Problems with an insulin pump user

Hospitalized patients with an insulin pump can present a challenge.

Issue: October 2008

ByRonald Tamler, MD, PhD, MBA

Our diabetes team was called to consult on a 50-year-old man with type 1 diabetes and chronic kidney disease. The patient was admitted after unsuccessful insertion of a catheter for peritoneal dialysis and was now awaiting arterio-venous fistula placement for hemodialysis.

In the meantime, he was “non per os” and was receiving dextrose 5% at 100 mL per hour. The patient had been diagnosed with diabetes 30 years ago and had been using an insulin pump with lispro to manage his diabetes for more than five years. He had discontinued pump infusion after feeling hypoglycemic the night before, and his blood glucose levels had been drifting up to 211 mg/dL on an aspart “sliding scale” every six hours.

The patient stated that he was not counting carbohydrates at home, but that his blood glucose levels were under reasonably good control. He did complain of occasional hypoglycemia in the evening and at night. He also felt very confident regarding his blood glucose management and had not seen an endocrinologist in about a year.

Other medical history was relevant for hypothyroidism, failed renal transplant, hypertension, diabetic retinopathy. Medications included levothyroxine, acetylsalicylate, minoxidil, losartan, cyclosporine, and mycophenolate mofetil. No known drug allergies; social history and family history non-contributory. Physical exam was remarkable for normotensive, tachycardic Hispanic male of normal weight (approximately 170 lb with normal build) with distended, non-tender abdomen, otherwise in no acute distress. Temperature mildly increased to 37.6 C.

Lab review was relevant for a HbA1c of 6.3%, a creatinine of 3.3 and a potassium level of 4.5. A review of the patient’s pump settings revealed that total insulin use on a typical day (when the patient was eating with a good appetite) was 20.1 U.

Basal rate accounted for the following:

Basal rate chart

What can you deduct about the patient’s blood glucose levels outside the hospital, and what is the best course of action for the patient in this situation?

The patient’s outpatient glycemic control is excellent, since his HbA1c is at goal. He should restart his pump regimen as before.

The patient is hyperglycemic because of the dextrose infusion. The infusion should be stopped, and the patient should be maintained on sliding scale insulin.

The patient is receiving a disproportionate amount of basal insulin with his outpatient pump regimen. The insulin pump should be restarted at approximately 0.4 U per hour (10 U per day) and titrated as needed.

The patient is very knowledgeable about his pump and diabetes management in general and should bolus himself with 1-2 U as needed, depending on his blood glucose levels.

Since the patient had been hypoglycemic, he should be maintained off insulin.

CASE DISCUSSION

Ronald Tamler

This patient combines several challenges: First of all, he is a patient with type 1 diabetes who is not eating. While some residents will make the mistake of stopping all insulin due to fear of hypoglycemia (E), physicians with experience in diabetes management know that this omission will ultimately lead to hyperglycemia and diabetic ketoacidosis.

The same can be said for maintaining a patient on insulin sliding scale, which is a merely reactive form of glycemia management and increases the risk of hypo- and hyperglycemia (B). Moreover, the dextrose infusion can provide a safety net against hypoglycemia and, to some degree, catabolism, and should not be stopped.

Another challenge that this case presents is that some patients with long-standing diabetes are not open to learning new techniques, such as carbohydrate counting. In this case, the patient, while indeed experienced in using an insulin pump, was managing his blood glucose levels with an occasional bolus unrelated to food intake, with subsequently fluctuating blood glucose levels. In a sense, he was using his own insulin sliding scale (D).

Another interesting fact comes to light when analyzing the ratio of basal versus bolus insulin and the timing of the insulin infusion: First of all, it is evident that, in contrast to human physiology, the patient’s basal rate increases in the afternoon and evening, a time when he would eat the most. Second, the basal : bolus ratio is 4:1. This is clear evidence that the patient had been titrating his basal rate to cover meals at home and was now experiencing hypoglycemia, since he was not eating. The HbA1c, while at goal, does not account for the likely blood glucose fluctuations this patient must have been having (A).

The best option is therefore answer C, which employs the insulin pump in a patient who is not eating the same way one would use an insulin drip in the hospital. An alternative way of managing this patient would be to give the estimated basal daily insulin requirement subcutaneously with a long-acting insulin. Eventually, once discharged from the hospital, the patient may be convinced to take interest in how to adequately calculate the insulin bolus for meals and correction.

Ronald Tamler, MD, PhD, is an Assistant Professor in the Division of Endocrinology at Mount Sinai School of Medicine.