How to Use Blood in Critically Ill Patients

Andrew F. Shorr, MD, MPH


December 26, 2014

This feature requires the newest version of Flash. You can download it here.

This is Andy Shorr, from Washington, DC, with a critical care literature update.

I would like to point out the TRISS trial (the Transfusion Requirements in Septic Shock trial), which was recently published by Holst and colleagues[1] in the New England Journal of Medicine.

Nearly a decade and a half ago we had the TRICC (Transfusion Requirements in Critical Care) study, which began this long, evidence-based approach to understanding how to use blood in critically ill patients. The TRICC trial looked at two different triggers for transfusion—a hemoglobin level of 7 g/dL vs 10 g/dL—and saw overall no difference in mortality in these essentially stable, critically ill patients.

That study was criticized for several reasons; nonetheless, it altered our approach to blood utilization in the intensive care unit (ICU), and since that time, other studies have suggested that in nearly every population that has been evaluated, a lower transfusion threshold is associated with similar, if not improved, outcomes relative to a higher transfusion threshold. So, in essence, when we are giving blood, in many cases to treat a number, we are either wasting a resource, treating ourselves, or potentially harming the patient.

The TRISS trial focused specifically on patients in septic shock who were critically ill and unstable, because they are people about whom we have had caution, and pause on the basis of theories about oxygen delivery and the use of packed cells to achieve that. In this trial, which was conducted predominantly in Scandinavia, the investigators randomly assigned 1000 patients with septic shock to a transfusion threshold trigger of a hemoglobin level of 7 g/dL or a goal threshold of 9 g/dL. The study enrolled a group of patients all within about a day of their arrival in the ICU. These weren't only medical patients; they included surgical patients. They had the standard background comorbid conditions that you would expect, so they didn't exclude patients with underlying cardiovascular disease, and about two thirds of the patients were on mechanical ventilation at the time of their enrollment. The median Sequential Organ Failure Assessment (SOFA) score was 10, so that is a very high organ failure burden. The median Simplified Acute Physiology (SAP) score was in the 50s in both arms, suggesting a very critically ill population and, in fact, very different from some of the fluid resuscitation studies we have seen recently. For example, in the ARISE trial, the Acute Physiology and Chronic Health Evaluation II (APACHE-II) score was in the 15 range.

In this study, the protocol was followed and they saw that patients who were in the 7-g/dL hemoglobin group during their hospitalization received, on average, 1 unit of blood while they were in the ICU vs about 4 units if they were in the higher threshold group. When they looked at the primary endpoint, which was 90-day mortality, there was no difference between the two groups. When they looked earlier, at 14- or 28-day mortality, there were no differences between the two groups. When they looked at the secondary endpoints, which included adverse events, need for continued life support, or the duration of vasoactive support, again there were no differences. The outcomes were similar whether the patients received more units of blood or fewer units of blood.

When they stratified and looked at selected subgroups that were prespecified, there was no suggestion of a benefit one way or the other. Although, in the more severely ill population, there seemed to be some interaction between transfusion, mortality, and severity of illness, suggesting that in the higher-severity-of-illness population, there might actually have been a benefit. But this study wasn't focusing only on super-high-risk patients; this study was focusing overall on what was going on in the population in general.

This study confirms for us that a lower threshold for transfusion is as acceptable as a higher one and, in fact, saves a scarce resource. This gives us all more pause to go back and re-evaluate our practice, even in this cohort of patients in whom our approach to transfusion might not have changed as much as it had for other patients in the ICU.

This is another evidence-based example of where, in the ICU, less is actually more. I encourage you to read the TRISS trial recently published in the New England Journal of Medicine. This is Andy Shorr, from Washington, DC.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: