If a product feature or characteristic related to the difference is associated with a critical task for a combination product (for example, a difference in the instructions about inhaling a dose of medication), then this likely constitutes an “other design difference.” Importantly, even if a design change seems to be an improvement as compared to the reference product, the difference will not automatically be deemed minor or otherwise acceptable by the FDA. As indicated before, the objective is for a patient to switch from using the RLD inhaler to the generic inhaler without additional training or explanation.
Note that the FDA expects manufacturers to document a rationale for having characterized each difference as none, minor or other. The FDA will carefully review and potentially disagree with the provided categorizations. They will likely judge minor differences as acceptable if the threshold analyses are comprehensive and additional HF-related data is likely unnecessary. However, if the threshold analyses identify one or more “other design differences,” you should consider modifying the proposed inhaler’s components (including labeling) to minimize differences from the RLD inhaler. If no further modifications are possible or implemented, the FDA might request additional HF data via a CUHF study.
Comparative use human factors study
A CUHF study is a simulated-use study in which representative end users (i.e., patients with experience using the RLD inhaler) simulate using both the generic inhaler and RLD inhaler under close observation by human factors experts. This type of study is required for generic inhalers when the differences identified during the threshold analyses have been deemed unacceptable. The test personnel watch for use errors that arise during combination product critical tasks related to external critical design attributes and then calculate and compare the use error rates between the generic and RLD inhalers.
Overall, the study objective is to demonstrate that the use error rate for the generic inhaler is no worse than the corresponding use error rate for the RLD inhaler. Suppose that priming the MDI is a product combination critical task, and the threshold analyses identified a difference in the steps that users must perform when priming the generic compared to the RLD. If RLD-experienced participants make a similar number of mistakes (or fewer) in priming the generic MDI compared to the RLD, and the types of mistakes are similar, the use error rate would be considered acceptable.
To calculate the use error rates and non-inferiority measures outlined in the FDA’s guidance, one must engage a statistically significant sample size, which will likely call for recruiting a minimum of 50 and as many as 150 or more participants.