In the large, multicenter, randomized, controlled INFORM (Informing Fresh versus Old Red Cell Management) study, in-hospital mortality did not differ between patients who received transfusions of red blood cells (RBCs) stored for short periods of time and those who received RBCs stored long-term.1 This is the latest study to assess outcomes related to the length of time for which blood is stored, and, according to lead author Nancy M. Heddle, MSc, its findings should end the debate about the impact of RBC storage length.
“The results of the INFORM trial should reassure physicians that the routine use of fresher blood for hospitalized adults that require red [blood] cell transfusion is not justified,” Prof. Heddle, an associate professor of pathology and molecular medicine at McMaster University in Ontario, Canada, told ASH Clinical News.
INFORM included adult patients who were hospitalized at six institutions in four countries (Australia, Canada, Israel, and the United States); patients were randomized 1:2 to receive transfusions of the freshest RBCs in the inventory (short-term storage; mean storage duration = 13±7.6 days) or the oldest available RBCs (long-term storage; mean storage duration = 23.6±8.9 days; p<0.001). Patients were excluded if they:
- were expected to receive a massive transfusion (a request for ≥10 red-cell units at a time)
- required blood that had not been cross-matched
- required autologous or directed transfusion
- had an indication for fresh-only, red-cell transfusion
Some patients with alloimmunity to red-cell antigens were also excluded if it was difficult to find compatible blood. Patients undergoing cardiac surgery at one U.S. site and in the intensive care unit (ICU) at one Canadian site were excluded due to competing studies.
Demographics, diagnoses, duration of hospital stay, and vital status at discharge were collected from hospital electronic medical records, and information on transfused red cells (including blood type and storage duration) were extracted from hospitals’ electronic laboratory information systems. Patients were followed until in-hospital death or discharge.
In-hospital mortality (the study’s primary endpoint) was assessed only in patients with type A or O blood (intention-to-treat cohort), given the study’s requirement for a difference in the average storage duration to be at least 10 days, and this was difficult to achieve in patients with other less common blood types.
Between April 2012 and October 2015, 24,736 patients met the inclusion criteria: 20,858 had type A or O blood and were randomized to receive short- (n=6,936) or long-term (n=13,922) stored RBCs. The median duration of hospital stay was 10 days for both the short- (interquartile range [IQR] = 5-19 days) and long-term storage groups (IQR=5-20 days). See TABLE for transfusion data.
There were 634 deaths (9.1%) in the short-term storage group and 1,213 (8.7%) in the long-term storage group, for an odds ratio of 1.05 (95% CI 0.95-1.16; p=0.34). At 30 days, the cumulative probability of death was 6.9 percent in the short-term cohort versus 6.5 percent in the long-term cohort.
The authors added that, “when the analysis was expanded to include the 24,736 patients with any blood type, the results were similar, with rates of death of 9.1 percent [in the short-term group] and 8.7 percent [in the long-term group] (OR=1.04; 95% CI 0.95-1.14; p=0.38).”
When the authors conducted a subgroup analysis (based on hospital site, country, blood type, and diagnosis), no significant interactions were found. There also were no significant differences in mortality with short- or long-term storage of RBCs in high-risk subgroups:
- undergoing cardiovascular surgery (12.3% vs. 11.2%; p=0.08)
- in the ICU (13.3% vs. 12.8%; p=0.52)
- with cancer (8.4% vs. 8.8%; p=0.82)
The study is limited in its use of electronic databases that did not systematically record detailed information on coexisting illnesses, reasons for transfusions, co-interventions, non-fatal cardiovascular outcomes, and causes of death. In addition, the study outcomes may not be directly applicable to patients with type B or AB blood due to the study’s design and inclusion criteria.
“The study results support the practice [of using] the oldest red cell products in the hospital inventory to reduce the number of blood units being wasted,” Prof. Heddle told ASH Clinical News. “Being able to store red blood cells for a longer period also helps blood suppliers to maintain an adequate supply of blood and ensure availability of product when patients need it.”
In an editorial accompanying the INFORM results, Aaron A.R. Tobian, MD, PhD, and Paul M. Ness, MD, commented on the study’s design and applicability to real-world practice.2
“Even though the results of the INFORM trial should end the debate regarding whether short-term or long-term storage of blood is advantageous, the question is still open as to whether the transfusion of red cells during the last week of storage (35 to 42 days) poses more risk than the transfusion of blood stored for the shorter intervals,” Drs. Tobian and Ness wrote. “The transfusion-medicine community needs to know whether the storage period should be reduced (e.g., <35 days) or new preservative solutions should be sought.”
- Heddle NM, Cook RJ, Arnold DM, et al. Effect of short-term vs. long-term blood storage on mortality after transfusion. N Engl J Med. 2016 October 24. [Epub ahead of print]
- Tobian AAR and Ness PM. Red cells—aging gracefully in the blood
|TABLE. Transfusion Data1|
|Median interval from randomization to issue of first red-cell unit for transfusion (IQR)||0.1
|Number of red-cell units transfused||25,466||50,890||0.57|
|Number of red cell units transfused per patient||0.57|
|Number of red-cell transfusion episodes per patients*||0.78|
|Duration of storage of transfused red cells in days||<0.001|
|Duration of storage of transfused red cells per patient in days||<0.001|
| Median of the mean
age of red cells
| Median of the oldest
red cells transfused
per patient (IQR)
|*Transfusion episode defined as all transfusions given on a single day.
IQR = interquartile range