Clinical Use and Evaluation of Insulin Pens

Abstract

Insulin pens are more accurate and easier to teach than other methods of insulin delivery. They also do not suffer from the risk of mismatch of insulin concentration and type of insulin syringe. The ISO standard used to test insulin pens, however, needs to be updated to reflect their clinical use.

In this volume, Abdel-Tawab and associates ¹ describe the comparative accuracy of 2 U100 pens with rapid acting insulin. They appeared to have similar accuracy and precision and both met the new ISO 11608-1:2012 requirements. The literature is filled with similar studies demonstrating the accuracy of insulin pens.^2,3 The article, however, raises 3 interesting questions:

• Is the ISO standards for pens clinically helpful in judging accuracy?

• Are insulin pens better than syringes in delivering insulin?

• What is the role of insulin pens in the era of 4-5 different insulin concentrations?

The ISO standard 11608-1:2012 ⁴ provides for the measurement of accuracy and precision at 3 doses, 1, 40 and 80 Units. Perhaps, we should add a dose that is commonly used in insulin therapy. The acceptable dose range for the 1 U dose is ± 1 U (0-2 U), for 40 U is ±5% (38-42 U) and for 80 U is also ±5% (76-84 U). The allowable error of ±100% at 1 U is greater than that allowed by the ISO standard for insulin syringes. The acceptable delivery of insulin by a 0.3 cc insulin syringe at 1 U is 0.53 to 1.47 U. The error of most insulin delivery devices is a combination of a fixed error and an error that is proportional to the delivered dose. In allowing a 1 U error at ±1 U, we are defining the acceptable fixed error at about ±1 U. Similarly, with an allowable error of ±5% at 40 and 80 U we are defining the acceptable proportional error at about 5%. In the current ISO standard, there is no testing done at 5-15 U, which are common doses for patients using short acting insulin. If the errors are linear, with a fixed dose error of 1 U and a proportional error of 5%, the allowable error at 5 U is ±25%, at 10 U is ±15% and at 15 U is almost ±12%. Pens are more accurate than that, but should demonstrate their accuracy at more meaningful doses. The dose of 40 U of rapid acting insulin is used by some patients with type 2 diabetes, but is rare in type 1 (although long-acting doses may be in this range). The 80 U dose is a very large dose, even for a patient with type 2 diabetes. Even though the standard is only 3 years old, it may be time for an update.

In Europe, the vast majority of patients with diabetes use insulin pens; less so in the United States, where insulin syringes have a larger market share. The pens described in this article are highly accurate. At a dialed dose of 1 U the mean delivered dose was of 1.04 U with a CV of 5-10%; at a dialed doses of 40 and 80 U the mean delivered doses were 39.7 and 79.5, respectively, with CVs of about 1%. Many studies show that below 10 U pens are far more accurate than syringes and at 10 U and above they have similar accuracy.^5,6 Since the pens are at least as accurate as syringes and are easier to teach and use, ⁷ they are frequently the choice for new insulin users. The argument that they do not allow custom mixing seems specious, as there is little evidence that custom mixes provide better blood glucose control than insulin pens.

The final issue deals with multiple insulin concentrations. Insulin does or will come in concentrations of U100, U200, U300 and U 500 and U400 is in clinical trials. I believe the use of syringes with all of these concentrations is dangerous. In the 1990s, a 10-year-old girl went to her local pharmacy to buy syringes. She had used U100 insulin her entire life and was unaware that there were any other concentrations. A new, young pharmacy clerk sold her U40 syringes that were kept for dog insulin, which is U40. They obviously look different; U100 syringes are orange and black, U40 red and black. She, however, did not know the difference and took her usual dose of 15 units of regular insulin with the U40 syringe. In actuality, she did not get 15 units, but rather 37 units and was admitted to the hospital in a coma. I am surprised that a similar accident has hot already happened with U500, which are routinely administered with U100 syringes. Insulin pens are safer. With an insulin pen the patient simply dials the dose and injects. Volume concerns are taken care of by the mechanics of the pen.

For all of these reasons, we need to move more and more of our patients onto insulin pens.