|Home | About | Journals | Submit | Contact Us | Français|
As epidemiological information is useful in planning the provision and assessing the efficiency of product use, we reviewed Spanish data on population, blood donation and blood component transfusion from 1997 to 2007, and the possible effect of universal leucoreduction.
Data on the Spanish population were obtained from the National Institute of Statistics, whereas data on blood donation and blood component transfusion were acquired from the Spanish Ministry of Health.
During the study period, the Spanish population increased by 5.6 million persons (14.4%), and blood donation by 28.1%, although the amount of red blood cells (RBC) obtained increased by only 21.5% whereas RBC transfusions increased by 28.3%. The RBC transfusion rate was significantly higher after the implementation of universal leucoreduction (2002 – 2006) than during the pre-leucoreduction period (1997 – 2001) (differ ence = 2.54 units/1,000 population/year; 95%CI 1.81 – 3.27; P<0.001). We also observed statistical ly, but not clinically, significant differences for platelet and plasma transfusions.
The increase observed in the RBC transfusion index after implementation of universal leucoreduction may have been due to a reduction of the haemoglobin content in the RBC units. Our data on blood use do, therefore, seem to add to the case against universal leucoreduction, which has led to an incremental cost for unknown, but probably slight, benefits for patients.
Allogeneic blood is an increasingly scarce and expensive resource. Epidemiological information on recipients and transfusions is useful for planning the provision and assessing the efficiency of use of this product. In this regard, in a recently published survey of the demographics of blood use, the authors gathered epidemiological information from different countries argued to be representative of the developed world1. However, according to another recent report on differences in blood collection rates in 18 countries, blood supply in Europe is uneven: some countries have plenty of blood donations (e.g., Denmark, 70 donations/1,000 population/year) while others are still struggling (e.g., Portugal 33 donations/1,000 population/year)2. Thus, the two European countries included in the above-mentioned survey (England and Denmark) may be representative of northern Europe, but are probably not representative of southern Europe. We present here official Spanish data on population, blood donation and blood component transfusion from 1997 to 2007, and the possible influence of the implementation of universal leucoreduction on blood component use.
Data on the Spanish population were obtained from the demographics and population section of the National Institute of Statistics (www.ine.es), whereas data on blood donations and blood component transfusions were acquired from the Spanish Ministry of Health (Dirección General de Salud Pública, Plan Nacional de Hemoterapia; www.msc.es). In order to estimate the red blood cell (RBC) use in the population ≥65 years old it was assumed that this population received 60% of all transfused RBC1,3,4. One plasma unit was considered to be equivalent to 250 mL of fresh frozen plasma (FFP)6, and one platelet dose was a pool of five whole-blood-derived platelet units1. Blood donation, RBC concentrates obtained, RBC transfusions, FFP transfusions and platelet transfusions were calculated as the number of units or doses per 1,000 population/year. Data are expressed as incidence (n), percentage (%) or mean ± standard deviation (SD) (n).
Parametric two-way ANOVA tests were used for the comparison of transfusion indices before and after the introduction of universal leucoreduction, which divided the whole study period into two equal 5-year periods (1997–2001 and 2002–2006). All statistical analyses were performed with SPSS 14.0 (Licensed to the University of Málaga, Spain) and a P value <0.05 was considered statistically significant.
The Spanish population increased by more than 5.6 million (14.4%) during the study period (from 39.5 million inhabitants in 1997 to 45.2 million in 2007 (Table I). During the same period, total blood donations increased by 368,678 units (28.07%), from 1.31 million in 1997 to 1.68 million in 2007. The total donation rate (or index) officially increased from 35.55 in 1997 to 37.6 in 2007, although this latter figure is the lowest in the last 6 years and far below the 39.22 recorded for 2003 (Figure 1).
Total RBC donations increased by 281,958 (21.47%) in the study period (from 1.31 million in 1997 to almost 1.6 million in 2007). Despite this increase, the adjusted donation rate in Spain (35.2 RBC units/1,000 inhabitants) is one of the lowest donation rates in Europe (Figure 1). The collection of plasma and platelets increased by 94,102 L (32.78%) and 188,053 units (21.38%), respectively (Table 2). In contrast, preoperative autologous blood donation decreased steadily by 13,706 units (− 48.1%), from 28,487 in 1998 to 14,781 in 2007; furthermore, when preoperative autologous blood donations were expressed as a percentage of total donations, the decrease was even greater (− 53.7%) and the percentage value fell from 2.01% in 1997 to 0.93% in 2007 (Table I) (Figure 1).
Between 1997 and 2007, the number of RBC transfusions increased by 332,965 units (28.3%), from 1.18 million to 1.51 million. When expressed as a percentage of red cell units obtained, RBC transfusions increased from 87.8% in 1998 to 94.7 in 2007, although the highest figure was 96.2% in 2004. During the same period, the consumption or RBC transfusion rate rose by 4.1 RBC units/1,000 inhabitants (14%), from 29.3 in 1998 to 33.4 in 2007. The average RBC transfusion rate was 31.2 ± 1.4 units/1,000 population/year (Table I) (Figure 1).
Interestingly, however, when we analysed two 5-year periods before and after the introduction of universal leucocyte reduction, we found that the RBC transfusion rate was significantly higher after the implementation of leucoreduction (2002 – 2006) than in the pre-leucoreduction period (1997 – 2001) (difference = 2.54 units; 95%CI 1.81 – 3.27; P<0.001) (Table III).
The use of FFP transfusions increased by 9,214 L (17.7%) and that of platelet transfusions by 150,545 units (28.2%). FFP transfusion rates, however, remained substantially unchanged, being 5.3 units/1,000 persons/year in 1997 and 5.4 units/1,000 persons/year in 2007 (difference, 1.9%). Platelet transfusion rates, in contrast, rose slightly (11.1%) from 2.7 doses/1,000 population/year in 1997 to 3.0 in 2007 (Table II). The use of both FFP and platelets transfusion was slightly, but statistically significantly, greater after the implementation of universal leucoreduction compared to during the pre-leucoreduction period (Table III).
During the study period, there was a steady increase in the Spanish population, as well as in blood donation and blood use, but variations in the use of different blood components were uneven. First, in line with the results of a survey on the use of preoperative autologous blood donation in Europe (1997–2000)7,8, this form of donation decreased steadily from 2.01% in 1997 to 0.93% in 2007. However, it must be borne in mind that “Allogeneic blood transfusion can never be risk-free, reflecting as it does the current state of health of society. Just as new illnesses and infections will influence the overall health picture, their impact on blood safety cannot be predicted”9. Thus, in accordance with Brecher and Goodnough10, “should another risk of transfusion-transmitted disease be identified, …, or if blood collections fail to keep pace with the demand for blood, …, we would expect a resurgence in interest in autologous blood”.
Secondly, the rate of RBC use increased progressively from 29.8 units/1,000 population/year in 1997 to 33.4 units/1,000 population/year in 2007 (Table 1), accounting for, respectively, 89.7% and 94.7% of the RBC units, thus with a decrease of 5% in the gap between donation and transfusion. A similar trend was observed in the USA: in 2001, the total USA blood supply was 10.4% greater than in 1999, the rate of transfusion per 1,000 total USA population increased from 45.5 units in 1999 to 50.0 units in 2001, the highest in 15 years of records, and the margin between transfusion demand and the total allogeneic supply was reduced from 9.1% in 1999 to 7.9% in 200111. In Spain this gap is only around 5%. Returning to the USA, out of 1,086 hospitals which responded to a questionnaire, 138 (12.7%) reported having cancelled elective surgical operations on one or more days during 2001 due to blood shortages, and this steady increase in demand continues to challenge the USA blood community11. For several weeks during the winter of 2008–2009, we suffered a similar situation in Spain with an important medical crisis due to insufficient supplies of blood.
In contrast to the situation with RBC transfusions, the use of FFP (mean 5.4 units/1,000 population/years) and platelets (2.8 doses/1,000 population/year) remained more constant during the period studied (Table III). Thus, compared to previously published data1,5, our figures more closely resemble those for the use of blood components in Australia or France than for the use in England or Denmark, highlighting the wide variations in blood collection and use between countries.
One possible reason for the increase in the RBC transfusion index might have been an increase in the elderly population, as patients aged = 65 years old receive around 60% of all RBC transfusions1,3,4. When the Spanish data were reanalysed considering the distribution of RBC transfusions according to the age of population (< 65 or = 65 years old), we found similar increases in RBC transfusions in both the group of patients = 65 years old (+ 9.1 units/1,000 population/year; 8.2% increase) and in the group < 65 years old (+1.8 units/1,000 population/year; 12.7% increase) (Table 1). However, as the increase in the Spanish population had been mostly due to immigration of young people, the proportion of older people remained relatively constant throughout the period studied (mean: 16.8%; min: 16.1%; max: 17.6%) (Table I). It, therefore, seems that the changes observed in the indices of transfused blood components cannot be attributed to demographic changes.
When the time-course of RBC transfusions was analysed, the maximal inter-annual variation was observed between 2001 and 2002 (+2.1 units/1,000 population/year). In our country, universal leucoreduction was introduced in 2002 (67% of units), and by 2006 was applied to 95% of all blood donations. We, therefore, reanalysed our transfusion data considering two periods: before (1997 – 2001) and after (2002 – 2006) the introduction universal leucoreduction. As shown in Table III, leucoreduction led to a significant 8.3% increase in the overall RBC transfusion index (+ 2.5 units/1,000 population/year), as well as in overall platelet index (7.4%) and FFP index (3.6%). Similarly, in the USA, the percentage of leucoreduced units increased from 35% in 1999 to 84% in 200111.
These data are concordant with those reported by Llewelyn et al.12 who performed a ‘before and after‘ study in patients undergoing elective cardiac or orthopaedic surgical procedures who received RBC transfusions. Patients undergoing surgery after universal leucoreduction (n=1,098) received more RBC units than those operated on before leucoreduction strategies were universally adopted (n=997) (3.3 ± 3.0 versus 2.8 ± 1.6 units/patient, respectively; p<0.0001), especially in cardiac surgery (4.0 ± 4.0 versus 3.1 ± 1.8 units/patient, respectively; p<0.0001)12. However, in contrast to the findings of the study by Llewelyn et al. 12, the increases in the use of FFP and platelets after the introduction of leucoreduction, although statistically significant, did not seem to be clinically relevant, suggesting that patients treated in the post-leucoreduction period were not more severely ill than those treated prior to the introduction of leucoreduction. In France, there has been a consistent increase in RBC transfusions since 2004, although this is 5 years after the introduction of universal leucoreduction.
Since there is a trend towards the use of restrictive transfusion protocols in Spain, we suspect that the increase in RBC transfusion index observed since the implementation of universal leucoreduction may have been due to a reduction of the haemoglobin content in the RBC units caused by erythrocyte loss during the leucoreduction procedures13,14. According to internal quality controls carried out in our institution, the haemoglobin content of RBC units decreased from 60 g/unit before leucoreduction to 52 g/unit after leucoreduction. These data are in accordance with those of the quality control program of the French Blood Service which reported a mean haemoglobin content between 53.6 and 54.9 g/unit depending on the year (2001 to 2005), although the requirement of 40 g/unit was reached for about 99% of units15. According to European regulations (Directive 2004/33/EC), the required minimum haemoglobin content for non-leucoreduced RBC concentrates is 45 g/unit. We, therefore, have to assume a significant reduction of at least of 11% in the haemoglobin content of a RBC unit simply because of leucoreduction. If this tendency is confirmed, it would force us to increase the blood volume drawn at each donation to avoid patients being exposed to higher numbers of donors.
With the advent of variant Creutzfeldt–Jacob disease, the United Kingdom introduced universal leucoreduction in 1999 for all patients as a precaution to prevent the transmission of prions.
This costly policy was followed by many other countries, including Spain, in the absence of much evidence of an actual health problem or of a more than presumed effectiveness of leucoreduction in preventing prion transmission. In fact, there are few medical or economic arguments for the introduction of universal leucoreduction rather than selective leucoreduction in European countries other than the United Kingdom. The central problem is legal accountability, underpinned by a society desire for zero risk. This irrational demand is matched by the offer of the industry, driving up the costs of blood, given for free by voluntary blood donors16. It must be stated that our study was observational in nature, comparing blood component use during successive periods of time. Our conclusions may, therefore, be affected by confounding factors, for example, an effect of increases in surgical and medical activity and/or complexity, although we do not have official data of health care activity. However, if our data on blood use were to be confirmed, they would seem to add to the case against universal leucoreduction, which has led to an incremental cost for unknown but probably slight benefits for patients17,18.
We gratefully acknowledge Dr. Arturo Pereira (Department of Haematology, Hospital Clinic, Barcelona, Spain) for his critical review of the manuscript.