Physicians make new push for better glycemic control
By Bonnie Darves
Endocrinologists have for several years known about the importance of tighter inpatient glucose control to ward off hyperglycemia—not just in diabetic or critically ill patients, but also for patients throughout the wards.
The problem? Awareness of the benefits of tighter glucose control in the hospital setting hasn't reached the majority of general internists or hospitalists.
"Many patients are not being treated aggressively enough," said Helena W. Rodbard, FACP, an endocrinologist in Rockville, Md., who is past president of the American Association of Clinical Endocrinologists (AACE). "We now have clinical practice guidelines for the intensive insulin therapy in hospitalized patients to achieve optimal glycemic control—and need a coordinated effort of primary care physicians, endocrinologists and hospitalists to ensure that every patient is treated appropriately."
New protocols call for using IV insulin preoperatively and for several different patient populations, including critically ill patients, before converting to subcutaneous insulin.
Standards and strategies used to control hyperglycemia now run the gamut from traditional sliding-scale insulin regimens to protocols crafted by individual physicians.
"That has been especially true in hyperglycemia management for non-ICU patients," said David E. Bybee, FACP, an endocrinologist in Louisville, Ky. "Even when hospitals have control protocols, physicians have typically taken a reactive, not proactive stance, waiting to see if patients develop hyperglycemia instead of making sure their blood sugar doesn't spike in the first place."
But a growing body of evidence now indicates that even intermittent, short-lived bouts of hyperglycemia can add up to substantial long-term morbidity, mortality and associated health care costs. That evidence is fueling a drive toward better inpatient glucose control.
The Institute for Healthcare Improvement's "100,000 Lives" campaign, for instance, which launched in 2004 to reduce hospital morbidity and mortality, includes measures for tighter glucose control. The American Diabetes Association (ADA) is pushing for better glucose-level monitoring and management of diabetic inpatients.
And the AACE has released standards for monitoring glucose and determining when levels are high enough to warrant clinical intervention—whether or not patients are known to be diabetic. Within the next several months, the AACE is expected to release guidelines on how to implement its recommended glucose-target levels.
"Poor glucose control is a big-dollar item, and the target value should be as close to 120 [mg/dL] as possible," said Dr. Bybee, who is Governor for ACP's Kentucky Chapter and helped spearhead an initiative to improve perioperative glucose control at Jewish Hospital in Louisville. "Most of the protocols published talk about 150 as the upper limit to try to achieve—and 200 as a call to action."
For critically ill patients, the AACE targets are even lower: 110 mg/dL for patients in ICUs. (Randomized trials have associated hyperglycemia in critically ill and post-surgical patients with longer hospital stays, poor wound healing, higher infection rates and higher mortality.) For patients not in critical care, the AACE target for maximal glucose is 180 mg/dL. (See "Glucose: How low do you need to go?")
The AACE standards—as well as different published protocols, including a well-known one developed by Providence Health System in Portland, Ore.—also call for the following interventions:
using anticipatory or scheduled nutritional insulin to prevent spikes, instead of using sliding-scale insulin regimens.
using IV insulin preoperatively and for several different patient populations, including those who are critically ill, before converting to subcutaneous insulin—at least until blood sugars stabilize.
monitoring capillary glucose hourly in patients on IV insulin until their levels are stable, then every two to four hours thereafter. Critically ill patients and surgery patients with diabetes may require hourly monitoring.
Such recommendations challenge long-held perceptions on inpatient hyperglycemia, according to Lakshmi Halasyamani, ACP Member, associate chair of medicine at St. Joseph Mercy Hospital in Ann Arbor, Mich. The old thinking, Dr. Halasyamani said, was that hyperglycemia events—whether caused by stress, steroids, or insulin or hyperalimentation mishaps—were tolerable as long as they resolved before or soon after discharge. But that view has changed.
"People have under-recognized hyperglycemia's significance in how patients do after hospitalization," said Dr. Halasyamani, who chairs the Society of Hospital Medicine's hospital quality and patient safety committee.
Data now show that hyperglycemia might not only make underlying illness worse, Dr. Halasyamani said, but could affect patients' "ability to recover and recuperate, possibly long beyond the hospitalization period." The new thinking is that hyperglycemia is and should be considered an independent risk factor for adverse outcomes, and as such should be aggressively managed.
Barriers to better control
But old perceptions about hyperglycemia have been bolstered by the lack of reliable data on how hyperglycemia affects patients who aren't critically ill or diabetic—and by the historic focus on hypoglycemia as the main cause for concern following interventions for diabetes.
Some physicians continue to resist standards and protocols. They say there is scant evidence beyond observational studies that hyperglycemia has serious long-term effects on non-critically ill patients on the general medicine and med-surgical wards.
According to Irl B. Hirsch, FACP, an endocrinologist at the University of Washington in Seattle, that view is not only outmoded but also shortsighted. Tight glycemic control should be the goal for all inpatients, not just those with diabetes, he maintained, because of the growing recognition that hyperglycemia is toxic for just about everyone and that inflammation appears to be comodulated between insulin and glucose.
'It's one thing to practice evidence-based medicine—it's another to practice ignorance-based medicine.'
—Irl B. Hirsch, FACP
"We're putting more effort into caring for these patients on the ICU side because that's where we have the data," said Dr. Hirsch, medical director of the University of Washington's diabetes care center. "But that doesn't mean we should ignore the people who are not sick enough for the ICU but not well enough to be discharged."
To critics who say, "we don't have the data," Dr. Hirsch had this reply: "It's one thing to practice evidence-based medicine—it's another to practice ignorance-based medicine." Studies to date, he pointed out—including one in the March 2002 issue of Journal of Clinical Endocrinology and Metabolism—have found that hyperglycemic events increased length of stay among all types of patients, and that new hyperglycemia in inpatients with undiagnosed diabetes was associated with a 16% mortality rate.
Hospitals face big problems trying to coordinate glucose testing and insulin use with meals, ordered tests and nursing time constraints.
But convincing physicians to get on the same protocol page is only one of several major barriers. In addition to dealing with wide variations in insulin order sets, hospitals face big problems trying to coordinate glucose testing and insulin use with meals, ordered tests and nursing time constraints.
That array of systems obstacles is one reason why the ADA's 2006 update on inpatient glucose control due out this month will emphasize gradual implementation of tighter glucose-control standards. The ADA is expected to recommend that institutions that have paid little attention to protocols for glucose control should start with a conservative approach.
Slow and steady
The University of Washington, for instance, has embraced a slow-and-steady course toward implementing better glucose control protocols. Since it began implementing targets, the medical center has incrementally lowered those targets as physicians and nurses become more comfortable with the hospital's protocol and less concerned about hypoglycemia.
"It makes it easier for physicians to get to target if you just continue to lower that target," says Dr. Hirsch, whose center has used IV insulin since 1992 and—based on a protocol he and his colleagues developed—subcutaneous insulin since 2003.
At Jewish Hospital in Louisville, Dr. Bybee and his colleagues first launched a glucose-control initiative in patients undergoing coronary artery bypass graft procedures. The success in that program, he said, has prompted the hospital to "commit resources to spread the better-control initiative throughout the hospital."
And at Doctors Medical Center in San Pablo, Calif., hospital staff first implemented tighter glucose control and standardized insulin protocols in the ICU before attempting to put those standards in place out in the wards, said internist James L. Naughton, ACP Member.
"That's been working well because that effort was led by the intensivists," said Dr. Naughton, whose practice—Alliance Medical Group in Pinole, Calif.—has been rotating several office-based internists in one-week hospitalist shifts to cover inpatients.
But getting the protocols used uniformly on the wards has been another story. "What we've found is that an awful lot of doctors on the floor write their own orders, and they use sliding-scale orders," he said. He added that standardization efforts should get easier when the hospital completes its transition to a full-scale hospitalist program over the next few months.
Big institutions, big payoffs
The University of California, San Diego (UCSD) Medical Center embarked on a tight-protocol program focusing on the non-critical care wards two years ago and is still working out the kinks, according to Gregory A. Maynard, ACP Member, chief of the division of hospital medicine. But the case at UCSD was compelling enough to get everyone's attention.
"We found that glucose control in the hospital was generally very poor and that insulin usage also was poor," recalled Dr. Maynard, who is now leading a Society of Hospital Medicine group working to develop a glycemic control toolkit for hospitalists that should be available later this year. "Seventy-five percent of our patients were on sliding scale only with no basal insulin when we started."
To tackle that problem, UCSD formed a large multidisciplinary team--physicians, nurses, diabetic educators, and nutrition and food-delivery staff, as well as an informatics professional. The physicians surveyed colleagues about their insulin-use order sets and yanked all the ones that promoted improper use, a thorny proposition eventually managed with help from a staff endocrinologist who championed the initiative.
After gathering baseline data on insulin use and glycemic events, the 400-bed center focused on improving and coordinating basal and scheduled nutritional insulin. The program also included close glucose monitoring and captured data on both hyper- and hypoglycemic events.
Working with dietary services to ensure an equal carbohydrate load in each meal and coordinating tray delivery with nurses' point-of-care glucose testing times and insulin administration have been even tougher than getting doctors to give up their old order sets, Dr. Maynard admitted.
That challenge, he added, was made easier by getting all those personnel on the program's team. However, these types of coordination problems will only get worse, he said, in hospitals that are moving to "concierge-style" meal service, where patients order meals at preferred times or even customize their meal selections.
Overcoming physicians' understandable concerns about the low targets and hypoglycemia potential was also a high hurdle. But that became easier, Dr. Maynard recalled, with data: Less than a year into the project, the number and severity of hypoglycemia events were improved from baseline.
"We've found that [hypoglycemia events] have gotten better, instead of worse, with standardization and with more basal insulin," said Dr. Maynard. "That reassured a lot of people."
And within two years, UCSD went from 11% of its hyperglycemic inpatients having at least one hypoglycemic event to 7%.
"That may not sound like much," he said, "but it represents about 145 fewer patients suffering from a hypoglycemic event every year." The results are even more impressive, he added, considering that the use of regimens including basal insulin jumped from 25% to 75% in the same time period—and glycemic control modestly improved across all wards and services.
He expects even bigger improvements with the institution's next-generation insulin protocol, which is being introduced this month. That protocol calls for replacing the basic standardized orders offering scheduled basal and nutritional insulin options to a more algorithmic approach. Physicians will enter patients' weight, body habitus and eating status to arrive at appropriate "default doses" to better tailor initial and correction-dose insulin regimens.
Standardization and better insulin use are also expected to yield a host of other benefits at UCSD. For one, it will allow for a far smoother transition between subcutaneous and infusion insulin (or vice versa), and between in-hospital regimens and the transition to home regimens.
"Standardization will be helpful for all three," said Dr. Maynard. UCSD is also, he added, launching a mandatory training and certification program in insulin management for housestaff, nurses and key pharmacists.
From the endocrinologist's point of view, maintaining tighter glycemic control and standardized insulin usage are worth the effort and resources they need to pull them off.
But according to Kentucky's Dr. Bybee, tackling the issue entails a concerted approach.
"This really has to be a hospital-system effort—and if the doctors are going to order a [protocol-based] treatment plan, the hospital must have the ability to carry out that plan," he said. "That's where the problems arise."
Bonnie Darves is a freelance writer in Lake Oswego, Ore.
The information included herein should never be used as a substitute of clinical judgment and does not represent an official position of ACP.
Retrospective studies have linked high inpatient glucose levels to, among other problems, higher rates of heart failure and mortality in heart attack patients and to slower recovery rates in stroke patients. The American Association of Clinical Endocrinologists has released the following upper glycemic targets for different hospitalized populations:
- ICU: 110 mg/dL
- Preprandial in noncritical care units: 110 mg/dL
- Maximal glucose in non-critical care units: 180 mg/dL
- Preprandial in pregnant patients, pre-labor: 100 mg/dL
- One-hour postprandial in pregnant patients pre-labor: 120 mg/dL
- Labor and delivery patients: 100 mg/dL
Other AACE recommendations include:
maintaining glucose levels as close as possible to euglycemic levels by using subcutaneous therapy or continuing IV insulin therapy in surgical patients transferred from the ICU to lower-acuity wards.
putting in hospitals' standardized insulin protocols specific instructions for identifying and monitoring patients at risk for hypoglycemia, as well as actions that should be taken to prevent or treat the condition.
Source: Endocrine Practice, January/February 2004.
Internist Archives Quick Links
Have questions about the new ABIM MOC Program?
One Click to Confidence - Free to members
ACP Smart Medicine is a new, online clinical decision support tool specifically for internal medicine. Get rapid point-of-care access to evidence-based clinical recommendations and guidelines. Plus, users can easily earn CME credit. Learn more