OBJECTIVE: To evaluate the precision, accuracy, and durability of an insulin pen injection device (NovoPen 3) at three preset doses (2 IU, 35 IU, and 70 IU) after exposure to various stress and durability tests that were intended to simulate daily use by patients. METHODS: Twenty-nine reusable NovoPen 3 insulin delivery devices were tested. The precision and accuracy of 10 insulin pen devices were evaluated after they were subjected to multiple thermal and vibration stress tests. Another 10 pen devices were subjected to a free-fall test. Nine other insulin pens were subjected to endurance testing that simulated 5 years of injections. RESULTS: The accuracy (as measured by the relative error of the delivered dose of insulin) of the insulin pen injection devices was within 1% of the preset dose after all stress or endurance tests. A free-fall test produced no indication of damage except for broken clips and snap catches on the caps, which did not affect the integrity or performance of the insulin pens. The precision of the pen devices (as measured by relative standard deviations of delivered volumes of insulin) was likewise high after thermal stress, vibration stress, free-fall testing, or 5-year endurance testing. CONCLUSION: Overall, this study showed that the insulin pen injection devices tested were durable under conditions of stress likely to be encountered in daily patient use. Neither a wide variety of repetitive stresses nor insulin injection cycles corresponding to 5 years of use affected the accuracy or precision enough to have clinical significance for reliable insulin delivery.
OBJECTIVE: To evaluate the precision, accuracy, and durability of an insulin pen injection device (NovoPen 3) at three preset doses (2 IU, 35 IU, and 70 IU) after exposure to various stress and durability tests that were intended to simulate daily use by patients. METHODS: Twenty-nine reusable NovoPen 3 insulin delivery devices were tested. The precision and accuracy of 10 insulin pen devices were evaluated after they were subjected to multiple thermal and vibration stress tests. Another 10 pen devices were subjected to a free-fall test. Nine other insulin pens were subjected to endurance testing that simulated 5 years of injections. RESULTS: The accuracy (as measured by the relative error of the delivered dose of insulin) of the insulin pen injection devices was within 1% of the preset dose after all stress or endurance tests. A free-fall test produced no indication of damage except for broken clips and snap catches on the caps, which did not affect the integrity or performance of the insulin pens. The precision of the pen devices (as measured by relative standard deviations of delivered volumes of insulin) was likewise high after thermal stress, vibration stress, free-fall testing, or 5-year endurance testing. CONCLUSION: Overall, this study showed that the insulin pen injection devices tested were durable under conditions of stress likely to be encountered in daily patient use. Neither a wide variety of repetitive stresses nor insulin injection cycles corresponding to 5 years of use affected the accuracy or precision enough to have clinical significance for reliable insulin delivery.