Robustness for gradient methods

[Lisa Parsons]

What does USP, ICH, or FDA say about performing robustness on gradient methods (if anything).

[tom jupille]

As far as I know, they make no distinction between isocratic and gradient methods with regard to robustness. In general, the gradient equivalent to %B would be gradient steepness (%B/min); the convenient parameter to use would be the gradient time. If you include instrument parameters as part of robustness, then the dwell volume (gradient delay volume) would be important as well, as it can affect instrument-to-instrument transfer.