Abstract
EDITOR,—P J Barrett and colleagues report that mefloquine is associated with an excess of neuropsychiatric adverse events compared with chloroquine plus proguanil (27.4% v 16.0%).1 While there are similarities in methodology between the authors' study and Steffen et al's larger study,2 there are crucial differences in assessment criteria. I propose that any differences in incidence reports are a function of the defining criteria and that this may account for the disparities in perceptions of the tolerability of mefloquine. Nevertheless, I wish to make …