The NO COVID-19 Study is a single center, two-arm, open label, group-sequential, pragmatic randomized controlled trial of hydroxychloroquine sulphate in adults hospitalized with COVID-19. Patients were randomly assigned to receive hydroxychloroquine sulphate (at a dose of 400 mg twice daily for seven days) in addition to standard care or standard care alone 12. Because of rapidly decreasing incidence of COVID-19 in Norway, the trial was prematurely stopped by the trial sponsor on May 25, 2020. The study protocol was approved by the Regional Committees for Medical Research Ethics (REC 121446) and the Norwegian Medicines Agency. The study was performed according to standard rules for Good Clinical Practice, with statistical methods and stopping rules described in the protocol and statistical analysis plan (see Supplementary Material).
All reverse transcriptase polymerase chain reaction (RT-qPCR) SARS-CoV-2 positive patients 18 years of age or older were eligible for study inclusion if they had moderately severe COVID-19 at admission (National Early Warning Score 2 [NEWS2]13 of 6 or less). For patients who tested positive for SARS‑CoV-2 before admission, SARS-CoV-2 status was verified with the external laboratory. Exclusion criteria included (1) the need of admission to intensive care unit on hospital admission, (2) history of psoriasis, (3) reduced hearing/tinnitus, (4) visual impairment, (5) known adverse reaction to hydroxychloroquine sulphate, (6) pregnancy, or (7) prolonged corrected QT interval (>450 ms). All study participants provided written informed consent before study inclusion.
Data on coexisting conditions were acquired from the hospital electronic patient records. Coronary artery disease was defined as history of myocardial infarction, coronary artery bypass grafting, or percutaneous coronary intervention. Diabetes was defined as history of diabetes mellitus type 1 or type 2, and the use of antidiabetic medication. Hypertension was defined as history of hypertension and the use of antihypertensive medication. Obstructive pulmonary disease was defined as history of chronic obstructive pulmonary disease or asthma. Obesity was defined as body mass index of 30 kg/m2 or above. Current smoking was defined as daily consumption of cigarettes.
The primary outcome was rate of decline in SARS-CoV-2 viral load in the oropharynx from baseline through the first 96 hours after randomization, using a single batch of swabs and a standardized sampling procedure to saturate them. Oropharyngeal swab samples were taken from patients at inclusion, at 48 hours and at 96 hours, by a selected group of study physicians. For analysis, total nucleic acids were extracted from 300 µl of each specimen using the Maxwell® RSC Viral total Nucleic Acid Purification Kit (Promega, Madison, Wisconsin, USA) according to the manufacturer’s instructions and eluted in 50 µl nuclease-free water. SARS-CoV-2 detection was performed in duplicate by RT-qPCR on 5 µl nucleic acid eluate in a total reaction volume of 25 µl on a QuantStudio™ 7 Flex Real-Time PCR System (Thermofisher Scientific, Waltham, Massachusetts, USA), according to the protocol published in January 2020 by Corman et al.14 that targets the viral E-gene of sarbecoviruses. For each patient, all samples in the time series were analyzed in the same extraction and PCR set-up. Single batches of all reagents for extraction and PCR were used for all samples in the study. SARS-CoV-2 RNA quantitation was calculated using a serial dilution of a the synthetic Wuhan coronavirus 2019 E gene RNA control comprising the viral region to be amplified, provided by the European Virus Archive Global (EVAg). Viral loads are expressed in log10 RNA copies/mL transport medium. The limits of detection (LoD) and quantitation (LoQ) of the assay are of 2.11 and 2.55 log10 RNA copies/mL, respectively. For data analyses, results below LoD (SARS-CoV-2 RNA not detected) were set to 0 log10 RNA copies/mL, and results below LoQ were set to the mean between LoD and LoQ values (i.e. 2.36 log10 RNA copies/mL). A qPCR assay targeting human β-globin was performed on all samples where no viral RNA was detected for assessment of sample adequacy.15 For the intention-to-treat population (n = 51), all actual samples were positive for human β‑globin analysis, indicating adequate sample quality. Further details regarding study sampling and analysis can be found in the Supplementary Methods.
The primary outcome was analyzed using a generalized linear mixed model, with subject-specific random intercept and slope. The complete statistical analysis plan can be found in Supplementary Methods. The statistical analyses were performed with STATA 16 (StataCorp LP, College Station, TX).