Anaesthesiology Intensive Therapy, 2011,XLIII,4; 183-186

Effect of red blood cell transfusions on the frequency of infections in the ITU

*Jarosław Mamak1, Lech Krawczyk2

1Department of Anaesthesiology and Intensive Therapy, Regional Specialist Hospital no. 2 in Jastrzębie-Zdrój

2Department of Emergency Medicine, Silesian Medical University in Katowice

  • Fig.1. Percentages of patients with infections in group N and T
  • Fig. 2. Percentages of patients who received transfusion within 48 h before infection (subgroup t) or without transfusion (subgroup n)
  • Table 1. Indications for ITU treatment
  • Table 2. Basic data regarding ITU hospitalisation
  • Table 3. Type and number of infections in patients without transfusions (group N) and with transfusions (group T)

Background. It has been suggested that red blood cell transfusions may affect the outcome and frequency of infection in patients treated in ITU settings. We have therefore retrospectively analysed the records of one hundred and sixty three patients treated in a large tertiary intensive care unit, in order to find any relationship between red blod cell transfusions and the frequency of nosocomial infections.

Methods. The patients were allocated to two groups: T – those who were transfused and N – those who were not. Among the transfused patients, two further subgroups were selected: those who were transfused within 48 h before any signs of infection were observed (n), and those in whom the transfusion was more than 48 h before infection was noted (t).

Results. There was a statistically significant difference between groups T and N (50.9% vs 38.1% infected patients), but not between subgroups t and n.

Conclusions. Nosocomial infections were more frequently observed in patients who required red blood cell transfusion. Red blood cell transfusion cannot be regarded a sole factor affecting nosocomial infection rate.

Transfusions of blood or its fractions can lead to many side effects of various aetiologies, dynamics and risks. A substantial group of complications is associated with biological properties of blood – a transfusion is the introduction of a foreign tissue, characterised by the definite, specific for a donor system of antigens to the recipient’s body. The immunoregulatory effect of transfusion may be associated with the risk of compromised resistance of patients to infections, including nosocomial infections. The clinical research data of surgical patients are inconclusive – some document the transfusion-related increases in the number of infections [1, 2, 3] while the others do not confirm this [4, 5, 6]. Only a few studies address the issues of adverse effects of transfusions in ITU patients.

The aim of the present study was to compare the incidence of nosocomial bacterial infections in patients treated in intensive therapy units receiving red blood cell transfusions versus those who did not receive them and to assess the effects of transfusions on the number of infections.


The medical records of consecutive patients treated in the 10-bed ITU of the 780-bed specialist hospital were retrospectively assessed. The analysis involved the course of treatment over at least 48 h in patients admitted to ITU directly from the emergency department or other hospital wards during the first hospitalisation day.  The infections which developed  48 h after admission were evaluated. If the patient developed more than one infection, the first one was analysed.

The causes of anaemia or reasons for transfusions were not distinguished. To determine the time relation between transfusion and post-transfusion infection, the criterion of the development of infection symptoms 48 h after red blood cell (RBC) transfusion was assumed  to exclude the patients in whom the clinically ‘silent’ infection had already been present during transfusion. Such a grace period was accepted based on the commonly used criteria to differentiate hospital-acquired and extra-hospital infections. Thanks to that, the risk of false positive diagnoses of infections instead of post-transfusion reactions was also reduced.

Patients treated for more than 24 h in the hospital setting, hospitalised in ITU for less than 48 h and those admitted with infections or with infections diagnosed within first 48 h of ITU stay, were excluded. To eliminate the effects of hospital treatment before ITU admission, additional exclusion criteria were applied: previous RBC transfusions or increased, difficult to estimate risk of hospital infections present on admission to ITU.

To determine the number of infections, the study population was divided into two groups: N – patients without transfusions red blood cells and T – those with transfusions.

The relationship between transfusions and incidence of infections was analysed based on two sub-groups: n – patients without transfusions before the development of infections plus those transfused RBC within 48 h preceding infections (or after its recognition); t – patients in whom the transfusion was more than 48 before infection.

In groups N and T as well as subgroups n and t, the percentage of patients with infections was calculated. In order to compare the groups according to the severity of general condition and anticipated risk of death, the SAPS II scale was used.

Statistical analysis was performed using the χ² Pearson’s test for non-measurable variables and t test for measurable variables; in the latter, arithmetic means of baseline values or logarithmic values between the subgroups were compared. The significance was set at p <0.05.


The total number of patients treated in ITU was 396, including 133 treated conservatively and 30 undergoing surgery (Table 1). After exclusions, medical records of 163 patients were analysed, 99 men (60.7%) and 64 women (32.5%); 53 patients had transfusions, i.e. 32.5% of the study population. The data concerning ITU hospitalisation were compiled in Table 2.

Infections with the onset 48 h after admission were observed in 60 patients; the most common infection was pneumonia (Table 3). In group T (with transfusions), the percentage of infected patients was significantly higher than in group N (without transfusions) (Fig. 1).

In 11 cases, red blood cell transfusions were performed after the development of infection or within the 48-hour period preceding the occurrence of symptoms – those patients were included in subgroup n (compared to group N, its size was larger – totally 121 patients). There were no differences found in the percentages of infections in subgroups n and t (Fig. 2).


Transfusions of blood and blood preparations have been controversial for many years. The key question whether haematocrit or haemoglobin concentration should be considered has not been answered. Adverse side effects and risk of various complications, including infections, were analysed in many studies concerning mostly surgical patients. However, only a few studies dealt with patients treated in the ITU setting. Heterogeneous groups of patients and multiple factors affecting the course of treatment and outcomes are the main reasons. In the study group analysed, a high percentage of patients was ventilated – 94.5% whereas in other studies this percentage was 61% with the mean period of ventilation being 7.4 days [7].

The study material showed that patients receiving RBC were twofold longer ventilated, on average. Longer hospitalisations of those patients were also observed in several multi-centre and single-centre research studies [7, 8, 9, 10]. Moreover, it was demonstrated that ¾ of blood units were transfused to patients undergoing mechanical ventilation [11]. 

The SOAP study revealed that more transfusions are given to patients hospitalised longer in ITUs[12]. In our study, analysis of the treatment throughout the ITU hospitalisation showed a significantly higher incidence of infections in patients receiving transfusions (50.9% vs 30%).

There was no difference in SAPS II scores between the group receiving transfusions and otherwise. The higher incidence of infections found in patients requiring RBC did not result from their severer general conditions, which could have been reflected in their lower resistance to infections. In the study population, a vast majority of infections were pneumonias, which should be associated with a high percentage of patients mechanically ventilated [13, 14, 15].

The retrospective analysis demonstrated a significantly higher incidence of infections in patients receiving transfusions (15.38% vs 2.92%), even when the severity of the ITU patients` conditions was accounted for. However, the methods applied in this study did not account for the cause-effect relation between a transfusion and an infection, or patients admitted to ITU with active infections or more than one infection in one patient. Moreover, the hospitalisation period shorter than that accepted for the hospital infection to develop was not considered (discharge or death  in less than 48 h after ITU admission) [8].

The findings concerning the correlation between RBC cell transfusions and nosocomial infections were different when the assumption was made to take into consideration previous blood transfusions. Our results suggest that in the population treated, red blood cell transfusions had no impact on increased incidence of nosocomial infections, which is conflicting with the data implying that a transfusion is an independent factor of development of ventilator-associated pneumonia [15]. However, the study mentioned did not include data whether there were differences in APACHE II and SOFA scores between patients receiving transfusions and those without them. Furthermore, according to another study, the number of bacteraemia cases was higher in patients transfused RBC. Higher incidences of bacteraemia were observed in patients with higher APACHE II and SOFA scores, the differences in scores between patients with and without transfusions, however, were not given [16].

The study evaluating all kinds of nosocomial bacterial infections in total revealed that in patients with higher chances of survival, the transfusion of RBC increased the risk of infections to a higher degree than in patients with poor prognosis. The findings of the study discussed prove the correlation between transfusions and incidences of hospital infections, yet the cause of this correlation is unknown [9]. Compared with our study, the duration of ITU treatment was much shorter in the study mentioned; moreover, it did not include data concerning the percentage of patients mechanically ventilated and the duration of ventilation.

Furthermore, the comparative assessment is hindered by the fact that patients receiving transfusions had higher APACHE II scores and their prognosis on admission was worse. In our study, there was no significant difference in SAPS II scores between patients receiving transfusions and those without them. Different scales accepted for evaluation of patients` conditions do not allow proper comparison of results.

The majority of studies assessing the effects of transfusions of blood and its preparations on the incidence of infections do not consider the severity of conditions of patients receiving transfusions [17].

Trauma and shock as well as their management may impair the patients` immunity, thus increase their susceptibility to infections. Therefore, physicians may tend to use transfusions more often in patients, whose incidence of infections would be higher even without transfusions [18, 19].

The representativeness of our findings is limited by retrospective analysis of medical records and a relatively small sample of patients, which, mostly resulted from the fact that the study was carried out in one medical centre. Moreover, the results may be influenced by the characteristics of the local population of patients and of hospital wards, sanitary and epidemiological conditions or specific therapeutic and diagnostic methods used. According to the accepted assumptions, analysis involved medical records of the mixed population of patients; specific factors predisposing to infections (diabetes, chronic obstructive pulmonary disease) were not considered.

The number of studies dealing with the issue in question in intensive therapy units is low. There are no prospective randomised clinical trials; only prospective-observational and retrospective studies are available, some of which are the secondary analyses of databases.

At present there are no conclusive data demonstrating adverse effects of transfusions on the incidence of infections in hospitalised patients [20]. To verify our findings, prospective randomised clinical studies should be carried out.


1. In the multi-profile ITU, the incidence of nosocomial infections is higher in patients requiring red blood cell transfusions.

2. In ITU patients, red blood cell transfusions do not increase the number of hospital infections. 



1.    Houbiers JG, van de Velde CJ, van de Watering LM, Hermans J, Schreuder S, Bijnen AB, Pahlplatz P, Schattenkerk ME, Wobbes T, de Vries JE, Klementschitsch P, van de Maas AH, Brand A: Transfusion of red cells is associated with increased incidence of bacterial infection after colorectal surgery: a prospective study. Transfusion 1997; 37: 126-134.

2.    Innerhofer P, Klinger A, Klimmer C, Fries D, Nussbaumer W: Risk for postoperative infection of white blood cell – filtered allogenic or autologous blood components in orthopedic patients undergoing primary arthroplasty. Transfusion 2005; 45: 103-110.

3.    Rovera F, Dionigi G, Boni L, Imperatori T, Tabachi A, Cacano G, Durni M, Dionigi R: Postoperative infection after oesophageal resection and blood transfusion. World J Surg Oncol 2006; 4: 80.

4.    Hill GE, Frawley WH, Griffith KE, Forestner JE, Minei JP: Allogenic blood transfusion increases the risk of postoperative bacterial infection: a meta – analysis. J Trauma 2003; 54: 908-914.

5.    Ali ZA, Lim E, Motalleb-Zadeh R, Callaghan CJ, Gerrard C, Vuylsteke A, Foweraker J, Tsui S: Allogenic blood transfusion does not predispose to infection after cardiac surgery. Am Thorac Surg 2004; 78: 1542-1546.

6.    Vamvakas EC: Meta-analysis of randomized controlled trials comparing the risk of postoperative infection between recipients of allogenic and autologous blood transfusion. Vox Sang 2002; 83: 339-346.

7.    Corwin HL, Gettinger A, Pearl RG, Fink MP, Levy MM, Abraham E, MacIntyre N, Pearl RG, Shabot MM: The CRIT study: anemia and blood transfusion in the critically ill – current clinical practice in the United States. Crit Care Med 2004, 32: 39-52.

8.    Taylor RW, Manganaro L, O’Brien J, Trottier SJ, Parkar N, Veremakis C: Impact of allogenic packed red blood cell transfusion on nosocomial infection rates in the critically ill patient. Crit Care Med 2002; 30: 2249-2254.

9.    Taylor RW, O’Brien J, Trottier SJ, Manganaro L, Cytron M, Lesko M.F,Arnzen K, Cappadaro C, Fu M, Plisco MS, Sadaka FD, Veremakis C: Red blood transfusions and nosocomial infections in critically ill patients. Crit Care Med 2006; 34: 2302-2308.

10.    Vincent JL, Baron J-F, Reinhart K, Gattinoni L, Thijs L, Webb A, Meier-Hellmann A, Nollet G, Peres-Bota D: Anaemia and blood transfusion in critically ill patients. JAMA 2002; 288: 1499-1507.

11.    Levy MM, Abraham E, Zilberger M, MacIntyre NR: A descriptive evaluation of transfusion practices in patients receiving mechanical ventilation. Chest 2005; 127: 928-935.

12.    Vincent JL, Sakr Y, Spring C, Harboe S, Damas P: Sepsis occurrence in actually ill patients. Anesthesiology 2008; 108: 31-39.

13.    Niederman MS, Craven DE, Bonten MJ, Chastre J, Craig WA, Fagon J-Y, Hall J, Jacoby GA, Koleff MH, Luna CM, Mandell LA, Torres A, Wunderink RG: Guidelines for the management of adults with hospital-acquired, ventilator-associated pneumonia. Am J Respir Crit Care Med 2005; 171: 388-416.

14.    Steinberg SM, Nicholas RL: Infections in the surgical critical care unit; in: Textbook of critical care; (Eds: Shoemaker WC, Ayres SM, Grenvik A, Holbrook PR);  W.B.Saunders Company, Philadelphia 2000.

15.    Shorr AF, Duh M-S, Kelly KM, Kollef MH: Red blood cell transfusion and ventilator- associated pneumonia: a potential link? Crit Care Med 2004; 32: 666-674.

16.    Shorr AF, Jackson WL, Kelly KM, Fu M, Kollef MH: Transfusion practice and blood stream infections in critically ill patients. Chest 2005; 127: 1722-1728.

17.    Vamvakas EC, Blajchman MA: Deleterious clinical effects of transfusion-associated immunomodulation: fact or fiction? Blood 2001; 7: 180-195.

18.    Hebert PC, Wells G, Martin C, Tweeddale M, Marshall J, Blajchman M, Pagliarello G, Sandham D, Schweitzer I, Boisvert D, Calder L: Variation in red cell transfusion practice in the intensive care unit: a multicentre cohort study. Crit Care 1999; 3: 57-63.

19.    Hebert PC, Wells G, Martin C, Tweeddale M, Marshall J, Blajchman M, Pagliarello G, Schweitzer I, Calder L: A Canadian survey of transfusion practices in critically ill patients. Crit Care Med 1998; 26: 482-487.

20.    Vamvakas EC: Pneumonia as a complication of blood product transfusion in the critically ill: transfusion-related immunomodulation (TRIM). Crit Care Med 2006; 34: 151-159.



*Jarosław Mamak

Oddział Anestezjologii i Intensywnej Terapii
Wojewódzki Szpital Specjalistyczny Nr 2
Al. Jana Pawła II 7, 44-335 Jastrzębie-Zdrój
tel: +48 32 478 43 40

received: 15.04.2011 r.
accepted: 20.06.2011 r.