College of American Pathologists
Printable Version

  Easy does it—showing caution with RBC transfusions


CAP Today




April 2009
Feature Story

Anne Paxton

The consensus is clear: Clinicians should take a much more cautious approach in transfusing blood products than what was once taught in medical schools. It’s now widely understood that excess transfusion of red blood cells can result in increased patient morbidity and mortality. But reversing decades of transfusion practice can be slow, and tracking progress is difficult when there ­isn’t much data on quality indicators.

Members of the CAP Quality Practices Committee aimed to fill some of that gap with their ­Q-Probes study titled “Utilization of RBC Transfusions,” the results of which were ­released in December 2008. The CAP study was designed to determine how well RBC transfusion practice conforms with institutional guidelines, to compare perfor­mance among institutions, and to find factors linked to better performance. It rep­resents a window into current transfusion practice at primarily small- to mid-size U.S. hospitals.

The study’s main finding: All indicators for transfusion, except for intraoperative he­morrhage, met the institutions’ guidelines about two-thirds of the time—from 63.7 percent to 68.3 percent. Intraoperative hemorrhage met the guidelines 37.2 percent of the time.

The findings are important because they provide a current benchmark against which institutions can compare and evaluate their own transfusion guidelines and practices, says study co-author Christopher M. Lehman, MD, co-director of hospital laboratories at the University of Utah Health Sciences Center, Salt Lake City, and a member of the CAP Quality Practices Committee.

Most of the 128 institutions that participated are in the United States; 49.1 percent of them are small hospitals with 150 or fewer beds. They reviewed a total of 5,830 transfusion events during the study period and found, overall, that 63.6 percent of the transfusion events met the institutions’ transfusion guidelines based on pre-transfusion hemoglobin (Hgb) level.

The study focused on three quality indicators: percentage of RBC transfusions meeting institutional guidelines during the study period, percentage of RBC transfusions reviewed for compliance with guidelines in the last calendar/fiscal year, and percentage of RBC transfusions in compliance with guidelines in the last calendar/fiscal year.

A total of 724,332 RBC units were transfused at the participating institutions over the last calendar or fiscal year, with 37.5 percent of them reviewed for compliance with institutional guidelines. Of those reviewed, 93.1 percent were in compliance with the transfusion guidelines.

The study demonstrated that a higher percentage of transfusions met institutional guidelines in institutions with 1) a lower percentage of transfusions with an indication of intraoperative hemorrhage, and 2) a committee that reviews RBC transfusion compliance that has a hematology/oncology physician as a committee member.

Perhaps surprisingly, there is little data to compare current average pre-transfusion Hgb levels with those of past years. This Q-Probes study found for the 5,830 transfusion events reviewed that the Hgb level averaged 8.1 g/dL, and “I would hope that is lower than it was 10 years ago,” Dr. Lehman says. “But it is difficult to find a comparable study from that era that looked at a global pre-transfusion Hgb level.”

The 8.1 g/dL average is a good sign, says Richard C. Friedberg, MD, PhD, chairman of the Department of Pathology at Baystate Medical Center, Springfield, Mass., and another co-author of the study, because it probably reflects a wider understanding that most hemodynamically stable patients who aren’t bleeding tolerate Hgb levels of 7.0 to 9.0 g/dL. “It’s certainly lower than 15 or 20 years ago.”

“I am encouraged by the data,” says Dr. Lehman. ”I hope that the changes are in response to studies that support more restrictive transfusion protocols in a variety of patient populations.”

The study’s authors suggest participants may want to further refine their RBC transfusion guidelines by lowering their “routine” threshold to 8 g/dL, and might wish to consider lowering the thresh­old to 7 g/dL for appropriate patient populations.

But the fact that 33.6 percent of labs still report at least one Hgb threshold of 10 g/dL or higher for patients without cardiovascular or pulmonary disease is cause for concern, the authors say, noting the developing consensus that such a threshold is too liberal and may in fact be harmful. “It may be a holdover of the 10/30 recommendations from the 1960s, which set the threshold at 10 g/dL or 30 percent Hct,” Dr. Lehman says. But “a hemoglobin greater than or equal to 10 g/dL is rarely an indication for transfusion, and institutions with that as a threshold should review their thresholds.”

It is significant, the Q-Probes study notes, that institutions that have a pathologist or transfusion medicine specialist actively involved in transfusion consultations were less likely to adopt a 10 g/dL threshold.

The study showed that intraoperative transfusions met the pre-transfusion Hgb guideline only 37.2 percent of the time, but the authors did not consider this surprising, since intraoperative RBC transfusion should be guided by ongoing blood loss and the hemodynamic status of the patient, rather than by Hgb levels.

The immediate postoperative Hgb value is a more useful estimate of the appropriateness of the transfusion, they point out, because a low postoperative Hgb suggests insufficient red cell volume replacement, while a high value suggests excessive transfusion.

However, the study also found that only 31 percent of reporting institutions review postoperative Hgb values. This is a “missed opportunity,” Dr. Lehman says, stressing that participants should evaluate surgical transfusion practice at their institutions and compare surgeons’ blood use within a surgical subspecialty. “Peer pressure among physicians can be a potent motivator to refine practice,” he points out.

Dr. Friedberg expresses reservation about whether such reviews should be done automatically for all patients. “It may make sense to check in the operative setting, but for the stable nonbleeding patient, I’m not sure if that number is high or low. I’ve heard people say we should get a hematocrit right after any transfusion to make sure the transfusion was effective. But the transfusion was to treat the patient, not the lab value. I’m just afraid we start treating numbers instead of patients.”

A large number of labs say there is “never” a review for guideline compliance before non-OR or trauma RBC units are released, which is unfortunate but not surprising, Dr. Lehman says. “Pretransfusion consultation —generally on the part of the pathologist—requires a significant commitment to be effective, and can become uncomfortable if the transfusing clinician becomes defensive. However, it’s important for institutions to remember that retrospective review has not been shown to be effective in changing transfusion practice.”

“Why do institutions have guidelines if they’re not going to see whether they’re complying with them?” Dr. Friedberg asks rhetorically. “If people are told ‘You need guidelines’ but they’re never enforced, it’s sort of contradictory.” He says it turns the guidelines into something more like suggestions.

On the other hand, he points out, 65.9 percent of institutions say they have changed their RBC transfusion guidelines in the past five years in response to published reports or reviews of studies on the risks and benefits of RBC transfusions. “The fact that so many have changed their guidelines tells me a lot have looked at them—and that’s good news,” Dr. Friedberg says.

Other responses to questions on practice characteristics found several areas for improvement. For example, only half of the institutions have physician trainees, but more than a third of those institutions do not give the trainees instruction on appropriate transfusion practice. Among all institutions, 27 percent said that institutional transfusion guidelines are not readily available for clinicians to consult when ordering transfusions. And 18.5 percent said they do not have a committee that reviews RBC transfusion compliance with institutional guidelines.

The study found that higher review rates of RBC transfusions for compliance with institutional guidelines in the last calendar year tended to be found in those institutions that do not have a teaching program, do send communications to transfusing clinicians requesting justification for out-of-compliance transfusions, do not use a local expert opinion as the source of the institutional transfusion guidelines, have lower volumes of plasma transfusions, and have lower volumes of cryoprecipitate transfusions.

Seventeen of the 128 participating institutions said they “never” send a communication, such as a letter, e-mail, or phone call, to the transfusing clinician requesting justification for the transfusion when the transfusion is judged to be out of compliance. Seven institutions said they never review RBC transfusions retrospectively.

The Q-Probes study authors were pleased to find that 60 percent of transfusion events involved one RBC unit, and that about 90 percent of institutions did not consider one unit an indication for utilization review.

“Years ago, there was a line that any single-unit transfusion was inappropriate, because if you only gave one, then either the patient didn’t need one or the patient needed two,” Dr. Friedberg says. “A lot of people still have that older school of thought in their heads, and I know places that used to review any single-unit transfusion because of it. But I think thinking has changed since then, and people are probably paying more attention to what’s appropriate.”

Single-unit transfusions are a “two-edged sword,” adds Dr. Lehman. “For many years, the transfusion community recommended against single-unit transfusions,” but recent outcome studies of RBC transfusions in a variety of patient populations support a more restrictive protocol that calls for the fewest number of units needed to maintain acceptable tissue oxygenation.

A plurality of participants in the study used the same thresholds for allogeneic and autologous transfusions, another finding the study authors considered encouraging. Past recommendations on autologous RBC transfusions proved to be misguided. “During the HIV era, autologous transfusions were pushed strongly, and some states mandated that they be made available. But over the past several years it’s become very clear that they’re rather inefficient,” says Dr. Friedberg.

“They require a lot of special handling, they’re much more expensive, and in many cases people were donating blood when their body didn’t have time to make the blood back. If all you were really doing was taking it out on Thursday and putting it in on Monday, you weren’t accomplishing anything,” because it takes a few weeks for the cells to replenish.

In the mid-1990s, Dr. Lehman points out, the American Society of Anesthesiologists suggested that indications for transfusion of autologous RBC units could be more liberal than those for allogeneic units. “In my experience, some surgeons followed this practice to the detriment of patient safety. For example, it was not uncommon to have patients’ hospital stays prolonged due to respiratory complications from volume overload because transfusion of autologous units was considered ‘safe.’” Other potential risks include bacterial contamination and inadvertent transfusion into an unintended recipient.

“Many physicians say, ‘It’s the patient’s blood and I’m just giving it back and there’s no risk of anything,’” Dr. Friedberg says. “But when you look at disease marker numbers and errors in transfusion rates, there’s a much greater risk of mistransfusion than of contracting a disease.”

Dr. Lehman sees it as a positive sign that institutions are not encouraging unnecessary transfusion of either allogeneic or autologous RBCs. The transfusion medicine community has played a role in teaching clinicians that most patients tolerate Hgb levels of 7–9 g/dL, he says. “But I think anesthesiologists, internists, and surgeons have also played a significant role by conducting clinical trials and retrospective reviews that evaluate the risks and benefits of RBC transfusion.”

The key message to derive from this Q-Probes study, says Dr. Friedberg: People should look at their transfusion practices and see if there is room for improvement. “The transfusion review committee or blood utilization committee should have active input from people who care about transfusion safety, and their policies should be reviewed carefully and accurately, instead of being automatically readopted. We want to make sure all of these institutions are not just functioning off of last year’s game plan” on transfusion practice, “but are actually sitting down to think about what makes sense.”

Anne Paxton is a writer in Seattle.