December 2008
Feature Story

Anne Paxton

About 80,000 times a day, physicians in the U.S. must decide whether to transfuse blood products into a patient. But often the decision is not an informed judgment based on established medical science, says anesthesiologist Timothy Hannon, MD, MBA, medical director of the St. Vincent Indianapolis Hospital Blood Management Program and president of Strategic Healthcare Group LLC, a blood management consulting company. It’s more likely to be a snap judgment based on physicians’ and nurses’ comfort levels. This pattern of “drive-by transfusion,” as he calls it, is hazardous to patients and why blood management programs are so important.

Until now, such programs haven’t been mandatory. But within the next few months, Dr. Hannon says, changing accreditation priorities, new Medicare payment policies, and economic realities will come together to create a “perfect storm” that will force hospitals to embrace blood management in a new way. “Blood management is not a new concept. But the environment now is going to produce the change that’s required.”

In a presentation at the CAP ’08 meeting last fall, “Blood Management Nuts and Bolts: The Role of Pathologists in Getting it Going,” Dr. Hannon and Arthur W. Bracey, MD, associate chief of pathology at St. Luke’s Episcopal Hospital in Houston, described their real-world experiences with using blood management programs to scale down blood use and improve patient outcomes.

Blood conservation comes naturally to Dr. Hannon, who did his medical training in the Navy. “The Navy and the Marine Corps work in austere environments and make good use of available resources, so it was sort of a natural extension to have a perioperative blood conservation program.” St. Vincent’s multidisciplinary transfusion committee and blood management teams applied this experience in the hospital and broadened the scope, and they have been able to reduce blood use by almost 30 percent and sustain the reduction for more than seven years. It’s a model for a strategic approach that other hospitals can employ, he says.

In the U.S. and elsewhere, most of the resources devoted to blood safety have focused on the bag of blood and on making it even safer than it is already, Dr. Hannon said. But transfusion safety is a “vein to vein” process, and the chief transfusion hazards are outside the bag: in the operating room and at the patient bedside. “So if we’re really going to improve patient safety and quality, we need to reengineer the process at the hospital level.”

What is blood management? Speaking as an MBA as well as an MD, Dr. Hannon said he employs the language of hospital administrators to define it as an evidence-based, multidisciplinary process designed to promote the optimal and efficient use of blood and blood-related resources throughout the hospital. It deals with “both sides of the blood bank window.”

The nation’s hospitals have considerable ground to cover before getting to “optimal,” he said. A 1998 study of 24 heart institutions across the country, in which they were asked what percent of patients they transfuse, generated answers ranging from 27 percent to 92 percent—a fourfold difference in transfusion rates (Stover EP, et al, for the Institutions of the MultiCenter Study of Perioperative Ischemia Research Group. Cardiothorac Vasc Anesth. 2000;14:171–176). And recently an international comparison found a 12-fold difference in transfusion rates for the management of very similar patients. (Snyder-Ramos SA, et al. Transfusion. 2008;48[7]:1284–1299. Epub April 14, 2008).

Even within institutions, practices are anything but standard. “If you look at the practices of cardiologists, surgeons, and intensivists within the same institution,” he said, “you will see a bell-shaped curve where some doctors are rarely transfusing and others are transfusing as if it’s going out of style. This sort of variance in a process that should be fairly well regulated and fairly well standardized doesn’t seem to make a lot of sense.”

Though blood use in the western world is stable or declining, for a variety of reasons it is still increasing by about two percent to three percent a year in the U.S., Dr. Hannon said. From 1999 through 2004, blood use rose by 16 percent in the United States while it declined by eight percent in the United Kingdom. “Currently we use 16 percent more blood than Europe and 44 percent more than Canada [measured by red cells per thousand of population],” he noted.

Added to that trend is a remarkable shift in where blood products are being used. At Dr. Hannon’s 750-bed community hospital in 2001, “35 percent of our blood was going to cardiac surgery patients, 22 percent to oncology patients. Now the numbers are almost reversed and about 25 percent of our transfusions are now given to outpatients. It’s a paradigm shift in the use of blood products—now there is as much given to medical patients as to surgical patients.”

To Dr. Hannon, however, the raw numbers are not that important. “I don’t care if we use blood more than other countries, but do we use it appropriately? That’s where it gets kind of ugly.” When experts sit down and review physician transfusion decisions within a given organization, they typically find that between 30 percent and 70 percent of transfusion decisions are inappropriate, he said. And these numbers haven’t changed for some time.

“It seems to be pervasive, not just at community hospitals but also academic hospitals,” that physicians receive little training in transfusion therapy. “Medical students don’t get this, even most anesthesia programs don’t adequately cover it. It’s mostly what doctors heard on call or from their staff. So the lineage or heritage of transfusion practice comes from wherever you trained.”

Nurses suffer the same training gap, he said, citing the experience of a nursing school graduate who reported getting four times more instruction on properly making a bed than about safely and properly administering a transfusion. “What drives this wedge even further,” he continued, “is that it takes about 12 years to integrate peer-reviewed article information into common practice. Physicians often fail to update information in a timely fashion, particularly for things they don’t perceive as important, such as plain old blood transfusion therapy.”

It doesn’t mean all 29 million transfusions ordered per year are inappropriate, he said—just that many are not evidence-based. The landmark study that remains the only prospective randomized trial of transfusion therapy effectiveness in adults was published in 1999 in the New England Journal of Medicine, he said (Hébert PC, et al. 340:409–417). “A restrictive strategy of red cell transfusion,” the article said, “is at least as effective as and possibly superior to a liberal strategy in critically ill patients, with the possible exception of patients with acute myocardial infarction or unstable angina.”

“That study changed transfusion practices worldwide,” Dr. Hannon said. “But I guarantee you that many of your physicians who transfuse every day are unfamiliar with it.”

Less well understood, but an equally important quality issue, is the age of blood products. “I’m not impugning the blood collection industry or the local blood banks, but I think we’re starting to understand better that there are downsides to storing something in a refrigerator for 42 days. Any biological substance we put in a refrigerator changes its properties over time, so this storage issue has become a bit of a hot topic.”

The 42-day outdates for blood are based on recoverability under ideal conditions, Dr. Hannon said. “But we do know that for every day we store blood, we get more downside having to do with the buildup of cytokines, plasma free hemoglobin, potassium, and cellular debris, collectively termed BRMs or biological response modifiers.”

These agents make the blood products more proinflammatory in the patient the longer they are stored—and unfortunately it also appears that red cells stored over time, whether because of nitrous oxide deficits or other storage lesions, become increasingly sticky and inflexible and less able to perfuse the capillaries. “So we know the risk-benefit situation changes with each day of blood storage, but we’re still not quite sure what to do about it.”

There is no study to show that blood should be discarded at 10 or 14 or 20 days, he added. “My guess is there is a statistically significant, but marginal, clinical difference in 10-day blood versus 20-day or 20-day versus 30. But I think the bigger issue remains not the age of the blood but the difference between one unit versus two, and two units versus three.”

Another hot topic affecting transfusion services is the use of massive transfusion protocols and trauma coagulopathy. As an anesthesiologist, he had to train Vietnam-era surgeons who were convinced there should be a set ratio between red blood cells and plasma for every patient in the OR. “It took a generation to get rid of that idea, but now we’re kind of back to where we started, with trauma surgeons in the ORs demanding that we reconstitute whole blood.” He’s not convinced the idea, even though it makes some sense, is being evaluated properly.

To keep wastage and utilization rates under control, trauma pro­tocols need to be monitored, he stressed. “To avoid collateral damage in transfusion protocols, clearly you need close coordination with the trauma team, the transfusion service, and the blood supplier, and I believe you need to limit decisionmakers very severely.” Point-of-care or near-patient testing should also drive decisions, he added. “The fact is only five percent of trauma patients present coagulopathic, so why give 100 percent of these patients this aggressive therapy?”

It’s inherently hazardous to transfuse large amounts of blood through patients’ veins, Dr. Hannon noted, citing the risks of TACO (trans­fusion-associated circulatory overload), associated with one in 350 transfusions, and TRALI (transfusion-related acute lung injury), which causes perhaps 500 deaths a year in the United States. “You know that; your blood bankers know that. It’s the people who order and administer the blood who do not know.”

He compares trans­fusions to che­motherapy. “Chemotherapy causes harm in every patient we treat, and blood products also cause some degree of harm in every patient we transfuse. We know in chemotherapy that if we give the right patient the right dose at the right time the benefit will exceed the harm. It is the same with transfusions—we need to balance every decision very carefully.”

A key problem is that patients’ immune systems change when they are transfused. “So unnecessary transfusions unnecessarily generate antibodies and make it challenging to cross match the next time the patient needs blood.” In addition, the immune system begins to be down-regulated. “The more we transfuse, the more we change the ratio of T-helper to T-suppressor cells. As a liquid transplant, when we receive blood our immune systems adapt; by tolerating these transfusions we’re able to use that blood. But unfortunately, this immune tolerance also makes us tolerate things like bacteria, viruses, and tumor cells.”

It’s a counterproductive but common fact in the field that many hospitals still fault doctors for single-unit transfusions, Dr. Hannon noted. “But the issue for nurses and physicians should be, if there is a difference in outcomes between one unit of blood and two, you should never ‘double down’ on blood products.” Even in the case of multiply transfused patients, where it might seem less important, “if the patient has received 30 units of blood, I would argue that you should think carefully about the 31st. And believe it or not, it’s very empowering to nurses and physicians to know that very little things like that can make a difference.”

Another high impact choice is to reduce the amount of blood drawn from patients in the ICU. “The average ICU patient is losing 40 to 70 ccs of blood per phlebotomy, so it adds up if they are sitting in the ICU for three or four days. Controlling that quantity can have important effects on patients’ lives.”

More and more hospitals will need to be looking into blood management programs like St. Vincent’s, Dr. Hannon stressed. In hospital accreditation and regulation, “there are a variety of risk management issues that have probably always been there—but they’re now being elevated.”

The blood bank, of course, is already fully familiar with regulatory standards. “Typically blood banks have a done a fabulous job of being on target because they’re always being inspected. The problem is that on my side of the blood bank window it’s the Wild Wild West. To date the clinical areas have been poorly regulated, but as early as next year, when blood management performance measures are added, blood issues in the hospital, which have always been near the bottom of the Joint Commission’s checklist, will move to the top.”

Among the major contributing factors to this perfect storm are the “never events” that were enforced as of Oct. 1. These are 28 events that should “never happen,” under new Medicare policy. And 11 of those 28 events will no longer be paid for.

These events started with fairly logical things, such as “never operate on the wrong patient” and “never administer the wrong unit of blood or wrong medication.” But they now include never having a bleeding complication or a hospital-acquired infection, Dr. Hannon noted.

“The hospital used to be able to upcode and charge for these events; it could enter a new diagnosis for a complication.”

But under this “stroke of evil brilliance by CMS,” as he puts it, events like an air embolism, which can occur during a massive transfusion, will not be paid for. Nor will bleeding complications in cardiac surgery, even though the Society of Thoracic Surgeons says good cardiac surgery programs have a bleeding complication rate of about two percent.

The true cost of transfusions is becoming clearer and inviting more “behavior modification” in the hospital, Dr. Hannon noted. It’s not simply the average charge from the American Red Cross and the independent blood banks—now $200 and up per unit of red cells. “The blood collection centers can pass that on, though we as hospitals can’t absorb those costs readily, since most patients we manage are under fixed-cost contracts or DRGs.” But the purchase cost per unit is only the tip of the iceberg. “When you actually transfuse the patient, you have to account for medical technologist time, RN time, reagents and supplies, plus allocated overhead. So the cost per unit is more like $700.”

Those numbers get much bigger when a blood transfusion causes adverse outcomes. For example, keeping a patient on a ventilator for a single day costs $4,800. Treating an infection: $28,000. Postoperative cardiac surgery bleeding: $30,000. “If you amortize the cost of these events across all the units of blood transfused, the cost is closer to $2,000 per unit.”

How much can these costs be modified? Not very much on the supply end. “If you really yell and scream you might get a two percent price break from your blood supplier.” For the hospital, when it curtails overtransfusing, the potential savings in dollar resources are far larger, perhaps 20 percent. According to some estimates, the blunt facts are these: Each unit of allogeneic transfusion increases the odds for nosocomial infections by 50 percent. So a two-unit transfusion doubles the risk of a hospital-acquired infection. It’s a reality that hospital administrators should be aware of, he said.

A comprehensive, strategic approach to blood management has three pillars, Dr. Hannon believes: broad educational efforts, change management strategies, and metrics with accountability.

On the education front, “doctors and nurses need to know the ‘why’ before the ‘how’: why blood management is important. The short answer is not that unnecessary transfusions are wasteful or expensive. Unnecessary transfusions are harmful to our patients.” Metrics are necessary not only to measure improvement but to engage hospitals in the process. “Once hospitals are convinced things are maybe not as good as they hoped, that gets them engaged in the change process.”

“You also need a strong overlay of change management,” said Dr. Hannon, a systems-based proactive approach, to change practices in a logical fashion. “The overall strategy is really not to eliminate transfusions per se, but to make sure when a nurse walks in a patient’s room that every transfusion is an appropriate decision.”

From his standpoint, the backbone of any blood management program is a strong transfusion committee. The process “must always be multidisciplinary,” he stressed. “This is not an anesthesiology, laboratory, or surgical issue; it’s a hospitalwide issue that has to join all stakeholders in this process.”

The integrated blood conservation program at Houston’s St. Luke’s Hospital was implemented over six months beginning in November 2007, and Dr. Bracey has already learned that pathologists are critical to educating users about its importance. “You won’t have to go very far in the hospital to find out just how little doctors and nurses know about it. It’s really striking how few of our clinical colleagues ‘get’ the importance of blood conservation.”

As pathologists, who serve as stewards of the blood supply, “we are really well situated to do something about this,” Dr. Bracey said. “In the past we focused on making sure the blood is safe, the safest product available. What we now have to do is go beyond the walls of the laboratory to look at the administration of blood and monitor it so we can assess outcomes.”

This “biovigilance,” as it is termed in Europe, is a new territory for pathologists but one for which they are uniquely suited. “CAP was one of the first organizations to put forth guidelines on blood utilization, an important part of blood management, back in 1989, and pathologists through the transfusion committees are in a very good position to work with their colleagues on key issues in blood inventory, blood ordering, and storage. All of these are important elements in managing the blood supply.”

As Dr. Bracey’s recent experience with disaster recovery in Houston has confirmed, demand and supply are critical factors. “There’s been a 30 percent increase since 1999 in the amount of blood used across the country, so there’s a tremendous demand. But we don’t have an endless supply of donors, so it’s very important that the blood we have is used effectively and efficiently,” he said.

Only about 37 percent of the total population is eligible to donate blood, and on an annual basis only eight percent actually donates. Many of the nation’s blood centers on average have about one day’s supply of whole blood available. “So we’re not swimming in blood,” Dr. Bracey notes. In disasters, “if you don’t have adequate inventories for four or five days, since donors will dry up, you can swiftly be sucked into a situation where you don’t have adequate supply.”

Underscoring Dr. Hannon’s message, Dr. Bracey said the Joint Commission’s involvement in blood management has finally brought the issue into the mainstream. “Any of you who have worked in a hospital know that hospitals snap to it when JCAHO becomes involved,” he said, citing the effectiveness of JCAHO’s recent emphasis on outcome measures involving deep vein thrombosis prophylaxis and prevention of bedfalls. “So blood management will become a major focus that we can’t ignore, and I think we as pathologists are well positioned, because we have the data to assess how we are using the blood.”

In canceled surgeries alone, hospitals can be hit hard by blood shortages. “In most cases, suppliers come through 90 percent of the time, but one canceled case has a tremendous economic impact on your hospital.”

More important, however, are the patient outcomes following transfusion. An enormous study in the United Kingdom (Murphy GJ, et al. Circulation. 2007;116:2544–2552) looked at the most important outcome—survival—in individuals who were transfused during their first cardiovascular surgery and those who were not, Dr. Bracey noted. “You can see early on, and continuing over the years, that there’s a separation in the survival rate.” Those who did not receive transfusions tended to live longer.

The surgeons in his hospital would be likely to say this occurred because patients who are transfused are sicker patients. But Dr. Bracey says data from a recent report suggest that the severity of a patient’s condition may not be the sole determinant of transfusion-associated outcomes. “They bear out that basically it’s not that the individuals who were sicker got blood, but as you got older blood, the survival decreased. So this is a signal that as you’re going up in days of storage, the probability of adverse outcomes increases, and in fact you can have negative attributes in terms of patient outcomes.”

In the U.S., he says, the cost of blood is not well understood, making it “an important figure” for pathologists to work on. “At our hospital, the OR’s cardiovascular surgery recovery rooms use about 64 percent of all transfusions, so for us that area is an incredibly target-rich environment. But what we know is there is a tremendous variation in practice.” A 1991 study found that 25 percent to 75 percent of patients are receiving blood for a coronary bypass procedure, and when anesthesiologists repeated the study in 1998 and 2008, the same data were reported, he says. “So clearly people are transfusing not based on physiologic parameters but based on habit. This shows room for improvement.”

Outlining the interdisciplinary team effort that went into St. Luke’s integrated blood management program, Dr. Bracey said the working groups in cardiovascular surgery and orthopedics identified opportunities in three areas: preoperation, intraoperation, and postoperation. “We started with routine screening of patients for Plavix or antiplatelet agents when they come to the hospital. In the OR, our efforts included expanded blood recovery, increased intraoperative hemodilution, evaluating circuit size, and auditing and enforcing hemoglobin thresholds.”

The postoperative program included implementing a transfusion algorithm to avoid unnecessary transfusions, expanded blood recovery efforts, plans to mitigate diagnostic blood loss through phlebotomy, and monitoring of outliers.

Through individual monitoring of surgeons’ blood use, “we saw a way to shock the surgeons into a sense of competition,” Dr. Bracey said. “We looked at blood use for aortocoronary bypass surgery of 11 surgeons operating in our facility, and there was extreme variation. One doctor used 0.8 red blood cell units per case, while one outlier used 13 in a single case.” The list was distributed to the surgeons. “We didn’t give the names, but this really sort of sparked the surgeons into thinking about blood conservation in a major way.”

The hospital also started active use of Cellsaver much more routinely because there was considerable variation among the surgeons in applying this technology. “After implementing this very simple step, we found a profound drop in the rate of blood transfusion. Intraoperatively, in 2006 transfusions dropped from 49 percent to 15 percent, and that carried over to postop with a 51 percent decrease in blood utilization.”

The results show that when pathologists take the data on blood use back to the users, and discuss variation in transfusion practices, the users become cooperative over time, Dr. Bracey said. Even so, when St. Luke’s compared its results to those of others in the University Healthcare Consortium, which the hospital recently joined, “we found there’s additional room to go.” That is another database Dr. Bracey recommends using as a tool in managing blood.

Around the time of the influential TRICC trial (Hébert PC, et al. N Engl J Med. 1999;340:409–417), which found no improvement in ICU patients treated with an aggressive transfusion strategy, St. Luke’s published a study of its own (Bracey AW, et al. Transfusion. 1999;39:1070– 1077) showing that individuals could tolerate hemoglobins of eight if they met four criteria: if they were first-time cardiovascular surgery patients, less than 65 years of age, with no congestive heart failure and hemodynamically stable. “But despite our own study, our patients were still being transfused with hemoglobins of 10, because that’s what somebody taught all our internists and all our cardiologists. Plus there’s a tendency to transfuse two units, because people like even rather than odd numbers.”

The pathologists took data from these trials back to the surgery section and developed an RBC transfusion protocol that called for transfusing one unit of RBCs if the patient has a hemoglobin of less than eight, then reassessing the patient. It also limits notification. “If a nurse calls a resident at 3 AM and says I have Mr. Jones with a hemoglobin of nine, Mr. Jones is probably going to get a transfusion, so what we did was restrict routine protocol-driven notification to the clinician if the hemoglobin was less than eight in stable patients. This is something we just started, and in time we hope to be able to show its impact on our transfusion rate.”

Daily transfusion rounds are another way the hospital prevents excessive transfusions. “Last Friday there was a renal failure case with a valve replacement who was already sitting in a chair eating breakfast,” Dr. Bracey said. “His hemoglobin was 8.6 and there was an order for two units of RBCs. But since we have a program where the nurse is going around the unit checking the orders, we were able to interdict the transfusion. So having a prospective review of transfusions is very important.” The program also sends an e-mail about transgressors to the cardiovascular surgery champion on the team, who helps by educating the surgeons on the transfusion guidelines.

Near-site testing is another important component of controlling blood use. “In our hospital, we have a near-site system that tests all patients with microvascular bleeding, applies PTT ratios, and assesses platelet function,” Dr. Bracey said, recommending that pathologists support such testing, as well as phlebotomy audits. “If we eliminate arterial lines or discards, eliminate standing orders, and use smaller tubes, we can clearly reduce blood utilization.”

The integrated blood management program is also tackling blood use in orthopedics. A substantial percentage of patients presenting for orthopedic surgery are anemic, and preoperative anemia is the most likely reason for a patient receiving a blood transfusion, Dr. Bracey said. “So we have a protocol in place where we look at iron stores and we will treat patients to get their hemoglobin up. We just got this element of the program approved by our orthopedic surgeons, and we hope to show it has helped avoid unnecessary transfusions.”

Improving patient and hospital staff awareness of blood conservation through brochures and online information is a final component of the program. “You have to let your patients know what it’s all about as well as the staff. It’s great to have a program, but if you don’t market it, people won’t use it,” he said.

Drs. Bracey and Hannon emphasized that pathologists must continue to play the lead role in blood management programs. “You have the information about how much blood is used and who’s ordering it, and we traditionally, as a specialty, have developed the guidelines for transfusion,” Dr. Bracey said. “So if we partner with our clinical colleagues in surgery and anesthesiology, we can really have great success in moving this important initiative forward.”