.3 in 2013 to 6.6 g/dL in 2019, fewer RBCs were given at pre-transfusion Hb concentrations ≥7 g/dL, there was a decrease in 2-unit RBC transfusions at Hb ≥7 g/dL, and there was a 34% reduction in the number of RBC transfusions annually [37].

The average cost of an RBC transfusion varies between countries and institutions, with a cost between $71 and $2388 reported [24,38,39]. Because the transfusion of RBCs is a complex process consisting of many separate stages, including official, administrative, laboratory, nursing, and medical activities [40], accurate calculation of the cost of a blood transfusion is not easy. In this study, we attempted to calculate average blood transfusion costs [41]. Although a patient insured by a public health care system in Poland has access to information about the costs of his treatment via the Online Patient Account, due to the complexity of calculating a highly specialized service such as a RBC transfusion, the patient does not have direct information about the cost of this procedure. In our study, the average cost of RBC transfusion in the successive years was €109 and €126, respectively. As the literature shows, costs vary within and between countries. The cost in France was €339.6 per RBC transfusion [39], €506 in Italy [42], and $2388 in the United States [38]. The estimated total cost savings attributed to the 6-year patient blood management program duration after full implementation in 2014 amounted to $2.1 million, leading to the conclusion that, overall, patient blood management implementation significantly decreased RBC transfusions and enhanced transfusion practices. The findings emphasize that successful patient blood management strategies do not always necessitate extensive resources or increased budgets but instead rely on intuitive methods, as evidenced by the present study [37].

In addition to concerns about patient safety and transfusion-related adverse outcomes, there is the issue of resource overuse, coupled with the high costs associated with RBC transfusion, which may contribute to financial problems for healthcare systems [43].

The overall length of hospitalization slightly decreased after the implementation of the RBC transfusion decision protocol in the present study. However, it is impossible to say with certainty that it was due to the new protocol.

The strength of this study project is that, to the best of our knowledge, it is the first to estimate the cost of RBC transfusion in the local healthcare system. Due to its retrospective nature, the study also had limitations. Some clinical information may have been lost due to needing to be recorded in the hospital’s information technology system. Therefore, despite a thorough analysis of daily patient hospitalization records prepared by attending physicians, with a specific focus on descriptions related to blood transfusions, there is a possibility that certain information was missing; therefore, specific transfusions could have been categorized as inappropriate, when in fact, they were clear indications for transfusion. Therefore, the additional costs that could be saved by reducing inappropriate transfusions are less than those reported in the study. Finally, the estimated cost of blood transfusion presented in this study is not the total cost because, due to the lack of available data, it needs to considers factors such as hospital maintenance costs and equipment depreciation; however, these can differ between institutions.

Conclusions

Implementing a RBC transfusion decision protocol based on a 2-factorial approach to transfusion decision making can lead to more than 2-fold reduction in costs of RBC transfusions in a low transfusion rate setting. Even after implementation of the procedure, further education of prescribers through different educational initiatives is required, as there may still be potential to reduce the cost associated with inappropriate transfusions and improve patient outcomes. The non-RBC acquisition cost of an RBC transfusion procedure approximates the cost of a blood component itself, each of these components coming close to €60 in the local healthcare system.