Hyperglycemic Control in Hospitalized Patients: Best Practice Susan S. Braithwaite Clinical Professor of Medicine University of North Carolina Glycemic Control Resource Room November 5, 2006

Objectives – to discuss: Relationships of medical outcomes and cost to presence of diabetes, glycemic control, and / or treatment modality Target range glucose Intravenous infusion of insulin Subcutaneous insulin (scheduled and correction therapy) Prevention of hospital hypoglycemia Diabetes patient self-management Moving toward discharge Managing complexity – orders sets completed by checking off boxes and filling in numbers, protocols activated by a single signature, and computerization

Background

Observational studies and comparisons of intensified regimens against historical controls show correlation between glycemic control and outcomes Stagnaro-Green. Mortality in hospitalized patients with hypoglycemia and severe hyperglycemia. Mt. Sinai J Med 1995; 62: 422. Umperriez. Hyperglycemia: an independent marker of in- hospital mortality in patients with undiagnosed diabetes. J Clin Endocrinol Metab 2002; 897: , Finney. Glucose control and mortality in critically ill patients. JAMA 2003; 290: Krinsley. Effect of an intensive glucose management protocol on the mortality of critically ill adult patients. Mayo Clin Proc 2004; 79: Pittas. Insulin Therapy for Critically Ill Hospitalized Patients: A Meta-analysis of Randomized Controlled Trials. Arch Int Med 1004; 164: 2005.

Capes. Stress hyperglycemia and prognosis of stroke in nondiabetic and diabetic patients: a systematic overview. Stroke 2001; 32: Bruno. Admission glucose level and clinical outcomes in the NINDS rt-PA stroke trial. Neurology 2002; 59: 669. Gentile. Blood glucose control after actue stroke: a retrospective study. Academic Emergency Medicine 2003; 10: 432. Baird. Persistent poststroke hyperglycemia is independently associated with infarct expansion and worse clinical outcome. Stroke 2003; 34: Leigh. Predictors of hyperacute clinical worsening in ischemic stroke patients receiving thrombolytic therapy. Stroke 2004; 35: Lindsberg. Hyperglycemia in acute stroke. Stroke 2004; 35: Gentile. Decreased mortality by normalizing blood glucose after acute ischemic stroke. Academic Emergency Medicine 2006.

Laird. Relationship of early hyperglycemia to mortality in trauma patients. J Trauma 2004; 56: Thomas. Early perioperative hyperglycaemia and renal allograft rejection in patients without diabetes. BMC Nephrology 2000;1:1. Thomas. Early peri-operative glycaemic control and allograft rejection in patients with diabetes mellitus: a pilot study. Transplantation 2001;72:1321. Melin. Protective effect of insulin on ischemic renal injury in diabetes mellitus. Kidney International 2002;61:1383. Weiser. Relation between the duration of remission and hyperglycemia in induction chemotherapy for acute lymphocytic leukemia... Cancer 2004;100:

Capes. Stress hyperglycaemia and increased risk of death after myocardial infarction in patients with and without diabetes: a systematic overview. Lancet 355:773-8, 2000 Foo. A single serum glucose measurement predicts adverse outcomes across the whole range of acute coronary syndromes. Heart 2003; 89: 512. Schnell. Intensification of therapeutic approaches reduces mortality in diabetic patients with acute myocardial infarction: the Munich registry. Diabetes Care 2004; 27: 455. Stranders. Admission blood glucose level as risk indicator of death after myocardial infarction in patients with and without diabetes mellitus. Arch Intern Med 2004; 164: 982. Meier. Plasma glucose at hospital admission and previous metabolic control determine myocardial infarct size and survival in patients with and without type 2 diabetes. (LAMBDA study). Diabetes Care 2005;28: Cheung. The hyperglycemia:intensive insulin infusion in infarction (HI- 5) study: a randomized controlled trial of insulin infusion therapy for myocardial infarction. Diabetes Care 29:2006;765.

Goyal. Prognostic significance of the change in glucose level in the first 24h after acute myocardial infarction: results from the CARDINAL study. European Heart Journal 2006;27: Barsheshet. Admission blood glucose level and mortality among hospitalized nondiabetic patients with heart failure. Arch Intern Med 2006;166: Chu. Early predictors of in-hospital death in infective endocarditis. Circulation 2004; 109: Thomsen. Diabetes and outcome of community-acquired pneumococcal bacteriemia. Diabetes Care 2004;27: 70. Falguera. Etiology and outcome of community-acquired pneumonia in patients with diabetes mellitus. Chest 2005; 128: 3233.

Pomposelli. Early postoperative glucose control predicts nosocomial infection rate in diabetic patients. J Parenteral and Enteral Nutrition; 1998; 22: 77. Golden. Perioperative glycemic control and the risk of infectious complications in a cohort of adults with diabetes. Diabetes Care 1999; 22: Latham. The association of diabetes and glucose control with surgical- site infections among cardiothoracic surgery patients. Infection Control & Hospital Epidemiology. 2001;10:607. Vriesendorp. Early post-operative glucose levels are an independent risk factor for infection after peripheral vascular surgery. A retrospective study. Eur J Vasc Endovasc Surg 2004;5 : 520

Szabo. Early postoperative outcome and medium-term survival in 540 diabetic and 2239 nondiabetic patients undergoing coronary artery bypass grafting. Annals of Thoracic Surgery 2002; 74: 712. Estrada. Outcomes and perioperative hyperglycemia in patients with or without diabetes mellitus undergoing coronary artery bypass grafting. Ann Thorac Surg 2003; 75: McAlister. Diabetes and coronary artery bypass surgery: an examination of perioperative glycemic control and outcomes. Diabetes Care 2003; 26: Gandhi. Intraoperative hyperglycemia and perioperative outcomes in cardiac surgery patients. Mayo Clin Proc 2005; 80: Furnary. Clinical effects of hyperglycemia in the cardiac surgery population: the Portland Diabetic Project. Endocrine Practice 2006;12(suppl 3):

American Diabetes Association. Economic costs of diabetes in the U.S. in Diabetes Care 2003; 26: Ahmann A. Reduction of hospital costs and length of stay by good control of blood glucose levels. Endocrine Practice 2004; 10 (suppl 2): Vora. Improved perioperative glycemic control by continuous insulin infusion under supervision of an endocrinologist does not increase costs in patients with diabetes. Endocrine Practice 2004; 2: Almbrand B. Cost- effectiveness of intense insulin treatment after acute myocardial infarction in patients with diabetes mellitus: results from the DIGAMI study. Eur Heart J 2000; 21:

Background

Heart surgery

Zerr. Ann Thorac Surg 1997;63: Portland: Deep sternal wound infection rate correlates with quartile of blood glucose on postoperative day 1

Furnary Ann Thorac Surg 1999;67: Percent of patients on POD 1 with mean BG < 200 mg/dL was increased by using insulin infusion after cardiac surgical procedures 1499 in study group 968 historical controls

Furnary Ann Thorac Surg 1999;67: Blood glucose differed among patients having deep sternal wound infections vs patients having no deep sternal wound infections after cardiac surgical procedures

Over time, average postoperative glucose has continued to decline among patients undergoing heart surgery Furnary 2006 Introduce Portland Protocol Year BG

Over time, mortality among patients with diabetes has decreased dramatically. Furnary 2006

At higher quintiles of blood glucose, the increase in overall mortality among patients with diabetes was overwhelmingly accounted for by an increase in cardiac-related mortality. Furnary 2003 Cardiac related mortality Non-cardiac related mortality > 250 BG < 150

Myocardial infarction

Malmberg. BMJ 1997;314: DIGAMI I Study: Mortality was reduced among suspected and known diabetes patients by intensive insulin after acute myocardial infarction (glucose-insulin infusion, followed by at least 3 months of SQ insulin)

Direct comparison of risk-adjusted 30-day mortality (A) and 1-year mortality (B) in patients with and without recognized diabetes across range of glucose values Kosiborod, M. et al. Circulation 2005;111: with no diabetes with diabetes Mortality at given BG level with no diabetes with diabetes Hyperglycemia predicts 30-d and 1-year mortality in elderly patients with myocardial infarction, particularly those without recognized diabetes

Meier. LAMBDA study. Diabetes Care 2005;28: In patients both with and without type 2 diabetes, hyperglycemia at the time of MI is associated with : reduced prospective survival in diabetes, higher A1C at higher tertile of BG, higher CK diabetesnondiabetes Causality cannot be inferred; the authors called for randomized trials of tight control

Surgical ICU

Leuven, Belgium: Morning blood glucose was reduced by intensive treatment in Van den Berghe’s ICU trial of insulin infusion ( 13% had previous diabetes ) Van den Berghe. N Engl J Med 2001;345: All in treatment group Subgroup on insulin mg/dL

Survival increased with intensive insulin therapy ( nondiabetic patients included ) targeting BG mg/dL among patients who remained in ICU > 5 days Van den Berghe. N Engl J Med 2001;345:

Van den Berghe, Crit Care Med 2003 Filled bars < 110 Shaded bars Open bars > 150 mg/dL Complications correlated with average blood glucose

Van den Berghe, Crit Care Med 2003 Although the Van den Berghe study was not designed to determine the threshold above which mortality increases, mortality was higher at glucose , compared to glucose < 110 mg/dL. < > 150 mg/dL

Hartford: Nosocomial infections in SICU N. Grey. Endocrine Practice vs 179 = av BG ( mg/dL) IV device, blood stream, both, surgical site, UTI, nosocomial pneumonia

General and Medical ICU

Hospital mortality rate in critically ill patients vs mean BG Mortality Rate (%) Mean Glucose Value (mg/dL) Krinsley. Retrospective review of 1,826 consecutive intensive care unit patients at The Stamford Hospital, Connecticut. Mayo Clin Proc. 2003;78:1471–1478.

Intensive management protocol, reducing mean BG from to mg/dL in medical-surgical adult ICU patients, reduced mortality Percent reductions were glucose > 200 mg/dL  56.3% new renal insufficiency  75% RBC transfusions  18.7% mortality  29.3% length of ICU stay  10.8% Krinsley. Effect of an intensive glucose management protocol on the mortality of critically ill adult patients. Mayo Clin Proc. 2004;79:

Average blood sugar depicted in critically ill surgical patients at Tufts-New England Medical Center in 1997, before implementation of successive protocols dedicated to aggressively treating hyperglycemia. By Spring of 2002, the mean glucose was < 130 mg/dL. Nasraway. Journal of Parenteral and Enteral Nutrition, Vol. 30, No. 3,

Hyperglycemia responded to aggressive institution of successive intensive insulin protocols from 1997 to By 2002, the desired endpoint of blood glucose < 130 mg/dL was achieved. During the same interval, patient mortality was observed to have significantly declined, in spite of increasing severity of illness. Nasraway. Journal of Parenteral and Enteral Nutrition, Vol. 30, No. 3,

Effect of Intensive Insulin Therapy on Morbidity in Medical ICU Van den Berghe, G. et al. N Engl J Med 2006;354:

Kaplan-Meier Curves for In-Hospital Survival Van den Berghe, G. et al. N Engl J Med 2006;354:

Implication: Hospitals should focus on prevention of hyperglycemia as an important patient safety factor

Target range blood glucose (AACE 2004) Preprandial < 110 mg / dL Peak postprandial:< 180 mg / dL Critically ill surgical patients: mg / dL The targets for critically ill patients were determined from the Leuven, Belgium study, in which whole blood glucose methodologies were used

Target range blood glucose (ADA 2005) Noncritically ill Premeal blood glucose as close to mg/dL as possible (midpoint 110 mg/dL) Postprandial blood glucose < 180 mg/dL Critically ill Blood glucose as close to 110 mg/dL as possible and generally < 180 mg/dL These patients generally will require IV insulin

Can benefits of glycemic control be shown for diseases sometimes managed outside of the ICU, with subcutaneous therapy ? What specific outcomes might be improved ? For hypoglycemia vs. tight glycemic control, what is the risk - benefit analysis in the setting of the general hospital ward ?

Stroke

Maintaining near normal BG levels after stroke may improve outcome (n = 765 with ischemic stroke) Gentile. Blood glucose control after acute stroke: a retrospective study. Academic Emergency Medicine 2003; 10: 432.

Persistent poststroke hyperglycemia is independently associated with infarct expansion Baird. Persistent poststroke hyperglycemia is independently associated with infarct expansion and worse clinical outcome. Stroke 2003; 34: 2208.

Pneumococcal Sepsis

With pneumococcal bacteremia, the 30- d mortality is higher in the presence of diabetes Thomsen. Diabetes and outcome of community-acquired pneumococcal bacteriemia. Diabetes Care 2004;27: 70. no diabetes diabetes mortality

Kidney Transplant

In 50 patients with diabetes undergoing their first cadaveric renal transplantation, poor perioperative glycemic control in the first 100 postoperative hours was associated with an increased incidence of acute rejection With mean BG < 200 mg / dL, only 3 of 27 patients had rejection episodes ( 11% ) With worse control, the majority had rejection episodes ( 58% ) Thomas. Early peri-operative glycaemic control and allograft rejection in patients with diabetes mellitus: a pilot study. Transplantation 2001;72:1321.

Among patients with vs without subsequent rejection: mean perioperative glucose in 1 st 100 hours differed BG 180 BG – 72h these patients had subsequent acute rejection these patients did not

In the 50 patients with diabetes undergoing their first cadaveric renal transplantation, postoperative infection ( POI ) occurred in 35, or 70%, at a mean of 10.8 days after transplantation Every patient with mean BG > 202 mg / dL over 1 st 100 h after surgery developed infection ( n = 23 ). BG values below are for postop day 1. Without POI,mean glucose 167 mg/dL With POI, mean glucose 248 mg/dL Thomas. Early peri-operative glycaemic control and allograft rejection in patients with diabetes mellitus: a pilot study. Transplantation 2001;72:1321.

Acute Lymphocytic Leukemia

Duration of complete remission / median survival were shorter among 103 hyperglycemic patients ( 37% ) with acute lymphocytic leukemia Weiser complete remission : 24 vs 52 mo median survival : 29 vs 88 mo Hyperglycemia: ≥ 2 BG of ≥ 200 mg / dL w / 30 d of induction chemo

Sepsis or any complicated infection (sepsis, pneumonia, or fungal) appeared more often in the hyperglycemic patients with ALL Weiser. Relation between the duration of remission and hyperglycemia during induction chemotherapy for acute lymphocytic leukemia with a hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone/methotrexate-cytarabine regimen. Cancer 2004; 110:1179.

Nosocomial Infections

Among 100 consecutive diabetic elective surgery patients, nosocomial infection rate within first 14 postoperative days (overall 25.8 %), correlated significantly with BG on postoperative day 1 (POD 1). Pomposelli. Early postoperative glucose control predicts nosocomial infection rate in diabetic patients. J Parenteral and Enteral Nutrition; 1998; 22: 77.

Early hyperglycemia ≥ 200 mg/dL in trauma patients was associated with higher infection and mortality rates independent of injury characteristics Laird. Relationship of early hyperglycemia to mortality in trauma patients. J Trauma 2004; 56: 1058.

Post- operative glucose levels in the first 48 hr were an independent risk factor for infection after peripheral vascular surgery Vriesendorp. Early post-operative glucose levels are an independent risk factor for infection after peripheral vascular surgery. A retrospective study. Eur J Vasc Endovasc Surg 2004;5 : 520 BG > 151 mg/ dL BG < 103 postoperative infection rate

Overall Mortality

Total inpatient mortality was highest with newly recognized hyperglycemia Umpierrez. Hyperglycemia: an independent marker of in-hospital mortality in patients with undiagnosed diabetes. J Clin Endocrinol Metab 2002; 897: , 2002.

Length of Stay and Cost

Length of stay was highest with newly recognized hyperglycemia Umpierrez. J Clin Endocrinol Metab 2002; 897: , 2002.

Economic Costs of Diabetes in the US in 2002 Diabetes Care 2003; 26: 917 Per capita medical expenditures per year Without diabetes: $ 2, 560 With diabetes: $ 13, 243 Expenditures in diabetes: Office visits 10.9 % Nursing home care 15.1% Inpatient days 43.9%

Deciding on a Treatment Plan in the Hospital to Maintain Glucose in the Target Range

Deciding whether to maintain the ambulatory treatment plan in the hospital

The Pattern of Insulin Requirement during Normal Health, Meal Plan, and Activity is Not Necessarily Reproduced in the Hospital

Scheduled subcutaneous insulin Which pattern of carbohydrate exposure describes the patient ? Discrete meals Negligible carbohydrate Continuous carbohydrate exposure Transitional meal plan / grazing Daytime grazing / overnight enteral feedings

Generalizations about subcutaneous insulin therapy in the hospital

Subcutaneous insulin in the hospital has 3 components basal nutritional (or prandial) correction therapy

“do not withhold” basal insulin for type 1 diabetes “hold” or “reduce” prandial insulin for poor oral intake “hold” nutritional insulin for interruption of dextrose-containing fluids, enteral nutrition, or other exposure to carbohydrate “hold” regular insulin or rapid acting analog for low BG Additional directions are attached to specific scheduled insulin orders, or are specified under protocols

Standardized order sets are completed with check- marks and numbers Subcutaneous scheduled daily insulin Check appropriate times or meal times to right, and fill in insulin doses below.  0600  1200  1800  2400 or  Breakfast  Lunch  Supper  Bedtime Rapid-acting insulin analog ________ units Short-acting insulin Regular ________ units Intermediate-acting insulin NPH ________ units Long-acting insulin analog ________ units Other insulin: ______________________ ________ units

Twilight of the “Sliding Scale,” and “Today’s Insulin” Dawning

Relative risk of hyperglycemic episodes (BG >300 mg/dL) among 171 hospitalized diabetes patients was greatest with sliding scale, when comparing antihyperglycemic drug therapies to nontreatment Queale 1997

A1C monitoring and use of scheduled insulin improved as a result of housestaff education on scheduled insulin combined with prohibition of sliding scale Baldwin. Eliminating inpatient sliding scale. Diabetes Care 2005;28:1008. Sliding Scale Use: 100 % Sliding Scale Use: 0 %

BG control improved as a result of housestaff education on scheduled insulin combined with sliding scale prohibition for treatment of hyperglycemic patients Baldwin. Eliminating inpatient sliding scale. Diabetes Care 2005;28:1008.

Correction Dose Algorithms Algorithm 1 Rapid-acting analog [ TDDI ≈ 24 ( < 28 ) units, or wt < 56 kg] Regular insulin [ TDDI ≈ 20 ( < 23 ) units, or wt < 46 kg] Algorithm 2 Rapid-acting analog [ TDDI ≈ 30 (28–36) units, or wt 56–73.9 kg] Regular insulin [ TDDI ≈ 25 ( 23–30) units, or wt 46–61.9 kg] Algorithm 3 Rapid-acting analog [ TDDI ≈ 45 ( 37–55) units, or wt 74–111.9 kg] Regular insulin [ TDD I ≈ 37½ (31– 46) units, or wt 62–93.9 kg] Algorithm 4 Rapid-acting analog [ TDDI ≈ 72 (56 –90) units, or wt 112–181.9 kg] Regular insulin [ TDDI ≈ 60 (47–75) units, or wt 94–151.9 kg] Algorithm 5 Rapid-acting analog [ TDDI ≈ 120 (91–144) units, or wt  182 kg] Regular insulin [ TDDI ≈ 100 ( 76 – 120) units, or wt  152 kg] Algorithm 6 Rapid-acting analog [ TDDI ≈ 180 ( > 144 ) units] Regular insulin [ TDDI ≈ 150 ( > 120) units] BG units       390 7

Admitting a hyperglycemic patient who is eating

Subcutaneous insulin therapy in the hospital for patients eating discrete meals

A preferred hospital regimen for use with a consistent carbohydrate diet Lispro / aspart / glulisine ( ~ 50 % ) Glargine ( ~ 50 % )

The patient who is eating and running high blood glucose... Use insulin analogs Use basal – prandial – correction therapy “Just say no” to sliding scale

Insulin analogs are the preferred therapy for patients who are eating Rapid- acting analogs (aspart, lispro, glulisine) are best reserved for prandial and correction dose therapy (not basal), whereas long- acting analog therapy (glargine, detemir) is best reserved for basal coverage (not nutritional)

A Bad Practice : 70/30 Insulin plus Sliding Scale Scheduled or routine BID, or 0800 & 1700 Sliding Scale Q4HRS, or 0200, 0600, 1000, 1400, 1800, 2200 This means 8 shots daily. There is a risk of stacking, & BG tests fail to synchronize: 0200, 0600, 0800, 1000, 1400, 1700, 1800, 2200

Tease out the basal and prandial components for patients admitted on premixed insulins meal omission is inevitable premixeds won’t work the argument for analogs is reduction of hypoglycemia

Apportion or reapportion the insulin About 50% basal About 50% prandial

“ Teasing out ” the basal and prandial components of insulin in type 2 diabetes for a patient admitted on premixed insulin Suppose Robert normally takes 70 / 30 insulin, 60 units at breakfast and 30 units at supper He sometimes has fairly severe lows A SQ regimen is needed to permit occasional meal omission

Basal insulin in type 2 diabetes is % of the usual total daily dose A regimen of scheduled insulin might be: –glargine 36 units qHS (“cut 50% if NPO”) –aspart 12 units WMEALS (“hold if NPO”) –aspart 2 units prn BG WMEALS, HS, 0300 –aspart 4 units prn BG >199WMEALS, HS, 0300

“ Teasing out ” and “locking in” the basal insulin component in type 1 diabetes

Requirement for exogenous insulin during prolonged fasting may disappear in type 2 DM, but even during prolonged fasting it is absolute in type 1 DM type 1 DM type 2 DM requirement for exogenous basal insulin vs time fasting

Corticosteroids

Procedures, anesthesia and surgery

Preoperative planning in the office....

For patients receiving insufficient prandial coverage, overdoses of long-acting analog (glargine or detemir) established as home therapy may spell hypoglycemia once the patient is in the hospital...

Glargine or determinr, overdosed to play catchup, plus sliding scale ! pm BG

During NPO status, among type 1 diabetes patients, a previously established glargine dose maintains normoglycemia Mucha, Diabetes Care: 2004 Fasting days Control days

Requirement for exogenous insulin during prolonged fasting or calorie restriction shows variability in Type 2 DM type 2 DM patient B patient A

Long -acting analog > 50% of total insulin? On the night preoperatively, cut long -acting analog dose to be no more than 50% of daily total insulin, and consider % further reduction ambulatory plan glargine 72 units qHS metformin sulfonylureas the night preoperatively  glargine 36 units

ambulatory plan glargine 12 units qHS glulisine 6 units WMEALS the night preoperatively glargine 12 units Long -acting analog < 50% of total insulin? On the night preoperatively, no dose reduction

Insulin by pump

A 35 year old woman with type 1 diabetes She is having bariatric surgery Normally she wears an insulin pump She is in your office asking for preoperative advice

Lispro or aspart For procedures or NPO status, pump patient can hold the prandial insulin but maintain true basal insulin

A caution on subcutaneous therapy Delayed response and late snowballing from repeated insulin doses may occur among patients with poor perfusion of subcutaneous sites, as with hypoalbuminemia edema hypotension pressors

Intravenous Infusion of Insulin: the only insulin treatment strategy specifically developed for use in the hospital

Indications for intravenous insulin infusion Diabetic ketoacidosis Non-ketotic hyperosmolar state Critical care illness Myocardial infarction or cardiogenic shock Post- operative period following heart surgery NPO status in type 1 diabetes General pre-, intra- and post- operative care Organ transplantation Stroke (possibly) Exacerbated hyperglycemia during high- dose glucocorticoid therapy Dose-finding strategy, anticipatory to initiation or reinitiation of subcutaneous insulin in type 1 or type 2 diabetes Labor and delivery Any illness for which prompt glycemic control is important to recovery

Intravenous Infusion of Insulin: a well- designed protocol damps the swings of BG

Using an algorithm that considers BG, insulin rate, and velocity of change, Goldberg and colleagues safely approached target Goldberg et al. Diabetes Care 27: 461, 2004

The BG results compared favorably with those of historical controls Goldberg et al. Diabetes Care 27: 461, 2004

Braithwaite et al. Diabetes Technology and Therapeutics 2006; 8 (4):476

Intravenous Insulin Infusion Targeting BG mg/dL Column 1Column 2Column 3Column 4Column 5Column 6 BG units/h <       Braithwaite et al. Diabetes Technology and Therapeutics 2006; 8 (4):476

23 patients, 1537 BG’s, 25 / 27 patient runs; 2 patients did not reach BG < 110 before interruption of infusion Blood Glucose (BG) Distribution after First Reaching BG < 110 mg/dL, Analysis Shown Using BG or Patient Run as the Unit of Observation %

in this series of 1537 BG measurements, among the 23 patients or 25 runs in which target BG < 110 was reached, there were 31 BG measurements < 70 mg/dL, and in this series none < 50 mg/dL. Prevalence of 60 ≤ BG < 70, 16 / 25 runs Prevalence of 50 ≤ BG < 60, 8 / 25 runs

Insulin to cover TPN Distribution of insulin, desired –67-80 % of the insulin in the TPN bag – % as correction therapy by SQ injection or IV infusion Redistribution strategy –add 2 /3’s of yesterdays correction therapy to today’s bag

Intravenous insulin infusion under basal conditions correlates well with subsequent subcutaneous insulin requirement Hawkins. Endocrine Practice: 1995 Total IV vs SQ 24h Insulin Requirements Units SQ Units IV

Proposed predictors of successful transition from intravenous insulin to SQ insulin after CABG Type 2 diabetes or nondiabetes ( hospital hyperglycemia ) Uncomplicated CABG and / or valve surgery Patient extubated, no pressor support, eating Transfer orders written Stable BG < 130 mg/dL Insulin infusion ≤ 2 units / h overnight Projected insulin dose  48 units / d

Possible predictors of hyperglycemia after early transition to SQ insulin after heart surgery Type 1 diabetes COPD Morbid obesity Bilateral internal thoracic artery harvesting Heart transplant Ventilator dependency, pressor support, IABP, NPO status Corticosteroids No transfer orders written Glucose > 130 mg / dL while on the insulin infusion Infusion rate > 2 units / h overnight Projected insulin dose > 48 units / d

Transition off intravenous insulin infusion....

Constructing a profile for scheduled subcutaneous insulin …. 6 pm 12 am 6 am 12 pm Glargine NPH SQ Regular Lispro / Aspart / Glulisine

The Time Arrives to Stop Insulin Drip Discrete meals? Glargine and rapid-acting analog NPH and regular insulin No Yes

In the following case of a patient who is eating, if transfer orders are written.... what is needed for glycemic management ?

Estimating initial SQ basal insulin from drip It is 8 am on the morning of postop day 3 after CABG. The patient is well, still on insulin drip. There were no overnight pressors or dextrose. Since yesterday, as prandial insulin, 4 units aspart are given with each meal. Based on the insulin requirement during 8 hr overnight, if nothing changes, what is the 24h basal requirement? am1.5 unit / hr 2 -4 am1.0 unit / hr 4 -6 am1.5 unit / hr 6 -8 am2.0 unit / hr

Answer: Average hourly insulin was 1.5 units / hr 24 hr basal requirement is 36 units Consider ~ % basal dose reduction each day during recovery Adjust prandial doses upward each day Eventual goal: 50% basal and 50% prandial

The hyperglycemic patient receiving continuous exposure to carbohydrate, or no carbohydrate....

Insulin requirement during continuous D5, TPN or enteral feedings

Prolonged NPO Status Regimens for patient while NPO, on IV’s, or receiving continuous enteral feedings

6 pm 12 am 6 am 12 pm Prescribe glargine Corpak falls out BG q 2 hr, dextrose if < 120 Long- acting insulin doses should not exceed basal requirements during continuous enteral feedings

6 12 R q 6 hr (~ 33 %, hold if low) NPH q 6 hr (~ 67 %) — Mixtures of NPH and regular insulin q 6 – 8 h maintain a flat- line, safe and rapidly reversible insulin effect for a patient while NPO, on IV’s, or receiving continuous enteral feedings

In the following case of a patient not eating, if transfer orders are written... what is needed for glycemic management ?

Estimating combined basal and nutritional coverage from drip A stable ventilator- dependent trauma patient has had a tracheostomy and PEG procedure. Over the past 24h he has received 72 units of insulin daily by intravenous infusion to cover enteral feedings, which have been at target for 24 hr. How might subcutaneous orders be written to permit transition off of the insulin infusion?

One acceptable answer : NPH 12 units q 6 h ( hold if tube feeds stop ) Regular insulin 6 units subcutaneously q 6 h ( hold if tube feeds stop, hold if BG < 100 ) Correction dose regular insulin “prn” –BG units –BG units –BG 250 or higher + 6 units Stop insulin drip 2 h after first dose of subcutaneous insulin

Grazing ( transitional meal plans, nutritional supplements )....

Grazing

Insulin regimen, no tube feeds, for patients attempting to eat ( transitional meal plan with “grazing”) Estimate daily insulin requirement and split 2/ 3 in AM, 1/ 3 in PM 2/ 3 of AM dose as NPH, 1/ 3 as regular 1/ 2 of PM dose as regular at / 2 of PM dose as NPH at 2200 Hold scheduled regular for BG < 90

Example of a “grazing” patient A postoperative patient receives nutritional supplements between meals and grazes on trays 36 units of insulin are required daily Sample insulin orders: –NPH 16 units before breakfast –regular insulin SC 8 units before breakfast –regular insulin SC 6 units before supper –NPH 6 units qHS –regular insulin 2 units prn BG WMEALS, HS –regular insulin 4 units prn BG > 199 WMEALS, HS

Overnight tube feeds....

Ovenight enteral feedings / daytime grazing

70 / 30 change dose select " pre- med" for priority change frequency to “QPM enter additional directions free text pre- med for TF’s, hold if no TF’s

Today’s Insulin Dose How much? How much to start with? How much for today?

No clue where to start ? start low.... basal insulin requirement  0.15 units /k basal + prandial insulin  0.3 units /k if eating Known preadmission TDDI (total daily dose of insulin) ? basal insulin requirement if no longer eating  % of preadmission TDDI Initial insulin calculation, hospitalized patient, type 2 diabetes

During continuous nutritional support or IV dextrose infusion, provide basal insulin 0.15 units per kilo provide nutritional insulin, initially estimated as 1 unit per 10 gm carbohydrate cover both needs continuously

If insulin is required at all the requirement for exogenous insulin may vanish during caloric restriction or development of organ dysfunction calculations on a “per kilo” basis may overestimate basal needs when overnutrition ceases For type 2 diabetes --

How much to start with ? Melinda: weight 80 kilos Problem: ovarian carcinoma, type 2 diabetes Status: postop, NPO, D5 in 0.45 at 100 cc/hr Basal insulin: ________ units per 24 hr Insulin to cover D5: ________ units per 24 hr Total insulin / 24 hr:________ units per 24 hr

Type 2 diabetes, NPO, continuous D 5 W Melinda: weight 80 kilos Status: postop, NPO, D5 in 0.45 at 100 cc/hr Basal insulin: 12 units per 24 hr Insulin to cover D5: 12 units per 24 hr Total insulin / 24 hr:24 units per 24 hr SQ insulin orders: –4 units NPH q 6h (hold if D5 stops ) –2 units regular insulin q 6h (hold if D5 stops, hold if BG < 90 ) –Correction dose algorithm for highs

Correction therapy signals the need for a change of scheduled therapy to prevent recurrent hyperglycemia

Dose- finding strategy Determine yesterday’s total insulin dose actually administered Review yesterday’s glycemic control Calculate today’s scheduled insulin dose (this example uses peak glucose targets appropriate during D5 or tube feeds) –some BG < 90  80 % of yesterday’s total –BG’s  100 % of yesterday’s total –some BG ≥ 180, no BG < 90  110 % of yesterday’s total

How much for today ? A patient on tube feeds received 6 units NPH and 3 units regular insulin as scheduled every 6 hr, total 36 units for the day, plus the following correction doses of regular insulin: 0600 BG unit 1200 BG units 1800 BG unit 2400 BG units

Assuming nothing else has changed, what is today’s dose of scheduled insulin? A.24 units B.34 units C.44 units D.48 units

Answer (doing it one doctor’s way) : Total dose given yesterday was 44 units 110 % of yesterday’s dose is 48 units Today’s insulin dose is 48 units, 8 N, 4 R at 0600, 1200, 1800, and 2400

Prevention of Hospital Hypoglycemia

An analysis of proximate causes of hospital hypoglycemia suggests that iatrogenic hypoglycemia usually is predictable and preventable (Fischer) :

Transportation off ward causing meal delay New NPO status Interruption of intravenous dextrose Interruption of TPN Interruption of enteral feedings Interruption of continuous renal replacement rx Insulin or secretogogue previously prescribed. Triggering Events for Hypoglycemia

Insulin previously prescribed. Triggering events for hypoglycemia ? Increase BG monitoring to q 2 hr for timeframe of action of the insulin, and: Give carbohydrate for BG < 120 mg/dL -- oral, IV dextrose infusion, or, for volume restricted patients, concentrated dextrose Except for basal insulin in type 1 diabetes and correction doses: Hold insulin No Insulin given? Yes

Moving toward Discharge

Was it just “hospital hyperglycemia?” Review A1C from first day of admission –> 6.0 % ? –< 5.2% ? Followup BG in outpatient setting Greci. Diabetes Care 2003

A1C- driven therapeutic modifications were followed by improvement of A1C ( 34 of 98 patients were evaluable after 12 mo ) Baldwin. Eliminating inpatient sliding scale. Diabetes Care 2005;28:1008.

How will change be promoted ?

Managing Complexity : Standardized Order Sets and Computerized Order Entry Consistent carbohydrate meal plan Intravenous infusion of insulin Diabetes hospital patient self-management Scheduled subcutaneous insulin Oral antihyperglycemic agents Combination insulin / oral antihyperglycemics Point-of-care glucose monitoring Correction dose insulin Standardized order sets can guide and teach

Summary

Hyperglycemia is a patient safety factor. At the time of transition from intravenous insulin therapy, “scheduled” or “routine” or “programmed” subcutaneous insulin is the mainstay of glycemic management. Good care involves discovery and ongoing revision of daily insulin dose, replacing “correction” therapy with “scheduled” subcutaneous insulin. Hypoglycemia in the hospital is mostly preventable (by means other than undertreatment of diabetes). Sliding scale is ineffective and dangerous and should be eliminated. Patients conscious, eating, and experienced in self-management should continue self-management in the hospital. Nation-wide opportunities for improvement include preadmission and preanesthetic planning, facilitation of insulin drip therapy, protocols for hypoglycemia prevention, standardization of diabetes order sets and correction dose algorithms, computerized order entry that guides and teaches, and enabling policies with safety provisions on patient self-management, patient education, and discharge planning.