Comparative Use Human Factors Study of a Manual vs. A Partially Automated Pen Injector for <scp>GLP</scp> ‐1 Receptor Agonist Delivery

Injectable glucagon-like peptide-1 receptor agonists are developed as drug-device combination products, with delivery facilitated by pen- or autoinjectors. As generic alternatives with different device designs emerge, evidence is needed to support therapeutic equivalence for Abbreviated New Drug Application approval. Comparative use human factors studies represent one FDA-accepted approach for this demonstration. This comparative use human factors study evaluated whether a proposed generic manual pen injector could substitute for a reference partially automated pen injector for critical task performance and dose delivery. We conducted a randomized, two-period, two-sequence crossover study evaluating use success rates in 100 adult patients experienced with partially automated pen injectors. The study population included 16 current reference device and 84 current surrogate device (similar pen injector design) users. Subjects performed critical tasks with both devices in randomized sequence across two simulated scenarios. Co-primary endpoints were use success rates for dose selection and injection in Scenarios 1 and 2, analyzed using noninferiority testing with a prespecified 10% margin. Both co-primary endpoints demonstrated noninferiority. For Scenario 1 (0.6 mg dose selection), both devices achieved 95% success rates with 0% difference (95% one-sided upper confidence limit: 5.7%, P = 0.0045). For Scenario 2 (3 mg dose selection and simulated injection), the reference device achieved 100% vs. 95% for the test device, with a 5% difference (95% one-sided upper confidence limit: 9.92%, P = 0.0475). All confidence bounds remained below the 10% noninferiority margin. This study provides evidence supporting substitutability of the proposed manual generic device for the partially automated reference device.