- Email: [email protected]
0096 : Adipocyte mineralocorticoid receptor activation leads to metabolic syndrome and induction of prostaglandin D2 synthase
Diabetes Mellitus and Hyperglycemia Management in the Hospitalized Patient Patricia A. Mackey, FNP-BC, BC-ADM, and Michael D. Whitaker, MD, FRCPC ABSTRACT
Diabetes mellitus and hyperglycemia are common in hospitalized patients. Uncontrolled hyperglycemia during hospitalization is associated with poor outcomes. A glucose goal of 140-180 mg/dL is recommended. Scheduled subcutaneous insulin with basal, prandial, and correction components is preferred for treating diabetes in nonecritically ill patients. The pharmacodynamics of insulins differ, and the type of insulin used should match daily glucose excursions. Varying hospital settings may warrant using a particular insulin type to achieve optimal glucose control. Herein we describe approaches to address hyperglycemia in the hospitalized patient on the basis of insulin pharmacodynamic profiles. Keywords: basal-bolus insulin, correction insulin, diabetes mellitus, hospitalized patient, hyperglycemia Ó 2015 Elsevier, Inc. All rights reserved.
O
ptimal glucose control is a challenge for hospitalized patients. Proper treatment of hyperglycemia while avoiding hypoglycemia should be the goal of multidisciplinary teams (endocrinologists, hospitalists, nurses, surgeons, advanced-level practitioners, pharmacists, and intensivists) working together to provide care for the patient with diabetes mellitus or hyperglycemia in the hospital setting. Hyperglycemia in hospitalized patients can represent previously known diabetes, undiagnosed diabetes, or illness-related hyperglycemia. Hemoglobin A1c values 6.5% suggest that diabetes preceded the hospitalization.1 Numerous studies have indicated that targeted glucose control in the hospital has been shown to improve clinical outcomes; the association between hyperglycemia in hospitalized patients (with or without diabetes) and the increased risk for morbidity and mortality have been well established.2 Challenges encountered in the hospital setting can make controlling glucose difficult. These challenges include a new diagnosis of diabetes; infection; a more rigid diet; inactivity; decreased appetite; variable renal and hepatic status; an unpredictable schedule of testing, procedures, and surgical interventions; the use of hyperglycemic-provoking www.npjournal.org
agents (glucocorticoids, octreotide, catecholamines, and calcineurin inhibitors); the administration of contrast agents with certain tests; enteral and total parenteral nutrition (TPN); and the stress induced by the hospitalization itself. GLYCEMIC GOALS
Although several organizations have issued guidelines for outpatient glucose management, no guidelines or protocols have been formulated for inpatient management. Maintaining glucose levels between 140 and 180 mg/dL is recommended for the majority of hospitalized patients.1 Individualized goals for younger patients without comorbidities (with previous stable glucose control before admission), or for the elderly, terminally ill, or those with extensive comorbidities (eg, congestive heart failure, cirrhosis, and renal failure), have been established for use in the outpatient setting, but no recommendations exist for inpatient glycemic goals for these different groups. Standardized glycemic goals for certain populations of hospitalized patients have suggested that targets < 110 mg/dL are not recommended and may lead to poor outcomes, especially in critically ill patients.1 Recent studies failed to show a significant The Journal for Nurse Practitioners - JNP
531
improvement in mortality with intensive glycemic control.3,4 Hyperglycemia in the hospital (blood glucose > 140 mg/dL) can increase the risk of infections, delay wound healing, and possibly prolong the length of hospital stay. Hyperglycemia has been associated with endothelial dysfunction, oxidative stress, acidosis, caloric and protein losses, electrolyte imbalances, impairment of neutrophil function, and potential exacerbation of myocardial and cerebral ischemia.2 Conversely, hypoglycemia (blood glucose < 70 mg/dL), if brief and mild, may not have long-term sequelae, but severe hypoglycemia (blood glucose < 40 mg/dL) could provoke neurologic effects or seizures, or could trigger arrhythmias or other cardiac events.5 Possible causes of hypoglycemia in the hospital include variability in insulin sensitivity related to the underlying illness, changes in counterregulatory hormonal responses to procedures or illnesses, prolonged nothing-by-mouth (NPO) status, variable doses of dextrose fluids or glucocorticoid therapy, unexpected decreases in food intake or emesis, interruption of enteral or parenteral nutrition, sepsis, concurrent malignancy, use of quinolone antibiotics, or worsening of hepatic or renal function. Providers should be proactive in reducing insulin doses in such settings. MANAGEMENT OF HYPERGLYCEMIA Oral Agents
Oral antidiabetic agents used in the hospital are difficult to titrate, have not been studied for safety and efficacy in hospitalized patients, and may predispose patients to hypoglycemia. Also, the use of noninsulin injectable medications, such as the glucagonlike peptide 1 analogs, may be contraindicated in many inpatient settings. Sulfonylureas act by increasing insulin release from islet cells in the pancreas, which can lead to severe, prolonged hypoglycemia, particularly in elderly patients and patients with poor appetite or impaired renal function.6 Metformin, which suppresses glucose production by the liver, should not be used in patients who have decompensated heart failure, renal insufficiency, or chronic pulmonary disease.1 Diagnostic tests, such as computed tomography scans, that involve IV contrast dye, can put the patient taking metformin at risk of 532
The Journal for Nurse Practitioners - JNP
renal failure or lactic acidosis. Despite being discontinued during the hospital stay, oral agents and non-insulin injectable medications may be resumed at discharge in patients with a history of good glycemic control before hospitalization, who are stable, and have no contraindications. Insulin
For optimal glucose control, scheduled subcutaneous insulin with basal, prandial, and correction components is the preferred treatment method for the nonecritically ill patient in the hospital.1 There is a paucity of accepted insulin algorithms to facilitate inpatient glucose management. Prudent treatment of hyperglycemia while avoiding hypoglycemia is the primary goal. With normal endogenous insulin secretion, the body’s insulin production increases after each meal with consumption of carbohydrates, and then normalizes between meals, bringing serum glucose levels back to within a normal range (Figure, part A). Even with prolonged fasting, blood glucose rarely falls below 50-60 mg/dL, due to hepatic glycogenolysis and subsequent gluconeogenesis. Moreover, endogenous insulin secretion is present to metabolize hepatic glucose production. Understanding this relationship and trying to mimic a natural, physiologic response requires an understanding of insulin pharmacodynamics. Using rapid-acting insulin at mealtimes and a long-acting basal insulin (programs with multiple daily injections) best mimics the natural physiologic responses of the body (Figure, part B). The pharmacodynamics of types of insulin have differences (Table). In the hospital setting, with many extraneous variables, the appropriate types and timing of insulin doses are paramount, as is consideration of the circumstances and the current clinical condition of the patient. Basal-Bolus Insulin Therapy
Basal insulin therapy. Long-acting basal insulins, such as glargine, detemir, or intermediate-acting neutral protamine Hagedorn insulin (NPH), are usually given once or twice daily. Basal insulin prevents ketosis in insulinopenic patients (patients with type 1 diabetes mellitus or post-pancreatectomy diabetes) and, if used properly, will manage fasting Volume 11, Issue 5, May 2015
Figure. (A) Normal glucose-insulin postprandial excursions. The normal postprandial blood glucose excursions in fasting glucose levels over a 24-hour period (solid blue line), with concurrent endogenous insulin responses (red dotted line) are shown. Hepatic glucose production sustains glycemia during the fasting state. (B) Multiple daily injections of insulin. The example shows a subcutaneous basal-bolus insulin regimen matching meal intake and fasting glucose hepatic production. Glargine and aspart are used as insulin examples, but detemir could be substituted for glargine, and any other rapid-acting analog could be substituted for aspart. (C) Insulin action and duration. The example shows a postprandial glucose excursion and superimposed insulin action compares rapidacting aspart insulin to short-acting regular insulin.
hepatic glucose production. Fasting glucose levels are the best indicator of an adequate basal insulin dose. Basal insulin is usually administered at bedtime HS, but administering it in the morning in patients with renal failure can prevent nocturnal hypoglycemia.6 It may be used twice daily in patients receiving continuous nutrition (parenteral or continuous tube feedings), or for enhanced absorption in those requiring large doses (> 100 units) of basal insulin daily. Because of its pharmacodynamic profile (Table), NPH has limited use in the hospital setting. However, it can be effective in hospitalized patients receiving nocturnal tube feedings. Compared with the long-acting analogs glargine or detemir, NPH has been associated with an increased risk of hypoglycemia. Combination insulins, such as Novolin or Humulin 70/30 mixed insulin (70% isophane insulin human suspension [intermediateacting basal insulin] and 30% regular insulin human www.npjournal.org
[short-acting insulin]), can be administered at the start of tube feeding in the hospital, or twice daily before breakfast and before dinner in hospitalized patients who are eating. A disadvantage of using combination premixed insulins is their lack of dosing flexibility. They could be an option at discharge for patients who are on an insulin program of multiple daily injections and who prefer 2 rather than 4 daily insulin injections, or for patients who have limited financial resources, medical insurance, and prescription drug coverage. Combination premixed insulins cost less than the analog mixes (Humalog Mix 75/25 and NovoLog Mix 70/30), or other analog insulins (lispro, aspart, glulisine, detemir, and glargine). Bolus insulin therapy. Bolus insulin (also referred to as prandial or nutrition insulin) is usually given just before meals or at mealtimes (AC), or at the time of bolus tube feedings. The rapid-acting The Journal for Nurse Practitioners - JNP
533
Table. Insulin Pharmacodynamicsa Type of Insulin
Trade Name
Ultrafast-acting
Technosphere insulin (Afrezza) (inhaled)b
Instant
15-20 min
2-3 h
Rapid-acting
Lispro (Humalog)b Aspart (NovoLog)b Glulisine (Apidra)b
10-15 min
45-60 min
3-5 h
Short-acting
Regular
30-60 min
2-4 h
4-8 h
Intermediate-acting
NPH
1-2 h
6-10 h
20-24 h
2h 2h
Small None
24 h 24 h
Long-acting
Onset
b
Detemir (Levemir) Glargine (Lantus)b
Peak
Duration
h ¼ hours; min ¼ minutes; NPH ¼ neutral protamine Hagedorn. a The metabolic effect of various insulins (peaks, onsets, and durations) are shown. Ranges account for patient variability. b
Insulin analog.
insulin analogs are preferred (aspart, glulisine, or lispro) over short-acting regular insulin because of their quicker onset and shorter duration and because they more closely match mealtime glucose excursions (Figure, part C). Prandial insulin analogs can be dosed by the amount of carbohydrates eaten, and they can be given before or immediately after the meal for patients who may not be sure before the meal how much food they will actually consume (eg, patients with nausea or gastroparesis). In the hospital setting, fingerstick point-of-care capillary blood glucose testing is usually ordered at mealtimes and at bedtime. Scheduled prandial insulin is ordered for mealtimes to cover the carbohydrates consumed. CORRECTION INSULIN THERAPY
Correction insulin (also known as sliding scale insulin) is administered to correct a glucose level above a certain target. Correction scales based on insulin sensitivity can be mild (for frail, elderly, thin, insulin-sensitive patients, or for patients with renal insufficiency), moderate (for most patients), aggressive (for obese, insulin-resistant patients on high-dose steroids), or individually providerdirected, based on the particular situation of the individual patient. Correction insulin, if needed, is usually added to the scheduled prandial dose for a single injection. Sometimes, in the hospital, correction insulin is also administered at HS. Correction insulin is intended to lower high glucose levels, not to cover nutritional glucose intake.7 Using correction insulin alone in 534
The Journal for Nurse Practitioners - JNP
the hospital setting is neither recommended nor effective.8 A physiologic insulin regimen is indeed superior to a standardized insulin sliding scale for management of hyperglycemia in hospitalized patients. Using correction insulin alone, without basal and prandial insulin components, puts patients with type 1 diabetes and insulin deficiency at risk for ketoacidosis, and it may be detrimental to glucose control. It is a reactive approach to treating hyperglycemia, rather than a proactive or an anticipatory approach. As a result of previous findings, many hospitals are not using correction insulin as a stand-alone therapy for treating diabetes and hyperglycemia in inpatients, and they instead are using a basal-bolus insulin regimen to improve inpatient glucose control.3,9 When correction scales are used alone for several days (one size fits all), glycemic control is often not assessed, which may result in wide swings in glucose levels throughout the hospitalization period. Correction scale insulin can, however, be useful in the first 24-48 hours of hospitalization. It can serve as an insulin doseefinding strategy before an individually appropriate basal-bolus program can be determined and started. A general guideline for calculating insulin sensitivity to select an appropriate correction scale is to use the “Rule of 1500,” which calculates the insulin sensitivity factor for regular insulin. The sensitivity factor shows how far blood sugar will decrease per unit of regular insulin: 1500 divided by the total daily dose (TDD) of insulin ¼ correction Volume 11, Issue 5, May 2015
factor, which equals the interval in a correction scale (also known as the insulin sensitivity).10 The “Rule of 1800” was later developed to show the sensitivity factor per unit of analog insulin (NovoLog, Humalog, or Apidra) because blood sugar will decrease faster with analog insulins. This rule is 1800 divided by the TDD of insulin ¼ correction factor/insulin sensitivity. As an example, when using the rule of 1500 with 40 units of TDD insulin per day (eg, NPH þ regular), the correction factor would be 1500 / 40 or 38. When using an analog insulin, and the rule of 1800 with 40 units of TDD (eg, NPH þ NovoLog), the correction factor would be 1800 / 40 or 45. TPN AND INSULIN
Correction insulin is usually administered intermittently (4-6 times daily), if necessary, for patients who are NPO on continuous tube feedings, or on parenteral nutrition or TPN. Checking the capillary glucose of patients on continuous TPN therapy every 4 hours, and using correction insulin if necessary at those times, may result in better glucose control. This is because of the loss of incretin effect with the use of TPN (IV nutrition bypasses the intestinal regulators of glucose metabolism), which is also often high in carbohydrates. Instead of a separate basal insulin injection administered concurrently with the TPN, the total daily basal requirement of insulin, once established, can be incorporated as short-acting regular insulin into the TPN bag. When initially adding insulin to the TPN bag for the first time, a good starting point would be to add 0.1 unit of regular insulin for every gram of carbohydrate in the bag.7 Any scheduled basal insulin that was received by the patient in the 24 hours before TPN was initiated can be incorporated into the TPN bag as well (as regular insulin). The scheduled subcutaneous basal insulin (eg, detemir, NPH, or glargine) is then discontinued. The correction scale insulin is continued (rapid-acting or short-acting) to supplement the insulin in the TPN bag. The TDD of insulin incorporated into the TPN bag is adjusted daily, depending on the additional rapid-acting or short-acting correction insulin used by the patient during the previous 24-hour period. www.npjournal.org
INITIAL START OF INSULIN
When an inpatient (who is not transitioning off an insulin drip) is started for the first time on a basalbolus insulin regimen in the hospital, a weight-based calculation can be used as a starting point. The TDD ¼ patient weight (in kilograms) 0.5, with 50% of the dose being the basal requirement and 50% of the dose being the prandial requirement. To start a 220-pound (100-kg) patient who is eating well on a basal-bolus insulin regimen, multiply weight (in kilograms) by 0.5 (100 0.5). The result is 50 units as the TDD of insulin. Then divide 50 units by 2, and one half of the TDD is the basal insulin requirement (25 units) and the other half is the prandial requirement (25 units). The prandial doses should be ordered as 8 units with each meal, that is, 3 times a day. If that same patient were to be made NPO, the scheduled prandial insulin should not be ordered; the patient would have just the basal insulin (25 units). Order an appropriate correction scale using rapid-acting insulin for AC and HS to be administered concomitantly with the scheduled prandial and basal insulins. Because of the current limited length of stay of hospitalized patients, basal insulin should be titrated daily by 10%-20% until the fasting glucose is at goal. The prandial insulins should also be titrated daily by 10%-20% until the pre-meal glucoses are at goal range, and bedtime glucose is < 180 mg/dL (preferably < 140 mg/dL), to optimize glucose control during the hospitalization. Using every opportunity throughout the patient’s hospitalization to make an insulin adjustment, if necessary, will help to determine the most accurate insulin doses at the time of discharge. IV INSULIN INFUSIONS
IV insulin infusions are often used in the intensive care unit to attain optimal glucose control in patients with glucose levels > 180 mg/dL. Indications for use of an IV insulin infusion occur in patients with uncontrolled blood glucose (> 400 mg/dL), diabetic ketoacidosis, and hyperosmolar hyperglycemic nonketotic syndrome; posteorgan transplant patients on high-dose corticosteroids in the immediate postoperative period; postemyocardial infarct patients or those having cardiac surgery; patients admitted on The Journal for Nurse Practitioners - JNP
535
insulin pumps undergoing certain types of surgery involving high doses of steroids (the insulin pump is disconnected); and post-pancreatectomy patients to determine insulin requirements in the immediate postoperative period. IV insulin infusions require hourly blood glucose testing and rate adjustment, if necessary. They can be effective at keeping glucose levels in the targeted ranges, and they allow for easy titration of the insulin. IV insulin algorithms are standardized guidelines for dose adjustment that are based on the glucose level. Appropriate doses may depend on the degree of illness, body weight, or medications. Many hospitals use predetermined algorithms, from mild (for frail, elderly, type 1 diabetic, post-pancreatectomy, or postebariatric surgery patients) to more aggressive (for morbidly obese or insulin-resistant patients or those on high doses of glucocorticoids). Some facilities are using Glucommander (Glytec LLC) as a method of maintaining IV glucose control. It is a safe, computer-directed algorithm that has been studied extensively, and it can be effective throughout the hospital, not just in the intensive care unit.11 If the patient is clinically stable, has reached a steady state of glucose control, is recovering from critical illness, is beginning to eat regular meals, or is transferred to a general nursing unit, he or she may be ready for the transition from IV to subcutaneous insulin. All patients with type 1 or type 2 diabetes mellitus should be transitioned to subcutaneous longacting or intermediate-acting insulin at least 2 hours before discontinuation of the IV insulin infusion. This overlap between the discontinuation of the insulin drip and administration of subcutaneous basal insulin is to prevent recurrent hyperglycemia during the transition period. When transitioning a stable inpatient off an insulin drip who has not been eating while on the drip, it is reasonable to add up the hourly infusion rates (per hour) over the previous 6 hours, and then multiply by 4 to calculate the TDD of basal insulin requirement. The premise is that, because of the steady state of IV insulin infusion over the previous 6 hours, this value will probably best reflect the current clinical state and insulin requirements of the patient for a 24-hour period. Another option would be to take the average hourly rate over the previous 6 hours and 536
The Journal for Nurse Practitioners - JNP
multiply that by 24 to obtain the TDD of basal insulin. Provider discretion and clinical judgment can dictate whether the TDD could be reduced by 10%-20%, depending on the clinical situation of the patient (eg, concomitant medication dose changes, nutritional status, and severity of illness). One half of the TDD of basal insulin is administered subcutaneously at the time of transition off the insulin drip, and the other half is administered at bedtime that evening. The insulin drip is typically discontinued 2 hours after the first subcutaneous injection of basal insulin. If the patient was eating while on the insulin infusion, the TDD value should be divided into 50% prandial insulin and 50% basal insulin doses. The 50% total prandial dose value would then be divided by 3 for a scheduled dose of insulin at each mealtime of the day. A correction insulin scale is also ordered for AC and HS. If the patient has not been eating 100% of their meals but has been eating some while on the IV insulin infusion, reducing the amount of scheduled prandial insulin by 10%-20% would be reasonable. Consider this example of transitioning a stable patient who has been NPO from an IV insulin infusion to subcutaneous insulin. If the patient has been receiving a stable rate of 4 units of insulin per hour over the previous 6 hours, multiply 6 (hours) 4 (units/hour) ¼ 24 units, and then multiply 24 (units) 4 ¼ 96 units to determine the total daily (24-hour) basal insulin requirement. Consider ordering about 80% of this amount if the patient is not as acutely ill and if doses are to be decreased for vasopressors, steroids, or any other medications that would affect glucose levels. If ordering 80% of the total basal insulin requirement, then 38 units of basal insulin would be ordered once (80% of 96 units ¼ 77 units, then divided by 2 ¼ 38 units), and the insulin infusion would be continued for another 2 hours. The nighttime basal insulin dose would also be 38 units. If this same patient had been eating 100% of their meals while on the insulin infusion, the 96 units TDD would be 48 units for basal and 48 units for prandial insulin requirements. If the total insulin requirement was not being decreased by 80%, in this case, the basal insulin would be 24 units once, another 24 units at bedtime, and the prandial insulin would be 16 units 3 times per day with meals. Volume 11, Issue 5, May 2015
INSULIN PUMP THERAPY
With the increased utilization of insulin pump therapy, many institutions are allowing patients to remain on these devices during their hospitalization, including during some types of surgery.12 Patients who use continuous subcutaneous insulin infusion in the outpatient setting can be considered candidates for continuation of their insulin pump throughout their hospitalization, barring no contraindications and if they have the physical and mental capacity to do so. The hospital should have a policy and procedures delineating the inpatient guidelines for continuous subcutaneous insulin infusion. Throughout the hospitalization, basal rates, assessments of infusion site, pump settings, and bolus insulin doses should be accurately documented in the patient’s medical record. CONCLUSION
Data continue to accumulate regarding adverse effects of poorly controlled diabetes and hyperglycemia in the hospitalized patient. Institutions worldwide are striving to develop safe and effective ways to manage diabetes and hyperglycemia proactively in the hospital setting. Frequent and effective glucose monitoring is critical, as is judicious treatment of hyperglycemia while avoiding hypoglycemia. Overall, glucose targets near 140 mg/dL are recommended as the most appropriate for the majority of hospitalized patients, and insulin remains the most appropriate agent for management of inpatient hyperglycemia. A subcutaneous basal, bolus, and correction insulin regimen is the preferred method of achieving and managing glucose control in nonecritically ill hospitalized patients.1 Continued use of correction insulin scales alone as a method of controlling hyperglycemia is not recommended. Glycemic management of hospitalized patients requires ongoing team efforts, education of all multidisciplinary team members, and careful implementation of standardized approaches to help ensure patient safety and promote the best possible outcomes.
www.npjournal.org
Follow-up after hospital discharge requires planning to provide a smooth transition to outpatient care. A home health referral should be considered for the patient, and resources in the community should be identified. Outpatient diabetes self-management is critical to minimizing risks of future complications. References 1. American Diabetes Association. Standards of medical care in diabetes: 2014. Diabetes Care. 2014;37(Suppl 1):S14-S80. 2. Umpierrez GE, Hellman R, Korytkowski MT, et al. Management of hyperglycemia in hospitalized patients in non-critical care setting: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2012;97(1):16-38. 3. Kosiborod M, Inzucchi SE, Krumholz HM, et al. Glucose normalization and outcomes in patients with acute myocardial infarction. Arch Intern Med. 2009;169(5):438-446. 4. Murphy DM, Vercruysse RA, Bertucci TM, et al. Reducing hyperglycemia hospitalwide: the basal-bolus concept. Jt Comm J Qual Patient Saf. 2009;35(4):216-223. 5. Ghosal S, Sinha B. Hypoglycaemia and CV risk: perceptions and reality. Diabetes Metab Syndr. http://dx.doi.org/10.1016/j.dsx.2013.10.021. [Epub ahead of print]. 6. Moghissi ES. Addressing hyperglycemia from hospital admission to discharge. Curr Med Res Opin. 2010;26(3):589-598. 7. Magaji V, Johnston JM. Inpatient management of hyperglycemia and diabetes. Clin Diabetes. 2011;29(1):36-39. 8. Cobaugh DJ, Maynard G, Cooper L, et al. Enhancing insulin-use safety in hospitals: practical recommendations from an ASHP Foundation expert consensus panel. Am J Health Syst Pharm. 2013;70(16):1404-1413. 9. Chen HJ, Steinke DT, Karounos DG, Lane MT, Matson AW. Intensive insulin protocol implementation and outcomes in the medical and surgical wards at a Veterans Affairs Medical Center. Ann Pharmacother. 2010;44(2):249-556. 10. Dinsmoor RS. Insulin sensitivity factor. Diabetes self-management. February 21, 2013. http://www.diabetesselfmanagement.com/Articles/Diabetes -Definitions/insulin_sensitivity_factor/. Accessed October 9, 2014. 11. Yamashita S, Ng E, Brommecker F, Silverberg J, Adhikari NK. Implementation of the glucommander method of adjusting insulin infusions in critically ill patients. Can J Hosp Pharm. 2011;64(5):333-339. 12. Cook CB, Beer KA, Seifert KM, Boyle ME, Mackey PA, Castro JC. Transitioning insulin pump therapy from the outpatient to the inpatient setting: a review of 6 years’ experience with 253 cases. J Diabetes Sci Technol. 2012;6(5):995-1002.
Both authors are affiliated with the Division of Endocrinology at the Mayo Clinic in Scottsdale, AZ. Patricia A. Mackey, FNPBC, BC-ADM, CDE, is an assistant professor of medicine, Mayo Clinic College of Medicine and can be reached at mackey. [email protected]. Michael D. Whitaker, MD, FRCPC, is a consultant in endocrinology, assistant professor of Medicine, Mayo Clinic College of Medicine. In compliance with national ethical guidelines, the authors report no relationships with business or industry that would pose a conflict of interest. 1555-4155/15/$ see front matter © 2015 Elsevier, Inc. All rights reserved. http://dx.doi.org/10.1016/j.nurpra.2015.02.016
The Journal for Nurse Practitioners - JNP
537
Diabetes mellitus and hyperglycemia are common in hospitalized patients. Uncontrolled hyperglycemia during hospitalization is associated with poor outcomes. A glucose goal of 140-180 mg/dL is recommended. Scheduled subcutaneous insulin with basal, prandial, and correction components is preferred for treating diabetes in nonecritically ill patients. The pharmacodynamics of insulins differ, and the type of insulin used should match daily glucose excursions. Varying hospital settings may warrant using a particular insulin type to achieve optimal glucose control. Herein we describe approaches to address hyperglycemia in the hospitalized patient on the basis of insulin pharmacodynamic profiles. Keywords: basal-bolus insulin, correction insulin, diabetes mellitus, hospitalized patient, hyperglycemia Ó 2015 Elsevier, Inc. All rights reserved.
O
ptimal glucose control is a challenge for hospitalized patients. Proper treatment of hyperglycemia while avoiding hypoglycemia should be the goal of multidisciplinary teams (endocrinologists, hospitalists, nurses, surgeons, advanced-level practitioners, pharmacists, and intensivists) working together to provide care for the patient with diabetes mellitus or hyperglycemia in the hospital setting. Hyperglycemia in hospitalized patients can represent previously known diabetes, undiagnosed diabetes, or illness-related hyperglycemia. Hemoglobin A1c values 6.5% suggest that diabetes preceded the hospitalization.1 Numerous studies have indicated that targeted glucose control in the hospital has been shown to improve clinical outcomes; the association between hyperglycemia in hospitalized patients (with or without diabetes) and the increased risk for morbidity and mortality have been well established.2 Challenges encountered in the hospital setting can make controlling glucose difficult. These challenges include a new diagnosis of diabetes; infection; a more rigid diet; inactivity; decreased appetite; variable renal and hepatic status; an unpredictable schedule of testing, procedures, and surgical interventions; the use of hyperglycemic-provoking www.npjournal.org
agents (glucocorticoids, octreotide, catecholamines, and calcineurin inhibitors); the administration of contrast agents with certain tests; enteral and total parenteral nutrition (TPN); and the stress induced by the hospitalization itself. GLYCEMIC GOALS
Although several organizations have issued guidelines for outpatient glucose management, no guidelines or protocols have been formulated for inpatient management. Maintaining glucose levels between 140 and 180 mg/dL is recommended for the majority of hospitalized patients.1 Individualized goals for younger patients without comorbidities (with previous stable glucose control before admission), or for the elderly, terminally ill, or those with extensive comorbidities (eg, congestive heart failure, cirrhosis, and renal failure), have been established for use in the outpatient setting, but no recommendations exist for inpatient glycemic goals for these different groups. Standardized glycemic goals for certain populations of hospitalized patients have suggested that targets < 110 mg/dL are not recommended and may lead to poor outcomes, especially in critically ill patients.1 Recent studies failed to show a significant The Journal for Nurse Practitioners - JNP
531
improvement in mortality with intensive glycemic control.3,4 Hyperglycemia in the hospital (blood glucose > 140 mg/dL) can increase the risk of infections, delay wound healing, and possibly prolong the length of hospital stay. Hyperglycemia has been associated with endothelial dysfunction, oxidative stress, acidosis, caloric and protein losses, electrolyte imbalances, impairment of neutrophil function, and potential exacerbation of myocardial and cerebral ischemia.2 Conversely, hypoglycemia (blood glucose < 70 mg/dL), if brief and mild, may not have long-term sequelae, but severe hypoglycemia (blood glucose < 40 mg/dL) could provoke neurologic effects or seizures, or could trigger arrhythmias or other cardiac events.5 Possible causes of hypoglycemia in the hospital include variability in insulin sensitivity related to the underlying illness, changes in counterregulatory hormonal responses to procedures or illnesses, prolonged nothing-by-mouth (NPO) status, variable doses of dextrose fluids or glucocorticoid therapy, unexpected decreases in food intake or emesis, interruption of enteral or parenteral nutrition, sepsis, concurrent malignancy, use of quinolone antibiotics, or worsening of hepatic or renal function. Providers should be proactive in reducing insulin doses in such settings. MANAGEMENT OF HYPERGLYCEMIA Oral Agents
Oral antidiabetic agents used in the hospital are difficult to titrate, have not been studied for safety and efficacy in hospitalized patients, and may predispose patients to hypoglycemia. Also, the use of noninsulin injectable medications, such as the glucagonlike peptide 1 analogs, may be contraindicated in many inpatient settings. Sulfonylureas act by increasing insulin release from islet cells in the pancreas, which can lead to severe, prolonged hypoglycemia, particularly in elderly patients and patients with poor appetite or impaired renal function.6 Metformin, which suppresses glucose production by the liver, should not be used in patients who have decompensated heart failure, renal insufficiency, or chronic pulmonary disease.1 Diagnostic tests, such as computed tomography scans, that involve IV contrast dye, can put the patient taking metformin at risk of 532
The Journal for Nurse Practitioners - JNP
renal failure or lactic acidosis. Despite being discontinued during the hospital stay, oral agents and non-insulin injectable medications may be resumed at discharge in patients with a history of good glycemic control before hospitalization, who are stable, and have no contraindications. Insulin
For optimal glucose control, scheduled subcutaneous insulin with basal, prandial, and correction components is the preferred treatment method for the nonecritically ill patient in the hospital.1 There is a paucity of accepted insulin algorithms to facilitate inpatient glucose management. Prudent treatment of hyperglycemia while avoiding hypoglycemia is the primary goal. With normal endogenous insulin secretion, the body’s insulin production increases after each meal with consumption of carbohydrates, and then normalizes between meals, bringing serum glucose levels back to within a normal range (Figure, part A). Even with prolonged fasting, blood glucose rarely falls below 50-60 mg/dL, due to hepatic glycogenolysis and subsequent gluconeogenesis. Moreover, endogenous insulin secretion is present to metabolize hepatic glucose production. Understanding this relationship and trying to mimic a natural, physiologic response requires an understanding of insulin pharmacodynamics. Using rapid-acting insulin at mealtimes and a long-acting basal insulin (programs with multiple daily injections) best mimics the natural physiologic responses of the body (Figure, part B). The pharmacodynamics of types of insulin have differences (Table). In the hospital setting, with many extraneous variables, the appropriate types and timing of insulin doses are paramount, as is consideration of the circumstances and the current clinical condition of the patient. Basal-Bolus Insulin Therapy
Basal insulin therapy. Long-acting basal insulins, such as glargine, detemir, or intermediate-acting neutral protamine Hagedorn insulin (NPH), are usually given once or twice daily. Basal insulin prevents ketosis in insulinopenic patients (patients with type 1 diabetes mellitus or post-pancreatectomy diabetes) and, if used properly, will manage fasting Volume 11, Issue 5, May 2015
Figure. (A) Normal glucose-insulin postprandial excursions. The normal postprandial blood glucose excursions in fasting glucose levels over a 24-hour period (solid blue line), with concurrent endogenous insulin responses (red dotted line) are shown. Hepatic glucose production sustains glycemia during the fasting state. (B) Multiple daily injections of insulin. The example shows a subcutaneous basal-bolus insulin regimen matching meal intake and fasting glucose hepatic production. Glargine and aspart are used as insulin examples, but detemir could be substituted for glargine, and any other rapid-acting analog could be substituted for aspart. (C) Insulin action and duration. The example shows a postprandial glucose excursion and superimposed insulin action compares rapidacting aspart insulin to short-acting regular insulin.
hepatic glucose production. Fasting glucose levels are the best indicator of an adequate basal insulin dose. Basal insulin is usually administered at bedtime HS, but administering it in the morning in patients with renal failure can prevent nocturnal hypoglycemia.6 It may be used twice daily in patients receiving continuous nutrition (parenteral or continuous tube feedings), or for enhanced absorption in those requiring large doses (> 100 units) of basal insulin daily. Because of its pharmacodynamic profile (Table), NPH has limited use in the hospital setting. However, it can be effective in hospitalized patients receiving nocturnal tube feedings. Compared with the long-acting analogs glargine or detemir, NPH has been associated with an increased risk of hypoglycemia. Combination insulins, such as Novolin or Humulin 70/30 mixed insulin (70% isophane insulin human suspension [intermediateacting basal insulin] and 30% regular insulin human www.npjournal.org
[short-acting insulin]), can be administered at the start of tube feeding in the hospital, or twice daily before breakfast and before dinner in hospitalized patients who are eating. A disadvantage of using combination premixed insulins is their lack of dosing flexibility. They could be an option at discharge for patients who are on an insulin program of multiple daily injections and who prefer 2 rather than 4 daily insulin injections, or for patients who have limited financial resources, medical insurance, and prescription drug coverage. Combination premixed insulins cost less than the analog mixes (Humalog Mix 75/25 and NovoLog Mix 70/30), or other analog insulins (lispro, aspart, glulisine, detemir, and glargine). Bolus insulin therapy. Bolus insulin (also referred to as prandial or nutrition insulin) is usually given just before meals or at mealtimes (AC), or at the time of bolus tube feedings. The rapid-acting The Journal for Nurse Practitioners - JNP
533
Table. Insulin Pharmacodynamicsa Type of Insulin
Trade Name
Ultrafast-acting
Technosphere insulin (Afrezza) (inhaled)b
Instant
15-20 min
2-3 h
Rapid-acting
Lispro (Humalog)b Aspart (NovoLog)b Glulisine (Apidra)b
10-15 min
45-60 min
3-5 h
Short-acting
Regular
30-60 min
2-4 h
4-8 h
Intermediate-acting
NPH
1-2 h
6-10 h
20-24 h
2h 2h
Small None
24 h 24 h
Long-acting
Onset
b
Detemir (Levemir) Glargine (Lantus)b
Peak
Duration
h ¼ hours; min ¼ minutes; NPH ¼ neutral protamine Hagedorn. a The metabolic effect of various insulins (peaks, onsets, and durations) are shown. Ranges account for patient variability. b
Insulin analog.
insulin analogs are preferred (aspart, glulisine, or lispro) over short-acting regular insulin because of their quicker onset and shorter duration and because they more closely match mealtime glucose excursions (Figure, part C). Prandial insulin analogs can be dosed by the amount of carbohydrates eaten, and they can be given before or immediately after the meal for patients who may not be sure before the meal how much food they will actually consume (eg, patients with nausea or gastroparesis). In the hospital setting, fingerstick point-of-care capillary blood glucose testing is usually ordered at mealtimes and at bedtime. Scheduled prandial insulin is ordered for mealtimes to cover the carbohydrates consumed. CORRECTION INSULIN THERAPY
Correction insulin (also known as sliding scale insulin) is administered to correct a glucose level above a certain target. Correction scales based on insulin sensitivity can be mild (for frail, elderly, thin, insulin-sensitive patients, or for patients with renal insufficiency), moderate (for most patients), aggressive (for obese, insulin-resistant patients on high-dose steroids), or individually providerdirected, based on the particular situation of the individual patient. Correction insulin, if needed, is usually added to the scheduled prandial dose for a single injection. Sometimes, in the hospital, correction insulin is also administered at HS. Correction insulin is intended to lower high glucose levels, not to cover nutritional glucose intake.7 Using correction insulin alone in 534
The Journal for Nurse Practitioners - JNP
the hospital setting is neither recommended nor effective.8 A physiologic insulin regimen is indeed superior to a standardized insulin sliding scale for management of hyperglycemia in hospitalized patients. Using correction insulin alone, without basal and prandial insulin components, puts patients with type 1 diabetes and insulin deficiency at risk for ketoacidosis, and it may be detrimental to glucose control. It is a reactive approach to treating hyperglycemia, rather than a proactive or an anticipatory approach. As a result of previous findings, many hospitals are not using correction insulin as a stand-alone therapy for treating diabetes and hyperglycemia in inpatients, and they instead are using a basal-bolus insulin regimen to improve inpatient glucose control.3,9 When correction scales are used alone for several days (one size fits all), glycemic control is often not assessed, which may result in wide swings in glucose levels throughout the hospitalization period. Correction scale insulin can, however, be useful in the first 24-48 hours of hospitalization. It can serve as an insulin doseefinding strategy before an individually appropriate basal-bolus program can be determined and started. A general guideline for calculating insulin sensitivity to select an appropriate correction scale is to use the “Rule of 1500,” which calculates the insulin sensitivity factor for regular insulin. The sensitivity factor shows how far blood sugar will decrease per unit of regular insulin: 1500 divided by the total daily dose (TDD) of insulin ¼ correction Volume 11, Issue 5, May 2015
factor, which equals the interval in a correction scale (also known as the insulin sensitivity).10 The “Rule of 1800” was later developed to show the sensitivity factor per unit of analog insulin (NovoLog, Humalog, or Apidra) because blood sugar will decrease faster with analog insulins. This rule is 1800 divided by the TDD of insulin ¼ correction factor/insulin sensitivity. As an example, when using the rule of 1500 with 40 units of TDD insulin per day (eg, NPH þ regular), the correction factor would be 1500 / 40 or 38. When using an analog insulin, and the rule of 1800 with 40 units of TDD (eg, NPH þ NovoLog), the correction factor would be 1800 / 40 or 45. TPN AND INSULIN
Correction insulin is usually administered intermittently (4-6 times daily), if necessary, for patients who are NPO on continuous tube feedings, or on parenteral nutrition or TPN. Checking the capillary glucose of patients on continuous TPN therapy every 4 hours, and using correction insulin if necessary at those times, may result in better glucose control. This is because of the loss of incretin effect with the use of TPN (IV nutrition bypasses the intestinal regulators of glucose metabolism), which is also often high in carbohydrates. Instead of a separate basal insulin injection administered concurrently with the TPN, the total daily basal requirement of insulin, once established, can be incorporated as short-acting regular insulin into the TPN bag. When initially adding insulin to the TPN bag for the first time, a good starting point would be to add 0.1 unit of regular insulin for every gram of carbohydrate in the bag.7 Any scheduled basal insulin that was received by the patient in the 24 hours before TPN was initiated can be incorporated into the TPN bag as well (as regular insulin). The scheduled subcutaneous basal insulin (eg, detemir, NPH, or glargine) is then discontinued. The correction scale insulin is continued (rapid-acting or short-acting) to supplement the insulin in the TPN bag. The TDD of insulin incorporated into the TPN bag is adjusted daily, depending on the additional rapid-acting or short-acting correction insulin used by the patient during the previous 24-hour period. www.npjournal.org
INITIAL START OF INSULIN
When an inpatient (who is not transitioning off an insulin drip) is started for the first time on a basalbolus insulin regimen in the hospital, a weight-based calculation can be used as a starting point. The TDD ¼ patient weight (in kilograms) 0.5, with 50% of the dose being the basal requirement and 50% of the dose being the prandial requirement. To start a 220-pound (100-kg) patient who is eating well on a basal-bolus insulin regimen, multiply weight (in kilograms) by 0.5 (100 0.5). The result is 50 units as the TDD of insulin. Then divide 50 units by 2, and one half of the TDD is the basal insulin requirement (25 units) and the other half is the prandial requirement (25 units). The prandial doses should be ordered as 8 units with each meal, that is, 3 times a day. If that same patient were to be made NPO, the scheduled prandial insulin should not be ordered; the patient would have just the basal insulin (25 units). Order an appropriate correction scale using rapid-acting insulin for AC and HS to be administered concomitantly with the scheduled prandial and basal insulins. Because of the current limited length of stay of hospitalized patients, basal insulin should be titrated daily by 10%-20% until the fasting glucose is at goal. The prandial insulins should also be titrated daily by 10%-20% until the pre-meal glucoses are at goal range, and bedtime glucose is < 180 mg/dL (preferably < 140 mg/dL), to optimize glucose control during the hospitalization. Using every opportunity throughout the patient’s hospitalization to make an insulin adjustment, if necessary, will help to determine the most accurate insulin doses at the time of discharge. IV INSULIN INFUSIONS
IV insulin infusions are often used in the intensive care unit to attain optimal glucose control in patients with glucose levels > 180 mg/dL. Indications for use of an IV insulin infusion occur in patients with uncontrolled blood glucose (> 400 mg/dL), diabetic ketoacidosis, and hyperosmolar hyperglycemic nonketotic syndrome; posteorgan transplant patients on high-dose corticosteroids in the immediate postoperative period; postemyocardial infarct patients or those having cardiac surgery; patients admitted on The Journal for Nurse Practitioners - JNP
535
insulin pumps undergoing certain types of surgery involving high doses of steroids (the insulin pump is disconnected); and post-pancreatectomy patients to determine insulin requirements in the immediate postoperative period. IV insulin infusions require hourly blood glucose testing and rate adjustment, if necessary. They can be effective at keeping glucose levels in the targeted ranges, and they allow for easy titration of the insulin. IV insulin algorithms are standardized guidelines for dose adjustment that are based on the glucose level. Appropriate doses may depend on the degree of illness, body weight, or medications. Many hospitals use predetermined algorithms, from mild (for frail, elderly, type 1 diabetic, post-pancreatectomy, or postebariatric surgery patients) to more aggressive (for morbidly obese or insulin-resistant patients or those on high doses of glucocorticoids). Some facilities are using Glucommander (Glytec LLC) as a method of maintaining IV glucose control. It is a safe, computer-directed algorithm that has been studied extensively, and it can be effective throughout the hospital, not just in the intensive care unit.11 If the patient is clinically stable, has reached a steady state of glucose control, is recovering from critical illness, is beginning to eat regular meals, or is transferred to a general nursing unit, he or she may be ready for the transition from IV to subcutaneous insulin. All patients with type 1 or type 2 diabetes mellitus should be transitioned to subcutaneous longacting or intermediate-acting insulin at least 2 hours before discontinuation of the IV insulin infusion. This overlap between the discontinuation of the insulin drip and administration of subcutaneous basal insulin is to prevent recurrent hyperglycemia during the transition period. When transitioning a stable inpatient off an insulin drip who has not been eating while on the drip, it is reasonable to add up the hourly infusion rates (per hour) over the previous 6 hours, and then multiply by 4 to calculate the TDD of basal insulin requirement. The premise is that, because of the steady state of IV insulin infusion over the previous 6 hours, this value will probably best reflect the current clinical state and insulin requirements of the patient for a 24-hour period. Another option would be to take the average hourly rate over the previous 6 hours and 536
The Journal for Nurse Practitioners - JNP
multiply that by 24 to obtain the TDD of basal insulin. Provider discretion and clinical judgment can dictate whether the TDD could be reduced by 10%-20%, depending on the clinical situation of the patient (eg, concomitant medication dose changes, nutritional status, and severity of illness). One half of the TDD of basal insulin is administered subcutaneously at the time of transition off the insulin drip, and the other half is administered at bedtime that evening. The insulin drip is typically discontinued 2 hours after the first subcutaneous injection of basal insulin. If the patient was eating while on the insulin infusion, the TDD value should be divided into 50% prandial insulin and 50% basal insulin doses. The 50% total prandial dose value would then be divided by 3 for a scheduled dose of insulin at each mealtime of the day. A correction insulin scale is also ordered for AC and HS. If the patient has not been eating 100% of their meals but has been eating some while on the IV insulin infusion, reducing the amount of scheduled prandial insulin by 10%-20% would be reasonable. Consider this example of transitioning a stable patient who has been NPO from an IV insulin infusion to subcutaneous insulin. If the patient has been receiving a stable rate of 4 units of insulin per hour over the previous 6 hours, multiply 6 (hours) 4 (units/hour) ¼ 24 units, and then multiply 24 (units) 4 ¼ 96 units to determine the total daily (24-hour) basal insulin requirement. Consider ordering about 80% of this amount if the patient is not as acutely ill and if doses are to be decreased for vasopressors, steroids, or any other medications that would affect glucose levels. If ordering 80% of the total basal insulin requirement, then 38 units of basal insulin would be ordered once (80% of 96 units ¼ 77 units, then divided by 2 ¼ 38 units), and the insulin infusion would be continued for another 2 hours. The nighttime basal insulin dose would also be 38 units. If this same patient had been eating 100% of their meals while on the insulin infusion, the 96 units TDD would be 48 units for basal and 48 units for prandial insulin requirements. If the total insulin requirement was not being decreased by 80%, in this case, the basal insulin would be 24 units once, another 24 units at bedtime, and the prandial insulin would be 16 units 3 times per day with meals. Volume 11, Issue 5, May 2015
INSULIN PUMP THERAPY
With the increased utilization of insulin pump therapy, many institutions are allowing patients to remain on these devices during their hospitalization, including during some types of surgery.12 Patients who use continuous subcutaneous insulin infusion in the outpatient setting can be considered candidates for continuation of their insulin pump throughout their hospitalization, barring no contraindications and if they have the physical and mental capacity to do so. The hospital should have a policy and procedures delineating the inpatient guidelines for continuous subcutaneous insulin infusion. Throughout the hospitalization, basal rates, assessments of infusion site, pump settings, and bolus insulin doses should be accurately documented in the patient’s medical record. CONCLUSION
Data continue to accumulate regarding adverse effects of poorly controlled diabetes and hyperglycemia in the hospitalized patient. Institutions worldwide are striving to develop safe and effective ways to manage diabetes and hyperglycemia proactively in the hospital setting. Frequent and effective glucose monitoring is critical, as is judicious treatment of hyperglycemia while avoiding hypoglycemia. Overall, glucose targets near 140 mg/dL are recommended as the most appropriate for the majority of hospitalized patients, and insulin remains the most appropriate agent for management of inpatient hyperglycemia. A subcutaneous basal, bolus, and correction insulin regimen is the preferred method of achieving and managing glucose control in nonecritically ill hospitalized patients.1 Continued use of correction insulin scales alone as a method of controlling hyperglycemia is not recommended. Glycemic management of hospitalized patients requires ongoing team efforts, education of all multidisciplinary team members, and careful implementation of standardized approaches to help ensure patient safety and promote the best possible outcomes.
www.npjournal.org
Follow-up after hospital discharge requires planning to provide a smooth transition to outpatient care. A home health referral should be considered for the patient, and resources in the community should be identified. Outpatient diabetes self-management is critical to minimizing risks of future complications. References 1. American Diabetes Association. Standards of medical care in diabetes: 2014. Diabetes Care. 2014;37(Suppl 1):S14-S80. 2. Umpierrez GE, Hellman R, Korytkowski MT, et al. Management of hyperglycemia in hospitalized patients in non-critical care setting: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2012;97(1):16-38. 3. Kosiborod M, Inzucchi SE, Krumholz HM, et al. Glucose normalization and outcomes in patients with acute myocardial infarction. Arch Intern Med. 2009;169(5):438-446. 4. Murphy DM, Vercruysse RA, Bertucci TM, et al. Reducing hyperglycemia hospitalwide: the basal-bolus concept. Jt Comm J Qual Patient Saf. 2009;35(4):216-223. 5. Ghosal S, Sinha B. Hypoglycaemia and CV risk: perceptions and reality. Diabetes Metab Syndr. http://dx.doi.org/10.1016/j.dsx.2013.10.021. [Epub ahead of print]. 6. Moghissi ES. Addressing hyperglycemia from hospital admission to discharge. Curr Med Res Opin. 2010;26(3):589-598. 7. Magaji V, Johnston JM. Inpatient management of hyperglycemia and diabetes. Clin Diabetes. 2011;29(1):36-39. 8. Cobaugh DJ, Maynard G, Cooper L, et al. Enhancing insulin-use safety in hospitals: practical recommendations from an ASHP Foundation expert consensus panel. Am J Health Syst Pharm. 2013;70(16):1404-1413. 9. Chen HJ, Steinke DT, Karounos DG, Lane MT, Matson AW. Intensive insulin protocol implementation and outcomes in the medical and surgical wards at a Veterans Affairs Medical Center. Ann Pharmacother. 2010;44(2):249-556. 10. Dinsmoor RS. Insulin sensitivity factor. Diabetes self-management. February 21, 2013. http://www.diabetesselfmanagement.com/Articles/Diabetes -Definitions/insulin_sensitivity_factor/. Accessed October 9, 2014. 11. Yamashita S, Ng E, Brommecker F, Silverberg J, Adhikari NK. Implementation of the glucommander method of adjusting insulin infusions in critically ill patients. Can J Hosp Pharm. 2011;64(5):333-339. 12. Cook CB, Beer KA, Seifert KM, Boyle ME, Mackey PA, Castro JC. Transitioning insulin pump therapy from the outpatient to the inpatient setting: a review of 6 years’ experience with 253 cases. J Diabetes Sci Technol. 2012;6(5):995-1002.
Both authors are affiliated with the Division of Endocrinology at the Mayo Clinic in Scottsdale, AZ. Patricia A. Mackey, FNPBC, BC-ADM, CDE, is an assistant professor of medicine, Mayo Clinic College of Medicine and can be reached at mackey. [email protected]. Michael D. Whitaker, MD, FRCPC, is a consultant in endocrinology, assistant professor of Medicine, Mayo Clinic College of Medicine. In compliance with national ethical guidelines, the authors report no relationships with business or industry that would pose a conflict of interest. 1555-4155/15/$ see front matter © 2015 Elsevier, Inc. All rights reserved. http://dx.doi.org/10.1016/j.nurpra.2015.02.016
The Journal for Nurse Practitioners - JNP
537