Advantages of Tight Glycemic Control in the Hospital Setting

Advantages ofTightGlycemicControlin the Hospital SettingRichard M. Bergenstal, MDExecutive DirectorInternational Diabetes Center at Park NicolletMinneapolis, MNPresident, Medicine & ScienceAmerican Diabetes Association

Disclosure StatementRichard M. Bergenstal, MDI have participated in clinical research and/or served asa consultant for:• Eli Lilly• Novo Nordisk• sanofi-aventis• MannKind Roche LifeScan / J&J Abbott Bayer• Medtronic• Intuity• DexCom / EdwardsInstitute for Research and Education•I have inherited Merck stock• Amylin• Merck• Pfizer• ResMed• Takeda• HygieiaRMB receives no personal compensation for any of theseactivities and all contracts are with the non-profit Park NicolletI am a volunteer officer of the American Diabetes Association

1986 ADArecommended 10%not 15% total error?But by 1993 this goalwas not achieved.

Complications Risk in DiabetesThe Impact of Intensive Glycemic Control

International Diabetes CenterPark Nicollet

DCCT was presented at ADA June 1993 and thisADA SMBG Consensus meeting was Sept 1993

I believe the American DiabetesAssociation feels that: SMBG is one important component of safelyimproving glucose control

2009

NHS Diabetes publishesSelf-Monitoring of Blood Glucose report26 February 2010

Status of Glucose Control in the USA1C Measures from NHANES10075% of IndividualsAchieving A1c

Severe Hypoglycemia Ratesin INT & STD GroupsPts w hypoglycemia events (%)INTintensive15% vs. 5%3-fold increaseSTDstandard

ReasonSelf-reported Antecedents toHypoglycemia Events% (n=875 events)Delayed/missed meal or ate fewer carbohydrates 58% (510)None 17% (148)Took incorrect dose of glucose lowering medication 9% (82)Cognitive decline 8% (70)Intercurrent illness 5% (44)Ingested alcohol 3% (26)Recent weight loss 3% (29)Started or increase of other medication 3% (28)

Why Can’t We ImproveGlycemic Control & Do it Safely?Maybe We Rely Too Much on the HbA1cAlone for Clinical Decision MakingHbA1cif elevatedSMBGif displayedto show a patterntells you a change in therapyis neededtells you what changein therapy to make

I believe the American DiabetesAssociation feels that: SMBG is one important component of safelyimproving glucose control There is room for improvement in both SMBGaccuracy and the clinical use of SMBG data– Patients in the hospital setting have the most potentialto benefit from improved meter accuracy. In addition,improved accuracy would add some benefit in mostoutpatient settings– Teaching professionals and patients how to effectivelyuse SMBG data needs at least equal emphasis The role of CGM (inpatient & outpatient) deserves intensestudy and continued innovation to define its role inmanagement in both settings.

Special thanks tomy colleaguesfor data and slidesSilvio E. Inzucchi MDEtie S. Moghissi, MDMary Korytkowski MDAnthony Furnary, MD

Diabetes and HospitalizationScope of the Problem Prevalence of diabetes is estimated at 12% to25% of hospitalized patients and may besignificantly underestimated 50% of patients admitted to CCU have diabetesor pre-diabetes 29%-50% of all cardiac surgery patients 1-3 days longer hospital stayACE Position Statement. Endo Pract. 2004;10:77-82.Clement S. Diabetes Care. 2004;27:553-591.Diabetes Care. 2006, 29; 1955-1962

Inpatient Glucose ManagementThe Past Acute hyperglycemia was considered to bebenign Sliding scale was the norm No published standards of care or guidancefrom medical societies Hyperglycemia in the hospital wasessentially ignored

Glycemic Control in the Hospital:An Elusive Goal“Stress hyperglycemia”D/C outpatient regimensIV D5/ TPN / PPNSteroids, pressors↓ Physical activityFear of hypoglycemia↓ NutritionMeal interruptionsMonitored complianceInsulin ‘stacking’Δ Mental status Metchick LN et al. Am J Med 113:317, 2003

The ‘Portland Protocol’:Glucose Control & Sternal InfectionsDSWI(%)4.03.02.0Insulin infusionPatients with DMPatients without DM1.00.087 88 89 90 91 92 93 94 95 96 97YearDSWI = deep sternal wound infection; CII = continuous insulin infusion.Furnary AP et al. Ann Thorac Surg 67:352, 1999

Intensive Insulin Therapy in theSurgical ICU: The Leuven Study1548 SICU patientsIntensiveIV insulin if BG >110 mg/dlTarget: 80-110 mg/dl39% insulinBG=153 mg/dl99% insulinBG=103 mg/dlVan Den Berghe G et al. N Engl J Med 345:1359, 2001

Intensive Insulin Therapy in theSurgical ICU: The Leuven Study100100Survival in ICU (%)96928884Intensive treatmentConventional treatmentMORTALITY ↓ 42%P

Hospital Mortality Rate & GlucoseControl in a Medical-Surgical ICUMortalityRate(%)N=1826 ICU patientsMean Glucose Value During ICU Stay (mg/dL)Krinsley JS. Mayo Clin Proc. 2003;78:1471–1478.

Hyperglycemia: A Predictor of MortalityFollowing CABG in Diabetics10BG 200P

Intensive Insulin Therapy in Patients withSevere Sepsis (VISEP Study) N = 537Multicenter, 2x2 factorial trialIntensive group vs conventionalgroup• Mean morning bloodglucose (112 mg/dL vs 151mg/dL; P

Tight Glucose Control in Critically Ill PatientsA Meta-AnalysisWiener RS et al. JAMA 300:933-944, 2008. 29 randomized controlled trials totaling 8432patients Tight glucose control vs usual care: no significantdifference in mortality Tight glucose control was significantly associatedwith:• Higher risk of hypoglycemia (≤ 40 mg/dL) (13.7% vs2.5%; RR, 5.13)

Responding to New EvidenceAACE/ADA Inpatient Glycemic task ForceEtie S. Moghissi, MD, FACP, FACE, AACE ChairMary T. Korytkowski, MD , ADA Chair Monica DiNardo, MSN, CRNP, CDE Daniel Einhorn, MD, FACP, FACE Richard Hellman, MD, FACP, FACE Irl B. Hirsch, MD Silvio E. Inzucchi, MD Faramarz Ismail-Beigi, MD, Ph M. Sue Kirkman, MD; Guillermo E. Umpierrez, MD, FACP, FACE

AACE-ADA Consensus Statement on Inpatient Glycemic ControlMay 2009ENDOCRINE PRACTICE 15: 1-17, 2009Diabetes Care 32: 1119-1131, 2009

NICE-SUGAR Study: Design Multi-centre, open label, randomized, controlled trial Examining the effects of blood glucose management on 90day, all cause morbidity and mortality 6104 ICU patients• 3054 in intensive control group (target: 81 to 108 mg/dL)• 3050 in conventional control group (target: ≤180 mg/dL) The two groups had similar baseline characteristics 40 centers of Australia, New Zealand, and Canada Recruitment April 2005 to November 2008 Last Follow-up November 2008The NICE-SUGAR Study Investigators. N Engl J Med. 2009;360:1283-1297.

The NICE-SUGAR StudyBlood Glucose Level, According toTreatment GroupProbability of Survival751829RR= 1.14IIT goal: 81 – 108 mg/dL(mean BG 118 mg/dL)CIT goal:

NICE-SUGAR Study Results:Treatment and Glucose MeasuresInsulin RxPatients treated withinsulinIntensiveControl GroupConventionalControlGroupP Value97.2% 69.0%

NICE-SUGAR Study: OutcomesOutcomemeasureIntensiveControlGroupConventionalControl GroupP Value28 Day Mortality 22.3% 20.8% 0.1790 DayMortality27.5% 24.9% 0.02Mech. Ventilation96%95.3%0.17(Mean days + SD)(6.6 + 6.6)(6.6 + 6.5)(0.56)Dialysis 15.4% 14.5% 0.34Bloodstreaminfections12.8% 12.4% 0.57The NICE-SUGAR Study Investigators. N Engl J Med. 2009;360:1283-1297.

NICE SUGAR Study: Conclusions This large, international, randomized trial, foundthat intensive glucose control did not offer anybenefit in critically ill patients Blood glucose target of < 180 mg/dL with theachieved target of 144 mg/dL resulted in lower (90day) mortality than did a target of 81-108 mg/dL There was increased hypoglycemia with lowerglucose targetsThe NICE-SUGAR Study Investigators. N Engl J Med. 2009;360:1283-1297.

Mean Glucose & In-Hospital Mortalityin Patients with AMI(Reference: Mean BG 100-110 mg/dl)N= 16,871Kosiborod M et al. Circulation 2008:117:1018

What Should WeTake Away from these Trial? Good glucose control, as opposed to near-normalcontrol, is likely sufficient to improve clinicaloutcomes in the ICU setting We need to have a renewed concern abouthypoglycemia in critically ill patients Importantly, the NICE-SUGAR results do notendorse a laissez-faire attitude toward inpatienthyperglycemia that was prevalent a decade ago It is time to rethink the targets and measurement

What glycemic targets can be recommended indifferent patient populations?TARGETS: Critically ill Insulin infusion to control hyperglycemia Starting threshold no higher than 180 mg/dL Maintain BG between 140-180 mg/dL• Possibly greater benefit at lower end of range Somewhat lower targets may be appropriate inselected patients Targets < 110 mg/dL not recommendedNot recommended< 110May beappropriate110-140Recommended140-180Not recommended>180Moghissi ES, et al; AACE/ADA Inpatient Glycemic Control Consensus Panel. Endocr Pract. 2009;15(4).

AACE/ADA Target Glucose Levelsin Non–ICU Patients Non–ICU setting:• Premeal glucose targets

Recommendations for Managing PatientsWith Diabetes in the Hospital SettingAntihyperglycemic TherapyInsulinRecommendedOADsNot GenerallyRecommendedIV InsulinCritically ill patientsin the ICUSC InsulinNon-critically illpatients1. ACE/ADA Task Force on Inpatient Diabetes. Diabetes Care. 2006 & 20091. Diabetes Care. 2009;31(suppl 1):S1-S110.

IV Insulin ProtocolsMultiple published protocols are available that are effectiveand safe. Some examples include: Yale Markovitz Leuven Portland Texas Diabetes Council DIGAMI University of Washington Luther Midelfort Mayo Health System Rush University Protocol Northwestern UniversityGoldberg PA et al. Diabetes Care. 2004;27:461-467.Markovitz LJ et al. Endocr Pract. 2002;8:10-18.Van den Berghe G et al. N Engl J Med. 2001;345:1359-1367.Malmberg K et al. Circulation. 1999;99:2626-2632.Malmberg K et al. Eur Heart J. 2005;26:650-661.Krinsley J. Mayo Clin Proc. 2004;79:992-1000.Ku SY et al. Jt Comm J Qual Patient Saf. 2005;31:141-147.Donaldson S et al. Diabetes Educ. 2006;32:954-62.DeSantis AJ et al. Endocr Pract. 2006;12:491-505.Texas Diabetes Council. Available at: http://www.dshs.state.tx.us/diabetes/pdf/algorithms/iv%20insulin%20infusion.pdf

Yale Insulin Infusion Protocol.:First 33 UsesGoldberg PA, et al. Diabetes Care. 2004;27:461-467.

The Ideal IV Insulin Protocol Easily ordered (signature only) Effective (gets to goal quickly) Maintains BG within a defined target range over severalhours Includes an algorithm for making temporary correctiveincrements or decrements of infusion rate Safe (minimal risk of hypoglycemia) Directions for treatment of hypoglycemia Easily implemented Can be executed by nursing staff in response to a singlephysician order

Recommendations for Managing PatientsWith Diabetes in the Hospital SettingAntihyperglycemic TherapyInsulinRecommendedOADsNot GenerallyRecommendedIV InsulinCritically ill patientsin the ICUSC InsulinNon-critically illpatients1. ACE/ADA Task Force on Inpatient Diabetes. Diabetes Care. 2006 & 20091. Diabetes Care. 2009;31(suppl 1):S1-S110.

RABBIT 2: Randomized Study of Basal-Bolus InsulinTherapy in the Inpatient Management of patientswith type 2 diabetes240220Mean blood glucose difference betweengroups = 27 mg/dL (P

“Basal-Bolus” Insulin TherapyRapid (Lispro, Aspart, Glulisine)InsulinLevelLong (Glargine, Detemir)0 2 4 6 8 10 12 14 16 18 20 22 24Hours

Maintaining Physiologic Insulin Deliveryin the Hospital --- 3 Part Insulin OrderCorrection Dose orSupplemental Dose ofInsulinInsulinMealtime insulin(bolus)Basal insulinBreakfast Lunch Dinner Bedtime

Estimating SC Insulin Dose1. Calculate estimated total daily dose of insulin asfollows:– T2DM: 0.5 to 0.7 U/kg– T1DM or type unknown: 0.3 to 0.5 U/kg2. Divide total daily dose of insulin into 50% basalas glargine or detemir and 50% prandial asaspart, lispro, or glulisine3. Divide prandial insulin into 3 equal doses to begiven with mealsDeSantis AJ et al. Endocr Pract. 2006;12:491-505.

Premeal Algorithm forCorrection-Dose InsulinAdditional InsulinPremeal BGmg/dLLowDoseMediumDoseHighDoseIndividualized150–199 1 unit 1 unit 2 units200–249 2 units 3 units 4 units250–299 3 units 5 units 7 units300–349 4 units 7 units 10 units>349 5 units 8 units 12 unitsTo be administered in addition to scheduled insulinto correct premeal hyperglycemia

Hypoglycemia Is a Concern

Addressing Safety Concerns of Inpatient Management ofHyperglycemia

What are areas for future research? Stress hyperglycemia (mechanisms, targets) Severe hypoglycemia (causes, outcomes, risks,costs) Comparisons of glycemic targets and outcomes innon-critically ill patient populations Effect of glycemic variability on patient outcomes Hospital systems and safety Monitoring (meter accuracy, CGM) Pediatric populations (targets)

Continuous Glucose Monitoringin the Inpatient Setting• Good control of blood glucose is recommendedin the ICU and medical floors• If glucose could be accurately measured morefrequently (continuously) hypoglycemia may beminimized and even tighter glycemic control mayprove beneficial• Research in underway by Edwards Lifesciences,DexCom, Inc. and others evaluating the accuracyand utility of frequent IV glucose monitoring in theinpatient setting. Initial results presented atscientific meetings are very encouraging.

How Will We Translate What We Are LearningInto Improved Practice and Outcomes? FDA, ISO & CLIA-POC are gatheringcritical data and writing important standardsor guidelines regarding glucose monitoring inthe hospital and outpatient settings. But where do people with diabetes, cliniciansand educators look for guidance on diabetesmanagement?

How Will We Translate What We Are LearningInto Improved Practice Patterns and Outcomes? FDA, ISO & CLIA-POC should find a wayto Partner with the American DiabetesAssociation (and others) to translate newscience and standards into changes inpractice patterns for patients and healthcare professionals.