Apparently, we are going to trial it as a preventative and not a treatment. I think katy's linked data is using HCQ as a treatment?
I am not sure how anyone measures a preventative ... unless they stick them in a room full of covid positive people while not wearing any PPE.
Otherwise it could be any number of reasons why a person does not catch the virus ... good PPE, good natural resistance, good physical distancing and hygiene. I mean, who is going to intentionally set themselves up to catch the virus?
Covid party, anyone?
Typically what they do (and they do this with vaccines as well) is divide people into two groups, age and sex matched, who have similar lifestyles and occupations. So, for example, they divide everyone who works in the CoVid ward into 2 groups.
1 gets HCQ and the other does not. Then the wait begins. If the prior infection rate was about 1 person per 30 testing positive every week (I'm making that up - I have no idea how many negative pressure rooms they have or what PPE they have, etc), then in the HCQ group it should go down (preferably no people at all) and in the untreated group, it should remain at 1 per 30. If the group of employees was only 30 to begin with, and people were getting CoVid at an undesirable rate, then the 15 on HCQ would be able to work for a longer period of time without contracting CoVid (if HCQ actually works). Relying on a small study of this kind is not great science.
The real problem is that people really do change their behavior. Perhaps the HCQ group will be more confident and a bit more relaxed about close interactions with highly infectious patients. Perhaps they will be more confident in their private lives. Usually, there are questionnaires or surveys about this kind of thing, often requiring that participants keep daily diaries of interactions.
The people without the HCQ may likewise be more cautious - but if 2 of them get it within a month or two, and in the HCQ group, zero get it, then the study goes on a bit longer. If after 3-4 months, no one in the HCQ group gets CoVid but half the other group does - well then, they have a nice study to publish.
What usually happens though, is that the participants who didn't get the drug, upon realizing that their companions seem immune, want the drug. If it's really only 15 or even 25 in a group and 2 of them get CoVid in the first month, the placebo group is going to want HCQ.
The solution to some of this is to make absolutely sure that no one knows what drug they're getting (but being medical professionals, they all know how to research what pills look like and they're very likely to figure out whether they have a bottle of HCQ or not).