Feature Story

Getting the scoop on nutrition testing

Physician nutrition care communication form

January 2002
William Check, PhD

It was a routine wager between a physician and a laboratory administrator. The physician declared that a patient weighing more than 350 pounds could not be malnourished. Elia Mears, MS, MT(ASCP), SM, suspected otherwise.

Mears, laboratory director at the Leonard J. Chabert Medical Center, Houma, La., was low-key. "I said, ‘What if I can prove to you he is malnourished?’ He said, ’There is no way he can be malnourished, just by his size. I’ll eat my hat if you prove that.’" In fact the obese patient was malnourished—chronically short of protein—and thereby at risk for a variety of complications.

The wager in Louisiana was not settled on the basis of appearances or hunches. It was settled with hard data from the medical center’s laboratory, a strategy Mears has been promoting for almost a decade. The test that resolved the debate: prealbumin.

As Mears and other researchers have affirmed, prealbumin is the best available marker to monitor malnutrition and how successfully it is treated. It’s synthesized in the liver and has a two-day half-life, so a physician and dietitian can tell if a patient’s nutrition status improves or deteriorates over a short time. Prealbumin doesn’t appear to be affected by a patient’s hydration. Its most attractive trait as a marker is how quickly it rebounds after 55 percent of a patient’s energy and protein needs have been satisfied.

Prealbumin is a precise measure in the sometimes subjective world of nutrition. "If we see that the prealbumin does not change or is below 11 g/dL," Mears says, "it gives the dietitian a chance to re-examine the entire patient, the nutrient concentration that they need in a very timely fashion. They can look at it and not waste our health care dollars by doing something three, four, or five days later." A solid numerical value also helps Chabert Medical Center’s dietitians prioritize their day: They see the least-nourished patients first.

Thanks to such efforts, patients are discharged sooner and healthier. Using nutrition testing, Mears has cut the overall hospital length of stay by two days. "What was more surprising," she says, "was that for patients at severe risk of malnutrition, we reduced length of stay by 12 days. I am talking about critical care patients, some of our ICU patients, who are there for a long period of time."

Hospital administrators have caught on. To them, nutrition testing by the lab makes sense. Inexpensive screening makes Chabert patients so much healthier that the hospital saves an impressive $635,000 annually. That includes revenues to be had with proper Medicare coding for malnutrition, sums in the range of $20,000 to $40,000 annually. "You’ll save dollars and be a hero to the community," Mears promises.

At least one physician sees her that way. Internist Thomas G. Ferguson, MD, chief of medicine at Chabert and a 24-year veteran of the hospital, says Mears’ tenacity and persistence are part of the reason she prevailed. "She kept rattling and showing data, and finally showing data that got everybody’s attention," Dr. Ferguson says.

But what clinched the case was the evidence. "It’s hard to argue with medical data and fact," says Dr. Ferguson, who can recall the days before routine nutrition testing. "We can show better outcomes—and it’s cheaper." The intensive-care specialist concedes that medical students are not overwhelmed with nutrition course work, but he himself sounds like a convert: "I can’t imagine doing without nutritional testing. It is crucial to patient outcome. You are what you eat."

In the end, he says, the change in physician practice patterns was not disruptive: "It’s easy to add a couple of markers with your routine blood work. Now you have very objective measurements of nutritional status."

Part of the shift needs to be attitudinal, Dr. Ferguson suggests. "A doctor has no qualms about testing someone for diabetes or hyperlipidemia. That’s routine," he says. "If you put malnutrition as a risk factor in the same league as diabetes or hyperlipidemia or hypertension, then you’re going to want that objective marker. You’re not going to want somebody up there squeezing a tricep fold or measuring abdominal girth."

The biological roots of the savings are not mysterious. Patients who have more protein available heal faster. They have fewer infections; they are less likely to be readmitted. Ideally, Mears says, patients in need would be given extra protein by dietitians, nurses, or pharmacists providing nutritional intervention. And the physician’s decisions about which patients to supplement, in turn, would rest on data from the laboratory.

Mears, in short, is advocating a little-known strategy to depart from more subjective techniques to measure malnutrition. (Her ideas are not new: Some of the articles in the literature are 25 years old.) Mears is proposing precise, quantitative nutrition standards that are monitored by the lab and implemented by a multidisciplinary team of physicians, nurses, pharmacists, and dietitians.

For those outside the lab, the billion-dollar clinical issue is protein-calorie malnutrition or protein-energy malnutrition. A body of 150 papers links PCM and PEM to increased morbidity, impaired wound healing, and compromised immune response. The condition may affect 30 to 55 percent of hospital patients. "We have to realize it is not a rarity," Mears says. Any patient admitted by internal medicine, surgery, or the ICU is at risk.

But neither PCM nor PEM has been studied as exhaustively as, say, heart disease. Good outcomes research in nutrition testing and intervention has not been done, Mears acknowledges. One reason: A control group would be needed, and no cohort of patients could safely be deprived of food.

For Mears, the hardest part of what she proposes may be calling attention to the lab’s pivotal role. "The lab has never really been included in this whole process," she says. "Without the numbers we offer, clinicians have no objective data. Our services and support are just as important as the nurse who can take the temperature or the physician who can offer the antibiotic."

But how can the remainder of the hospital staff be coaxed into recognizing the lab’s role in monitoring malnutrition? Education is part of the answer: Mears briefs incoming medical residents every month. New policies also help. In 1994, Mears was allowed to amend the Chabert Medical Center bylaws to require that most adult nonmaternity patients be screened for malnutrition on admission.

She recalls being rebuffed more than once. It never bothered her. "We are here and we are an integral part of the health care for this patient," she says. "We’re not just trained to run an instrument. We have education and a lot of knowledge we can share."

That knowledge draws on some of the articles in the literature that support the conventional view that the elderly and poor are at greater risk for malnutrition. But the literature also overturns other preconceptions. PCM, for example, is not confined to the elderly, or to the HIV positive, or to those with kidney disease. Young, healthy patients can be malnourished. Only testing can reveal who needs a protein boost.

Why does protein malnutrition persist in a well-fed nation? Largely because of the so-called anthropometric method of identifying it. "It’s very subjective," says Mears, citing the pinching of skin, the time-consuming questionnaires, the measuring of the patient’s upper arm. The anthropometric technique is common—and makes sense—in Third World nations when relief agencies and charities try to assess mass starvation.

But in the United States, Mears contends, the method should not be considered part of modern medical practice. "If you ask any individual, ’Are you eating appropriately?’ the reply is, ’Yes, I eat.’ Most of the time, the elderly patients cannot answer the question appropriately. You get ambiguous responses, general responses that do not allow the dietitians to get an accurate picture of what is going on."

Enter the laboratory and more scientific methods. Mears’ lab already had a nephelometer to do prealbumin testing. With a small sum of money, she started out with a pilot study at Chabert, a 147-bed facility affiliated with Louisiana State University and located 50 miles southwest of New Orleans.

Those early results indicated a significant share of Chabert’s patients were malnourished: 65 percent. "Granted, our hospital is in an area of south Louisiana where a lot of patients do not readily seek health care," Mears says. Still, it was an eye-opener. From the literature, she expected half to be malnourished. Demographics, she believes, was a piece of the puzzle. "We have a lot of indigent patients," she says.

The laboratory was able to perform the prealbumin analysis in-house, which came as a surprise to Chabert’s dietitian. "They had read about it, but thought that the laboratory in-house could not perform it," Mears says.

Mears’ data showed that albumin testing alone would not have identified patients at medium or high risk for malnutrition. Only prealbumin would have worked. "Those [albumin] patients would have fallen through the cracks," she says. "They would not have had the supplementation and intervention they needed. That scared the dietitian. She didn’t realize such a high percentage of our patients were at risk."

At Chabert, the prealbumin work in the lab costs $3.13 per test, assuming 6,000 tests are performed annually. The potential savings are especially appealing for a hospital in which at least 30 percent of the patients lack insurance.

Oddly, the physicians were the only real obstacle to prealbumin testing. "The physicians were the hardest ones to convince," Mears says. Part of the resistance was logical. Why bother to test if the result arrives too late for clinicians to act on it at the bedside? Says Mears, referring to nutrition testing done by reference labs: "If you are dealing with a result that is four days old, what’s the point? It has already changed."

But there was more to physician resistance. "Physicians do not get the background they should," says Mears, referring to the absence of nutrition education in the medical curriculum.

Territorial or psychological prerogatives also came into play. Mears pauses. "I love physicians to death. I come from a family of physicians, but they are very hesitant in having a nonphysician tell them how to practice medicine. They don’t take being told what to do very well."

A tip for any lab is to not merely deposit in a chart a warning that a patient is malnourished. Such a note subtly pressures a physician to take action and probably will not be well received.

Instead, nutrition testing and treatment options, and the benefits of both, need to be explained. In that way, Mears says, doctors can come to understand that they still control their patients’ destinies. The role of the lab and the dietitian is to provide solid, timely data and a range of options in a field—nutrition—that physicians may approach with shaky confidence.

Once the physician is onboard, time is critical. The earlier the laboratory can supply the patient’s status, the earlier the surgeon or internist can make decisions about nutrition. And the sooner a physician can make the decision to supplement a patient’s protein, the less the hospital stay will cost. A nutritional milkshake is less expensive than a feeding tube, which is cheaper than a central line. "If you can address the patient with an enteral nutrition product, or something that is oral," says Mears, "you’re going to be saving big bucks. If these patients are addressed earlier, even an ICU stay will be less."

If cost and quality are not sufficient incentives, there’s one more reason to do nutrition testing: accreditation. The Joint Commission on Accreditation of Healthcare Organizations requires that patients be assessed for malnutrition within 24 hours of admission, but the rules do not specify the techniques. Says Mears: "They’re not telling you how to do it; they’re telling you you have to do it. So everybody should be doing it if they’re Joint Commission-accredited. But you can be doing it à la 1960s or 1970s and you’re okay."

In Delaware, Mears’ ideas have long been appreciated by registered dietitian Linda Brugler. Brugler, manager of medical nutrition therapy at St. Francis Hospital, Wilmington, is not shy about saying she was the first to replicate Mears’ results.

Brugler, in fact, has done more than follow Mears’ work. She has expanded on it, publishing prolifically on her own. Her nutritional interventions have saved the 150-bed St. Francis Hospital about $1.2 million annually, or about $1,000 per high-risk patient. In the two most severely malnourished categories of patients, mortality dropped 57 percent and 44 percent with laboratory-enabled nutritional intervention. Major complications fell 77 percent; the readmission rate within 30 days dropped 57 percent.

"By treating these deficiencies," says Brugler, "patients have fewer complications. They have better wound healing. You’re able to get these people well and moving on to the next level of care."

At first, Brugler encountered as much noncooperation as Mears. The morning after Brugler began inserting notes about nutrition risk in patient charts, she got a call from a distressed physician. He said, "It looks like you were on your broomstick last night, flying through the hospital."

Years later, Brugler laughs at the memory. "He was uncomfortable with what we were doing," she says. "The concern was, ’You have said my patient is at risk for malnutrition. That requires me to do something, but I’m not sure what to do.’ Other physicians had similar concerns." As she sympathetically explains, patients often must not eat prior to tests or surgery. The doctors are in a bind: A patient must not eat. Likewise, nutritionists must dutifully file dire bulletins about protein deficiencies that sometimes cannot be addressed until much later.

Making matters worse was that, in the early days, St. Francis’ prealbumin tests were sent to a reference lab. "The results were coming back maybe three or four days later," Brugler says. "That wasn’t doing our patients any good. It wasn’t timely, and it wasn’t used because of that. So we had to approach the lab with the idea of bringing the test in-house. We worked out the cost-effectiveness of that. They brought it in on a limited basis, leasing the equipment initially and later transferring the method to a previously purchased instrument." In-house, a prealbumin test costs $5.50, a significant reduction from what the reference lab charged.

Once the prealbumin data were more current—same day instead of three or four days old—the next order of business was to reassure the physicians. "If you wrote malnutrition in the chart, the doctors would think it implied they were doing something wrong," Brugler says. "That’s not true. There are medical situations and conditions that are caused by illness that affect a person’s ability to eat enough. It’s not the doctor’s fault; it’s not the patient’s fault."

Part of Brugler’s success was simply accepting the slow pace of change. "Practice patterns change because relationships develop, and you show physicians you’re here to help them," Brugler says. "They’re still leading the care team—they’re coordinating the care. However, nutrition care can be facilitated by other caregivers."

To help cement those relationships, Brugler and her colleagues conducted focus groups with physicians to determine what they wanted. "They [the physicians] came up with a solution that has worked extremely well. They said, ’We want the down and dirty. We want you to tell us exactly what you want us to do. And we want you to put it in the chart so that we cannot miss it.’"

Presto! A bright green form was born. The form (page 68) has easy-to-check boxes that show doctors what the nutritionist is recommending.

Brugler concedes physicians may have skimmed or skipped nutrition notes in charts in the past. Now they rarely do, partly because the forms are inserted in the "orders" section of the chart, the exclusive purview of the physician.

At CAPTODAY press time, Brugler was set to start a CDC-funded study to develop a gold-standard pathway of tests for malnutrition. She is hoping to expand the study into a larger National Institutes of Health-funded project.

For some patients, a routine complete blood cell count may provide a few hints of nutrition status. Even a traditional albumin may be informative. For nursing home residents, albumin may be useful as a general tipoff to long-term malnutrition.

But as a researcher, Brugler is more interested in backup or companion tests to prealbumin. Two leading candidates are retinol-binding protein and C-reactive protein. "The retinol-binding protein has a very short half-life. It’s 12 hours," Brugler says. "In theory, you can rapidly monitor a response to nutrition treatment. If you get someone on an adequate amount of kilocalories and protein, in 12 hours, by measuring the amount of retinol-binding protein, you’ll be able to see if they are in an anabolic state or not, whether they are building protein or not. If it is not improving, you can question whether your therapy is working."

As for C-reactive protein, it can be monitored in concert with prealbumin to overcome a limitation of prealbumin. "Prealbumin is decreased with inflammatory response," says Brugler. "You might have someone on enough nutrition, but if they have inflammation because of illness, their ability to make prealbumin is suppressed. You wouldn’t want to change your nutrition therapy. You’d want to know there is another condition affecting their prealbumin."

Brugler adds that 98 percent of the adult nonmaternity patients at St. Francis undergo some form of nutrition testing within the first 34 hours of their hospital stay. That’s a big improvement from 36 percent in 1993. Because subclinical malnutrition is not physically obvious, and because her own research has convinced her she can’t pick out every at-risk patient on admission, Brugler does her best to leave no stone unturned.

"Every five days we check," she says. "If a patient doesn’t have deficiencies upon admission and is still here after five days, and we’re not seeing him or her for some other reason, we’ll check." The ordeal of being in the hospital is itself a cause of malnutrition, even for patients who are well nourished when admitted.

Brugler, like Mears, does not think age is a good indicator of potential protein malnutrition. "Our work here shows that age is not a good identifier," Brugler says. "You have people at all ages who have nutritional deficiencies, either because of economic or social circumstances that compound their health issues. If you just say everybody over 70 is at risk for malnutrition, you will be evaluating a lot of the wrong people."

Overall, nutrition testing is as much a morale-booster for the laboratory as it is for her. "Laboratorians feel disconnected from patient nutrition care," Brugler says. "A program like this brings them in touch with the patient and is a clear example of how what they do impacts the patient."

Mark Uehling is a writer in Chicago.

Mears and Brugler will conduct a workshop on nutrition testing at the 2002 American Association for Clinical Chemistry meeting, July 28-Aug. 1, in Orlando, Fla..