International Hospital Diabetes Meeting

Summary of presentations October 8-9, 2010, San Diego,California

Randi Kington, APRN, CDE

Inpatient Protocols:

Effects of Hyperglycemia on Hospital Outcomes:

  1. Consequences of hyperglycemia; Dungan,K, Braithwaite,S, PreiserJC, Lancet 2009l 373:1798 There is a chart/graphic that shows Oxidative stress, Intracellular signaling, tissue effects and complications both acute and chronic.

Acute: renal failure, polyneuropathy, prolonged mechanical ventilation, sepsis/wound infection, ischemia/infarction, hemodynamic compromise, arrhythmia, transfusion requirements.

Chronic: nephropathy, neuropathy, retinopathy, cva/ami.

  1. Hyperglycemia is not just known diabetes. Higher mortality has been shown in “new hyperglycemia” compared to diabetes at 16% vs 3% Umpierrez, JCEM 87: 978-982, 2002 The higher the glucose in the new hyperglycemia patients, the higher the odds of inpatient mortality.
  2. Ongoing discussion of how tight the control. Published in CHEST 2010:137(3); 544-551 compared the larger of the recent published data. Cardiac surgery patients benefit from tight control but the ‘medically ill’ patients do not benefit from tight control. The meeting consensus was 100-120 on the low end and 180-200 on the upper end. Current AACE 140-180 for ICU; 110-140 for CABG and not to target <=110. ADA & AACE agree 140-180 in general.
  3. Research design, difficulty achieving substantial differences in glucose levels, the changing standard of care over time, and small studies are all contributing to conflicting and confusing results. All studies have been done in ICU/CCU type settings. No clear evidence for glucose controls in ‘ward’ patients.
  4. Finger stick glucoses in the Critical Care settings are inconsistent; sources of cap/ven/art, compromised tissues, acidotic and oxygenated states variable.

Effects of Hypoglycemia on Hospital Outcomes:

  1. Spontaneous hypoglycemia of <= 40 increases mortality by 2.5% Autonomic dysfunction leading to hypoglycemia persists after hospitalization. Is this dysfunction the cause of the increased mortality.
  2. How much of an impact comes from the glucose variability caused by correcting hypoglycemia. How much IV or oral glucose is given? How high and for how many hours is the patient then

hyperglycemic. Studies were presented that said 10-12 hours were needed to normalize the glucose again.

Page 2

  1. Similar to hyperglycemia, cellular mechanisms of inflammation, oxidative stress and endothelial dysfunction are now known to occur with hypoglycemia. There is potential for cns dysfunction and malignant arrhythmias (QT prolongation).

Glucose Variability: moderator and presenter, Irl Hirsch, MD

  1. This definition includes hyperglycemic spikes and hypoglycemia usually associated with insulin therapy.
  2. More oxidative stress than hyper or hypo alone.
  3. Progressive worsening of inpatient outcomes as glycemic variability increases; mortality increases with increased variability. Not known, is the variability the driving force?
  4. NEJM 2010, 363, 410 an Australian study was published the day before the conference. Inflammation & hypoglycemia can fuel microvascular disease.
  5. MAGE (mean amplitude of glycemic excursions) quantitates the propensity of a patient to have wide swings of blood glucose. Each glucose level is evaluated independently. In assessment ofcomplexity, each blood glucose level is assessed in relation to its neighbors in time. Some evidence supports concepts that high variability and low complexity are associated with unfavorable outcomes. (Susan Braithwaite,MD)
  6. Higher mean glucoses and higher variability had the highest mortality. Krinsley,JS Crit Care Med 2008: 36: 3008-3013.
  7. A 2010 article regarding the indices of glycemia was referenced Col Robert Vigorsky, MD as, Diabetes 49: 1585-1590.

Insulin infusion dosing in the Hospital:

  1. IV insulin infusion therapy has been enhanced by several electronic support systems that are commercially available or locally developed.
  2. The systems respond to input data including current and previous glucose levels, rate of change, and previous insulin infusion rates.
  3. There is more success when patient condition change is slower. Rapidly changing conditions can negatively affect the glycemic control.
  4. Interruptions in tube feedings, insulin added to parenteral nutrition, hemodialysis, are a few examples of conditions that electronic support would be challenged.
  5. The “Glucomander” and the ”Yale Protocol” are the two programs approved by FDA.
  6. Essentials of IV insulin algorithms: Input; current bg, previous bg, test times of current and previous bg’s, and previous insulin infusion rate. Output; nest insulin infusion rate and next test time.
  7. Multiplier formulas use the above and apply parameters appropriate to given medical functions. “Glucomander” was the first to use this.

Page 3

  1. Most of the current data available applies to ccu/icu patients.
  2. None of the systems are perfect, but can be successful.

Pitfall suggestions:

  1. Adequate point of care monitoring to avoid the hyper and hypoglycemia. Remember to order for the non DM hyperglycemia patients, high dose steroid patients, parenteral nutrition patients.
  2. Use D10 when TPN or tube feeds interrupted.
  3. EMR crossovers for p.o.c. monitoring and insulin drip orders when initiating NPO, TPN/PPN, tube feeds, etc .

Transitioning to subcutaneous dosing:

  1. Basal and bolus doses must be ordered.
  2. EMR crossovers when food is ordered.
  3. Experiential ordering is valuable when the transition is made at ‘non typical basal dose times’.
  4. Protocols may be used with and without EMR.
  5. Protocols need to steer the orders for both basal and bolus orders. Ranges for supplemental or correction doses may be set up as a check off.
  6. Cross over orders when NPO for a procedure should guide for continued basal doses at same or reduced level.
  7. Crossover orders when nutrition order is changed to “clear liquids” also guided for continued basal, but perhaps reduced bolus doses.
  8. Suggestions for management of continuous tube feeds ranged from 25-50%of total dose as basal. The popular choices were NPH every 8 or 12 hours. But some used Lantus. Correction bolus doses every4-6 hours were used with all options.
  9. Both Bode and DeSantis have published transition protocols used to determine basal and bolus doses.

Insulin Pump Protocol

Considerations from Outpatient to Hospital

  1. Competency of patient:
  1. Admission nursing assessment
  2. Bill and document under patient education for competency assessment
  3. Patient-hospital agreement
  4. Able to compute insulin carb ratios and corrections
  5. Able to test own glucose
  6. Able to change equipment

Page 4

  1. Agreement between patient and Hospital
  1. Patient uses own pump supplies
  2. Hospital meter is used for required tests
  3. Hospital insulin is used in the pump and for sc boluses if needed
  4. Change to hospital insulin within 24 hours of admission
  5. Patient will record all carbs, glucoses, boluses, basal changes on hospital form
  6. Staff inspects sites of insertion
  7. Staff determines competency for use of pump
  8. Contact number(s) for pump company will be recorded on hospital form
  9. Hospital designates where pump/supplies are to be stored when disconnected from inpatient
  10. Hospital designates where pump/supplies are to be stored when disconnected from outpatients; ie locker.
  11. Participation of patient designated caregiver for pump management
  1. Insulin via infusions or subcutaneous injections will replace pump infusion
  1. Patient request
  2. Cognitive impairment with anesthesia, sedation, pain medications
  3. Progression of labor impairing self management of pump
  4. Sepsis, trauma, critically ill, altered state of consciousness
  5. Requires insulin drip
  6. Radiological or MRI procedures (may be temporarily disconnected)
  7. Depressed/suicidal
  8. Patient or caregiver unable or unwilling to provide self care
  9. Non adherence to agreement for patient use
  10. Malfunction of pump
  11. Lack of supplies
  1. Insulin orders for pump use
  1. Diabetes/Endocrine team
  2. Designated hospitalist per prior agreement
  3. Use of U500 insulin
  1. Ongoing evaluations
  2. Site evaluation by nursing staff
  3. Competency
  4. Glucose management/insulin adjustments
  5. Adherence to hospital policies
  6. Evaluate adverse events for patients remaining on pump
  1. EMR suggestions
  2. Order sets for Endocrinology/diabetes Team consults
  3. Order sets for basal and bolus rates, correction factors and carb ratios

Page 5

  1. Crossovers from insulin order set to education documentation screen for competency
  2. Crossovers when NPO ordered to query insulin drip or sc insulin orders
  3. Crossovers for MRI’s and other radiology exams to discontinue pump
  4. Crossover when pump discontinued to insulin drip or sc insulin orders
  5. Crossover from sedation orders to discontinue pump.
  6. Orders to change sites if glucose >250
  7. Orders for sc bolus if glucose >250

Curtiss Cook, MD and Mary Korytkowski, MD have published data regarding pump use in the hospital.

Mary Korytkowski, MD University of Pittsburgh and Portland Oregon Medical Center have pump policies published.

U500

  1. All agreed this requires experience. Protocols need to be established as to who can order it.
  2. There was mixed response as to how supplied; RPh measured and sent to floor vs RN on unit.

Education on Nursing Units

  1. Any new protocol, policy or procedure needs to involve the nurse from step one of planning to have any success. Nurse representation from any area involved is highly recommended. Buy in is directly related to successful implementation.
  2. Ongoing support is needed for ongoing success.
  3. Staff nurses were expected to oversee pump management, all teaching related to diabetes self management, etc. In these hospitals there was CDE support by a hospital based RN or APRN.

Monitoring on Units

  1. Continuous monitoring not there yet. Both companies were focusing on ICU type patients. They are quite time consuming and as monitoring currently is not approved in the critical care units, my best guess is two years.
  2. Point of care monitoring will soon be compromised by a requirement for cleaning between patients. An 8-10 minute process is going to be imposed. Issue came from contaminated body fluids on gloves that were holding the monitors and thus the surface was contaminate and carried to the next patient.