Although a restrictive transfusion strategy significantly reduces the risk for 30-day all-cause mortality in critical care patients, its benefits are less clear in perioperative patients, according to a meta-analysis by a Canadian research team.
“Although packed red blood cells are arguably the best therapy to use when you have a problem with oxygen-carrying capacity, they come with a couple risks, such as the potential for transfusion reactions and infectious disease transmission,” said Matthew Chong, MD, a resident at Western University, in London, Ontario. “Furthermore, they’re also costly, at $300 to $500 U.S. dollars, to give a patient a unit of blood. Finally, the products have a limited shelf life and supply.
“With that in mind, we wondered if there is a difference in the time course of the pathophysiology of the oxygen demands for critically ill patients versus the perioperative patients that we care for every day as anesthesiologists,” he said. “Therefore, we sought to perform a meta-analysis looking at whether clinically important outcomes differ between critical care patients and perioperative patients when they are given these transfusion triggers.”
Liberal Versus Restrictive Strategies
To that end, Dr. Chong and his colleagues performed a meta-analysis of randomized controlled trials of adult surgical or critically ill patients receiving a liberal versus restrictive transfusion strategy. The primary outcome of interest was 30-day all-cause mortality. Secondary outcomes included 90-day mortality, morbidity, the blood volume transfused and hospital length of stay. “Finally, we wanted to look at economic measures, such as hospital length of stay,” Dr. Chong added.
As Dr. Chong reported at the 2016 annual meeting of the Canadian Anesthesiologists’ Society (abstract 152971), the systematic review retrieved 6,055 citations, of which 25 randomized controlled trials (comprising 10,617 patients) met the inclusion criteria. Of these, 11 trials were in critically ill patients and 14 in perioperative patients.
Not surprisingly, patients who received the liberal protocol received more blood than their counterparts (1.5 units; 95% CI, 1.1-1.8 units; P<0.001). “It turns out that no matter how you slice it, whether it’s a proportion of patients transfused or the amount of blood given, the restrictive-strategy patients got less blood,” he said.
It was also found that in critical care patients, the restrictive transfusion strategy resulted in significantly reduced 30-day mortality compared with a liberal transfusion strategy (odds ratio [OR], 0.82; 95% CI, 0.69-0.99). “The number needed to treat for this finding is 33,” Dr. Chong added. “So that’s a reasonable magnitude of effect, and comparable to ACE [angiotensin-converting enzyme] inhibitors and beta-blockers.” In surgical patients, however, the restrictive transfusion strategy led to the opposite direction of effect for 30-day mortality (OR, 1.33; 95% CI, 0.96-1.84).
“Furthermore, critically ill patients also experienced less cerebrovascular events, less transfusion reactions and slightly shorter length of stay with the restrictive strategy,” Dr. Chong noted.
The researchers subsequently performed a subgroup analysis of perioperative patients by surgery type, finding a higher risk for myocardial infarction among non–cardiac surgery patients receiving a restrictive transfusion strategy (OR, 1.66; 95% CI, 1.01-2.70). “However, the number needed to treat was quite high, so the clinical significance of this finding may be somewhat less concerning,” Dr. Chong said.
“So we can reasonably conclude that critical care patients do benefit from a restrictive strategy, insofar as our primary outcome of 30-day mortality was lower,” he said.
“The benefits of a restrictive strategy for the critically ill were also reflected in our secondary outcomes, where patients had fewer cerebrovascular events, fewer transfusion reactions and slightly shorter hospital lengths of stay.”
No Mortality Benefit in Perioperative Patients
“In contrast, the perioperative patients in our analysis did not have a mortality benefit at 30 days, and indeed there is an equivocal signal of harm,” he added. “This uncertain state of the evidence is important because it means that we cannot advocate for any one strategy in perioperative patients at this time. Our findings are in stark contrast to the ASA [American Society of Anesthesiologists] guidelines, which do recommend a restrictive strategy in perioperative patients.”
Bruce D. Spiess, MD, professor and associate chair of anesthesiology at the University of Florida College of Medicine, in Gainesville, told Anesthesiology News that Dr. Chong’s analysis of randomized controlled trials is important because it supports findings that have been found in retrospective database research.
“Blood transfusion has evolved in medicine as a culture and a religion, a human behavior not based on hard science,” he commented. “The belief that people will be better with more blood transfusions has been refuted over and over again.
“Those on the front lines making the hard decision to transfuse any given unit to a patient still do not have adequate decision trees for when to invoke this, the most common therapy in American medicine,” Dr. Spiess continued. “The review of randomized trials should awaken us all to the fact that the most common therapy in medicine seems to be making people die more often the more we give it!
“We need research across all of medicine to define when critical oxygen delivery is threatened and when transfusion could help.”