Transfusion policy in critically ill patients

Abstract

Anemia is a common and early condition in critically ill patients. 29% patients have Hb < 10g% at ICU admission (1). 95% patients have anemia in the 3rd ICU day (2). We studied the incidence of anemia in 132 critically ill patients with long ICU stay (>7 days) and found it presence in 110 (83.33%) patients at ICU admission and in 126 (95.45%) patients at ICU discharge (3). Causes of acquired anemia in intensive care patients are blood loss due to surgery, trauma, gastrointestinal bleeding or medical procedures (mainly phlebotomy), hemodilution, decreased red blood cell production or increased distruction and others (4). Phlebothomy may result in 30-70ml blood withdrawal/day (1,2). In critically ill patients an increase of 3,5ml/day withdrawn blood doubles the risk of transfusion (5). The amount of blood withdrawn by phlebotomy decreases over time during ICU stay, mainly due to decreased number of ABG (5). The critical illness is frequently associated with inflammatory status. The proinflammatory cytokines (TNFα, IL-1, IL-6) result in failure of circulating EPO to appropriately increase in response to Hb reduction (relative EPO deficiency), inhibition of erythrocyte production (relative EPO resistance) and decreased iron availability (6,7,8). Corwin terms this status as anemia of critical illness. Iron metabolism is severely altered in during critical illness. The main features are decreased iron availability for erythropoesis and increased iron storage (9). Markers of iron metabolism alterations during inflammation are decreased serum Fe, decreased serum transferrin, decreased transferrin saturation and increased serum ferritin. This status is termed functional iron deficiency (FID) (10). Critically ill patients with FID have a longer ICU stay compared with those without FID (11). Taking into account that severity of iron metabolism disturbances parallel the inflammatory status (12), can ferritin be used as a prognostic marker in critical illness? In ICU patients with long ICU stay higher the SOFA score, higher the maximal ferritin level and nonsurvivors have higher ferritin level compared with survivors (13). Mortality of ICU patients significantly corelates (p<0.01) with an increase of serum ferritin >300ng/ml during ICU stay (13). During ICU stay Hb has a steady tendency to decrease over time. Despite the fact that Hb values on ICU admission may vary widely, after 14 days of ICU stay the Hb values tend to converge (14). Blood transfusion is a common occurrence in ICU patients. Three landmark studies are dedicated to transfusion policy in critically ill patients: TRICC study (Canada, 838 patients, 25 ICUs, over 3 years) (15), ABC study (Europe, 3534 patients, 146 ICUs) (1) and CRIT study (SUA, 4892 patients, 284 ICUs) (2). The incidence of blood transfusion in critically ill patients is as high as about 40-65% (1,2,5). In our ICU the incidence is about 60% in patients with long ICU stay (14). In our study over 50% of transfusion are given in the first ICU day, the incidence decreases along ICU stay and a transfusional event increases Hb level with an average of 0,89g% (3). The reported indications for blood transfusion vary widely: active bleeding, low Hb level, postoperative status, need to increase oxygen delivery and others. Reported Hb transfusion trigger vary between 7-8,6g% (1,2,5,15). In our study Hb transfusion trigger was 7.83 ± 2.30g% (3). However the actual transfusion policy relies not on a specific Hb figure, but on the need to increase tissue oxygen delivery. Transfusion policy in critically ill patients changed over time. Former recommendations encouraged blood transfusion in ICU patients with Hb <10g%. Due to several risks and side effects, mostly due to immunodisreggulation, blood transfusion was placed under scrutiny. Starting with the reported results of TRICC study, which documented that transfused ICU patients had a higher mortality compared with non-transfused patients, despite the fact that mean Hb level was much lower in this later group (15), restrictive transfusion policy was promoted and several studies were dedicated to investigate the relation between different blood transfusion policies and mortality. In our study mortality was significantly higher in ever transfused (65%) compared with never transfused (20%) critically ill patients with long ICU stay (3). Multivariate analysis identified SOFA scores and transfusion status as independent risk factors for mortality (3). Despite the fact that a lot of papers demonstrated the increased mortality rate in transfused critically ill patients, a recent paper published by JL Vincent from the European SOAP study showed no correlation between mortality and transfusion status (16). In conclusion, when is blood transfusion indicated in ICU? The indication according to a specific Hb transfusion trigger is no longer accepted. The decision to transfuse should be made according a case-to-case judgment, which should take in account the specific features of each patient (actual status of tissue oxygen delivery and its match with oxygen consumption, coexisting diseases, risk of ischemic events, transfusion risk/benefit ratio, s.o.). General recommendations for clinical practice: if Hb is <6g%, blood transfusion is probably indicated, but not mandatory; if Hb is >10g%, transfusion is usually contraindicated; when Hb is 6-10g%, perform an evaluation of tissue oxygenation, take into account other means to increase oxygen delivery and indicate blood transfusion only after weighting risk/benefit ratio.