Abstract

Abstract 3346

The rate of septicaemia was measured following blood transfusion in patients admitted to hospital greater than 48 hours. The data was collected prospectively and analysed retrospectively. The aim of the analysis was to disprove that red cell transfusion increased the rate of hospital acquired septicaemia. Data was extracted on patients who had a transfusion and an episode of septicaemia between 1999 and 2008 at a university teaching hospital, John Hunter Hospital, Newcastle, Australia (JHH). The database contained information on 20161 transfusion events of 102,600 units of packed red cells. The septicaemia database contained 8375 patients with blood culture proven septicaemia. Blood was issued using a computerised cross matching system. Analysis of the issuing of blood found no bias in the age of blood issued based on age, sex, number of units or by year of issue. All patients having a septicaemic event recorded following transfusion were analysed and compared to septicaemia occurring in all admissions greater than 48hrs in which no transfusion occurred. All patients who were diagnosed with septicaemia in the 6 month prior to transfusion were excluded as were patients who received five or more units of packed red cells to exclude the bias from the sickest patients. All patients whose septicaemia occurred greater than 16 days after transfusion were excluded as were patients whose septicaemia occurred before the second day after transfusion. A total of 258 patients were reviewed to test the hypothesis. Table1. There was a statistically significant increase in septicaemia following red cell transfusion. The null hypothesis was rejected. The data base was examined by the age of red cells transfused and its effect on nosocomial septicaemia. There was a statistically significant effect of older blood on the rate of septicaemia. Packed red cells transfused less than 14 days of age had no effect on the septicaemia rate. Blood that was 14–28 days of age increased the rate of nosocomial septicaemia by 1.65. Red Cells that were between 29 and 35 days of age increased the rate of septicaemia by 2.5 times. Blood that was between 36 and 42 days of age increased the rate by approximately 4.4 times, with the absolute risk of developing septicaemia within 15 days being approximately 4%. Table2. The time course of septicaemia was examined for patients receiving at least one unit of 28 day old blood. It was found that the risk of sepsis lasted less than 15 days p<0.001. The conclusion of our analysis suggests that packed red cells older than 14 days increase the risk of septicaemia in hospitalised patients, this effect continues to rise until 42 days post collection.

Table 1
 Odds Ratio (CI) 
Total cases of nosocomial septicaemia 1684  
    • Admissions greater than 48 hours 175,325  
Prior to exclusions     • All transfusions  
    • No previous sepsis 790 10.2 (10.1–10.3) 
Study Group   
    • Less than 5 units Transfusion.   
    • No previous sepsis.   
    • Septicaemia >2 days <16days 258 2.02 (1.92–2.12)   
 Odds Ratio (CI) 
Total cases of nosocomial septicaemia 1684  
    • Admissions greater than 48 hours 175,325  
Prior to exclusions     • All transfusions  
    • No previous sepsis 790 10.2 (10.1–10.3) 
Study Group   
    • Less than 5 units Transfusion.   
    • No previous sepsis.   
    • Septicaemia >2 days <16days 258 2.02 (1.92–2.12)   
Table 2
Age of oldest red cell pack Less than 15 days of age 15 to 28 days of age 29–35 days of age 36–42 days of age 
Percentage transfused 24% 49% 14% 12% 
Number of transfusions < 5 units 3524 7161 2115 1802 
Septicaemia observed 33 102 52 71 
Odds ratio 0.98 1.54 2.55 4.4 
Confidence Interval. 0.65–1.31 1.34–1.74 2.26–2.84 4.16–4.64 
Age of oldest red cell pack Less than 15 days of age 15 to 28 days of age 29–35 days of age 36–42 days of age 
Percentage transfused 24% 49% 14% 12% 
Number of transfusions < 5 units 3524 7161 2115 1802 
Septicaemia observed 33 102 52 71 
Odds ratio 0.98 1.54 2.55 4.4 
Confidence Interval. 0.65–1.31 1.34–1.74 2.26–2.84 4.16–4.64 

Disclosures:

No relevant conflicts of interest to declare.

This icon denotes an abstract that is clinically relevant.

Author notes

*

Asterisk with author names denotes non-ASH members.