I just read this critical review by Derek Lowe and I added the most important references for your info.
Essentially, the “encouraging” results of Chloro and Aza by Gautret et al (1st pdf see Fig 1 and 2 p 20-21) are heavily criticized by a statistical re-analysis (2nd pdf).
Moreover, a similar study in Paris shows essentially no effect (4rd pdf) and a study in NYC warns for frequent Q-T prolongations, a well-known side effect of CQC (5th pdf).
Randomized control trials are urgently needed is the least you can say. But, to be honest: we have seen such inconsistencies before in various setting with drugs that were not really well validated before clinical use. It would be one of the first times in the medical history that such a drug “survives” a well-designed placebo-controlled double blind trial. But you never know….