The fact that despite "the presence of QTc prolongation in 11-19%, this finding was not associated with death or torsade de pointes, as that may be induced by QTc prolongation" might imply CQ was not the real culprit. If one looks careful at the demographic comparison, one would be very surprised to find: 1) the high dose CQ group had 12.2% > 75 years in age (vs low dose had 0%). 2) then the same high dose group had more hypertension, diabetes, renal disease than the low dose; and, 3) alas, 17.9% high dose had heart disease vs 0% in the low dose group. Three of 5 patients (60%) in the high-dose group with underlying heart disease died. One would view this is not "randomization" but rather "pre-selecting" the high risk for the high dose. This study will teach us when not to use CQ: as if we are looking for disaster, we might what we wish for.