Bioequivalence and Bioavailability Forum

Hi John,

❝ Can someone give me some hints on running a BE study involving injectable device? i.e., design, sources of variability etc?

This is a situation very similar to what you encounter when you develop generic inhalers:
Train volunteers very well to do the self-administration, or, sometimes a better and approvable option, let the staff do the administration. At any rate, the clinical staff has to be trained very well too.

If using the self option: Staff may be able to literally make an "x" on musculus vastus lateralis or whereever with a marker and the task for the volunteer is to hit the "x". Note that a lot of healthy volunteers cannot in practice self-inject even after signing icf and will chicken out. If the product is an epi-pen generic then originator training devices are available. Your own formulation should be accompanied by training dummies too.

Device training takes quite a bit of resources but isn't per se difficult - it just need to be done thoroughly.

Training should be done before each dosing and individually for each subject. A relevant goal is to make the volunteer show the staff three to five consecutive successful administrations just prior to time zero. If the staff are to administer the drug the staff can demonstrate the same to your own clinical trial manager/staff/someone in the presence of monitor.

Relying on simple pre-trial training e.g. the evening before first period will likely give you results that are all over the place.


The chicken factor can be very pronounced (can be your biggest source of variation) and even if this is a drug that has few side-effects volunteers might opt out after period 1 in case of crossovers. If a parallel trial is an option then that might be better in practice.

Thus my initial proposal is a parallel group design with training the evening prior to dosing and with individual training before drug administration in the morning.