Safe and Effective Insulin Infusion Therapy Judith Jacobi, PharmD, FCCP,BCPS Critical Care Pharmacist Indiana University Health Methodist Hospital Indianapolis, IN [email protected]

Disclosure • No Disclosures

Guidelines Committee Judith Jacobi, PharmD Nicholas Bircher, MD James Krinsley, MD Michael Agus, MD Susan S. Braithwaite, MD Clifford Deutschman, MD Amado Friere, MD, MPH Douglas Geehan, MD Benjamin Kohl, MD

Stanley Nasraway, MD Mark Rigby, MD, PhD Karen Sands, APRN-BC, ANP, MSN Lynn Schallom, RN, MSN,CS Beth Taylor, MS, RD, CNSD Guillermo Umpierrez, MD John Mazuski, MD Holger Schunemann, MD

Objectives • Define elements needed for insulin infusion therapy • Describe elements of a safe insulin infusion protocol • Review literature on treatment of hypoglycemia • Describe potential metrics for insulin infusion therapy

Questions to Address Intravenous Insulin Therapy

• What is the impact of hyperglycemia on outcome of CC patients? – What is the glucose trigger for intensive insulin therapy?

• What is the impact of IIT on patient outcome? • What is the optimal glucose range during IIT? • What are the population-specific considerations for glycemic control?

Questions to Address • How should glucose be monitored in ICU patients? • How often should glucose be monitored? • What other interventions may impact glucose control? • What is the impact of hypoglycemia? • How should hypoglycemia be treated?

Questions to Address • What metric should be used to assess the achievement of glycemic goals? • What is the economic/workload impact of tight glycemic control? • When/how should patient’s transition off an insulin infusion?

What is the impact of hyperglycemia? • Hyperglycemia is common

– Stress response – Drug therapy (corticosteroids) – Nutrition (overfeeding, IV dextrose)

• Hyperglycemia contributes to organ dysfunction – – – – –

Mitochondrial injury WBC dysfunction Reduced respiratory chain complex activity Oxidative injury Elevated cytokine levels and NFκB

Glucose vs. Mortality

N= 259,040 ICU patients

Falciglia Crit Care Med 2009; 37:3001–3009

Intensive Insulin Therapy – Big Impact • Mortality reduced from 8% to 4.6% with target mean morning glucose 180-200 (1011.1 mmol/L) vs. 80-110 mg/dL (4.4-6.1 mmol/l) – RCT, adequate power – Mechanical ventilation

• 2/3 CV surgery- most with short LOS

– Confounders present

• Single center • One glucose per day- mean morning glucose • Role of IV dextrose/TPN- clamp to high glucose levels

Van den Berghe N Engl J Med 2001; 345:1359-1367

IIT in Critically Ill Patients: SICU * * *

*P 150 mg/dl (>8.3 mmol/l) Glu 110-150 mg/dl (6.1 – 8.3 mmol/l)

Subset > 5 days in the ICU Glu 25: 31% Finfer N Engl J Med 360;13: 1283-1297

The NICE-SUGAR Study Blood glucose

3054 received IIT goal: 81 – 108 mg/dL (mean BG 118 mg/dL) 3050 received CIT goal: 10% severe hypoglycemia rate – Risk of failure to detect a change

• Continuous monitors future solution Diab Care 2010: 33, Suppl 1, s11-s61 Quinn Pharmacotherapy 2006; 1410-1420

Insulin Infusion Potential Benefit in TPN

Marik Chest 2010; 137:544-551

Glucose Meters in the ICU • Not developed for ICU use • Significant error risk in critically ill – Anemia (Hct < 25-30%) – Drug interactions

• Avoid fingerstick if poor perfusion, on vasopressors, significant edema

– Whole blood (arterial/venous) preferable – Potential to double-correct for plasma equivalent

• Lab standards: Over 95% of discrepancy is < 15mg/dL (< 75 mg/dL = 4.2 mmol/l) or up to 20% for glucose ≥ 75 mg/dL (CLIA) Mahoney Diab Technol Ther 2007; 9:545-552

Glucose Meter – Capillary Blood

73% within goal for glucose meter analysis of capillary blood Kanji Crit Care Med 2005; 33:2778–2785

Confounders in POC Glucose Assay Patient variables Hypotension Hct pO2 Temperature Drugs

GO= glucose oxidase GD= glucose dehydrogenase Dungan Diab Care 2007; 30:403-409

Glucose Meter – Arterial Blood

88% within goal for glucose meter analysis of arterial blood Kanji Crit Care Med 2005; 33:2778–2785

Insulin Titration Protocols • Over 30 publications describing protocols • Many derivatives of a few protocols • Standard approach reduces variability • Challenge to balance prescriptive (complex) with less defined (simpler) • Can 1 protocol work in all populations?

Leuven Protocol

Van den Berghe Int J Obesity 2002: 26, Suppl 3, S3–S8

Portland Protocol

Ref:

PortlandProtocol.org

Ideal Solution • Computer-directed insulin infusion – Complexity is moved to the computer – Standardization is achieved

• Insulin = (Glucose – 60) x Multiplier – Multiplier = 0.02 baseline • Increase 0.01 for each value above target • Decrease 0.01 for each value below target

Davidson PC, et al. Diabetes Care. 2005; 28:2418-2423. 33

Protocol Compliance

Lipshutz, A et al. Strategies for Success: A PDSA Analysis of Three QI Initiatives in Critical Care, The Joint Commission Journal on Quality and Patient Safety, August 2008 ,Volume 34 Number 8.

Algebraic Protocols

Chart Copyright Georgia Hospital Association Osburne Diabetes Educator 2006; 32:392-403

Insulin Titration & Concomitant Therapies • Critical thinking skills • System support for safe insulin therapy – Double checks – Avoid intermittent dextrose administration

• Multidisciplinary communication of patient specific goals • ∆ Clinical status • ∆ Nutritional support • ∆ medications

• Protocol driven assessment guidelines • If change, reassess more frequently

Treatment of Hypoglycemia During Insulin Infusion • ½ to 1 amp 50% Dextrose – Incomplete order

• Guidelines suggest 15 gm oral glucose – Average change 38 mg/dL in 20 minutes – Retest BG in 15 minutes

• Canadian Guidelines: with IV access – 10-25 gm (20-50 ml) 50% Dextrose over 1-3 minutes

Diabetes Care Jan 2004 Supplement www.diabetes.ca.cpg2003

IV Dextrose During Insulin Infusion • Give a metered dose of dextrose and titrate based on BG • Dose 50% Dextrose (ml) = (100-BG) x 0.4 – Example: Glucose 50 mg/dL = 20 ml – Glucose = 65 mg/dL = 14 ml

• Repeat BG in 15 minutes – Corrects BG from ≤ 50 to 106.4 mg/dL in 20 minutes1 – Corrects BG from 49 10 to 83 10 mg/dL in 33 minutes2 1. Junega Diabetes Technol Ther 2007;9:203-211 2. Davidson Diabetes Care 2005;28:2418-2423

Metrics • Reliable documentation and data review – Transcription errors – Protocol deviaitons

• Potential measures – % time in goal range – Rate of severe hypoglycemia < 40 mg/dL (