Accepted answer
On the detail: persistent vomiting is a different problem from nausea and needs a different response.
Extending the interval before the next escalation is the intervention with the best evidence. Trials titrated at four-week intervals for exactly this reason.
Trial incidence for nausea in this class runs to roughly a quarter to a half of participants depending on agent and dose, concentrated in the escalation phase, with discontinuation for it in low single-figure percentages.
Tachyphylaxis of the gastric-emptying effect with continued exposure is documented for the long-acting agents and is the mechanistic basis for tolerance.
Research-use material of unverified content makes any dose-response reasoning unfounded from the start.
Smaller meals, less fat, stop at first fullness, fluids between meals.
edited 30 Oct 2024 by elke_brunner — added the method parameters