Focus on Inpatient Diabetes Management: Optimizing Glycemic Control

CARLOS R. HAMILTON, JR., MD, FACE: Hello, I'm Dr. Carlos Hamilton and welcome to the AACE Clinical Conversations. In this program, we will discuss inpatient diabetes management, focusing on optimizing glycemic control.

Joining me today is Dr. Andrew Ahmann, Medical Director of the Diabetes Center and associate professor of medicine in the Division of Endocrinology, Diabetes and Clinical Nutrition at Oregon Health and Science University. And Dr. Etie Moghissi, associate clinical professor of medicine at the University of California, Los Angeles, and Co-Chair of the AACE Inpatient Glycemic Task Force. Thank you both for joining us.

We know of the challenges of caring for our diabetic patients in the inpatient environment. Etie, what are some of the goals for controlling the blood sugar in these patients?

ETIE S. MOGHISSI, MD: Well, established glycemic targets currently are glucose of less than 110 in the ICU, less than 110 for patients who are on the floor before meals and a maximum of 180. These are, for the most part, very challenging goals to achieve.

It is also important to understand that this is an area that is evolving very rapidly. More studies are coming in and it appears that there may be different glycemic targets for different patient population. So I think it is important to consider these targets, modify them when it's clinically appropriate -- for example, for patients who are prone to hypoglycemia. It may not be unreasonable to set higher targets.

CARLOS R. HAMILTON, JR., MD, FACE: So I gather that, in the intensive care unit, we would like to see the blood sugar in the range of 110. And then, after meals or when they're out of the intensive care unit, slightly higher values but still values that represent what we would consider to be good levels of blood sugar, is what we should strive for in the inpatient setting, is that correct?

ETIE S. MOGHISSI, MD: That's correct. And, of course, even in the ICU, for patients who are very prone to hypoglycemia, it may be appropriate to set higher targets, as long as it's -- we are not talking about, you know, glucoses of 180 and 200.

CARLOS R. HAMILTON, JR., MD, FACE: Now we all know, Andrew, and I'm familiar with your opinion about the role of the sliding scale orders that are often written for inpatients. I assume that you would much prefer some other sort of method of controlling blood sugar.

ANDREW J. AHMANN, MD: Right. I think if we could change one thing that would make a big difference across the country, it would be to get rid of sliding scales, which has been a traditional way of people approaching diabetes control in the hospital and totally inappropriate. The sliding scale, on a basic premise, is a reactive phenomenon. You're waiting for the blood sugar to get high instead of preventing it from getting high.

Secondly, it doesn't -- as we have used it historically, we don't allow for differences in insulin sensitivity between different patients and we have this phenomenon of 2, 4, 6 and 8 and we may vary where we start it, but it's always that. And, as an article that just came recently from University of Colorado suggests and similar to previous ones, that 80-some percent of the time, we never change that initial order, which was based on no information at the time that we wrote the order.

So what -- as a consequence, sliding scales will not get to any of the targets we're talking about, even if we were to go to more lax targets and it's probably gonna result in more blood glucose variability and possibly even, in some cases, more hypoglycemia as well as the potential for diabetic ketoacidosis, if it's ever applied to a type 1 patient.

CARLOS R. HAMILTON, JR., MD, FACE: Well, this is, unfortunately, what may happen when you're not using some sort of a basal insulin treatment program. When you're only using intermittent short-acting insulin, is that, if you skip a dose, you can go for a rather extended period of time with basically no insulin on board at all and, in our ketosis-prone diabetics, that is something we certainly want to avoid. How do you go about managing this in the inpatient environment?

ANDREW J. AHMANN, MD: So the primary tool that we would use for these patients who are particularly ill would be to use insulin infusions and I think that, with an insulin infusion, that's the obvious therapy or at least the standard therapy at this point for patients with hyperglycemic emergencies like diabetic ketoacidosis and hyperosmolar state. It's going to be used for patients on steroids initially, trying to get them under control. It's going to be used in most postop cases where you need to have quick control and you need to be able to respond to the rapid changes in insulin sensitivity that are going with the physiologic stress state.

We use it commonly in labor and a number of other indications where continuous insulin infusions are really the way to go.

CARLOS R. HAMILTON, JR., MD, FACE: And in the recovery room or in the intensive care unit, patients are almost routinely not taking in nutrition orally, so that the continuous intravenous infusion is a good way to manage these patients.

ANDREW J. AHMANN, MD: Right. It's a relatively safe way, a very consistent way, a very predictable way. And with the kind of nursing care that you have in those parts of the hospital, you can do it very safely.

CARLOS R. HAMILTON, JR., MD, FACE: And we all, in our different hospitals, have various protocols that we try to use to educate the staff and our other physicians that are involved in the care. Etie, where might one go about getting information for these protocols, if some of our viewers are particularly interested in instituting such a program in their own hospital?

ETIE S. MOGHISSI, MD: Well, there are at least eight or ten published protocols and they all have been used effectively and safely. The trouble is that these protocols or these IV insulin infusion order sets have not been compared with each other, so no one can really claim that one is better than the other. So we have multiple options. These are, as I mentioned, published and the references are available on our ACE website.

We just recently created an inpatient glycemic resource center, which can be accessed via ACE online. On this site, there are examples of different kind of protocols available. There are other informations and reference lists are available as well. So individuals who are trying to implement glycemic control in their hospital can use this resource very easily to sort of access the information that they need and then they can modify -- these are just examples, modify it and then monitor, to make sure that they -- they are actually effective and then move on from there.

CARLOS R. HAMILTON, JR., MD, FACE: Yeah, Etie, in my own experience with various protocols that we have tried in our hospitals, generally speaking, the simpler the program is, the more likely it is to be accepted by the nursing staff and by other physicians that would be involved in this. And many of these continuous insulin infusion protocols require varying the blood -- the amount of insulin or the rate of the infusion, depending on what the blood sugar is. Is that generally the way these things are arranged?

ETIE S. MOGHISSI, MD: Yes. The protocols use the column method, which is based on different insulin sensitivity. These are -- generally are very easy to use. Not a whole lot of calculation is necessary by the nursing staff. They can just move from one column to the next, provide more insulin to get the patient to target. Many of these can be computerized.

There are others that need a little more calculation and maybe a little bit of a challenge by the nursing staff. I think that the successful protocols are the ones who are, you know, complicated enough to be effective. At the same time, simple enough to be really driven by the nursing staff, because they are the one who are driving this protocol in the intensive care unit and, therefore, they have to be accepted by the nursing staff. Of course, education is necessary to make sure that they're familiar.

CARLOS R. HAMILTON, JR., MD, FACE: The protocol has to be one that the nurses are comfortable with, that they're not anxious about the effects that these changes are going to have and that they will be -- you know, create more problems than they already have.

Now, Andrew, we know that these patients that are in the recovery room or in the intensive care unit will be there for a relatively short period of time and they will come out of the intensive care unit or the recovery room back to other parts of the hospital, such as a regular ward or inpatient environment. What sort of recommendations do you give when you move a patient from a more intensive area to a less intensive area of the hospital?

ANDREW J. AHMANN, MD: Yeah, there's -- on one level, one of the possibilities is to use a continuous IV infusion of insulin on the floor, but there's only a few institutions who really feel comfortable with that and are doing that routinely. And, even when it's done, it has to be a simpler method that's going to require less frequent monitoring and where you set different goals.

So, aside from those few institutions that can do that, you're thinking really about transitioning to subcutaneous insulin. And when you transition to subcutaneous insulin, you try to use the information that you gathered from your recent use of the IV insulin to make that transition a smooth one. Unfortunately, there's many cases where we lose control at that time of transition, where, if it's not done properly, the blood sugars go very high and the patient then ends up having to go revert back and go on to an insulin infusion, I suppose, in a few cases, even heading back to the intensive care unit.

So there's really two methods that I've seen that are routinely used to accomplish this. One is -- first of all, you kind of calculate what your basal amount of insulin requirement's going to be, based on the end -- the last portion of your IV insulin infusion, so you might take four hours or six hours, extrapolate that to a 24-hour requirement. The reason for taking that smaller period of time is that this is a dynamic period where the stress response can be decreasing, where you want to catch the most recent time to make those determinations.

Then you usually put in a safety factor and so you might reduce it by 20% or reduce it by 40%. Some of that will depend on whether the patient was a known diabetic before this or whether they're hyperglycemic in a new setting and you don't know how insulin resistant they're going to be. Some of those individuals who are truly stress-hyperglycemia alone will revert fairly quickly.

So you make that calculation and then you either -- one you start that injection of a basal insulin to accomplish that sort of similar delivery of insulin over a 24-hour period, you would go and either overlap the IV for an hour and a half to three hours, people use various times, depending on -- including which insulin they're going to and try to smooth it. Or, in a few cases, they're actually taking 10% of the amount that was calculated as the basal insulin and giving it as a single bolus injection of a rapid-acting analog to accomplish that same thing.

CARLOS R. HAMILTON, JR., MD, FACE: But when you plan to discontinue the intravenous infusion, you write the orders for them to have both a basal insulin dose and -- and an intermittent dose of shorter-acting insulin that will hopefully be associated with -- with meals and so forth. Is that generally correct?

ANDREW J. AHMANN, MD: Right. So we get back to the concept of the basal/bolus insulin regimen.

CARLOS R. HAMILTON, JR., MD, FACE: That's right.

ANDREW J. AHMANN, MD: And, in particular, if the patient's eating, we're going to all three components, the three components being the basal insulin I just described, then a nutritional component to cover the food, which usually is sort of equivalent -- for the 24-hour period, is equivalent to the basal amount, but you divide it between the three meals and you may have to adjust for whether the patient's actually able to eat that meal. So, in many cases, that basal -- or that bolus, rapid-acting insulin is given immediately after the meal rather than before so we can make that determination.

And then there's still this correction component. We don't do a great job. We need to allow for that. We sometimes have, in a sense, a scale, of supplementary insulin that helps out to get back to goal quicker and learn from that and try to keep improving our baseline or scripted insulin to prevent blood glucoses from going up in the future.

CARLOS R. HAMILTON, JR., MD, FACE: Obviously the care of diabetic patients in the inpatient environment is very much of a team approach. Etie, I know you've had a lot of experience in educating teams in terms of the care of these patients. Tell us a little bit about that and especially what the role of the endocrinologist is.

ETIE S. MOGHISSI, MD: This is a very important aspect of implementation of tight glycemic control in the hospital setting. We know that patients who are in the hospital are coming into contact with different clinical team members, different specialties. So it is important to create a multidisciplinary team at the beginning when we are trying to sort of change a practice. This multidisciplinary team needs to have members from medical staff, nursing staff, dietary and nutrition, pharmacy, even from departments such as the medical information system need to be part of this multidisciplinary team.

CARLOS R. HAMILTON, JR., MD, FACE: And, Etie, even the patient needs to be a part of this team, because many of our patients are very experienced in managing their own blood sugars and want to be a part of the team, is that not correct?

ETIE S. MOGHISSI, MD: Very true. Very true. Obviously, the -- the patient is a very important part of this team. What I'm referring to, that, at the beginning, when we are trying to bring about practice changes, creating this team so we get buy-in from multiple stakeholders and the endocrinologist can easily be the champion of this change and lead this steering committee to assess the current processes of what is really the average blood glucose in this institution, where do we want to go, create a roadmap, create protocols, order sets. And then, finally, a very important aspect of this would be educating everyone in regard to importance of tight glycemic control.

CARLOS R. HAMILTON, JR., MD, FACE: So there really are major challenges in the care of our diabetic patients in the hospital, but if we do this and do it well, we can very significantly improve their outcomes and -- and have very beneficial effects on the whole healthcare system.

Well, thank you very much. Andrew and Etie, I appreciate both of your joining us today. It's been a very informative discussion.

And thank you very much for watching. I'm Dr. Carlos Hamilton and thank you very much.