Management of the Hospitalized Diabetic Patient
University of California, San Francisco
Last Author Revision: 2010
Over the past decade, there has been a revolution in the care of diabetic patients. Long-term prospective research studies, demonstrating that improved glycemic control decreases microvascular disease, [1][2][3][4][5][6][7][8][9][10][11][12] have led to a new standard of care for ambulatory patients. Recently, there has been a heightened interest in improving the quality and safety of the management of diabetes and hyperglycemia in the hospital [13]. While observational data strongly suggest an association of hyperglycemia with morbidity and mortality in adults on general medicine and surgery units, clinical research has not yet defined the best practices for managing hyperglycemia in the hospital outside the ICU, and there is controversy over the best course for management for the patient in the ICU. [14][15] As a consequence, many physicians do not have a well-formulated approach to managing hyperglycemia in either the non-critically ill or the critically ill hospitalized patient, and the use of insulin therapy to attain targeted blood glucose (BG) control is often subject to practice variability, leading to suboptimal glycemic outcomes.
Practical “guidelines” for the management of this common clinical problem have been formulated by experts in the field, based on understanding of the physiology of glucose and insulin dynamics, the characteristics of currently available insulin preparations, and clinical experience. In 2004, the American Diabetes Association published a technical review promoting the use of physiologic ("basal-nutritional-correction dose”) insulin regimens in the hospital to achieve targeted glycemic outcomes [16]. This approach has been disseminated via review articles [17], and more recently, a randomized, controlled trial demonstrated that hospitalized type 2 diabetes patients experienced better glycemic control when treated with a physiologic insulin regimen than when treated with sliding scale insulin alone [18]. These guidelines had also extended to the ICU, but the initial excitement that tight control of glucose levels in the ICU setting would decrease morbidity and mortality has been tempered by more recent studies that have not duplicated these results and have led to new suggested guidelines [19] .
Contemporary inpatient diabetes protocols have been successfully employed to manage diabetic patients who are not eating (NPO). New protocols have also been implemented to manage hypoglycemia [20][21][22]. Notwithstanding the availability of these inpatient protocols, and the general revolution in outpatient diabetes management, in most cases inpatient management remains a neglected area. The customary protocols utilized in most institutions reflect the diabetes management tools of the 1970s or earlier, rather than these new approaches. [23]
Finally, patients with diabetes are often highly educated about their diabetes and the reasons for good metabolic control. Poor inpatient management of their diabetes and resultant poor glycemic control creates a frustrating experience and sends a confusing message to the patient about the importance of glycemic control.
In this review, the current evidence that improving inpatient diabetes management changes hospitalization outcomes will be shown. Strategies for improving diabetes inpatient management, implementing these changes, and measuring these changes are also discussed.
The emphasis on inpatient diabetes care has traditionally centered on treatment of diabetic ketoacidosis (DKA) and hyperosmolar nonketotic coma, and the newly diagnosed diabetic patient. This focus has been misplaced as these are rarely related to any of the reasons the patient has been hospitalized. In general, poor glycemic control, as measured by glycosylated hemoglobin level, has been associated with an increased risk for hospitalization.[24][25] Patients with diabetes have a 2.2 to 4-fold increase in hospitalization rate compared to the nondiabetic population. [26] [27][28]
In the year 2003, there were 5.1 million hospitalizations with diabetes as a listed diagnosis, a 2.3-fold increase over 1980 rates [29]. Acute care inpatient hospital costs represent approximately 50% ($65.2 billion in 2007) of the annual health care cost for diabetes [30]. The majority of this expense is for hospitalized patients with a secondary, not primary, diagnosis of diabetes. Thus, management of patients hospitalized for heart disease, infections, surgery, etc. most often characterizes inpatient diabetes care. In fact, almost half of all health care expenditures attributed to diabetes come from higher rates of hospital admission and longer average lengths of stay. [31]. In addition, hospitalized patients who are found to have previously unrecognized hyperglycemia and/or diabetes at the time of their hospitalization have been an overlooked population. It is estimated that 38-50% of hospitalized patients have hyperglycemia and of these, 25-50% of these patients were not aware of the hyperglycemia before their hospitalization. [32][33][34]
The major factors that need to be overcome in order to achieve glycemic control in the hospitalized diabetic patient are:
Epidemiological and small retrospective studies show diabetes increases morbidity and mortality for myocardial infarction [35][36][37][38][39][40][41][42][43][44][45][46][47][48][49][50][51][52][53][54], and coronary bypass surgery [55][56][57][58][59][60][61][62][63][64][65][66][67][68][69][70][71]. In general, hyperglycemia is associated with increased mortality and morbidity in patients with acute cerebral ischemia. There has been up to double the mortality risk in patients admitted with a stroke. [72][73][74][75][76][77][78][79][80][81][82][83][84][85][86][87][88][89][90][91][92][93][94][95][96]
In an extensive retrospective review of 259,040 critically ill patients conducted by the Veterans Affairs Inpatient Evaluation Center based in Cincinnati, for all patients reviewed, hyperglycemia was an independent predictor of mortality starting at 111 mg/dl. [97] However, in diabetes patients, the increase in mortality risk was not seen until mean glucose was >146 mg/dl. Moreover, there were distinct differences in the association of hyperglycemia and mortality depending on the underlying diagnosis. The correlation was most significant for acute myocardial infarction (1.7-6 fold increase in predicted mortality), unstable angina (1.4 to 3 fold increase in predicted mortality), and stroke (1.8 to 29 fold increase in predicted mortality). A significant but weaker effect was seen in patients with sepsis, pneumonia, and pulmonary embolism. Hyperglycemia was not found to be associated with mortality in diseases such as COPD and hepatic failure, hip fractures.
More specifically, when glucoses are elevated (generally greater than 200 mg/dl) there may be:
1. Fluid and electrolyte abnormalities secondary to osmotic diuresis
2. Decreased WBC function [98][99][100][101][102][103][104][105][106][107][108][109]
3. Delayed gastric emptying [110][111][112][113][114]
4. Increased surgical complications including:
a. Relative risk for “serious” postoperative nosocomial infections increased by a factor of 5.7 when glucose >220 mg/dl [115]
b. Relative odds of wound infection increased to 1.17 with glucoses that were 207-227 mg/dl and to 1.78-1.86 when glucoses were >253 mg/dl. [116]
5. Delayed hospital discharge:
a. Length of stay was compared for patients in “control” (glucose 60-250 mg/dl) and those with “fluctuating” glucoses (glucose <60 or >250 mg/dl). [117]
|
Length of Stay (days) |
|
|
Glucoses in Control |
Fluctuating Glucoses |
Acute Myocardial Infarction |
4.1 |
6.7 |
CABG |
6.3 |
8.2 |
Community Acquired Pneumonia |
4.5 |
6.3 |
The above studies indicated that high glucose levels are associated with a poor outcome. Illustrative data showing that improving glycemic control improves outcomes is listed below.
Incidence of deep wound infections decreased from 2.4 to 1.5% [121]
2.0 to 0.8% [122]
Furnary and colleagues have administered IV insulin to their post-CABG diabetic patients with glucose goals decreasing from 150 mg/dl in around 2000 to now the low 100 mg/dl range. Their uncontrolled, retrospective analyses have shown decreased mortality with glucoses under 150 mg/dl. [124][125][126][127][128] See Figure 1.
Figure 1. Effect of glucose control on Mortality after open heart surgery.
[129]
§ Mortality at 1 year was 26% in the control patients compared to 19% in the study patients.
§ Mortality at 3.4 years was 44% in the control patients compared to 33% in the study patients
§ After this study, it was not clear if the benefit was due to the IV insulin, the prolonged SQ insulin, or another intervention. So a second study was performed.
DIGAMI 2 (Diabetes Mellitus, Insulin Glucose Infusion in Acute Myocardial Infarction 2) In DIGAMI 2, three treatment strategies were compared: group 1, acute insulin-glucose infusion followed by insulin-based long-term glucose control; group 2, insulin-glucose infusion followed by standard glucose control; and group 3, routine metabolic management according to local practice [137][138]. (see Figure 2)
§ There were no significant differences in morbidity expressed as non-fatal reinfarctions and strokes among the three groups.
§ DIGAMI 2 did not support the hypothesis that an acutely introduced, long-term insulin treatment improves survival in type 2 diabetic patients following myocardial infarction when compared with a conventional management at similar levels of glucose control or that insulin-based treatment lowers the number of non-fatal myocardial reinfarctions and strokes. However, an epidemiological analysis confirms that the glucose level is a strong, independent predictor of long-term mortality in this patient category, underlining that glycemic control seems to be an important part of their management.
5. Decreased morbidity and mortality in critically ill patients – an area of controvery.
a. Post-operative patients on ventilators: A comparison of patients with plasma glucoses between 80-110 mg/dl (intensive treatment), and those given insulin only if glucoses were >215 mg/dl and then maintained 180-200 mg/dl (control) [140] indicate that:
Intensive insulin treatment decreased overall mortality from the control rate of 8% to 4.6% (p<0.04). In addition, the intensive treatment decreased sepsis by 46%, need for dialysis by 41%, and these patients were less likely to require prolonged ventilatory support and intensive care. However, on critical review of this paper, there are questions regarding the generalizability of the data. The subjects received significantly more glucose daily than is ususal care in the US. In addition, the benefit was mainly seen in the surgical patients who remained in the ICU for greater than 5 days, also a small percentage of patients. Finally, on secondary analysis, there was only a small difference between the long-term mortality outcomes between the group that had glucoses <110 mg/dl and the group that had glucoses 110-150 mg/dl.
b. Medical ICU patients:
1. A large prospective, single-center comparison of patients with plasma glucoses between 80-110 mg/dl (intensive treatment), and those given insulin only if glucoses were >215 mg/dl and then maintained at 180-200 mg/dl (control) [141] indicate that: Intensive insulin therapy significantly reduced morbidity but not mortality among all patients in the medical ICU. Patients were preselected as having the potential to stay in the ICU for more than 3 days. In sub-analysis, the risk of subsequent death and disease was reduced in patients treated for three or more days, but these patients could not be identified before therapy.
2. VISEP: A large multicenter, prospective comparison of intensive IV insulin treatment compared to less tight control. [142] Study showed no benefit to the tighter glucose control and was stopped early due to increased risk of hypoglycemia. Authors concluded: these studies (this one and the Van den Berghe study above) establish that intensive insulin therapy has no measurable, consistent benefit in critically ill patients in a medical ICU, regardless of whether the patients have severe sepsis, and that such therapy increases the risk of hypoglycemic episodes.
3. NICE-SUGAR: A large multicenter randomized trial comparing the effects of intensive glucose control (target glucose 81-108 mg/dL) and conventional glucose control (target glucose < 180 mg/dL) on death from any cause within 90 days of randomization. 6104 patients were randomized, all adults admitted to the ICU in the 24 hours prior, and all expected to require treatment in the ICU for 3 or more days. More of those randomized to intensive control died than those randomized to conventional control (27.5% vs. 24.9%, p = 0.02), severe hypoglycemia was reported in more of those randomized to intensive control than to conventional control (6.8% vs. 0.5%, p < 0.001), and there was no significant difference in length of stay in the ICU, median days of mechanical ventilation, or renal-replacement therapy.[143]
4. META-ANALYSIS. A 2008 meta-analysis identified 29 studies that included 8432 patients and reported 1869 deaths. The overall relative risk of death for those treated with tight glucose control was 0.93 (95% confidence interval, 0.85-1.03). Additionally, subgroup findings for trials that targeted a blood glucose concentration of less than 110 mg/dL and subgroups selected according to the type of ICU (surgical, medical, or medical-surgical) did not demonstrate a significant reduction in mortality. [144]
5. COIITSS Study. In this multicenter study of hydrocortisone-treated patients with septic shock, intensive insulin therapy did not improve outcomes.[145]
6. Intraoperative glucose management. There has been very little attention to actual outcomes from intraoperative glucose control. An early retrospective analysis showed that lower glucoses during cardiovascular surgery were associated with improved outcomes. [146] However, when tested with a randomized, prospective study, maintaining glucoses less than 110 mg/dl, there was no apparent benefit to intensive insulin therapy during cardiac surgery. It did not reduce perioperative death or morbidity. In fact, there was an increased incidence of death and stroke in the intensive treatment group (though not statistically significant) and raises concern about routine implementation of this intervention. [147]
The general goal for glucose control is to avoid the undesirable short-term effects of hypo- and hyperglycemia. From a practical standpoint, this means that glucoses should be between 100 to 200 mg/dl. Glucoses under 200 mg/dl should decrease the short-term diabetes complications that may occur during a hospitalization, while lower glucoses (with a HgA1c goal of about 7%) are desired for the long-term prevention of microvascular and macrovascular disease. Hypoglycemia should be avoided to prevent counterregulatory hormone-induced cardiovascular complications. From the data discussed above, specifically the large retrospective ICU reviews, the actual benefit of specific goals in patients with different underlying diagnoses remains unclear [150].
General Medical and Surgical Wards:
At present, recommended glycemic targets for non-critically ill hospital patients are based entirely on expert opinion, as there have been no clinical studies directly comparing different glycemic targets in this patient population. However, the American College of Endocrinology, the American Association of Clinical Endocrinologists, and the American Diabetes Association do provide recommendations about glycemic targets for inpatients (Table 1) . [151]Although controversial, these recommendations make it clear that uncontrolled hyperglycemia is no longer the accepted standard of care for hospitalized patients. Of note, most hospitals adopt glycemic targets that are less stringent than those shown in Table 1, recognizing the challenges of controlling blood glucose levels in hospitalized patients, and the potential risk for hypoglycemia when lower BG targets are used. Each hospital’s glycemic control champions must reach consensus on a target blood glucose range for their institution. In practice this range has been 90-110 mg/dl for the lower BG limit and 140-180 mg/dL for the upper BG limit. It is also important to note that the recommendations from professional organizations emphasize the need to individualize BG targets, based upon the clinical circumstances of each patient. In our institution, our stated goal is for glucoses of 80-150 mg/dl with the understanding among the institution's experts that our actual goal is 80-180 mg/dl. We decided from a practical standpoint that, if the higher numbers were our general stated goal, then even higher glucose levels would be felt to be reasonably close, a situation we attempted to avoid.
ICU:
Postoperative ICU and Medical ICU glucose control trial data is discussed in detail above. As described, there is now significant controversy about the generalizability of the initial studies that suggested goals of 80-110 mg/dl for this arena. In our institution we had and continue to believe that present evidence supports maintaining glucoses < 140-150 mg/dl with the avoidance of hypoglycemia. AACE and ADA have updated their reccomendations and moved from a goal of 80-110 to now 140-180 mg/dl.
On the basis of the available evidence, insulin infusion should be used to control hyperglycemia in the majority of critically ill patients in the ICU setting, with a starting threshold of no higher than 180 mg/dl (10.0 mmol/l). Once IV insulin therapy has been initiated, the glucose level should be maintained between 140 and 180 mg/dl (7.8 and 10.0 mmol/l). Greater benefit may be realized at the lower end of this range. Although strong evidence is lacking, somewhat lower glucose targets may be appropriate in selected patients. Targets <110 mg/dl (6.1 mmol/l), however, are not recommended. Use of insulin infusion protocols with demonstrated safety and efficacy, resulting in low rates of occurrence of hypoglycemia, is highly recommended. [152]
Table 1. Recommendations for inpatient glycemic targets [153]
Organization |
ICU |
Non-ICU, Preprandial* |
Non-ICU, Maximum* |
AACE/ACE
|
140-180 mg/dl |
<140 mg/dl |
180 mg/dl |
ADA |
140-180 mg/dl |
<140 mg/dl |
180 mg/dl |
University of California, San Francisco |
100-160 |
80-180 mg/dl |
180 mg/dl |
*Specific glycemic targets for non-critically ill hospital patients are based entirely on expert opinion, as there have been no clinical studies directly comparing different glycemic targets in this patient population. ACCE = American College of Clinical Endocrinologists, ACE = American College of Endocrinology, ADA = American Diabetes Association, ICU = intensive care unit
The current protocols for both intravenous insulin infusions and subcutaneous insulin injections require frequent glucose measurement with immediate decisions and changes in insulin dosing. Fundamental to all areas of diabetes management, but specifically for inpatient management, is the assumption that the measured glucose level is accurate. This assumption may not be borne out by the reality of current technology. The ADA has previously recommended accuracy goals, in 1986 (target accuracy of +/- 15%) and in 1993 (target accuracy of +/- 5%) These goals have never been met by any manufacturer of a glucose monitoring device. A number of alternative standards have been suggested by national standards organizations in the U.S., Canada, and Europe. An international standard ISO DIS 15197 is under development recommending accuracy of +/- 20 mg/dl for glucose values under 100 mg/dl and +/- 20% for higher glucose values. [154] In addition, care must be taken as to the location the blood sample is taken from as “pseudohypoglycemia” has been reported in fingertip samples in patients with Raynaud's phenomenon [155] and in patients with altered circulation from shock. [156]
Confounding variables in point of care glucose measurement are shown in Table 2.
Table 2— Confounding variables in point of care glucose measurement
|
*Change relative to venous plasma measured at central laboratory. GO, glucose oxidase.
Adapted from: (139)
Although oral antihyperglycemic agents are used frequently in the outpatient setting, there are many potential disadvantages to using these medications in acutely ill hospitalized patients (Table 3). [157][158] In general, oral agents are difficult to titrate quickly to effect, and contraindications, adverse reactions, and many patients’ variable nutrition can limit their use in the hospital. A commonly-encountered example is that of metformin, which can lead to lactic acidosis if used in clinical situations which predispose the patient to lactate production (e.g., renal failure, exposure to contrast agents with subsequent risk of contrast nephropathy, circulatory failure, or hypoxemia), situations that arise frequently in the hospital. Patient’s with lactic acidosis with metformin usually have had normal renal function but then develop acute renal failure just prior to the onset of the lactic acidosis. [159] Another example is that of sulfonylureas, which can cause hypoglycemia if used by a patient with inconsistent nutrition.
___________________________________________________________
Table 3. Potential disadvantages associated with specific oral antihyperglycemic agents in the hospital setting.
CHF= congestive heart failure, GLP = glucagon-like peptide, DPP-IV= dipeptidyl peptidase IV
___________________________________________________________
There are some situations in which it is appropriate to continue oral antihyperglycemic medications in the hospital. In hospitalized patients who are clinically stable and who have normal nutritional intake, normal blood glucose levels, and stable renal and cardiac function, it is likely that the medications can be continued safely. Oral agents may be started or resumed in the hospital if they are to be included in the discharge medication regimen, once the patient is clinically stable and once it has been ensured that contraindications to their use no longer exist.
A physiologic insulin regimen is one that mimics normal endogenous insulin activity by providing appropriate types and doses of insulin at appropriate times. It consists of three separate components: basal, nutritional, and correctional insulin. (see figure 3) Pharmacodynamics of insulin are fully detailed in Insulin Pharmacology, Types of Insulin and Adjustments.
Basal insulin: Basal insulin is the insulin normally released continuously by the pancreas, even when fasting, which serves to suppress glucose and ketone production. Exogenous basal insulin is provided as a long- or intermediate-acting, low- or non-peaking insulin (e.g., glargine or detemir) and allows for a consistent level of basal coverage. NPH can be used as a basal insulin, but it has a peak that may exceed the level of insulin actually required for basal needs. As will be reinforced below, NPH should be used with caution in a patient who is NPO and reduced by 1/3 to 1/2 to avoid hypoglycemia at the time of its peak. Whichever preparation of insulin is used, basal insulin is provided even when the patient is not receiving any nutrition. It generally represents about half of a patient’s total daily dose (TDD) of insulin.
Figure 3. Nutritional and basal insulin
Nutritional insulin: Nutritional insulin is insulin normally secreted in response to the absorption of glucose and amino acids into the blood after ingestion of a meal. Exogenous nutritional insulin must be provided in a way that matches the nutrition provided to the patient. For example, a patient who is receiving nutritional boluses (i.e., meals or bolus tube feeds) can be given rapid-acting insulin (e.g., aspart, lispro, or glulisine) along with each ingestion of nutrition. The rapid-acting analog insulins can be given at the end of a meal or bolus tube feed in cases where it is not clear if the nutrition will be well-tolerated; a reduction in the insulin dose proportionate to the amount of nutrition actually taken can then be made to decrease the risk of subsequent hypoglycemia. Regular insulin can also be given for nutritional coverage, although its later peak requires that it be given 30 minutes before a meal is ingested, the timing of which is a challenge for most hospital nursing units and which requires confidence in advance of a meal that an ill patient will indeed eat most of the meal served. Patients not receiving any nutrition should not receive any nutritional insulin. Nutritional insulin generally represents about half of a patient’s TDD of insulin in a patient who is eating.
Correctional insulin: Correctional insulin is insulin given in addition to basal and nutritional insulin to correct hyperglycemia. It is usually provided as rapid-acting or regular insulin, given in a dose specifically designed to reduce a patient’s elevated blood glucose back into the target range. It is usually given at the same time as the nutritional insulin in patients receiving nutrition, and of the same type as the nutritional insulin, so that the nutritional and correctional doses may be combined in one injection. For patients not receiving nutrition, it is usually given every 4-6 hours. Correctional insulin is usually ordered in a “stepped” format to provide a variable amount of insulin for variable blood glucose values. A consistent requirement for or a requirement for high doses of correctional insulin suggests a need to modify the basal and/or nutritional insulin ordered for a patient.
As described in the insulin chapter, newer basal bolus insulin regimens using rapid-acting insulin analogs and basal insulins such as glargine or detemir are often prefered. It should be pointed out however, that in the only inpatient study comparing Detemir plus aspart vs NPH plus regular, there was no difference in glucose control or hypoglycemia. [160] Numerous outpatients studies have shown both improved control and decreased hypoglycemia when the newer analogues are used.[161]
While “Sliding Scale” insulin protocols remain the mainstay for inpatient glucose control, this method does not work. “Sliding Scale” methodology dates to diabetes monitoring by urine glucose levels. The tape that was used for the test would change colors depending on how much glucose was in the urine. Insulin was then given based on the change in color. This was called “rainbow coverage.” Unfortunately, whether urine or plasma glucose is used, there is no physiologic basis for this form of insulin therapy. Patients therefore tend to have “roller coaster glucose control.” Under this protocol, the patient would not receive insulin when their glucose level is normal. A few hours later their glucose level increases because no insulin had been given. Insulin is then administered for the elevated glucose level and a few hours later the glucose level returns to normal. This cycle is repeated again and again. (See figure 4)
Figure 4. Roller Coaster Effect of Insulin Sliding Scale
Under treatment with a “Sliding Scale,” patients with Type 1 DM will go into DKA. If a patient with Type 1 diabetes does not receive insulin for their “normal glucose,” they will be insulinopenic by their next glucose check, and likely be ketotic. As many patients who are thought to have Type 2 diabetes may actually have Type 1 diabetes or may be relatively insulinopenic, it is difficult to predict who is at risk for DKA
Finally, in the typical “Sliding Scale,” no insulin is given for a glucose <150 mg/dl. For a glucose of 151-200 mg/dl 2 or 3 units of insulin may be given, for 201-250 3 or 4 units, on so on. Imagine a typical insulin resistant Type 2 diabetic patient who has been well controlled on a total of 80 units per day (in split doses). Is it any wonder that this patient will have marked hyperglycemia upon admission to the hospital and treatment with the “Sliding Scale” described above. One early study [162] of the use of insulin “Sliding Scales,” concluded that there was no benefit in their use. In the accompanying editorial, Sawin wrote:
Routine multiple measurements of capillary blood glucose levels, along with sliding scale insulin doses, offer no benefit to sick patients with diabetes, and when such patients come to the hospital, they need to follow their previous treatment of insulin or an oral hypoglycemic drug. The burden of proof is on those who continue to use a sliding scale regimen. [163]
More recently, when use of “sliding scale” insulin was compared to basal-bolus insulin in a randomized prospective multicenter study and in single center studies, the use of “sliding scale insulin was associated with not only higher average glucoses but actually an increased risk of hypoglycemia.[164] [165]
The use of pre-printed order sheets for subcutaneous insulin and continuous insulin infusions provides for better patient safety than the writing of de novo orders for each individual patient, as pre-printed order sheets are complete, guideline-based, and standardized so that they are familiar to providers. For example, when University Hospital in Pittsburgh implemented a pre-printed order sheet for correctional insulin orders, prescribing errors decreased from 10.3 per 100 applicable patient-days at baseline to 1.2 per 100 applicable patient-days at one year (p = 0.03), and the number of hyperglycemic episodes decreased from 55.9 to 16.3 per 100 applicable patient-days (p = 0.0001). [166]
Hospitals’ carbohydrate-controlled or “diabetic diet” trays typically consist of 60-75 grams of carbohydrates. To specify “no concentrated sweets” is prudent, as inappropriate juices will not be served. If it is important for an individual patient’s nutritional insulin dosing, a hospital kitchen should be able to provide an itemized list of the carbohydrate values for the components of a meal tray.
Generally, if a patient’s blood glucose is less than 70 mg/dL and the patient is taking food or drink by mouth, give 20 grams of oral fast-acting carbohydrate, either 4 glucose tablets (5 grams glucose/tablet) or 6 oz. of fruit juice. If a patient is not taking food or drink by mouth, give a 25 mL D50 IV push. Check blood glucose every 15 minutes and repeat the above treatment until blood glucose is 100 mg/dL or above. From a practical standpoint, orders for treatment of hypoglycemia must be integrated into the standard preprinted orders for insulin administration. Please see the order sheets for specifics.
The hospital discharge is a challenging transition for the diabetic patient, as many patients’ diabetes medication regimens will be altered during their hospitalizations. Whether a given patient returns to his or her pre-hospital regimen or goes home with an altered regimen should depend on the adequacy of the pre-hospital regimen (often assessed by the patient’s hemoglobin A1C at or near the time of admission, and the patient’s report of self-monitoring at home), and on whether circumstances that have altered the patient’s insulin requirement persist (e.g., whether an infection persists or has resolved). For example, on the one hand, a patient with diabetes well-controlled on a single oral agent prior to admission should usually resume use of that oral agent at discharge, with the discontinuation of his or her inpatient insulin regimen. On the other hand, a patient with uncontrolled diabetes prior to admission requires a new regimen at the time of discharge home, which might involve oral agents or insulin.
The American Diabetes Association recommends that a diabetes education plan including “survival skills education” and follow-up be developed for each patient, and the Joint Commission specifies that patients with newly diagnosed diabetes or educational deficits have educational plans of care including medication management, nutritional management, exercise, signs, symptoms, and treatment of hypoglycemia, treatment of hyperglycemia, the importance of blood glucose monitoring, sick day guidelines, contact information in case of an emergency or unanswered questions, and a plan for post-discharge education or self-management support.[167]
If a hospital has a diabetes educator or nurse specialist available, that person should be involved in a patient’s discharge planning. Advance planning for eventual discharge also includes identification of whether a patient already owns a glucose meter and whether he or she will need lancets, glucose strips, syringes/needles, insulin, glucose tablets, and/or a glucagon kit included in the discharge prescription. A patient cannot just be discharged with a prescription for insulin without determining whether or not that insulin may or may not be covered by their medical insurance company. Finally, communication of the diabetes regimen must be made to the patient’s outpatient health care provider.
Many patients with diabetes that is diet-controlled at baseline will be hyperglycemic in the hospital due to high levels of insulin antagonists such as catecholamines or glucagon, such as occurs in the setting of infection, and thus will require insulin in the hospital.
These step-wise instructions are described in detail on the our institution’s pre-printed insulin order sheets, included at the end of this chapter.
As discussed above, there are many potential disadvantages to using oral antihyperglycemic medications in acutely ill hospitalized patients, including difficult quick titration, contraindications, adverse reactions, and variable patient nutrition. Thus, insulin is the mainstay of most diabetes management in the hospital.
These step-wise instructions are described on our institution’s pre-printed insulin order sheets, included at the end of this chapter.
When a patient on insulin at home is admitted to the hospital, an assessment of the adequacy of the home glycemic control will guide inpatient management significantly. For example, a patient on a basal-bolus regimen at home with a hemoglobin A1C of 6.5% may be ordered the same regimen in the hospital, with adjustments made based on appetite and renal function. A patient with a hemoglobin A1C of 11%, however, will presumably need more insulin than he usually takes at home.
These step-wise instructions are described on the our pre-printed insulin order sheets, included at the end of this chapter.
As hospitalized patients may be made non per os (NPO) for short procedures or due to difficulty tolerating enteral nutrition, questions frequently arise regarding whether and which aspects of the inpatient insulin regimen should be held or adjusted in the fasting state.
These step-wise instructions are described on our institution’s pre-printed insulin order sheets, included at the end of this chapter.
The hospitalized patient who is NPO for an extended period of time, including the hospitalized patient who is to undergo an operation of more than approximately 3 hours’ duration, or the patient under close monitoring in the ICU, is usually best managed with a continuous insulin infusion. The predictable delivery and short biologic effect of IV insulin allows for rapid dose adjustment and more stable glucose levels. Continuous insulin infusions are described in a separate section below.
Patients receiving bolus tube feeds are typically treated like patient who are eating meals, with a bolus of nutritional insulin with each bolus of nutrition.
As continuous tube feeds represent nutrition on a continuous basis, a patient receiving this type of nutrition must receive nutritional insulin in a way that provides for some continuous coverage. One might argue that providing all of the patient’s total daily dose of insulin (basal plus nutritional) in the form of a long-acting insulin without a peak, like glargine, might be most physiologic. However, if this strategy is used and the tube feeding is interrupted (e.g., turned off due to residual formula in the stomach, or discontinued due to a problem with the tube), the patient will be in danger of hypoglycemia for the duration of the action of the long-acting insulin. It is for this reason that many endocrinologists recommend administering separate basal and nutritional insulin.
Related step-wise instructions are described on our institution’s pre-printed insulin order sheets, included at the end of this chapter.
Standard TPN contains 25% glucose, which, if administered at 100 mL/hour, yields 25 grams of carbohydrates hourly. Regular insulin can be mixed with the parenteral nutrition to provide safe and effective glycemic control. This has the advantage that if the parenteral nutrition is interrupted, (e.g., discontinued due to a problem with the central line), the insulin will be discontinued as well, and there will not be a risk of hypoglycemia as there would be if an anticipatory bolus of nutritional insulin had already been administered. When starting parenteral nutrition, use of a separate IV insulin infusion can help in the assessment of the TDD of insulin required on the parenteral nutrition regimen, and then that can be used to determine the amount of insulin added to the parenteral nutrition. If the insulin added to the parenteral nutrition accounts for basal as well as nutritional insulin needs and then the parenteral nutrition is discontinued, separate administration of subcutaneous basal insulin will be required. Alternatively, the insulin added to the parenteral nutrition can account for nutritional insulin needs only, with separate subcutaneous basal insulin at all times. Subcutaneous correctional insulin is usually ordered with a rapid-acting analog or with regular insulin.
The transition from a continuous insulin infusion to a subcutaneous insulin regimen is one that is accompanied by a risk of adverse events including DKA, if it is undertaken inappropriately.
These step-wise instructions are described on our institution’s pre-printed insulin order sheets, included at the end of this chapter.
In diabetic patients, glucocorticord use adversely impacts glucose control. In addition, “steroid diabetes” can occur in up to 25% of patients previously thought to be nondiabetic [168]. The glucose elevation is predominantly postprandial hyperglycemia with a relative lack of fasting hyperglycemia, a pattern that reflects the steroid-induced inhibition of glucose uptake into fat and muscle, with a general insulin resistance, but with much less effect on gluconeogenesis.[169]
Continuous subcutaneous insulin infusion (CSII) therapy, often referred to as insulin pump therapy, is a method of delivering intensive insulin therapy for tight glycemic control. As it becomes more common in the outpatient setting, it is encountered more frequently in the acute care setting. Challenges to the use of CSII therapy in the hospital involve the requirements for patient participation (potentially limited by patient level of consciousness or critical illness), and for provider comfort (often limited due to lack of exposure to and training about CSII therapy). Many hospitals, then, choose a policy that CSII therapy be converted to standard subcutaneous insulin injection therapy for hospitalized patients, which is usually straightforward and involves the administration of one or two basal insulin injections calculated from the pump basal rate, the administration of nutritional insulin calculated from the usual ratios of insulin administration to carbohydrate intake, and the administration of stepped correctional insulin. If a hospital is to allow a patient on established CSII therapy to continue the technology while in the hospital, institutional-level policies and procedures should be established:[170]
The predictable delivery and short biological effect of intravenous insulin make it the therapy of choice for patients requiring insulin who are NPO for an extended period of time, including patients who are to undergo an operation of more than approximately 3 hours’ duration, and patients under close monitoring in the ICU. As described in sections above, it can also be used to establish insulin requirements for patients who will be transitioned to basal-bolus regimens or to insulin administration with parenteral nutrition. Indeed, though now more controversial, there is evidence that aggressive glycemic control with insulin infusions in ICU patients can improve clinical outcomes.[171] However, there is wide variability in institutional practice of insulin administration by infusion, due to general lack of consensus about approach.[172] General principles for which there is consensus include the following:
These step-wise instructions are described on our pre-printed insulin order sheets below.
The authors’ institution employs the following pre-printed insulin order forms in its inpatient care for adults:
There is more to inpatient diabetes management than copying an order sheet such as one of the ones at the end of this chapter and then placing it on a patient’s chart. That is a formula for failure. Successful implementation requires an institutional approach, with the following components:
o Physicians: Endocrinologist, Hospitalist
o Clinical Nurse Specialists: Diabetes, education
o Nurses: ICU Manager, at least one manager from medical floor (or their representative)
o Clinical Pharmacist
o Administration presence – from level of quality assurance or similar title
o Discharge Coordinator – not required for initial discussions and implementation, but needed later
o Nutritional services – not required for initial design and implementation of forms.
o Formulary
o Clean up insulin
o Clean up oral agents
o Nursing Issues
o Policy on IV insulin use
o Policy on frequency of glucose monitoring
o Forms
o Design forms
o IV insulin forms
o SQ insulin forms
o ?DKA treatment forms
o Pharmacy and Therapeutics
§ Formulary issues
§ Oral agents
§ Insulins
§ Insulin Forms – iv, sq
o Forms
§ Insulin forms – iv, sq
o Quality Improvement
§ Need buy in at this level to achieve administrative support
o Smaller Hospitals
§ CEO
§ Chief of Staff
o Larger Institutions
§ Chairs of Medicine, Surgery
§ Heads of training programs from Medicine, Surgery
§ Chief of Staff, Chief Medical Officer, CEO
§ Chairs of other Departments
§ Chief Residents
§ Dean for Education
o Education
§ Nurses
§ Physicians
o Continual evaluation
o Measurements
It is important that the all the steps listed above are explored and taken seriously. An institution should expect this process to take years, not months for implementation and true change to occur.
The best way to measure and compare institution wide glucose levels has yet to be determined. Common sense dictates that, in order to compare two or more quantitative measures, they must be reported in the same unit of analysis. For example, one could not conclude that men were taller than women from comparing the mean height of men in a population to the median height of women. To date, both published clinical and QI studies report variable outcome measures, making direct comparisons and the establishment of benchmarks difficult. Most studies report mean glucose as the primary outcome, despite the fact that glucose values in hospitalized patients with hyperglycemia are heavily skewed (Figure individual glucose values), as shown by a simple histogram of individual glucose values obtained at our institution in August 2006. In this sample of 6871 glucose measurements performed in 311 patients, the mean glucose was 174.2 (+/- SD 70.5) mg/dL and the median glucose was 158 mg/dL.
Figure 3. Graphic Representation on Glucometrics
The term “glucometrics” was coined to describe methods for compiling and analyzing inpatient glucose data to determine performance. [173] In order to assess the quality of inpatient glucose management, at least three different units of analysis (population, patient, patient-day) may be used. A cartoon of this process is shown in figure – glucometrics. In table, glucometrics, these three methods of analysis are described in detail. Goldberg et al. concluded that rates of hypo- and hyperglycemic events were sensitive based on the model chosen, and analyses of blood glucose data at our institution have confirmed these observations
TABLE 3. GLUCOMETRICS EMPLOYING THE THREE CANDIDATE QUALITY PERFORMANCE MEASURES ON A
GENERAL MEDICAL WARD, APPLIED TO A DATABASE CONTAINING ONLY FINGERSTICK BG VALUES
|
Model |
||
Measure |
By population |
By patient-day |
By patient |
BGs per unit of analysis |
|
BGs Mean number of BGs for each patient-day |
Mean number of analysis for each patient during entire hospital stay |
Mean BG (mg/dL) |
Mean of all BGs for the entire population |
Mean of mean BGs for each patient-day |
Mean of mean BGs each patient during entire hospital stay |
Median BG (mg/dL) |
Median of all BGs for the entire population |
Median of mean BGs for each patient day |
Median of the mean BGs for each patient during entire hospital stay |
% BGs in range |
% of all BGs between 80–139 mg/dL |
% of patient-days where mean BG was 80–139 mg/dL |
% of patients with mean BG 80–139 mg/dL for their entire hospital stay |
% hypoglycemic events |
% of all BGs <60 mg/dL |
% of patient-days with any BG <60 mg/dL |
% of patients with any BG |<60 mg/dL during entire hospital stay |
% hyperglycemic events |
% of all BGs >300 mg/dL |
% of patient-days with any BG >300 mg/dL |
% of patients with any BG >300 mg/dL during entire hospital stay |
Repeat Measurements
Hypoglycemia is usually the limiting factor in implementation of tight inpatient glucose control and represents the most clinically significant potential adverse effect of aggressive anti-hyperglycemic therapy{Cryer, 2002 2 /id}. Standardized hypoglycemia protocols provide guidance to nursing and physician staff regarding the thresholds for treatment and prevent delays in treatment by providing anticipatory orders{Braithwaite, 2004 36 /id}. Such protocols may also avoid over-treatment of hypoglycemia, which can result in hyperglycemia. Many hypoglycemia protocols, including the one used at our institution, mandate repeat measurements of blood glucose until euglycemia is achieved.
When we performed analyses of glucometrics both including and excluding these repeat measurements, we found that inclusion significantly overestimated the incidence of hypo- and hyper-glycemia
Implementation may take several years. There are several resources that may be of assistance.
Workbook on how to implement inpatient diabetes changes prepared by the Society for Hospital Medicine
http://www.hospitalmedicine.org/ResourceRoomRedesign/GlycemicControl.cfm
Diabetes Mellitus: A practical approach to impatient management. This is a complete interactive course to train you on all aspects on inpatient care. It is free at:
http://www.rushakoff.com/inpatient%20diabetes%20course/index.htm
Training course for nurses on inpatient management:
http://www.rushakoff.com/rushakoff/rushakoff1/page_1.html
Inpatient Diabetes Toolkit developed in Georgia:
http://www.gha.org/pha/health/diabetes/index.asp
Not enough attention has been placed on the management of diabetic patients hospitalized with diabetes as a secondary diagnosis. Old ineffective protocols remain in widespread use. Using the forms and protocols described in this chapter should help improve inpatient diabetes management and bring it in line with outpatient care.
In a separate retrospective review of 1,509,890 patients in the University HealthSystem Consortium database (UHC) and the prospective Mayo Clinic Acute Physiology And Chronic Health Evaluation III critical care database (Mayo), critically ill adults with DM did not have an increased mortality compared with that seen in patients without DM, and may have a decreased mortality [174].
Footnotes