50/50 Rule Ties Blood Transfused To Increasing Mortality

Empty blood units piled on the operating room floor after a trauma.

A large retrospective analysis has uncovered a previously unnoticed linear relationship between mortality and units of blood transfused, namely, that there is a 10% increase in mortality for every additional 10 units of blood transfused—what has been dubbed “the 50/50 rule.” Morbidities were also assessed.

Trauma patients brought to the ER are often in hemorrhagic shock and require massive blood transfusions. Between 20% and 40% of these in-hospital deaths are potentially avoidable with rapid blood transfusion and improved resuscitation (Ann Surg 2015;261:586-590). Although increasing transfusion volumes are associated with increasing morbidity and mortality, the extent of such outcomes following high-dose transfusions has not been well characterized.

Steven M. Frank, MD, director of Johns Hopkins Health System’s Interdisciplinary Blood Management Program and director of the Center for Bloodless Medicine and Surgery in the Department of Anesthesiology and Critical Care Medicine at the Johns Hopkins Hospital, in Baltimore, together with colleagues from the Departments of Anesthesia, Critical Care and Transfusion Medicine, studied the clinical outcomes of more than 3,500 patients who had received massive transfusions (defined as ≥10 red blood cell [RBC] units per admission) from 2009 to 2014. This retrospective electronic medical records analysis totaled more than 272,000 medical and surgical patients, excluding those with hematologic malignancies.

“As anesthesiologists in the OR [operating room], we were giving coolers and coolers of blood, and we often wondered whether the patients survived their hospital stay,” Dr. Frank said. “We didn’t know the outcomes because we only dealt with OR. We decided to see if they survived, and also looked at the relationship between blood dose and in-hospital mortality. Perhaps more importantly, we also looked at five different morbid events,” he explained.

The researchers found a linear and dose-dependent increase in mortality, with a 10% increase in mortality for every additional 10 units of blood transfused (Figure). “At 50 units there was a 50% mortality, which we refer to as the 50/50 rule; it’s like flipping a coin,” Dr. Frank said. “We never knew this before because no one has ever described it,” he said.

 

Independent predictors of mortality were transfusion dose (odds ratio [OR], 1.034; 95% CI, 1.026-1.041), Charlson Comorbidity Index (OR, 1.183; 95% CI, 1.031-1.093), history of congestive heart failure (OR, 1.506; 95% CI, 1.08-2.092), and a high plasma- and platelet-to-erythrocyte transfusion ratio (OR, 1.062; 95% CI, 1.031-1.093).

The five morbid outcomes were hospital-acquired infections and thrombotic, renal, respiratory and ischemic events. These were assessed both as composites and by individual occurrence rates with multiple logistic regression analysis. Composite morbidity increased in a curvilinear fashion with increasing RBC dose, and exceeded 50% after 10 RBC units. For massively transfused patients, hospital-acquired infections and thrombotic events were four to five times more prevalent than renal, respiratory and ischemic events.

“We did not expect to see hospital-acquired infections and thrombosis to be four to five times more common than other morbid events,” said Dr. Frank, who is also associate professor of anesthesiology and critical care medicine at Johns Hopkins University. “The implications of this are that we ought to be more vigilant to prevent, diagnose and treat these [morbid events] after massive transfusions in order to improve outcomes.”

The study, “Dose-Response Curves for Morbidity and Mortality After Massive Transfusion and the 50/50 Rule,” is scheduled to be presented at the American Association of Blood Banks’ annual meeting in Anaheim, Calif., in late October (abstract SP283).

“We also wondered whether there was a point at which it becomes futile to continue giving massive transfusions,” Dr. Frank added. “It turns out the answer is no. A 50% survival rate is pretty good compared to other conditions, such as pancreatic cancer surgery, where the survival rate is about 20%. We don’t see a point of futility, and our recommendation is not to give up.”

‘There’s a Reason’

“If you’re giving massive amounts of blood, you’re not just giving it because nothing happened,” commented Peter J. Papadakos, MD, director of critical care medicine and professor of anesthesiology and surgery at the University of Rochester School of Medicine & Dentistry, in New York. “The primary problem is the initial physical insult. Shock in itself can activate a systemic inflammatory response. The more a patient bleeds and the more blood that is given is like putting gasoline on a fire in terms of an inflammatory response,” Dr. Papadakos told Anesthesiology News.

“Once the systemic inflammatory response is activated, be it from sepsis, severe trauma or neurological catastrophe, it itself is a major contributor to mortality,” said Dr. Papadakos, who was not involved in the study.

Jonathan H. Waters, MD, professor of anesthesiology at the University of Pittsburgh School of Medicine, said the association between massive transfusion and morbidity and mortality doesn’t mean that the transfusion itself is causative but, rather, could reflect the underlying condition that led to the transfusion. “What is striking is that as our blood management program has become increasingly effective and we reduce transfusions, we have seen associated reductions in morbidity,” said Dr. Waters, who was not involved in the study.

Because hospital-acquired infections are so common following massive transfusions, Dr. Frank believes they should be exempted from ever being classified as a preventable “never event,” for which Medicare does not reimburse providers. “We are talking about the sickest patients in the hospital,” Dr. Frank said. “After 50 units of blood, about 50% of patients would get a hospital-acquired infection and 50% would have thrombotic events. There should be an exception to the rule for hospital-acquired events for massively transfused patients.”

One outcome not included in the study’s published findings was that surgical patients had better prognoses than medical patients in terms of mortality. “Our hypothesis is that surgical massive bleeds are more amenable to treatment than medical,” Dr. Frank said.

“This is an important paper,” concluded Dr. Papadakos. “It shows that you get 50% mortality after giving 50 units of blood.” He added, “I’m surprised that it’s that low.”