Blood transfusion is a common hospital procedure, with 5.7% of hospitalized patients receiving a red blood cell (RBC) transfusion during 2014 in the United States.¹ To monitor the numbers of blood collections and transfusions in the United States, national surveys have been conducted since 1971 with support from the federal government. Since 1997, the National Blood Collection and Utilization Survey (NBCUS) has been the primary source of data related to annual blood collections and transfusions occurring in the United States.^2–5

Since 2008, data reported to the NBCUS have identified continuing declines in blood products collected and transfused. Decreases in collection have likely been a result of a decrease in demand for blood due to broad implementation of patient blood managements.^1,4,6 Patient blood management (PBM) refers to evidence-based medical and surgical programs to minimize blood loss and ensure appropriate hemoglobin levels to minimize the need for transfusion.⁷ These programs have been demonstrated to reduce the amount of blood transfused while improving patient outcomes.^7,8 While declines were steepest between 2008 and 2013, the rate of decline slowed and appeared to stabilize between 2015 and 2017, particularly among larger U.S. hospitals.⁴ In addition, the price paid for blood components has also decreased during this time. The median price paid by hospitals for a leukoreduced RBC unit declined from $221 in 2013 to $207 in 2017.⁴ This combination of declining transfusions and reductions in prices paid for blood products suggests financial pressures on blood collection organizations in the United States.^6,9,10

Here, data from the 2019 NBCUS are presented and include estimates on blood component collection, distribution, and transfusion in the United States.

The survey methods and questionnaire design for the 2019 survey were consistent with previous NBCUS surveys.^2,3,11 The 2019 survey included 44 questions, of which 18 were intended for blood collection facilities and 26 were directed toward transfusing hospitals.

Blood collection facilities were identified using the US Food and Drug Administration’s Blood Establishment Registration database, through which 53 community-based (e.g., non-hospital-based) and 90 hospital-based blood collection centers were identified and sent a survey (Figure 1). Community and hospital-based blood centers that were included in the database in 2017 but not included in 2019 were contacted to ascertain reasons for no longer being included in the blood establishment database. Transfusing hospitals were identified using the 2017 American Hospital Association annual survey database. As with previous surveys, hospitals with the following characteristics were excluded for any of the following reasons: performing <100 inpatient surgeries annually, location in a US territory, operated by the military or Department of Justice, or classified as rehabilitation, acute long-term care, or psychiatric facilities. After these exclusion criteria were applied, 3808 hospitals were included in the sampling frame. Of these, 100% of hospitals performing ≥1000 inpatient surgical operations a year (n = 2139) were surveyed and 40% of hospitals that performed 100 to 999 inpatient surgical operations per year were randomly selected to be surveyed (n = 669/1669).

For 2013–2017, the NBCUS surveys were administered with the MrInterview (IBM Corporation, Armonk, New York) software package. The 2019 survey was conducted using REDCap (Research Electronic Data Capture, Nashville, Tennessee).¹²

Following survey dissemination, facilities were contacted via mail, email, and/or phone to increase participation. Due to challenges posed by the COVID-19 pandemic, the survey deadline was extended by 2 months to allow hospitals and blood centers additional response time.

National estimates were calculated for the number of units of blood and blood components collected, distributed, transfused, and outdated in 2019. These estimates were rounded to the nearest 1000 units. To assist with weighting and imputation, blood collection centers were stratified based on expected levels of collection or transfusion in 2019. Community-based blood centers were stratified into the following categories derived from the number of whole blood or RBC units collected in 2017: fewer than 50,000, 50,000 to 199,999, 200,000 to 399,000, and 400,000 or more units. Hospital-based blood centers were stratified into the following categories derived from the number of inpatient surgical operations performed in 2017: fewer than 1000, 1000–7999, and 8000 or more inpatient surgical operations. Transfusing hospitals were stratified into the following categories derived from the number of inpatient surgical operations performed in 2017: 100–999, 1000–1399, 1400–2399, 2400–4999, 5000–7999, and 8000 or more surgical operations. Transfusing hospitals are categorized by the number of inpatient surgeries because this is the most accurate proxy variable to predict hospital RBC utilization rates.¹³

To account for non-responses within each stratum, responses were weighted by dividing the total number of eligible participants by the total number of respondents. Blood collection centers with an expected collection volume of more than 400,000 units were designated a weight of 1.0. The Taylor Series method was used to calculate confidence intervals (CIs) for national collection and transfusion estimates.¹⁴

To account for missing data, a multiple imputation method was applied. Imputed variables were all continuous and non-normally distributed. A two-step imputation process was used to accommodate distributions skewed toward zero using established imputation factors from previous surveys.^11,15,16 The following variables were weighted and imputed: whole blood and apheresis RBCs collected, distributed, rejected, outdated, and transfused; and apheresis platelets (PLTs), plasma, and cryoprecipitate units collected and transfused. To express whole blood-derived PLTs as apheresis equivalents, the numbers of whole blood-derived PLTs were divided by the median reported pool size, which was five.

Non-weighted data were used to calculate and report on mean and median cost per unit of blood and blood components paid by transfusing hospitals. Due to occurrence of outliers, medians were preferred over means for comparing differences in unit costs.

National rates of whole blood and RBC collection per 1000 population were calculated by dividing the total estimated number of units collected before the removal of rejected units by the 2019 US population aged 16–64 years. This denominator was used to maintain consistency with previous NBCUS reports and to correspond with the general age range of blood donors (Figure 2C). National rates of whole blood and RBC transfusion per 1000 population were calculated by dividing the total estimated number of units transfused by the entire 2019 US population. All population estimates were determined using state-specific and age-specific estimates from the US Census Bureau for 2019.¹⁷ A subset of transfusing hospitals completing both 2017 and 2019 surveys was created to determine if differences in collection and utilization estimates between 2017 and 2019 surveys were influenced by sampling and response rates. This matched subset of NBCUS respondents from 2017 to 2019 was used for sensitivity analyses to determine if differences observed between these 2 years were uniform or varying when holding the responding facilities constant. All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).

Since 2008, steep declines in blood utilization have been reported in the United States.⁴ During the same time period, US blood collection organizations have reported fewer donations and collections of blood products.⁴ The findings of the 2019 NBCUS suggest that transfusions have not decreased on an annual basis in the United States for the first time since 2008. Declines in blood transfusion have been largely attributable to adoption of patient blood management programs, decreased surgical blood loss, and other scientific and medical advancements in clinical care.^1,4,6 The trends in RBC transfusion differed by hospital surgical volume and by the location within the healthcare facility the transfusion occurred. Similar to 2017, hospitals with lower surgical volumes reported greater decreases in red blood cell transfusion than hospitals with larger surgical volumes.⁴ However, a greater median percent difference between 2017 and 2019 compared to difference between 2015 and 2017 was noted across all surgical volume categories. This suggests that larger hospitals are, on average, no longer experiencing a decrease in RBC utilization, and the decline in utilization among smaller hospitals continues to slow. Although further study is required, one reason may be that larger hospitals have fully implemented patient blood management programs and reached a nadir in blood use, while smaller hospitals continue to develop and implement these initiatives. An additional notable finding in the present study is that RBC utilization in high acuity settings such as critical care units and emergency departments increased between 2017 and 2019. While reasons for this increase are unclear, one explanation may be that many patient blood management program strategies (limiting phlebotomy for testing, treating preoperative anemia, implementing red cell recovery techniques and other surgery-associated strategies) are likely to have greater impact in other clinical settings, but not in the delivery of critical and emergency care.⁷ However, RBC utilization also increased in outpatient and nonacute settings, demonstrating a variety of clinical settings have reported an increase in RBC use.

Many other countries have seen changing trends in RBC utilization. Among countries reporting to the European Blood Alliance, between 2016 and 2017 most reported decrease in RBC units issued (2.2% decrease in the total among all reporting countries), but between 2017 and 2018, 7 of 18 countries reported an increase and only 11 reported a decrease (1.1% decrease in total).¹⁸ Other regions and countries that have reported slowing of declines or increases in red blood cell demand include New Zealand, Australia, and Quebec.^19–21 Many countries continue to have lower rates per population of red blood cell utilization compared to the United States (33.1 red blood cell units transfused per 1000 population). In 2019, reported rates of red blood cell units issued per 1000 population include 20.9 in New Zealand, 24.2 in Quebec, 24.7 in Australia, and 24.6 in England.^19,20,22 The cause for these differences is unclear but suggests that the United States could further decrease its rate of RBC utilization.

Between 2017 and 2019, RBC collections continued to decrease while transfusions increased slightly, indicating a narrowing between blood supply and demand. This suggests the potential for reduced elasticity in blood supply, which may impact blood sufficiency if blood demand increases or blood supply decreases during public health emergencies.²³ The potential for reduced elasticity has resulted in both governmental and non-governmental stakeholders to express concerns about the adequacy of the blood supply.^6,10,23 In June 2019, the U.S. Congress passed the Pandemic and All-Hazards Preparedness and Advancing Innovation Act, which directs the U.S. Department of Health and Human Services to perform periodic assessments about the safety and adequacy of the blood supply during disasters and public health emergencies.^6,23,24 In response, the Advisory Committee on Blood and Tissue Safety and Availability developed recommendations, including (1) funding a national blood campaign and supporting studies on attracting and maintaining blood donors, (2) developing a system to collect real-time national blood collection and utilization data to determine if blood supply is meeting demand, and (3) creating a blue ribbon panel to address funding and reimbursement issues that might pose challenges in maintaining an adequate blood supply.²³ CDC and the Office of the Assistant Secretary of Health (OASH) will continue to monitor the impact of the current COVID-19 pandemic on blood demand and the adequacy of the blood supply using both existing hemovigilance data (as part of the National Healthcare Safety Network Hemovigilance Module) and as part of the next NBCUS survey.

Since 2015, the number of collected and transfused apheresis PLT units has increased. An aging US population may lead to an increase in the need for PLT transfusions because of an increase in the incidence of cancer and other chronic diseases.^1,25 Between 2017 and 2019, PLT transfusions increased in inpatient, critical care, and outpatient and non-acute inpatients settings, consistent with an increase in demand for PLT transfusions associated with clinical conditions in an aging patient population. The increasing demand for PLTs may pose future challenges to maintaining an adequate PLT supply.^25,26 The average age of apheresis PLT donors is increasing,²⁵ with blood centers reporting difficulties with recruiting and retaining younger apheresis PLT donors. Most PLTs in the United States are collected through apheresis procedures, which generally require more time than whole blood donations and can discourage younger donors from donating.²⁵ PLTs also have a shorter shelf life than red blood cells and plasma.²⁶ Potential strategies to increase the PLT supply could include improving recruitment and retention of PLT donors, adopting technology to increase the shelf-life of PLTs (e.g., cold storage), and utilizing whole blood-derived PLT units.^25,26

The findings are subject to several limitations. First, 2019 NBCUS data were collected during the period when hospitals were experiencing significant disruptions due to the COVID-19 pandemic. Data are self-reported and minimal verification of hospital data was performed to avoid overburdening facilities. Although the deadline for completing the 2019 NBCUS was extended, the hospital response rate was lower than the 2017 NBCUS hospital response rate. Additionally, there is potential for novel pandemic-related data quality issues that cannot be easily quantified and would not be encapsulated by the confidence intervals presented. Second, imputation and weighting were used to generate national estimates. Changes in sampling and response rates could affect comparisons to previous NBCUS estimates. Finally, similar to previous iterations of the NBCUS, certain hospital types were excluded (e.g., smaller hospitals, military hospitals, and outpatient facilities), potentially resulting in underestimates.

In conclusion, the 2019 NBCUS suggests that the decline in utilization rates of RBC in the United States might have ended. Between 2017 and 2019, RBC collections declined while RBC transfusions did not significantly change, suggesting a narrowing between blood supply and demand. However, apheresis PLT collection and utilization have increased since 2015, and additional strategies may be needed to ensure a sufficient PLT supply. CDC and OASH will continue to monitor national blood collection and utilization data to support efforts to ensure the safety and availability of the blood supply.