Should Insulin be added to Parenteral Nutrition?

ESPEN Congress Gothenburg 2011 Educational Session - Pharmaceutical session (in collaboration with ASPEN) Should Insulin be added to Parenteral Nutrition? Jay M Mirtallo

Should Insulin be added to Parenteral Nutrition? Jay M Mirtallo, MS, RPh, BCNSP, FASHP Associate Professor of Clinical Pharmacy The Ohio State University, College of Pharmacy Pharmacy Practice and Administration Division President, The American Society for Parenteral and Enteral Nutrition

Insulin Indications for PN Manage hyperglycemia Protein accretion (anabolism) Anti-inflammatory activity

Insulin in PN: Outline Indication: Manage Hyperglycemia Frequency of hyperglycemia in PN patients Association of adverse outcomes with hyperglycemia in PN patients Variables associated with poor glucose control in PN patients Issues with insulin in PN Criteria for adding medications to PN

PN: Frequency of Hyperglycemia Reference Criteria N (%) Comment NSS Dodds et al. NCP 2001 @ least 1 value > 200 mg/dl 762 (28) Only 2 pts developed symptoms complication Yes Weinsier et *al. JPEN 1982 > 300 mg/dl after at least 48 hrs of PN 47 (47) No symptoms observed Dextrose based PN Yes but not used, guidelines, flow sheets, order sets in place ChrisAnderson * et al JPEN 1996 As per Weinsier 23 (22) (41) (39) No effect of NSS in prospective trial -Non NSS -NSS Use of 3 in 1 (TNA) Yes, consult with recommendations only(64% compliance), substantial staff education Rosmarin et al, NCP 1995 > 200 mg/dl 0 (0) 5 (7) -Dext infusion < 4 -Dext infusion 4-5 Yes, dextrose based diet 18 (43) -Dext infusion >5 No risk factors present Pleva et al, NCP 2009 >200 mg/dl >150 mg/dl 22 (44) 45 (90) Resulted in 1.4 events per patient course Risk factors: diab, pancreatitis, Steroids Yes, pharmacist management *Same institution

PN and Hyperglycemia: Adverse Outcomes Relation between blood glucose levels and outcomes Myocardial infarction Stroke Cardiothoracic surgery Critical illness General hospitalized patients Cheung NW et al Diabetes Care 2005; 28: 2367-2371

PN and Hyperglycemia: Adverse Outcomes Risk of any complication 1.58 (p <0.01) Infection Septicemia Acute renal failure Cardiac complications Death Quartile analysis Risk level increased at high quartile vs low quartile group OR of 4.3 for complication, 10.9 for death Cheung NW et al Diabetes Care 2005; 28: 2367-2371

Severity of Hyperglycemia Stronger predictor of adverse outcomes than history of diabetes Majority of PN patients who become hyperglycemic are not diabetic Excluded from Rosmarin study 12% of Pleva study population 27% of Wah Cheung study population Evidence that hyperglycemia in itself is harmful. Cheung NW et al Diabetes Care 2005; 28: 2367-2371

Blood glucose, mg/dl 450 400 401 350 300 250 200 158 150 100 112 129 50 47 0 0 435 869 1304 1738 Quartile Values for Patient Population Mirtallo PN Pleva et al Nutr Clin Pract 2009; 24:626-634

Variables Associated with PN Hyperglycemia Caloric dose Type of calorie provided Hidden Sources of CHO Target glucose range Impact of controlling glucose to this range Responsibility for glucose management Use of sliding scale insulin FEAR of hypoglycemia

Variables Associated with PN Hyperglycemia No one method known to be effective in achieving target glucose Lack of consensus for insulin use Long-acting insulin Sliding scale insulin Insulin drip Insulin in PN Any combination of the above Practice varies widely among patient populations, disciplines and individual clinicians Overall, management of hyperglycemia is most important Interdisciplinary nutrition care Experience and skill of staff managing PN View as a process

Insulin in PN Criteria for medications added to PN Stable and compatible Evidence supports clinical value of medication administered in PN Frequency of dosage adjustment no more than every 24 hours Insulin is associated with frequent harmful events in PN Mirtallo et al. JPEN 28 (suppl) S39-S70, 2004

Management of Hyperglycemia: Alternative to Insulin-Hypocaloric PN Reference Criteria Comment Choban PS et al Am J Clin Nutr 1997 McCowen KC et al Crit Care Med 2000 Ahrens CL et al Crit Care Med 2005 Hypocaloric 75 Cal:g nitrogen Normal 150 Cal: g nitrogen 2 g Pro/kg IBW/d Study: 1000 kcal, 70 g pro Control: 25 kcal/kg, 1.5 g/kg pro Low Cal: 20 NPC/kg/d Standard Cal: 30 NPC/kg/d Obese patients 12 pts received insulin (11/12 diabetic) Less insulin days in NIDDM hypocaloric group No difference in Frequency of hyperglycemia non diabetics insulin use Average glucose worse nitrogen balance in hypocaloric group Excluded underweight/morbid obese Used sliding scale insulin Insulin in PN if >50% values > 200 Fewer hyperglycemic events and lower severity Mean glucose lower (118 vs 172)

Insulin Availability from PN Range: 10-95% Composition of PN Lipids, trace elements, vitamins Final concentration of insulin Assay for insulin Laboratory simulation of clinical practice Adequate monitoring of patient clinical response Seres DS; NCP 1990;5: 111-116

Evidence Supporting Insulin Use Diabetic Patients Pre hospital insulin dose Reduced daily dextrose dose to start 100 g Type 1 150 g Type 2 Accept modest hyperglycemia to avoid hypoglycemia Sliding scale insulin: glucose > 250 Mean glucose around 200 mg/dl No hypoglycemic episodes Insulin in PN Significant calories from enteral nutrition or tube feeding Insulin separate from PN Source of dextrose determines route of insulin Dialysis Hongsermeier T et al. JPEN 17:16-19, 1993

Evidence Supporting Insulin Use: Insulin protocol NSS: primarily pharmacist Capillary Blood glucose (CBG) every 6 hrs Criteria: glucose > 140 mg/dl Insulin dose per g Carbohydrate (CHO) PN induced hyperglycemia 1 U/20 g CHO Diabetes/glucocorticoids CBG <11.1 mmol/l (200 mg/dl) 1 U/10 g CHO + 0.15 U/kg/d CBC > 11.1 mmol/l 1 U/5 g CHO + 0.25 U/kg/d 2/3 insulin dose in PN, 1/3 separate as long-acting insulin Jakoby MG et al. JPEN @ http://pen.sagepub.com/content/early/2011/08/06/0148607111415628

Evidence Supporting Insulin Use: Insulin protocol Mean CBG < in protocol group by 21 mg/dl Higher CBG in diabetic group but better control with protocol Hypoglycemia (CBG < 80) more frequent in protocol group (3 vs 1%) No episodes of severe hypoglycemia (CBG < 40) Jakoby MG et al. JPEN @ http://pen.sagepub.com/content/early/2011/08/06/0148607111415628

Evidence Supporting Insulin Use: Computer-assisted, Critically Ill Nurse centered computerized decision support for insulin administration step-up rule Graded increases in amount of PN administered For glucose < 10 mmol/l (180 mg/dl) End-point: achieve full PN at 24 hours along with glucose control during introduction period Goal: 25 kcal/kg/d, max = 2500 kcal Use of insulin drip Desired caloric intake achieved within 24 hr Glucose levels 6.6 (119 mg/dl) to 7.6 (137 mg/dl) mmol/l (ave 7.4 (133)) Insulin drip rate of 1.1-2.0 U/h Hoekstra M et al. JPEN 34: 549-553. 2010

Should Insulin be Added to PN? It depends Critically ill: separate insulin infusion (drip) Significant calories from enteral or tube feeding: separate insulin as sliding scale or long-acting Minimize Hidden sources of glucose Others: definitely use insulin in PN Evidence that better than using sliding scale insulin Reasonable glucose control with minimal hypoglycemia Consider insulin dose per gram of carbohydrate in PN Adjust dose daily with sliding scale insulin 2/3 previous days insulin dose

Systems Issues Establish target glucose Interdisciplinary involvement Assign responsibility for glucose control Provide algorithm or protocol to follow Evaluate success in achieving target glucose values

A.S.P.E.N. Guideline Insulin use in PN should be done in a consistent manner according to a method that healthcare personnel have adequate knowledge Mirtallo et al. JPEN 28 (suppl) S39-S70, 2004

Algorithm Steps 1 and 2 Step 1: Risk Assessment Does the patient have risk factors* for hyperglycemia during PN? Risk Factors* Diabetes Pre-existing hyperglycemia (> 150) Pancreatitis Corticosteroids Octreotide No Yes Routine Glucose Monitoring Monitor 5 AM blood glucose daily Order Accuchecks Q6H Order Accuchecks Q6H with sliding scale insulin Start sliding scale at 150 mg/dl and correct with 2-4 units for every 50 mg/dl above 150 No Does patient have >2 blood glucose >150 mg/dl in 24 hr? Yes Target Serum Glucose Continuous infusion: 100-150 mg/dl Cyclic: 100-200 mg/dl Step 2: Minimize glucose from other sources Is the patient hyperglycemic while on sliding scale insulin? If yes, then minimize... Maintenance IV with dextrose Medications prepared in dextrose Oral diet Tube feeding

Glucose Algorithm Step 3 Step 3: Adding insulin into PN What is the patient s glucose level prior to initiation? Euglycemic (for diabetic patients) Insulin dose: 0.1 units per gram of dextrose in formula (i.e. 15 units insulin per 150 grams/l dextrose) Serum glucose > 300 mg/dl PN contraindicated Normalize serum glucose prior to starting PN Hyperglycemic If glucose is 150-200 mg/dl Insulin dose: 0.1-0.2 units per gram of dextrose in formula (i.e. 15-30 units insulin per 150 grams/ L dextrose) If glucose > 200 mg/dl Start PN at 100 grams/l dextrose Insulin dose: 0.1 units per gram of dextrose in formula (i.e. 10 units insulin per 100 grams/l dextrose)

Glucose Algorithm Step 4 Step 4: Insulin monitoring Monitor Q6H Accuchecks Is the patient s blood glucose within goal range? Yes No Increase PN to goal rate When glucose is controlled at 40 ml/hr Adjust insulin in PN by adding 75% of insulin dose used via sliding scale in previous 24 hours Monitor Q6H Accuchecks Is glucose within range? No No Monitor Q6H Accuchecks Is glucose within range? Yes Yes Continue current insulin regimen Increase PN to goal rate When glucose is controlled at 40 ml/hr Yes Monitor Q6H Accuchecks Is glucose within range? No Adjust insulin in PN by adding 75% of insulin dose used via sliding scale in previous 24 hours

Glucose Algorithm - Notes PN may be cycled if glucose is controlled on continuous PN while at goal rate Taper insulin in PN when glucose < 100 mg/dl for 3 of 4 Accuchecks in a 24-hr period Insulin limit in PN is 60 units/l. If patient needs more insulin, then discontinue insulin in PN and begin insulin drip.

Summary The use of insulin in PN is a controversial topic Primary indication: hyperglycemia associated with PN Original issues with bioavailability from PN Little evidence evaluating outcomes of insulin use in PN Considerable variability in types of patients and PN practices

Conclusion Evidence and clinical practice suggests insulin is clinically effective in PN when dosage adjustments are suitable on a daily basis

References Shizgal HM, Posner B. Insulin and the efficacy of total parenteral nutrition. Am J Clin Nutr 50:1355-63, 1989. Cheung NW, Zaccaria C, Napier B, Fletcher JP. Hyperglycemia is associated with adverse outcomes in patients receiving total parenteral nutrition. Diabetes Care 28:2367-2371, 2005 Dodds ES, Murray JD, Trexler KM, Grant JP. Metabolic occrurences in total parenteral nutrition patients managed by a nutrition support team. Nutr Clin Pract 16:78-84, 2001 Weinsier RL, Bacon J, Butterworth CE. Central venous alimentation: a prospective study of the frequency of metabolic abnormalities among medical and surgical patients. J Parenter Enter Nutr 6: 421-425, 1982 ChrisAnderson D, Heimburger DC, Morgan SL et al. Metabolic complications of total parenteral nutrition: effects of a nutrition support service. J Parenter Enter Nutr 20:206-210, 1996 Rosmarin DK, Wardlaw GM, Mirtallo J. Hyperglycemia associated with high, continuous infusion rates of total parenteral nutrition. Nutr Clin Pract 11:151-156, 1996 Pleva M, Mirtallo JM, Steinberg SM. Hyperglycemic events in non-intensive care unit patients receiving parenteral nutrition. Nutr Clin Pract 24: 626-634, 2009 Mirtallo J, Canada T, Johnson D, et al. Safe practices for parenteral nutrition. J Parenteral Enter Nutr 28 (suppl):s39-s70, 2004 Choban PS, Burge JC, Scales D, Flancbaum L. Hypoenergetic nutrition support in hospitalized obese patients: a simplified method for clinical application. Am J Clin Nutr 66:546-550, 1997 McCowen KC, Friel C, Sternberg J, et al. Hypocaloric total parenteral nutrition: effectiveness in prevention of hyperglycemia and infectious complications a randomized clinical trial. Crit Care Med 28:3606-3611, 2000 Ahrens CL, Barietta JF, Kanji S et al. Effect of low-calorie parenteral nutrition on the incidence and severity of hyperglycemia in surgical patients: a randomized, controlled trial. Crit Care Med 33:2507-2512, 2005\ Seres DS, Insulin adsorption to parenteral infusion systems: case report and review of the literature. Nutr Clin Pract 5:111-117, 1990 McMahon MM. Management of parenteral nutrition in acutely ill patients with hyperglycemia. Nutr Clin Pract 19:120-128, 2004 Hongsermeier T, Bistrian BR. Evaluation of a practical technique for determining insulin requirements in diabetic patients receiving total parenteral nutrition. J Parenter Enter Nutr 17:16-19, 1993 Jacoby MG, Nannapaneni N. An insulin protocol for management of hyperglycemia in patients receiving parenteral nutrition is superior to ad hoc management. J Parenter Enter Nutr accessed August 17, 2011. avaialble at: http://pen.sagepub.com/content/early/2011/08/06/0148607111415628 Hoekstra M, Schoorl MA, Iwan CC, et al. Computer-assisted glucose regulation during rapid step-wise increases of parenteral nutrition in critically ill patients: a prood of concept study. J Parenter Enter Nutr 34:549-553, 2010