Delayed publication of an early terminated late treatment RCT with low-risk (no mortality) outpatients in the USA, showing no significant differences with HCQ.Authors do not provide symptom onset data, but the subgroup analysis suggests that more patients may have been in the 5+ days group (the estimate for the 5+ days group has a smaller confidence interval, and the overall mean/median for HCQ is much closer to the 5+ days group). Treatment was started one day after enrollment according to Table S1 (authors report "commonly 1 day after randomization" in the text). This suggests that most patients were treated 6+ days after onset.Subgroup analysis for <5, ≥5 days is provided only for viral shedding duration, and shows improved results for earlier treatment. Adherence was only 66% (Figure 1).Publication was 21 months after the trial ended. Registered outcomes were modified November 2022, December 2022, and January 2023, all over a year after completion of the trial. For example, in January 2023, the household acquisition outcome at 28 days was deleted, leaving only 14 days.There are 7 versions of the SAP, all dated after the start of the trial, and 5 dated after the completion of the trial [clinicaltrials.gov]. Many outcomes in the SAP are missing, including 6 month mortality and hospitalization, QOL, and KM for hospitalization/mortality.Notably, authors provide the age subgroup analysis for symptom scores and transmission, but they do not provide the time from onset analysis.Lack of symptom onset details, lack of onset subgroup analysis for clinical outcomes, and authors' incorrect claim that none of the RCTs to date show "meaningful clinical outcomes" suggests significant bias (all 57 RCTs can be found at [c19hcq.org]).