Treatment with subcutaneous insulin lispro is a safe and effective
treatment for uncomplicated diabetic ketoacidosis, new research
suggests. Because this therapy can be delivered in a non-ICU
setting, it may represent a more cost-effective alternative to
intravenous regular insulin.

Insulin lispro is an analog of human insulin with a rapid onset of
action. The drug is marketed as Humalog by Indianapolis-based Eli
Lilly & Co, which helped fund the study.

As reported in the American Journal of Medicine for September 1st,
Dr. Guillermo E. Umpierrez, from Emory University in Atlanta, and
colleagues assessed the outcomes of 40 diabetic ketoacidosis
patients who were randomized to receive subcutaneous insulin lispro
or intravenous regular insulin.

All of the subjects in the regular insulin group were treated in the
ICU, whereas half the insulin lispro subjects were treated on a
medicine ward and half were treated in a step-down unit.

The patients' biochemical parameters on admission were similar in
each group, the researchers point out. Moreover, the time needed to
correct the biochemical abnormalities was comparable in each group--
about 7 hours were required to correct the hyperglycemia and around
10.5 hours were needed to resolve the ketoacidosis.

The groups were also similar in terms of length of hospital stay,
hypoglycemia rate, and the amount of insulin used before
ketoacidosis resolution. No deaths occurred in either group.

The average hospitalization charge for diabetic ketoacidosis treated
in the ICU setting was $14,429, almost 40% higher than the charge
for treatment with lispro in non-ICU settings, at $8801 (p < 0.01).

The findings "raise interesting questions regarding the most cost-
effective and labor-saving methods for treating a common disease in
the hospital setting," Dr. Romana M. Haas and Dr. Andrew R. Hoffman,
from Stanford University in California, note in a related editorial.

"Since many cases of uncomplicated DKA are now being treated in
emergency departments or overnight treatment units, it will be
important to perform similar analyses in these settings."

Am J Med 2004;117:291-296,357-358