You have a patient on total parenteral nutrition (TPN). His every 6-hour 'finger stick' blood glucoses have been running over 170 and were covered with sliding scale insulin. The rest of his electrolytes, albumin, and other labs are within normal limits. The intern wants to add some insulin to the TPN bag. He suggests 5-8 units per bag per day.

Do you concur or not? Why or why not?
 

Probably not. In this instance, it's NOT that the patient doesn't need some insulin. With blood sugars on TPN consistently running above 110-120 mg/dL or so, I don't think anyone would disagree he needs some insulin to maintain his blood sugars closer to the normal range. The question is how best to do this? A couple of things need to be discussed here.

Adding insulin to the TPN bag itself - this is chemically compatible and has been done for years. The debate comes in when we start to discuss how much actually makes it to the patient. Study results have been variable. The amount of insulin adsorbed (i.e. stuck to) the glass bottle, plastic bag, or plastic IV tubing can be as high as 80%. Some factors affecting this adsorption include type of container, solution, administration set, previous exposure of tubing to insulin, etc. The binding appears to happen within the first 30-60 minutes. Some in vitro studies have been conducted to assess the
effect of "priming" the line with a dilute insulin solution. Priming the tubing with a dilute solution or running through and wasting the first aliquot of the insulin-containing solution increases delivery of insulin from 38% to 85% at 2 hours. Once the priming is done, the amount of insulin delivered remains pretty constant.

Other options - although costly and controversial, adding albumin in small concentrations may help deliver higher amount of insulin by decreasing the amount of insulin available to bind to the container and/or tubing. Adding 0.3 gram/100ml seems to decrease adsorption. Flushing the tubing with the insulin-containing solution two hours before administration seems to saturate the binding sites and minimizes further adsorption. Giving the insulin as a separate IV infusion is another delivery option but adsorption should be considered in this setup as well.

IV insulin infusions are best titrated according to patient response. Because there's so many factors influencing how much insulin is actually delivered to the patient, dosing should be adjusted according to
patient response and labs. Keep in mind you'll be starting from scratch every time you start a new bag and/or esp new IV tubing set! These IV insulin patients should be closely monitored with frequent blood
glucose checks and assessments. Recent literature shows intensive insulin therapy (maintaining blood glucoses < 110 mg/dL decreased mortality, bloodstream infections, acute renal failure, and transfusions compared to the "usual" standard of keeping blood glucoses between 180 and 200 mg/dL.

REFERENCES:

-    "Insulin drip protocol" www.fammed.washington.edu/netowrk/sfm/insulindrip.htm

-    "Should insulin be added to TPNs?" www.uic.edu/pharmacy/services/di/insulin.htm

-     J Clin Endo Metab 87:978-982, 2002

-    Malmberg K. BMJ. 1997;314:1512-15.

-    Diabetes 47(sup1):A87, 1998

-    Arch Int Med 157:545- 552, 1997

This Pearl is meant for academic and educational purposes only. This Pearl is meant to raise important points regarding the safe and cost-effective pharmacotherapy of patients. It is not meant to be the definitive reference for the treatment or prophylaxis of various diseases. Although every effort is taken to ensure this Pearl is correct and factual, errors may occur. The Pharmacoeconomic Center assumes no liability for incorrect information or harm that may occur from the use of the information included in this Pearl.