1. Dallas, TX November 2–4, 2012 Effective Glycemic Control Outside of the Critical Care Unit Christopher A. Newton, MD, FACE canewto@emory.edu Division of Endocrinology Grady Memorial Hospital Emory University November 3, 2012

2. Dallas, TX November 2–4, 2012 Objectives Describe acute care populations that are at high risk for hyperglycemia Identify essential components to achieving glycemic control in the acute care setting

3. Dallas, TX November 2–4, 2012 DEFINITIONS

4. Dallas, TX November 2–4, 2012 Accepted Glycemic Ranges

5. Dallas, TX November 2–4, 2012 Glycemic Thresholds In-hospital hyperglycemia is defined as an admission or in-hospital blood glucose >140 mg/dL Hypoglycemia is defined as any blood glucose <70 mg/dL A patient with an HbA1c >6.5% can be identified as having diabetes

6. Dallas, TX November 2–4, 2012 Comparison of sensitivity and specificity achieved for the diagnosis of diabetes based on fasting plasma glucose at various levels of HbA1c from data in NHANES III and 1999- 2004 NHANES J Clin Endocrinol Metab, July 2008, 93(7):2447–2453

7. Dallas, TX November 2–4, 2012 INCIDENCE OF HYPERGLYCEMIA

8. Dallas, TX November 2–4, 2012 Distribution of Patient-Day- Weighted Mean POC BG Values DATA from ~49 million POC-BG testing (12 M ICU; 37 M non-ICU) from 3.5 million patients (653,359 ICU; 2,831,436 non- ICU). The mean POC-BG was 167 mg/dL for ICU patients and 166 mg/dL for non-ICU patients. ICU Non-ICU Swanson et al. Endocrine Practice, October 2011

9. Dallas, TX November 2–4, 2012 Distribution of Patient-Day- Weighted Mean POC BG Values Data from ~37 million BG readings from 2,831,436 non-ICU patients - mean POC-BG: 166 mg/dL Swanson et al. Endocrine Practice, October 2011

10. Dallas, TX November 2–4, 2012 Hyperglycemia Is Common No Diabetes 26% 50 40 30 20 10 0 <110110-140140-170170-200>200 Diabetes 50 40 30 20 10 0 <110110-140 78% 140-170170-200>200 Mean BG, mg/dL Patients, % Kosiborod M, et al. J Am Coll Cardiol. 2007;49(9):1018-183:283A-284A.

11. Dallas, TX November 2–4, 2012 IGT and Undiagnosed T2DM Are Common in Acute MI and Stroke Norhammar A, et al. Lancet 2002;359:2140−4. Matz K, et al. Diabetes Care 2006;792−7. 2-hour OGTT

12. Dallas, TX November 2–4, 2012 CLINICAL OUTCOMES ASSOCIATED WITH HYPERGLYCEMIA

13. Dallas, TX November 2–4, 2012 Hyperglycemia: Independent Marker of In-Hospital Mortality Umpierrez GE et al, J Clin Endocrinol Metabol 87:978, 2002 * P < 0.01 *

14. Dallas, TX November 2–4, 2012 Hyperglycemia and Pneumonia Outcomes * * * * * P<0.05 vs BG<198 mg/dL N=2471 patients with community acquired pneumonia McAllister et al, Diabetes Crae 28:810-815, 2005

15. Dallas, TX November 2–4, 2012 30 Day Mortality and In-hospital Complications among Non-cardiac Surgery Patients A Frisch et al. Diabetes Care, May 2010 † p = 0.1 * p = 0.001 # p =0.017 † # * * * * *

16. Dallas, TX November 2–4, 2012 INSULIN PROTOCOL DEVELOPMENT

17. Dallas, TX November 2–4, 2012 Key Elements – BG Targets Glucose Target in non-ICU setting: –Premeal glucose targets <140 mg/dL –Random glucose <180 mg/dL –To avoid hypoglycemia, reassess insulin regimen if glucose levels fall below 100 mg/dL –Occasional patients may be maintained with a glucose range below and/or above these cut- points Moghissi ES, et al; AACE/ADA Inpatient Glycemic Control Consensus Panel. Endocr Pract. 2009;15(4). http://www.aace.com/pub/pdf/guidelines/InpatientGlycemicControlConsensusStatement.pdf

18. Dallas, TX November 2–4, 2012 Key Elements - Monitoring Glucose monitoring is an obvious, but crucial, element of success Monitoring glucoses: –Provides assessment of current glucose –Are interpreted for adjusting medications based upon the trends in the glucoses Frequency depends upon treatment regimen utilized –Quicker interventions need more frequent assessments

19. Dallas, TX November 2–4, 2012 Key Elements - Personnel Physician / Physician Assistant / Nurse Practitioner Nurses Pharmacy Staff Laboratory Staff Administration Many other people –Patients?

20. Dallas, TX November 2–4, 2012 Key Elements - Medication ACE/ADA Task Force on Inpatient Diabetes. Diabetes Care. 2006 & 2009 Diabetes Care. 2009;31(suppl 1):S1-S110.. Antihyperglycemic Therapy SC Insulin Recommended for most medical-surgical patients OADs Not Generally Recommended

21. Dallas, TX November 2–4, 2012 Key Elements - Medication AACE/ADA Consensus Statement: Non- insulin therapies in the hospital? –Sulfonylureas are a major cause of hypoglycemia –Metformin contraindicated in setting of decreased renal blood flow and with use of iodinated contrast dye –Tyiazolidinediones associated with edema and CHF –Alpha-glucosidase inhibitors are weak glucose lowering agents –GLP1-directed therapies can cause nausea and have a greater effect on postprandial glucose Moghissi ES, et al; AACE/ADA Inpatient Glycemic Control Consensus Panel. Endocr Pract. 2009

22. Dallas, TX November 2–4, 2012 Key Elements - Insulin Sliding scale short-acting insulin (SubQ) Subcutaneous basal/bolus therapy –NPH and Regular –Long-acting and Rapid-acting analogs Subcutaneous continuous infusion Intravenous insulin

23. Dallas, TX November 2–4, 2012 Key Elements - Insulin Study Type: Prospective, multicenter RCT Population: 130 subjects insulin naïve T2DM Basal-Bolus Protocol: –Starting total daily dose (TDD): 0.4 unit/kg/day for BG between 140-200 mg/dL 0.5 unit/kg/day for BG between 201-400 mg/dL –½ TDD as insulin glargine and ½ as glulisine Glargine – once daily at same time each day Glulisine – three equally divided doses with meals Umpierrez et al, Diabetes Care 30:2181–2186, 2007

24. Dallas, TX November 2–4, 2012 Sliding Scale Insulin Before meal: Supplemental Sliding Scale Insulin (# of units) Bedtime: Give ½ of Supplemental Sliding Scale Insulin Blood Glucose (mg/dL)Insulin SensitiveUsualInsulin Resistant >141-180246 181-220468 221-2606810 261-30081012 301-350101214 351-400121416 >400141618 Umpierrez GE et al. Diabetes Care. 2007;30:2181-2186.

25. Dallas, TX November 2–4, 2012 Rabbit 2 Trial: Changes in Glucose Levels With Basal-Bolus vs. Sliding Scale Insulin Umpierrez GE, et al. Diabetes Care. 2007;30(9):2181-2186. Sliding scale regular insulin (SSRI) was given 4 times daily Basal-bolus regimen: glargine was given once daily; glulisine was given before meals. 0.4 U/kg/d x BG between 140-200 mg/dL 0.5 U/kg/d x BG between 201-400 mg/dL Days of Therapy BG, mg/dL 100 120 140 160 180 200 220 240 Admit 1 Sliding-scale Basal-bolus b P<.05. a a a b b b b 2 345678910 a P<.05.

26. Dallas, TX November 2–4, 2012 Persistent hyperglycemia (BG>240 mg/dl) is common (15%) during SSI therapy Days of Therapy BG, mg/dL 100 120 140 160 180 200 220 240 Admit 1 Sliding- scale Basal- bolus 260 280 300 3345672421 Rabbit 2 Trial: Treatment Success With Basal-Bolus vs. Sliding Scale Insulin Hypoglycemia rate:  Basal Bolus Group:  BG < 60 mg/dL: 3%  BG < 40 mg/dL: none  Sliding Scale Group:  BG < 60 mg/dL: 3%  BG < 40 mg/dL: none Umpierrez GE, et al. Diabetes Care. 2007;30(9):2181-2186.

27. Dallas, TX November 2–4, 2012 Basal-Bolus vs Sliding Scale Insulin in Hospitalized Patients with T2DM The mean insulin daily dose was significantly higher in the basal-bolus group than in the sliding scale group Insulin Type Mean Insulin Dose, units / day Basal-Bolus Group SSI Group Basal insulin 22 ± 2 - Rapid-acting insulin 20 ± 1 - Regular insulin - 12.5 ± 2 Umpierrez GE, et al. Diabetes Care. 2007;30(9):2181-2186.

28. Dallas, TX November 2–4, 2012 Study Type: Prospective, multicenter, randomized, open-label trial in general surgery (non-ICU) Population: 211 subjects with T2DM on diet and/or oral hypoglycemic agents or low dose insulin (<0.4 units/kg/day) Primary Outcomes: Differences between groups in mean daily blood glucose and composite of hospital complications (wound infection, pneumonia, respiratory failure, acute renal failure, bacteremia Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

29. Dallas, TX November 2–4, 2012 RABBIT 2 Surgery Treatment on AdmissionAllSSIGlar+GluP-value Diet alone, n17116NS Oral antidiabetic agents, n1538073NS Insulin + oral antidiabetic agents, n20119NS Type of surgeryAllSSIGlar+GluP-value Cancer764036NS GI-GU benign592831NS Vascular311516NS Trauma382018NS Others752NS Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

30. Dallas, TX November 2–4, 2012 RABBIT 2 Surgery: Basal-Bolus Regimen D/C oral anti-diabetic drugs on admission Starting total daily dose (TDD): 0.5 unit/kg/day TDD reduced to 0.3 unit/kg/day in patients >70 years old or with creatinine >2 mg/dL ½ TDD as glargine and ½ TDD as glulisine –Glargine – once daily at same time of day –Glulisine – three equally divided doses with meals Goal glucoses: 100-140 mg/dL Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

31. Dallas, TX November 2–4, 2012 Basal-Bolus Dose Adjustment * Daily insulin adjustment was primarily focused on glargine dose. * The treating physicians were allowed to adjust prandial (glulisine) insulin dose, and to use the total supplemental dose, patient’s nutritional intake, and results of BG testing to adjust insulin regimen. Blood glucose levelsChange in Daily Insulin Dose* Fasting and pre-meal BG between 100-140 mg/dl in the absence of hypoglycemia no change Fasting and pre-meal BG between 141-180 mg/dl in the absence of hypoglycemia Increase by 10% Fasting and pre-meal BG between >181 mg/dl in the absence of hypoglycemia Increase by 20% Fasting and pre-meal BG between 70-99 mg/dl in the absence of hypoglycemia Decrease by 10% Fasting and pre-meal BG between <70 mg/dlDecrease by 20% Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

32. Dallas, TX November 2–4, 2012 RABBIT 2 Surgery: Glucoses During Therapy * p<0.001 † p=0.01 ‡ p=0.02 R = Randomization * † ‡ * † † Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

33. Dallas, TX November 2–4, 2012 RABBIT 2 Surgery: Mean Glucose During Day * * * * *p<0.001 Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

34. Dallas, TX November 2–4, 2012 RABBIT 2 Surgery: Postoperative Complications * Composite of hospital complications: wound infection, pneumonia, respiratory failure, acute renal failure, and bacteremia. P=0.003 P=NS P=0.05 P=0.10 P=0.24 Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

35. Dallas, TX November 2–4, 2012 RABBIT 2 Surgery: Impact on Need for ICU Post-surgical ICU AdmissionICU Length of Stay P=0.16 P=0.003 Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

36. Dallas, TX November 2–4, 2012 RABBIT 2 Surgery: Hypoglycemia p<0.001 p=0.057 There were no differences in hypoglycemia between patients treated with insulin prior to admission compared to insulin-naïve patients. Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

37. Dallas, TX November 2–4, 2012 RABBIT 2 Surgery: Insulin Doses SSI: range of daily regular insulin= 9.7 to 14.4 units after 24hr treatment 88.5% of patients received <20 units and 39.4% <10 units per day. Insulin Type Mean Insulin Dose, units / day Basal-Bolus Group SSI Group Total daily dose 33.412.3 Basal insulin 21.8 ± 8.6 - Rapid-acting insulin 14.8 ± 7.6 - Regular insulin - 12.3 ± 6.5 Umpierrez et al, Diabetes Care 34 (2):1–6, 2011

38. Dallas, TX November 2–4, 2012 Study Type: Prospective, randomized, open-label trial Population: 130 subjects with T2DM on oral hypoglycemic agents or insulin therapy Study Sites: –Grady Memorial Hospital, Atlanta, GA –Rush University Medical Center, Chicago, IL Umpierrez et al, J Clin Endocrinol Metab 94: 564–569, 2009

39. Dallas, TX November 2–4, 2012 DEAN Trial: Changes in Mean Daily Blood Glucose Concentration BG, mg/dL Duration of Therapy, d Data are means  SEM. Detemir + aspart NPH + regular Basal-bolus regimen: detemir was given once daily; aspart was given before meals. NPH/regular regimen: NPH and regular insulin were given twice daily, 2/3 in AM, 1/3 in PM. Umpierrez GE, et al. J Clin Endocrinol Metab. 2009;94(2):564-569. P=NS 100 120 140 160 180 200 220 240 Pre-Rx BG 0123456-10

40. Dallas, TX November 2–4, 2012 DEAN Trial Blood glucose (mg/dL) Detemir + Novolog NPH + Regular Umpierrez GE, et al. J Clin Endocrinol Metab. 2009;94(2):564-569.

41. Dallas, TX November 2–4, 2012 DEAN Trial: Insulin Doses The mean total daily insulin dose was not significantly different between treatment groups Insulin Type Mean Insulin Dose, units / day Detemir-NovologNPH-Regular Total Units/day 56 ± 45 45 ± 32 Basal insulin/day Detemir: 30 ± 28 NPH: 27 ± 20 Rapid-acting insulin Novolog: 27 ± 20* Regular: 18 ± 14 * P < 0.05 Umpierrez GE, et al. J Clin Endocrinol Metab. 2009;94(2):564-569.

42. Dallas, TX November 2–4, 2012 DEAN Trial: Hypogylcemia NPH/Regular –BG <40 mg/dL: 1.6% –BG <60 mg/dL: 25.4% Detemir/Aspart –BG <40 mg/dL: 4.5% –BG <60 mg/dL: 32.8% To determine risk factors for hypoglycemic events during SC insulin therapy Umpierrez GE, et al. J Clin Endocrinol Metab. 2009;94(2):564-569.

43. Dallas, TX November 2–4, 2012 Interim Summary Treatment with basal bolus improved glycemic control and reduced hospital complications compared to SSI in medicine and surgery patients with T2DM Basal-bolus insulin regimen is preferred insulin regimen over Sliding Scale Insulin in the hospital management of non-ICU patients with T2DM

44. Dallas, TX November 2–4, 2012 Hypoglycemia Risk Factors p-value Variable (univariate)BG <60 mg/dLBG <70 mg/dL Age0.0360.001 Weight0.0270.001 HbA1c0.5210.658 Creatinine0.0110.002 Enrollment BG0.1660.319 Previous treatment0.005<0.001 Previous insulin treatment<0.001 Treatment group<0.001 *p-values are from Wilcoxon Two-Sample Test Umpierrez et al, ADA Scientific Meeting, Poster #516, 2009

45. Dallas, TX November 2–4, 2012 CAN HYPOGLYCEMIA FROM INSULIN BE AVOIDED?

46. Dallas, TX November 2–4, 2012 Hypoglycemia: Triggering Events Transportation off ward, causing meal delay Failure to measure blood glucose before insulin doses New NPO status Interruption of –IV dextrose therapy –TPN –Enteral feedings –Continuous venovenous hemodialysis Especially in setting of continued/unchanged insulin dosing

47. Dallas, TX November 2–4, 2012 SubQ Basal Insulin Action Glargine NPH n=20 T1DM Mean ± SEM sc insulin 4.0 3.0 2.0 1.0 0 24 20 16 12 8 4 0 0 4 8 12 16 20 24 Time (hours) mg/Kg/min mol/Kg/min µ Adapted from Lepore M. et al., Diabetes 49:2142-8, 2000

48. Dallas, TX November 2–4, 2012 Intravenous Insulin By-passes the delay associated with subcutaneous insulin administration Insulin from an IV infusion or IV bolus will disappear from bloodstream in 7 minutes –With sufficiently frequent monitoring, can decrease the insulin dose prior to onset of hypoglycemia Majority of medical centers limit this option to intensive/critical care settings

49. Dallas, TX November 2–4, 2012 IV versus SubQ Insulin Long-acting subcutaneous insulin –Slow steady release of insulin into blood stream –Can be mimicked by continuous infusion Rapid-acting subcutaneous insulin –Faster absorption of insulin from subcutaneous space (doesn’t last) –Similar to a temporary increased infusion rate (not the same as IV bolus)

50. Dallas, TX November 2–4, 2012 Successful Insulin Infusion Protocols Reaches and maintains BG successfully within a prespecified target range Includes a clear algorithm for making temporary corrective changes in the IV insulin rate as patient requirements change Incorporates the “rate of change” in BG, not just the absolute values Incorporates the current IV insulin rate Minimizes hypoglycemia; provides specific directions for its treatment when it occurs Provides specific guidelines for timing and selection of doses for the transition to SC insulin

51. Dallas, TX November 2–4, 2012 IV Insulin in Non-ICU Retrospective review in 200 patients –90 General Medicine /110 General Surgery –Mean glucose 322 mg/dL Targeted glucose <150 mg/dL for 85% –67% achieved glucose <150 by day 2 –Mean glucose during infusion: 170 mg/dL Smiley D, et al. J Hosp Med. 2010;5:212-217.

52. Dallas, TX November 2–4, 2012 Hypoglycemia on IV Insulin Smiley D, et al. J Hosp Med. 2010;5:212-217.

53. Dallas, TX November 2–4, 2012 Transition from IV Continuous Insulin Infusion to SC Insulin Therapy  We recommend that all patients with type 1 and type 2 diabetes be transitioned to scheduled sc insulin therapy at least 1–2 h before discontinuation of CII.  We recommend that sc insulin be administered before discontinuation of CII for patients without a history of diabetes who have hyperglycemia requiring more than 2 unit/h.  We recommend POC testing with daily adjustment of the insulin regimen after discontinuation of CII.

54. Dallas, TX November 2–4, 2012 Transition From Intravenous to Subcutaneous Insulin Known Diabetics:  Calculate total daily insulin requirement: based on insulin rate during the last 4-hours of infusion, (e.g., 2 units/hour = 48 U/day)  Start SC insulin as follow:  ½ dose as basal (Glargine, Detemir)  ½ dose as prandial (Lispro, Aspart, Glulisine)  If patient not able to eat: give basal but hold prandial insulin Smiley et al. Ann. N.Y. Acad. Sci 1212:1-11, 2010

55. Dallas, TX November 2–4, 2012 Transition From Intravenous to Subcutaneous Insulin No History of Diabetes (stress hyperglycemia):  If HbA1c >7%, treat as diabetes  If HbA1c <6.4%  If insulin requirements during CII is 140 mg/dl  If requirements >2 U/hr during CII, start SC insulin:  ½ dose as basal (Glargine, Detemir)  ½ dose as prandial (Lispro, Aspart, Glulisine) Smiley et al. Ann. N.Y. Acad. Sci 1212:1-11, 2010

56. Dallas, TX November 2–4, 2012 CONTINUOUS SUBCUTANEOUS INSULIN INFUSION

57. Dallas, TX November 2–4, 2012 Keys to CSII Use in the Hospital AACE/ADA Inpatient Hyperglycemia –Candidates for inpatient CSII use are those using CSII as outpatients Must have mental and physical capacity to do so –Nursing personnel must document basal rates and bolus doses on regular basis –Hospital personnel with expertise in CSII therapy is essential

58. Dallas, TX November 2–4, 2012 Potential Issues with CSII Use Many nurses (and physicians) are unfamiliar with the technology and thus uncomfortable with allowing its continued use –Knowledge scores 67% for those with prior experience with CSII user vs 17% (p<0.01) –Agreed CSII effective strategy for managing diabetes in the hospital –Only 27% thought they could safely care for a patient using CSII Noschese et al. Diabetes. 2006;55:846-P.

59. Dallas, TX November 2–4, 2012 Policy for Continued CSII Use List of suggested contraindications –Altered state of consciousness –Critically ill –Risk of suicide –“other reason” deemed appropriate by MD Set of rules to guide medical staff Requirement of signed informed consent detailing conditions for CSII use Bailon et al. Endocr Pract. 2009;15:24-29.

60. Dallas, TX November 2–4, 2012 Procedures for Patients Admitted to Hospital on CSII Presence of insulin pump, brand of pump and insulin type are identified Blood or capillary glucose level is determined Contraindications for continued use of insulin pump are assessed Physician order for alternate insulin therapy is obtained if CSII must be discontinued Patient’s consent for CSII is obtained Admitting physician writes initial order for insulin pump therapy using the preprinted order form Endocrinology, diabetes education, and nutrition consultations are ordered by admitting physician Insulin pump basal-bolus blood glucose record flow sheet is placed at the patient’s bedside Bailon et al. Endocr Pract. 2009;15:24-29.

61. Dallas, TX November 2–4, 2012 Insulin Pump Therapy: One Institution’s Experience Frequency of hypoglycemic and hyperglycemic events among hospitalized patients receiving continuous subcutaneous insulin infusion (insulin pump) therapy Leonhardi BJ, et al. J Diabetes Sci Technol. 2008;2(6):948-962

62. Dallas, TX November 2–4, 2012 Pump “On” vs Pump “Off” Bailon et al. Endocr Pract. 2009;15:24-29. BG > 200 mg/dLBG < 70 mg/dL

63. Dallas, TX November 2–4, 2012 Bailon et al. Endocr Pract. 2009;15:24-29. Hypo and Hyperglycemia With and Without CSII

64. Dallas, TX November 2–4, 2012 Pitfalls to Continued CSII Use Limited experience –Published reports suggest 1-2 patients/mo Supplies –Tubing needs to be changed at most Q3days –Different pumps need different reservoirs Determining who is “in charge” and tracking the insulin dosing Pumps and MRI’s don’t mix Note, this is “continued” not “initiating”…

65. Dallas, TX November 2–4, 2012 CONCLUSIONS

66. Dallas, TX November 2–4, 2012 Summary Hyperglycemia is common in hospitals Evidence on the management of hyperglycemia in non-ICU settings is increasing –Vast majority of studies utilize subQ insulin Intravenous insulin can be implemented in non-ICU’s –Has been most often studied in ICU’s SubQ insulin infusion can be continued