Since its emergence in the early 1990s, tight glycemic control could almost have been called an orphan protocol. Despite evidence that it sharply lowers patient mortality and morbidity and hospital length of stay, the use of insulin to keep patients blood glucose at or near normal levels has spread slowly beyond a few pockets of fervent support, and has yet to sweep the nation’s hospitals. As Curtiss B. Cook, MD, associate professor of medicine in the Division of Endocrinology at Mayo Clinic Arizona puts it: "A lot of people haven’t quite bought into the concept yet."
But in recent weeks, with a major new controlled study and a public statement from leading professional associations in diabetes care, tight glycemic control may have reached its tipping point. The American Association of Clinical Endocrinologists and the American Diabetes Association are now advocating widespread adoption of tight glycemic control protocols not only in intensive care units but also in other areas of the hospital.
Their position statement, released Feb. 1, calls for implementing "structured protocols for aggressive control of blood glucose in both intensive care units and other hospital settings."
One day later, Belgian researchers led by Greet Van den Berghe, MD, PhD,
reported in the New
England Journal of Medicine (354:449-461) that intensive insulin
therapy, or tight glycemic control, significantly reduced morbidity among
all patients in the medical ICU—a benefit that was demonstrated
earlier on patients in surgical ICUs.
The findings are not quite as exciting as those from the earlier study by the same research team, says David Baldwin, MD, director of endocrinology at Rush University Medical Center, Chicago. But they are likely to ramp up enthusiasm for making tight glycemic control the standard of care.
That trend already got a shot in the arm last July, when LifeScan Inc., a Johnson & Johnson company that makes blood glucose testing systems, released its "Tight Glycemic Control Adoption Report" at the American Association for Clinical Chemistry annual meeting. The report’s central message: Tight glycemic control is shaping up as one of the most cost-effective inpatient interventions ever studied.
In a pharmacoeconomic analysis of the critically ill patients in Van den Berghe’s first study, Life Scan reported the total cost of care was more than 30 percent lower for those on a tight glycemic control protocol compared with conventionally treated patients. A meta-analysis of four peer-reviewed studies estimated that the intervention cost only $7,000 per life saved.
LifeScan also noted that its 2005 survey found the number of its customer hospitals using tight glycemic control protocols had jumped to at least 320—more than triple the number of hospitals that had reported employing the protocols one year earlier.
One thing is clear, says Atul Malhotra, MD, of Brigham and Women’s Hospital
and Harvard Medical School, in an editorial accompanying the Feb. 2, 2006
England Journal of Medicine article: "The days of ignoring blood
sugar levels or tolerating marked hyperglycemia in the ICU (which was
commonplace even five years ago) are over."
"In the past, most doctors believed it was okay to let somebody’s sugars run way up in the 200s if they were in the hospital and it wouldn’t matter," Dr. Baldwin says. "A lot of people were suspicious of that general philosophy, but there wasn’t any really good prospective controlled study with very compelling results."
Once the Belgian clinical trial was published (N
Engl J Med. 2001;345:1359-1367) and Rush started tight glycemic
control, the issues became more ones of logistics. "To give IV insulin,
you almost have to have the patient in the ICU because it requires measuring
blood sugar every hour. In the ICU you’ll typically have one nurse per
one or two patients, and they have enough time to accommodate the intense
monitoring of blood sugar."
"But most hospitals can’t afford to have so many nurses out on the floor, and can only do tight glucose control in the ICU, although a few centers have step-down units between the ICU and the floor."
What Rush did was end the "sliding scale" method of glucose control on the floor and start training the internal medicine residents in tight glycemic control. The belief was, "We can’t send board-certified doctors in internal medicine out into the world without being competent to deal with the worst disease in America—diabetes mellitus—which will be the dominant disease for the rest of their careers."
Until recently, Dr. Baldwin notes, there weren’t any good tools for tracking the impact of tight glycemic control except through getting the data by hand. But significant software improvements are changing that.
For example, with the Abbott software package that manages the Rush University Medical Center blood sugars, "now we have the data sliced and diced any way you want, by floor, by doctor, over the whole institution, including exactly how many sugars are less than 70, or from 120 to 160, and so on. It’s amazing; you can really drill down and see how units are doing from month to month and where to focus attention to make quality improvement."
Pinnacle Health System, a four-hospital health system in Harrisburg, Pa., has found similar benefits from Medical Automation Systems’ Rals-Report Glucose Benchmarking Service.
Since 2002, Pinnacle Health has steadily closed in on its goal to get patients’ blood glucose down to 110 mg/dL for the ICU and below 150 for the rest of the nursing units.
Today, "90 percent of the time when insulin orders are written, they use our preprinted supplemental order form," says Amy Helmuth, RN, MS, ONC, Pinnacle Health’s director of performance improvement. In the cardiac ICU, 99 percent of patients are managed on an insulin infusion to keep glucose under control.
With more than 3,000 diabetic patients in the system each year, the protocol, developed by Pinnacle Health endocrinologist Renu Joshi, MD, is used for most diabetic patients, she says. It requires that all diabetes patients be on basal insulin or oral medication depending on their blood glucose while they’re in the hospital.
Parts of the project have been easy, Helmuth reports. For example, "three years ago we were concerned that physicians would continue writing their own supplemental insulin orders, but there’s been overwhelming compliance with the preprinted form." It took physicians only three months to get to 90 percent compliance with using preprinted orders. But getting nurses and physicians to understand that a blood sugar of 89 is not something to worry about and that 160 is not a good blood sugar has been one of the challenges.
The Rals-Report shows where progress stands, Helmuth says: an average blood glucose level for all patients of 154, and for the diabetic population an average of 167 to 168, which is consistent with American College of Endocrinology guidelines. "We’re trying to get that number lower, obviously, but we started with an average of nearly 200. Our glucose control in the ICU has improved to an average of 134, but it could be better. We’d like to see a higher percentage of patients under 110."
The patients themselves haven’t posed a problem. "The patients seem to be tuned in to it. They make comments like, ’I don’t usually take insulin at home for that level.’ And the nurse will say, ’It’s important for good healing, we’re trying to get your blood sugar down lower than at home—it gets you out of the hospital faster.’"
The nursing staff has been positive, Helmuth says. "Our diabetes clinical initiative has many staff nurses on it, not just managers. That’s been one of the keys to our success, because the physician writes the order—but the nurse needs to implement it. If you know the patient’s blood glucose is still high, you need to make a phone call to the physician to switch to a different regimen."
The forms are reevaluated and "tweaked" every six months, Helmuth says. But having the preprinted forms means there are no problems with illegible handwriting, and nurses like knowing there are standard insulin protocols. "It is easier for the nurse who knows that all patients on what we call the usual protocol receive four units when the blood glucose is 155 versus having a bunch of diabetic patients with different orders."
It’s taken a considerable amount of education of the nursing staff just to change the idea of what a normal or desirable blood glucose level is, she adds. "Five years ago nobody worried about a level of 188. That was okay. So just changing that mindset takes a lot of work. If the average blood glucose in your organization is 200, don’t think that overnight you’ll get it down to 110. You have to take baby steps."
The kind of diplomatic mission required to gain cooperation on tight glycemic control requires tenacity, she admits. Pinnacle Health launched a preoperative insulin protocol about a year ago to increase the number of insulin infusions used for bigger procedures. But the main challenge has been getting anesthesiologists to sign on.
"Anesthesiology was hesitant to use insulin drips intraoperatively. They are not comfortable finessing the insulin drip. It’s more work for them to manage, and it’s not something they were taught in school. They fear they won’t be able to manage it properly, the patient will be hypoglycemic, and there will be a negative outcome."
"It is adding one more thing; it’s not just a matter of hanging an insulin drip and not worrying about blood glucose levels. But we’ve been able to demonstrate that we haven’t had a negative incident with tight glycemic control if the protocol is followed properly, and they need to get used to the idea."
Each physician receives quarterly feedback on glucose control practices. "We discuss tight glycemic control at our quality council, but we also send out letters to each physician with their performance data on key quality indicators like glucose control, AMI, and heart failure."
Usually they’re very receptive, she adds. "They’ll say, ’Thank you for letting me know. I didn’t know I wasn’t doing this,’ and they’ll ask for tools to help." Sometimes they’ll explain why they didn’t do something, and that’s helpful, Helmuth says, because then she can identify the "barriers to performance."
Dr. Cook, who has been extending tight glycemic control to patients at Mayo Clinic outside critical care since January 2004, says the recommendations very likely apply outside the ICU. "Most of the randomized controlled studies have been in the ICU environment, but our feeling is that high blood sugar is bad whether you’re in the ICU or not."
At Mayo, a pilot study is in pro gress that will document and quantify what the perceived barriers are to tight glycemic control. "To my knowledge, nobody’s formally explored the barriers to glucose management in the hospital," he says, noting that Mayo Clinic has just completed its preliminary survey of its residents, and he is developing a series of lectures targeted at teaching residents the importance of tight glycemic control and giving them tips about how to achieve it.
"In addition, I’m having discussions with the information services people on developing glucose "report cards" for various specialties that see patients in the hospital."
Dr. Cook is cautious about advocating one protocol over another. "There are a number of different protocols out there that achieve the same thing. One protocol really hasn’t been compared with another. Getting glucose down is the goal, and whether one way is better than another is not necessarily the critical issue."
Kathryn Jeanne Zerr, RN, MBA, coordinator of cardiac information for Providence Heart Services in Portland, Ore., worked with the team led by endocrinologist Stephen Bookin, MD, and cardiovascular surgeon Anthony Furnary, MD, to develop the Portland Protocol at Providence St. Vincent Hospital.
When it was implemented in 1991, "it was considered radical," she says. "Even our director of critical care was concerned about its potential for causing hypoglycemia. That’s why we utilized it first in critical care, as the nurse-patient ratio was much tighter there."
Her hospital didn’t set out to be on the cutting edge, she says. "We were just responding to a known infection problem in diabetics who had heart surgery. Dr. Bookin suggested that more-aggressive control of blood sugar would be helpful, so we developed the insulin protocol."
Over the years, the protocol has become much tighter. "The first iteration was not as aggressive. The acceptable blood sugar levels were much higher than what we try to achieve now—but we had to have a [gradual] process that allowed the clinicians to become more comfortable with the concept."
Other areas in the hospital have been slow to adopt the protocol. "Among the clinicians, there’s still some reluctance. It’s pretty well accepted that in cardiac surgery this is of proven benefit, but for patients in other sections of the hospital, it is not considered to be the standard of care. Physicians always want to see hard evidence before they change their practice; they’re not easily swayed without data. They do what they feel is safe for patients."
Have the fears of adverse outcomes from hypoglycemia become reality? "People with hypoglycemia get confused and can go into shock if it goes too far. But we’ve never had that happen," Zerr says. "If the blood sugar dips down, it’s caught quickly, and we haven’t had any adverse outcomes."
It doesn’t appear to Zerr that new residents are up to speed on the benefits of tight glycemic control, but the research in progress at her hospital has a physician-training component attached to it.
"This is a project that depends on cooperation and coordination of physicians and nurses," Zerr says. Before the Portland Protocol was implemented, the insulin orders were written as a sliding scale, "a reactive response rather than the proactive response we have now," she says.
Getting a tight glycemic control program in place can be frustrating, but the payback is worth it, says Amy Helmuth. Since Pinnacle Health started its protocol in 2002, "it’s changed six or eight times as we looked at outcomes. It takes a lot of time to change practices and people. Just set a goal and work toward that goal, and keep revising your practices to achieve it."
As for the long-term prospects of tight glycemic control, those hoofbeats you hear in the distance are the Joint Commission’s, Dr. Baldwin believes. "This is the way of the present and the way of the future, and anybody not doing it now had better start getting their act together."
He doesn’t know if it will be in two years or five years, but he predicts it will be mandated.
Anne Paxton is a writer in Seattle.