College of American Pathologists
Printable Version

  New game plans for taming
  blood use


cap today

The blood bank software of your dreams

October 2001
Cover Story

Anne Paxton

Fletcher Allen Health Care had just completed its budget for the upcoming fiscal year and was preparing to submit to the state when the announcement was made: The American Red Cross would have to raise prices substantially on its blood products. Suddenly, the 500-bed Burlington-based hospital, the largest in Vermont, had its projections knocked in a cocked hat. "It meant $750,000 in unplanned expenses for blood products," says Thomas Wadsworth, MBA, MT(ASCP) SBB, the hospital’s administrative director of pathology and laboratory medicine.

Hospital laboratories throughout the country are scrambling to cope with the same shortfall. Starting July 1, blood products, laboratories’ second biggest expense after salaries, have risen in price by 10 percent to 50 percent or even higher for hospitals supplied by the Red Cross.

In response, Fletcher Allen and other blood services are exploring how they can reduce their dependence on the Red Cross. But many had already started to address the problem from the standpoint of consumption: by adopting new strategies to control blood usage. A few years ago, "We thought the price of red cells would likely double or triple in two to five years," says Patrick James, MD, director of pathology and clinical laboratories for Health Midwest, Kansas City, Mo. "So we realized effective utilization is really the only answer, not only for patient safety and effectiveness, but also cost."

Though steep, the price hike wasn’t unexpected. As the Red Cross moves toward universal leukoreduction, continues nucleic acid testing, and readies broader deferral policies to forestall Creutzfeldt-Jakob disease, prices for blood products could go in only one direction. Transfusion audit committees are required under most accreditation standards, and monitoring of blood usage has long been routine. But hospitals are scrutinizing blood usage now with a fresh intensity.

A shifting consensus on laboratory values that should trigger transfusions has been part of the picture. Research emerging in the last three or four years has suggested that a significant proportion of red cell transfusions are not clinically indicated and place the patient at unnecessary risk, according to Robert Westphal, MD, former medical director of the Vermont-New Hampshire region of the American Red Cross.

For example, a recent study comparing "conservative" and "liberal" transfusion strategies in critically ill patients found that transfusions given to maintain the hemoglobin level between 8-10 g/dL were at least as safe, and possibly safer, than transfusions given to maintain the hemoglobin level above 10g/dL (Hebert PC, et al. N Engl J Med. 1999;340:409-418). In other comparative research, the risk of mortality of almost 20,000 Jehovah’s Witnesses, who were undergoing surgery for many types of procedures and declined transfusions, was shown to increase significantly only when hemoglobin fell below 7 g/dL (Carson JL, et al. JAMA. 1998;279: 199-205).

In 1998, Dr. Westphal helped form a cooperative effort of Vermont hospitals, the Vermont Transfusion Project, designed to educate physicians throughout the state about the appropriate use of blood products and their risks.

The transfusion project and an active Vermont lab group, Clinical Laboratory Managers of Vermont, give Vermont hospitals a united front in coping with the slated price increase—which is a double-edged sword, says Wadsworth. His hospital was paying about $117 for non-leukoreduced red cells, which are used for 60 percent of patients. Now the Red Cross, which has explained that $30 per unit of the price increase is attributed to universal leukoreduction, is not only charging $182 for ULR red cells but also no longer offers non-leukoreduced.

The CAP believes that leukoreduction is not needed for many patients and should be used at the physician’s discretion. Taking the same position, Vermont hospitals are looking at alternatives to supplement or replace the Red Cross. They have just completed a feasibility study of setting up a statewide donor system or bringing in another company to work with, and they plan to make a decision by the end of October.

But it may be better for some hospitals to focus their efforts elsewhere, Dr. James suggests. "It seems in many communities, hospitals and blood centers have a strictly customer-and-vendor relationship and the hospital tries to beat the stuffing out of the blood centers on pricing-when, in fact, utilization is a much more powerful tool."

The experience of Health Midwest shows how effective new utilization strategies can be.

"We knew the way to change behavior was through valid data," Dr. James explains. "And we knew the data would be most powerful if we could reference it to peers in the community. We took a five-year snapshot of blood utilization trends, then started taking our analysis through a typical quality improvement process. Looking at our blood expense by product, it turned out red blood cells made up two thirds of the expense, while platelets were 15 percent. Those were the areas with opportunity."

"In our experience, about 80 percent of our utilization was elective" rather than emergency, Dr. James says. "That gave us the opportunity to affect decision-making, because in those cases physicians were not making decisions in the heat of the moment. We drilled into the procedures that drove red-cell utilization and elected to pick three of them: open heart surgery, orthopedics, and oncology."

"In open-heart surgery, early on it seemed like some physicians were using more red cells than their colleagues in the region. By sharing data and getting them involved along with anesthesiologists and perfusionists, we showed them their outcomes are every bit as good as those using less blood."

The result: Health Midwest succeeded in cutting its blood usage by eight percent in 2000, going against the trend throughout Kansas City. In the same year, the general community’s blood usage rose seven or nine percent. "So our ’delta’ with the community was about 15 percent," Dr. James says.

Numerous hospitals have joined forces in hopes of attaining similar success. At least 50 that are members of the Premier Inc. purchasing cooperative are participating in a 12-month initiative launched in June to improve blood usage by integrating the use of quality plans, the surgical blood order schedule, and technology. A "rapid-cycle improvement model" designed by Premier’s partner, the Institute for Healthcare Improvement, will make it possible for network members to share best practices.

One Premier participant, Danville (Va.) Regional Medical Center, has its own donor center and is less affected by outside price increases. But it has benefited from other Premier outcomes initiatives on acute MI, stroke, and congestive heart failure, says Patricia Prevette, RN, Danville Regional’s senior review analyst, medical quality management.

"Premier’s data make it possible for our hospital to track the impact these initiatives have on patient outcomes and compare itself with peer groups," she says. Similarly, under the blood initiative, Danville Regional will be able to compare its blood usage with that of peer groups.

Laboratorians at SSM Cardinal Glennon Children’s Hospital, however, see a benefit of a different sort. The St. Louis hospital joined the Premier initiative because of worries about price increases, shortages, and patient safety. "We really don’t use a lot of whole blood in pediatrics. That’s why it’s more difficult for us to benchmark against other facilities," says Helen Gagen, MA, MT(ASCP), administrative director of diagnostic services.

"But we still have to focus on anything to reduce utilization. And since we’re the teaching arm for pediatrics at St. Louis University, it’s good practice to make sure the medical students and interns and residents are appropriately educated."

Gagen is hoping the initiative will help Cardinal Glennon adopt more standardized ordering practices. While there are standard ordering lists for adult surgeries, "they’re not used as much in pediatrics, but they should be." Right now the physician tells the laboratory how much blood will be needed. "We’re seeing if we can tweak our surgery orders somewhat" to reduce those numbers, she says.

St. Louis has some of the highest blood costs in the United States, so even a mild improvement in usage will help the laboratory’s budget. Says Gagen, "We don’t discard a lot of blood, but even discarding one unit of a platelet pheresis is excessive when the donor supply is going down and you’ve paid $400 to $500 for it."

The new transfusion audit system at Mary Hitchcock Memorial Hospital, launched in March, replaced a system that had remained unaltered for well over a decade, says James P. AuBuchon, MD, professor and chairman of the Department of Pathology at Dartmouth-Hitchcock Medical Center, Lebanon, NH. And there were several problems with it.

"In the past we had to have an apparently lenient transfusion audit trigger to avoid overwhelming the system. While most transfusions were occurring at a hemoglobin of 8-9 g/dL, the audit criterion was set at 10 g/dL." Some physicians may have interpreted this trigger as the guideline for transfusion, he speculates. "This obviously does not reflect current data indicating that a more conservative approach to red cell transfusion yields less patient morbidity and mortality."

Transfusion committee members were also concerned that their review of transfusion indications was always retrospective and forced them to second-guess decisions made in real time. "Any system like that has difficulty capturing the flavor of the moment and understanding what clinicians were facing," says Dr. AuBuchon, who is vice chair of the CAP Transfusion Medicine Resource Committee. "The ’retrospectoscope’ may have an entirely different view of the world than that of the physician standing at the bedside."

Since the audit was directed at the physician legally responsible for the patient’s care, it also usually bypassed the house officers who were most closely involved in the decision to transfuse. Further, since transfusion committee meetings occur quarterly, they frequently came so long after the event that they had little educational value. "Often the physician would remember little of the patient, let alone the transfusion," Dr. AuBuchon says.

That time frame has telescoped under the new audit procedures. For example, an audit is triggered automatically for red cell transfusion except in the following cases: patient age between 60 and 80 and hemoglobin < 9.0 g/dL, patient age 40-60 years and hemoglobin < 8.0 g/dL, patient age 4 months to 40 years and hemoglobin < 7.0 g/dL, patient age < 4 months and hematocrit < 24 percent, or patient age > 80 years. For plasma, audits are automatic except where PT = 20 seconds or PTT = 44 seconds (1.5 times the upper limit of the reference range).

Each morning, the laboratory information system generates a list of transfusions triggering an audit, and generally during morning rounds in the laboratory the pathologist attending on transfusion service reviews the list to select out any cases that may have had extenuating circumstances. The ordering physician for each transfusion selected for audit then receives an e-mail note with standardized wording reviewing the literature and advising why, in most circumstances, transfusion in that situation is not clinically necessary, but explicitly noting that there may have been clinical justification for that particular transfusion.

With this more automatic system, "We can now bring the trigger closer to what the committee felt were justifiable transfusion indications," Dr. AuBuchon explains. "It has increased the volume of interactions with clinicians more than tenfold-but it’s allowed a 90 percent reduction in the amount of time I spend on audits, because I’m no longer reviewing charts and preparing cases for presentations." To date, he says, the rate of transfusion appears not to be increasing with the new system, and it may be decreasing through broader dissemination of the current literature on transfusion triggers.

The procedure at Palomar Pomerado Health System in Escondido, Calif., is comparable. Jerry Kolins, MD, medical director of laboratory services at Palomar as well as part-time medical director of the American Red Cross for Southern California, said the clinical information system at the two hospitals checks the pretransfusion hemoglobin, platelet count, and coagulation studies on every patient who received red blood cells, pheresis, or fresh frozen plasma.

"If there is no laboratory value given, then the physician has to explain the transfusion. We believe there should be evidence—and generally speaking it should be laboratory evidence—to support transfusion of blood products."

Most of the time, with the occasional exception of frozen plasma, pretransfusion laboratory data are available. Although employees trained in quality management review the chart before it goes to the blood utilization committee and may find the proper studies, "There are always some cases of outliers where the hospital staff found the computer information was accurate. Then the doctors on the committee make a decision about the transfusion, send it to the doctor involved, and a copy is placed in the physician’s credentialing file if indeed there remains a difference of opinion."

Each quarter the committee has to evaluate about three to five cases for platelets and about 10 for plasma. In one recent case, a physician gave platelets to a patient with a normal platelet count and argued it was justifiable because the patient had uremia. The committee wrote a letter saying "We are aware of qualitative platelet defect in uremia, but we are not aware that platelets will correct it. Here’s a list of articles supporting our position that there is no value in transfusing a patient with a normal platelet count," Dr. Kolins relates.

"That would be a good example of an exchange between the committee and the doctor," he adds. "If the doctor doesn’t respond, there is an acknowledgment that a platelets transfusion occurred that was inappropriate and it is tracked to see if it occurs again, and it can be used during the recredentialing process."

The response isn’t always warm. He hoped that by having all the members of the committee sign the letter the medical director of the laboratory would not receive the "nasty-gram" from the member of the medical staff. "But I was wrong. I still get the nasty-gram," Dr. Kolins says.

Evaluating transfusion of red blood cells is more difficult, he believes. "In terms of cost savings, the window of opportunity for improving utilization is almost shut with red blood cells, in my opinion. A great deal of literature supports the concept that the transfusion trigger is a figment of the imagination of quality assurance reviewers. More often than not, when you think the trigger hasn’t been reached, the clinical situation on further review would justify a transfusion." By contrast, he says, "the indications for use of plasma and platelets are a little clearer and easier to get your arms around."

While its blood usage guidelines have remained unchanged for some for some time, Los Angeles County-USC Medical Center has sought to increase the involvement of clinical services in monitoring red blood cells, fresh frozen plasma, platelets, and cryoprecipitate usage. Under a new system launched last January, each of 12 clinical services is assigned a month when it is responsible for reviewing its physicians’ use of a particular blood product. The transfusion committee chair will also conduct random review of the cases that are forwarded to the clinical divisions.

"All the divisions know which month they will be expected to review their transfusions," says Sunita Saxena, MD, director of comprehensive health center laboratories for the network and chair of the blood use review committee. "For example, in January anesthesiology might be assigned to review all red cell transfusions that occur. So over a four-year period we would have reviewed all transfusions of all products."

"Once we forward the flagged transfusions to the representative for that division, they order the chart and review it based on the review guidelines that have been approved by the transfusion committee. Then they give a report at the transfusion committee of how many cases they reviewed, how many were considered justifiable, and if any were considered nonjustifiable, what corrective actions they took," explains Dr. Saxena, who is also chair of the quality improvement committee for the LAC-USC Medical Center Department of Laboratories and Pathology.

One benefit of this system is that the clinical services get feedback on how their faculty uses blood. At the last Joint Commission survey, while reviewing physicians’ credentialing files, the surveyors asked for information on justifiable and nonjustifiable transfusions ordered by the attending staff. "The new systems will allow us to start capturing some of that information, but the information will be only for the cases reviewed," Dr. Saxena says. "My eventual goal is to capture physician information for each transfusion ordered. Once the hospital information system allows this, I think we will be able to link almost all orders for transfusions to an attending staff and not just the ones that were selected for review, and that would be very helpful for compliance and improving blood utilization practices."

In addition to tracking blood usage, it is important to monitor and prevent blood product wastage. Says Ira A. Shulman, MD, director of transfusion medicine at University of Southern California: "Each unit wasted is a cost to the facility because we can’t bill the patient for waste that’s not the patient’s fault. So the hospital eats the cost. It doesn’t matter whether we are billing fee-for-service or not; it’s a cost we can’t recapture."

Managing inventory is "partly art and partly science," adds Dr. Shulman, who chairs the CAP Transfusion Medicine Resource Committee. "There’s always going to be a certain amount of blood product inventory that gets wasted just to make sure there’s always enough blood in the inventory to meet patient needs. If you try to maintain too tight of an inventory, you run the risk of not having enough."

Proper management of inventory can minimize another kind of waste: that which occurs when blood is dispensed for use but something happens after the blood is no longer under the blood bank’s direct control. "The nurse goes to administer the blood product and pokes a hole in it by mistake, or the product leaks or drops on the floor and bursts, or it was sent to the spot where the patient is but the patient’s condition improves miraculously, or the patient dies, and the blood product isn’t returned to the laboratory within the required time frame," Dr. Shulman says.

In the latter cases, it’s desirable to return units to the lab so they can go back into storage without expiring—but that can be complicated. "Different products that get issued occasionally get returned to the laboratory unused," he notes. "So what do you do? Take a $300 to $400 product and throw it away, or do you have rules to restock it?" (This issue and related topics are discussed in the blood bank forum Dr. Shulman moderates, at

Albert Einstein Medical Center in Philadelphia has used elaborate statistical review to make a dramatic impact on utilization. By conducting a real-time evaluation of blood usage, as well as retrospective review to ensure physicians are following the medical center’s policies, it has reduced red blood cell usage by 49 percent and platelet usage by 59 percent.

Ierachmiel Daskal, MD, PhD, chairman of the Department of Pathology and Laboratory Medicine at Albert Einstein, says the system has increased control over costs but has had a more important impact on product availability. "We can intervene, for example, to change the surgery schedule so as not to have the same blood type follow case after case and ensure an adequate supply in the blood bank," he says.

When there are blood shortages, the laboratory may actually intervene directly by reviewing the surgery schedule the night before. "Politically this can be extremely complicated," Dr. Daskal concedes. "Every surgeon wants to be the first case in the morning. But there’s no other way to rationally deal with occasional shortages."

In general, he says little progress can be made in improving blood utilization "unless you have a wonderful working relationship with the other side. You need to cultivate constructive dialogue with anesthesia and surgery—and that’s easier said than done."

One strategy he has found helpful is depersonalizing the review process. "The clinicians’ response is not always very favorable because they perceive it as someone looking over their shoulder. But the message we send to the chairman of surgery is the following: Here are your statistics. I don’t give names. I just give symbols on a histogram. You have nine surgeons and two exceed your own mean, so take a look. If you present absolutely pure statistical data, showing outliers when compared to the other surgeons, it’s easier to swallow."

At Mary Hitchcock Memorial Hospital, clinician reaction has also varied. "Some of them understand and appreciate what we’re doing, some have mistaken this for the previous kind of interaction and attempted to justify their decisions, and, predictably, some misunderstood the communication and have taken umbrage at being questioned about their transfusion practices," Dr. AuBuchon says. "However, I have responded to those e-mails by pointing out the necessity of collegial discourse in an academic medical center in order to achieve continuous improvement."

The blood bank can’t improve utilization alone, concurs Suzanne Butch, MA, MT(ASCP)SBB, chief technologist with the blood bank and transfusion service at the University of Michigan Health Systems, Ann Arbor. "Unless you have a cardiac surgeon, a transplant surgeon, or hematology/oncology physician who is interested in blood conservation, you are doomed."

Butch’s blood bank tries to work cooperatively with physicians on the best ways to sort blood utilization data. After providing those data on bone marrow transplants since the program began, "we’ve seen a significant decrease in the average and median blood use in our bone marrow transplant patients, even though we’re doing a lot more allogeneic transplants and they traditionally use more blood than autologous transplants."

Her review program simply provides quarterly information to the heads of selected clinical services about their blood usage. "We make it their job to look at their own practices and assess their performance," she emphasizes. "There’s not that animosity of ’you’re telling me how to practice.’ We say, ’here’s your practice, you say whether it’s appropriate and if there’s anything wrong with the data.’"

The trick is "trying to agree on what patient belongs in what group," and trending those data in the most useful way, she says. With cardiac surgeons, for example, "it’s been a slow process, because we realized that the patient groups were a lot different than in the past. Although we were sorting out the coronary artery bypass from some other thoracic or cardiac procedures like aortic root repair, even within the coronary bypass category there are groups that are different."

Aneurysm repairs may also differ in seriousness and difficulty. "So you can’t lump the data all together and say, ’your average is here,’" Butch says. When the blood bank saw that the bone marrow transplant services were doing fewer autologous transplants and more allogeneic transplants, for example, it started separating the trend data into different categories.

Because it is a tertiary care center that transfuses more than 100,000 blood components a year, the University of Michigan Health Systems may be difficult to compare to other institutions. Says Butch: "I don’t think it’s as important to benchmark against another facility as to look at current practices and see if they can be improved, regardless of the performance at other institutions. The only thing benchmarks do is identify potential areas to investigate. In and of itself, getting to match somebody else’s figures is not the goal."

Autologous blood accounts for the majority of units that have to be discarded at most blood banks. "We looked at waste in autologous blood a number of years ago, and we were amazed to find eight out of 10 units we were collecting were never transfused," says Dr. AuBuchon. Merely by sharing this information with surgeons, along with evidence that it was much safer for patients to get allogeneic blood rather than autologous, "we brought our proportion of use to 50 percent which is the national average."

Most of the waste of autologous blood at Health Midwest is in orthopedic surgery, Dr. James says. "The criteria for autologous are less strict in general. Our concern is that the orthopedic population is mostly elderly, many are not in the greatest health, and if they’re drawing several units of blood it often makes them borderline anemic and not in the best state for surgery. And sometimes they’re not getting the units at all."

But orders for autologous blood can be difficult to control. There is 80 percent wastage of autologous blood at USC’s Kenneth Norris Cancer Hospital. "But I don’t really call it wastage," Dr. Shulman says. "Patients make autologous donations in the hope they won’t need any blood, but they’ll only get their own back if they do. It’s an insurance policy you don’t want to use, and we teach doctors to only give a transfusion when it’s really needed."

Autologous blood is "probably the most emotional subject the transfusion committee deals with," Dr. Kolins says, "because many doctors—and they seem to be heavily concentrated in orthopedic surgery—feel that a patient is entitled to receive back his own blood even if they have 10 grams of hemoglobin. Of course, the committee simply asks them to weigh the risks and benefits. We try to ask what are you treating when you return that blood, and we’ll list some of the complications of autologous transfer, including endotoxic shock, septicemia, and even death."

Even with only a 50 percent discard rate for autologous blood, the cost is probably between $100 and $200 per unit, but the "fully loaded cost is close to double that because you’re throwing half out," he adds. "It’s not an insignificant amount. But California’s blood safety act places hospitals in a difficult position. Every doctor performing a procedure on a patient has to give the patient the opportunity to donate his own blood or even arrange directed donation. It’s actually against the law to proceed with an elective surgery if you did not provide that opportunity to the patient."

"What happens is the doctor knows he must be in compliance, so he outlines the options for the patient," Dr. Kolins says. "The patient says ’I’d better do it, otherwise why would he ask?’ so he donates two units. If they’re not used, the patient feels good, the doctor feels good, and everybody thinks everything’s going smoothly—except the chief financial officer who wants to know why they’re throwing out blood."

Decreasing waste in autologous blood "was one of our goals for the Premier initiative," says Prevette of Danville Regional Medical Center. At one time, Danville was discarding more than 70 percent of its autologous blood. In some cases, physicians were ordering donations that exceeded the maximum surgical blood ordering schedule. For example, "three or four units of autologous blood were being donated for a surgical procedure that should, without complications, require a maximum of only two units," she says.

Danville reduced its rate of discarded autologous units to 57 percent for year 2000 by making physicians more aware of "the amounts of discarded autologous units of blood." Data were collected and reported to involved physicians and to the medical staff as part of quality improvement efforts.

Donation criteria were an area identified for improvement. Before Aug. 1 of this year, "our minimum autologous donation criterion for hemoglobin had been 11.5 g/dL, which is low normal. It was changed to 12.5 g/dL, which is the level for all of our blood donors," Prevette explains. Danville found that patients who donated autologous blood when they were at 11.5 g/dL often returned with their hemoglobin too low for further donation or, worse, were found to be anemic at the time of their surgery. "They weren’t able to replenish their blood—thus, many of these patients would receive not only autologous units but require additional units from the blood bank," she says.

Another measure was to inhibit single-unit autologous donations. "As a general rule, patients should not require single-unit transfusions. In most patients, a single-unit blood will increase the hemoglobin 1 g/dL; therefore, single-unit transfusions are not considered standard of practice and should be avoided." Danville Regional Medical Center’s goal "is not to refuse the patient’s or physician’s request to donate autologous blood but to discourage donations which will more than likely be wasted," Prevette says.

It may seem difficult to position improved blood utilization high on the list of laboratory priorities. "We’re all so busy that when things seem to be going right, we don’t want to explore them. But the value of any initiative that looks for improved effectiveness is that it forces all of us to take a second look at what we’re doing," Dr. Kolins says. "The real truth is that just because it’s fixed doesn’t mean you can’t make it better."

More uniform practice in blood usage is a worthwhile goal, Dr. Daskal emphasizes. "As the new wave of outcome-based medicine takes hold in all other disciplines, there’s no reason why it shouldn’t be in blood banking." And there may be a broader benefit to new strategies to control blood use, Dr. James says: that through ongoing improvement in managing data, the laboratory can continue to expand its usefulness to the hospital at large. One of the values of laboratories is their "boundarylessness," he says. "We’d like to make ’lab’ a verb rather than a noun, to really see it as an activity that is generating clinically useful information."

Anne Paxton is a writer in Seattle.