I know this committee wants to look at the most recent research. My research is a bit historical now. The most recent research, which I think Dr. Libman will agree with, is the Cochrane review. I did the first Cochrane review of mefloquine, published in the British Medical Journal in 1997. It has now been updated four times. The most recent review, which looks at all the randomized controlled trials and tries to extract that type of data, came out in 2017.
To answer the question, that review looked at 20 different randomized controlled trials of mefloquine. It found that comparing it with, say, atovaquone-proguanil, three times more people taking mefloquine were likely to stop taking the drug because of side effects.
That really makes it not as effective. I know Dr. Libman is looking very doubtful, but the relative risk is 2.86, which I interpret as meaning you are three times more likely, if you're taking mefloquine, to stop your drug. If you stop your drug, you risk getting malaria.
Within that analysis, they are comparing mefloquine and atovaquone-proguanil. They find that 6% of mefloquine users discontinue the drug, 13% get insomnia, 14% get abnormal dreams, 6% get anxiety and 6% get depressed mood. That gives you a flavour of the types of figures one can expect, bearing in mind that these studies, these randomized controlled trials, tend to be done in perfectly healthy, unstressed populations. For soldiers, those figures are likely to be comparable or perhaps worse.
When you look at the comparison with doxycycline, the figures are even worse. Of mefloquine users, 31% get abnormal dreams, whereas only 3% of doxycycline users get abnormal dreams.