The burden of blood transfusion: a utilization and economic analysis – a pilot study in patients with chemotherapy-induced anemia (CIA)

Abstract

Objective:

The objective is to measure the burden of blood transfusion of Packed Red Blood Cells (PRBCs) in patients with chemotherapy-induced anemia (CIA) on the institutional outpatient transfusion center.

Methods:

This is a retrospective chart review (starting July 1, 2010, working backwards until 120 evaluable patients are accrued) at a single institutional transfusion center in the US. The mean and standard deviation (SD) were calculated for patient’s age, pre-transfusion Hgb level, and other transfusion-related activities.

Results:

One hundred and twenty records were reviewed. The majority included patients who were female (71%), African American (61%), and had either Medicare (48%) or private insurance (39%). The mean patient age was 59 years and the average pre-transfusion Hgb was 7.9 g/dL. The average patient visit to facility ranged from 213 min for one PRBC unit to 411 minutes for three PRBC units. The mean staff time for patient evaluation was 66 minutes. Actual time for transfusion was ∼100 min for each PRBC unit; 90% of patients received two PRBC units. Staff was engaged in direct patient care for an average of 322 min for two PRBC units. The labor cost of transfusion (in 2011 $US) ranged from $46.13–$49.33 per PRBC unit. The estimated fully loaded bundled cost was $596.49 for transfusion of one unit of PRBC. Limitations of the study include: the site included in this study may not be applicable to all sites in practice and the evaluated patient population was varied, with the exception that all patients were treated for some type of malignancy; and the review of blood bank records for 120 patients was not 120 independent events and, as such, may not have adequately captured actual variability.

Conclusions:

This analysis quantifies expense in terms of time for administration of the transfusion, as well as costs associated with outpatient blood transfusions.

Keywords::

• Transfusion, PRBCs, Chemotherapy-induced, Anemia, cost

Introduction

Anemia has come to be viewed as a relatively common condition in cancer patients, with the actual incidence largely dependent on the type and extent of the malignancy, as well as the treatment employed. Up to 60% of patients with solid tumors and lymphoma may experience anemia, and those receiving myelosuppressive chemotherapy and/or radiation therapy have an even higher incidence (70%–90%)1. Individuals who are treated with cytotoxic therapy for lymphomas, lung tumors, and gynecologic (ovarian) or genitourinary tumors experience a particularly high rate of anemia2,3. Co-morbid conditions and advanced age, which commonly co-exist, may increase susceptibility to anemia and decrease a patient’s ability to tolerate anemia-related symptoms1.

The incidence and severity of chemotherapy-induced anemia (CIA) is further dependent on a variety of factors such as the type, schedule, and intensity of chemotherapy administered, and whether the patient has received prior myelosuppressive chemotherapy and/or radiation therapy4. Platinum-based treatments (e.g., cisplatin, carboplatin), commonly used in lung, ovarian, and head and neck cancers, are well known to induce anemia due to combined bone marrow and kidney toxicity. Antimicrotubular agents (e.g., taxanes, vinorelbine) and camptothecins (e.g., irinotecan, topotecan) are also considered particularly myelosuppressive4. Dose intensity, the increasingly widespread practice of administering higher doses of chemotherapy over a shorter period of time, is associated with an increased risk of myelosuppression as well. Newer biologic agents are not cytotoxic or myelosuppressive in the traditional sense, but have nevertheless also contributed to the incidence of anemia5.

The association between uncorrected anemia before or during chemotherapy and poorer patient outcomes has been reported in a number of studies6–9. Anemia may lead to a multitude of symptoms (e.g., fatigue, tachycardia, cognitive impairment, shortness of breath, depression, dizziness) and may also have an adverse impact on co-morbid conditions such as cardiac and pulmonary disease. The economic burden of anemia is substantial, with costs arising from inpatient and outpatient care, as well as decreased productivity. As such, timely diagnosis and correction of CIA is essential, and is addressed by regularly updated guidelines released by several organizations (e.g., National Comprehensive Cancer Network; NCCN). Although some of the basic guideline recommendations (e.g., diagnosis, assessment) have not changed markedly over the past several years, many of the key therapeutic positions have been significantly revised, primarily as a result of emerging studies addressing the safety of erythropoiesis-stimulating agents (ESAs) and the accompanying regulatory/labeling revisions. As a reflection of these changes, there has been renewed interest in the risks and benefits of red blood cell (PRBC) transfusions.

Second only to surgical patients, oncology patients are among the largest consumers of blood products10,11. Most oncology patients requiring RBC support are transfused in outpatient blood transfusion centers. If such a center is not available, it may necessitate a patient admission to a 23-h short-stay room or to the emergency department for transfusion. Because of the large number of oncology and surgery patients as well as those presenting to the emergency department secondary to accidents, there is increased burden, not only in regard to staff and institution time/space, but also in terms of increasing demand for scarce blood products12,13. Transfusion of blood involves significant time and expense for both providers and patients, as evidenced by studies conducted prior to the Centers for Medicaid and Medicare Services (CMS) National Coverage Decision (NCD) regarding ESA use in 200714–16. In today’s busy practice environment, time is a commodity, and any opportunity to free up ‘chair time’ by avoiding routine procedures contributes to a more efficient provider delivery system.

Studies regarding the safety of ESAs have prompted hematologists and oncologists to reconsider routine use of these agents for correction of CIA. Consequently, clinicians are once again relying on blood transfusion for managing CIA17,18. A more recent study examining resource utilization and provider charges/costs for a surgical patient population after the NCD determination regarding ESA use indicated a higher than previously documented total cost for transfusion. However, since the NCD in 2007, no known studies of transfusion costs in patients with CIA have been conducted. As such, results observed by Shander et al.16 warrant a new investigation examining the burden of blood transfusion, specifically in patients with CIA. Because of the intense competition for blood products and the staff required to evaluate and manage patients receiving blood transfusions, it is reasonable to attempt to quantify the impact of RBC transfusion on available staff resources and costs to the healthcare system when alternative therapies (i.e., ESAs) are available.

The purpose of this pilot study was to gather preliminary data and determine feasibility of assessing the time components needed to understand the burden of blood transfusion at an oncology practice site/outpatient transfusion department. The primary objective of this pilot study is to understand the burden of blood transfusion in patients with chemotherapy-induced anemia on the private practice and the institutional outpatient transfusion department, blood bank or other area where blood is administered. The burden of blood transfusion will be characterized by quantifying time and personnel resources associated with the administration of blood (i.e., Packed Red Blood Cells: PRBCs) transfusions. A secondary objective will be to quantify cost (in terms of charges to payer) for staff activities and supplies consumed by office staff in the scheduling of and administration of blood transfusions.

This study will not, however, evaluate the suitability or advisability of one therapy vs another, but rather simply estimate the burden of blood transfusion on the healthcare system. The majority of the costs associated with blood administration are for allocation of staff time for screening and processing. The billing charges are pre-arranged and negotiated with payers, and are not necessarily correlated with actual hours/minutes of time. Should the results of this pilot warrant further scrutiny, a multi-center trial should be conducted to include a variety of currently existing practice/blood transfusion models and contrast charge/work units across all sites.

Methodology

This pilot study describes blood transfusion at a single site (private oncology practice/local institutional transfusion center), involving a record review of blood bank/transfusion center records for packed RBC (PRBC) transfusion (120 records). The source of the data for this portion of the study came from the blood bank administration records for oncology patients who have received PRBCs. Blood records were screened, starting July 1, 2010, working backwards until 120 evaluable patients were accrued.

The study population included oncology patients referred by a private oncology practice to a transfusion center for blood transfusion. The source of the retrospective data was blood bank/transfusion center blood administration records.

Inclusion criteria

• Oncology patient with anemia receiving PRBC transfusion due to anemia associated with CIA; and

• Age >18 years.

Exclusion criteria

• Age <18 years; and

• Primary diagnosis other than cancer.

Medical chart abstraction

Potential eligible records were identified through a database query of electronic medical records (EMR) and blood bank or a billing/claims database for oncology patients receiving PRBCs on or before July 1, 2010. The first 120 patients were selected from those who had a diagnosis of cancer and received PRBCs. When the list of all potential patients was obtained from the database query, records were selected sequentially, starting with the most recent (July 1, 2010) and working backwards, and reviewed for inclusion/exclusion criteria. When all criteria were satisfied, records were reviewed until 120 records were completed, or until all possible records were exhausted.

Data elements

Blood Bank Record Review:

• Age, gender;

• Reason/cause of anemia and PRBC transfusion;

• Hemoglobin (Hgb) level prior to transfusion as recorded in transfusion record;

• Primary payment source;

• Number of units of PRBCs transfused;

• Adverse Events (AEs) related to blood transfusion;

• Treatment modalities to treat AEs related to blood transfusion; and

• Total time for PRBC infusion/AEs related to blood transfusion.

Study outcomes (end-points)

1. Patient visit to facility;

2. Mean time for administration of a PRBC transfusion (including blood transfusion related AEs if any); and

3. Cost:

   • Evaluated and quantified in terms of estimates of fully loaded bundled cost; and

   • Salary calculations based on ‘Office Infusion Nurse Salary in Lexington County, South Carolina’, http://www.indeed.com/salary/q-Office-Infusion-Nurse-l-Lexington-County,-SC.html. Accessed September 7, 2011.

End-point 2 was provided as a means to determine and quantify burden to the practice in terms of staff labor. End-point 3 provided a quantification of cost to the healthcare system by quantifying bundled cost accrued by administration of PRBCs.

Statistical analysis

Exploratory and descriptive analyses were:

• Gender and race/ethnicity are described by N’s and percentages.

• Patient age, Hgb value and number of PRBC units are described by mean and standard deviation.

• Continuous end-points (e.g., time required for PRBC transfusions) are described by their mean, standard deviation, and range of values.

• Cost of blood transfusion per patient in terms of charges billed to insurers.

Results

The retrospective record review included 120 records, of which the majority included patients who were female (71%), African American (61%), and had either Medicare (48%) or private insurance (39%). The mean patient age was 59.9 years, the mean Hgb at time of transfusion was 7.9 g/dL, and the mean number of PRBC units transfused was 2.0. Baseline demographics are included in Table 1.

Table 1. Patient characteristics.

	n	Mean	(SD)
Age	120	59.9	13.8
Hgb	120	7.9	0.8
PRBC units transfused	120	2.0	0.4
Gender: N (Pct)
Female	85 (71)
Male	35 (29)
Race/ethnicity N (Pct)
Caucasian	40 (33)
African American	73 (61)
Other or unknown	7 (6)

Patient duration of visit to facility

The average duration of a patient visit to facility for transfusion ranged from 213 min (3.6 h) for one PRBC unit to 411 min (6.9 h) for three PRBC units (Table 2). The mean staff time for patient evaluation for transfusion was 66 min (Table 2).

Table 2. Transfusion characteristics*.

	n	Mean	(SD)	Range
Mean time for Patient Evaluation
		66	34	24–176
Transfusion of 1 PRBC unit, n = 10
Time: Transfusion of unit 1	10	108	11	85–120
Time: Patient visit to facility	10	213	45	135–290
Transfusion of 2 PRBC unit, n = 98
Time: Transfusion of unit 1	98	101	13	75–130
Time: Transfusion of unit 2	98	101	18	48–205
Time: Patient visit to facility	98	322	44	240–465
Transfusion of 3 PRBC unit, n = 12
Time: Transfusion of unit 1	12	98	8	85–110
Time: Transfusion of unit 2	12	101	9	90–125
Time: Transfusion of unit 3	12	98	15	70–130
Time: Patient visit to facility	12	411	49	340–490

*Note all times measured in minutes.

Retrospective chart review

Actual time for transfusion was ∼100 min (1.7 h) for each PRBC unit; 90% of patients received two PRBC units (Table 2). Staff was engaged in direct patient care for an average of 322 min (5.4 h) to infuse two PRBC units (Table 2).

Transfusion-related cost

The labor cost of transfusion ranged from $46.13–$49.33 per PRBC unit (Table 3). The fully loaded bundled cost estimates were $596.49 per unit transfused (Table 4).

Table 3. Transfusion-related labor cost.

All transfusions	n	Mean cost	Range
Transfusion of 1 PRBC unit
Transfusion of unit 1	10	$49.33	$38.82–$54.81
Transfusion of 2 PRBC units
Transfusion of unit 1	98	$46.13	$34.25–$59.38
Transfusion of unit 2	98	$46.13	$21.92–$93.63
Total time for 2 units	98	$92.13
Transfusion of 3 PRBC units
Transfusion of unit 1	12	$44.76	$38.82–$50.24
Transfusion of unit 2	12	$46.13	$41.11–$57.09
Transfusion of unit 3	12	$44.76	$31.97–$59.38
Total time for 3 units		$135.65

Notes: $57,000/y, 2080 contract hours, $0.467/min. Does not include benefits. Also cost analysis does not include overhead or any laboratory medicine costs.

Table 4. Estimated fully loaded bundled cost for transfusion of one unit of PRBCs in a Tertiary Community Medical Center.

	Testing		Product and Product Preparation		Infusion		Total PRBC Unit Cost
Overhead	9.3%	$ 56.46	19.1%	$115.95	16.9%	$102.59	45.3%	$275.00
Direct materials	1.0%	$ 5.37	18.0%	$ 96.69	5.2%	$ 27.93	24.2%	$130.00
Labor (variable and direct)	9.7%	$ 60.90	12.9%	$ 80.99	7.9%	$ 49.60	30.5%	$191.49
Totals	20.0%	$122.73	50.0%	$293.63	30.0%	$180.13	100.0%	$596.49

Discussion

Assessing time components and costs associated with the process of blood transfusion in outpatient oncology and institutional outpatient transfusion centers represents an important approach in understanding the burden of blood transfusion on the healthcare system. Our current study suggests significant expense in terms of total office time, administration of transfusion, as well as costs in association with outpatient blood transfusions. The average patient visit to facility for transfusion included 213 min (3.6 h) for one PRBC unit, 322 min (5.4 h) for two PRBC units, and 411 min (6.9 h) for three PRBC units. Since 90% of patients received two PRBC units, staff was engaged in direct patient care for over 5 h in the majority of cases. The labor cost of transfusion was considerable, ranging from $46.13–$49.33 per PRBC unit. When testing, product and product preparation are included with transfusion, the estimated fully loaded cost for the transfusion of one unit of PRBC is $596.49.

Our results are consistent with previous research that highlight substantial transfusion-associated time and expense for both providers and patients12–14. In a study examining the impact of 100 outpatient transfusion visits by 36 subjects on hospitals and their costs, the mean transfusion duration was 103 and 102 min for the first and second PRBC units, respectively, and for two-unit visits, mean duration between units was 18 min. Mean total time for a two-unit transfusion was 223 min (3.7 h)14. Transfusion visits required a mean of 67 min of labor (49% for clinical assessment and 55% during pre-transfusion phase). Total administration costs for one- and two-unit transfusions were $166 and $307, respectively16. Another trial that analyzed resource utilization by patients receiving RBC transfusion in a tertiary-care cancer center found that an estimated 71,504 min were consumed by all PRBC transfusions, constituting 5.4% of all outpatient clinic chair-time events. The authors concluded that RBC transfusion consumed a significant proportion of chair time, which could otherwise have been utilized for chemotherapy administration13. Finally, another study utilizing a costing model to determine the cost of supplying surgical patients with blood transfusions found that all major process steps, staff, and consumables to provide transfusions to surgical patients (including usage frequencies, and direct and indirect overhead costs) contributed to per-PRBC-unit costs between $522–$118316. These estimates exceed previously reported estimates and were 3.2–4.8-fold higher than blood product acquisition costs.

More, recently, the Red Cross raised the cost of blood by as much as 35%, depending on the region, to an average of $215 per pint. Added to this estimate is the expense of processing, which increases the cost of one unit of whole blood to as much as $42519. Many health insurance companies do not cover the first two or three pints of blood needed in a medical procedure and the cost is transferred to the patient. In instances where there is no insurance, hospitals are bearing the brunt of this cost. It is estimated that hospitals may spend up to an additional $3 million per year on blood, which leads to other cutbacks within facilities19.

While our current study provides important clinical insights regarding the burden of transfusion in outpatient settings, several limitations should be considered. While every attempt was made to select a ‘typical’ oncology practice and model for blood transfusion, results from this site may not be applicable to all sites. For example, some sites may actually have a transfusion or infusion center adjacent to or within the actual oncology practice, while others do not. It is highly likely that different total times may be obtained from these alternate models than those obtained from this study model. Also, the evaluated patient population was varied, with the exception that all patients were treated for some type of malignancy. Another potential limitation comes from the fact that review of blood bank records for 120 patients was not 120 independent events (likely few staff are involved) and may not have adequately captured the variability. However, administration of PRBC transfusion, the only event measured, was likely to be reliable and consistent across these patients, with the expectation that occurrence of AEs related to blood transfusion may have introduced variability.

Conclusions

The results of this pilot study are consistent with previous research of blood transfusion-associated time and expense. The findings suggest significant expense in terms of total office time, administration of transfusion, as well as costs in association with outpatient blood transfusions. Additional research to further quantify expenses in terms of time for administration of the transfusion, as well as further details on costs associated with outpatient blood transfusions and identification of other patient and center-specific variables is warranted.

Transparency

Declaration of funding

Amgen was the research sponsor for the study. PC-L and SS are employees of Amgen and they are co-authors of this article.

Declaration of financial/other relationships

PC-L and SS are employees of Amgen. FJK, KF, and AvB have no conflicts to disclose. JFR is a principal of the RJM Group LLC, which received research support from Amgen for the conduct of this study. CMRO peer reviewers may have received honoraria for their review work. The peer reviewers on this manuscript have disclosed that they have no relevant financial relationships.

Acknowledgments

The authors wish to thank Marissa Seligman, PharmD, for her editorial assistance in the preparation and submission of this manuscript. Financial support was provided by RJM Group LLC.