Study design and patients

The FLOSAN trial recruited patients with mild-to-moderate COVID-19 from 12 hospitals in China, 3 hospitals in Thailand, and one hospital each in Vietnam and the Philippines between February and December 2022 (Additional file 1: Table S1). The full version of study protocol has been published recently (See details in Online Supplement) [16]. Briefly, eligible patients were aged 18-70 years, had mild-to-moderate COVID-19 (according to World Health Organization criterion) [17], tested positive to either rapid antigen test (RAT) or nucleic acid amplification test (NAAT), had an interval between symptom onset and screening of within 4 days, and had at least three major symptoms (stuffy or runny nose, sore throat, cough, shortness of breath, low energy or tiredness, myalgia, headache, chills or shivering, feeling hot or feverish) occurring within 12 hours prior to screening. We excluded patients who had: (1) known co-morbidities of other infections; (2) poorly controlled systemic diseases; (3) alcohol or drug abuse within one year; (4) participated in other trials within one month; (5) become pregnant, breastfeeding or within two weeks of delivery. The FLOSAN trial was conducted in accordance with the Declarations of Helsinki. Ethics approval has been obtained the ethics committee of each participating site, based on Good Clinical Practice. All patients signed written informed consent.

Table 1 Baseline demographic and clinical characteristics of the full-analysis set

Randomization and masking

We randomly assigned patients (1:1) to receive treatment with LHQW or matching placebo (manufactured by Shijiazhuang Yiling Pharmaceutical Co. Ltd., Shijiazhuang, China) based on the randomization numbers generated with the SAS package (SAS Inc., Cary, USA). The block size was 4 with no stratification. With competitive recruitment scheme, the sub-site investigators allocated patients in an ascending order. The study medications had an identical color, odor and appearance, except that the placebo did not contain any active ingredient of LHQW. Patients, the study investigators and other staff were masked to treatment allocation until database lock.


After randomization, patients took LHQW (4 capsules [0.35g/capsule], thrice daily) or matching placebo for 14 consecutive days following hospitalization in designated hospitals (in mainland China) and out-patient recruitment (in the Phillipines, Thailand and Viet Nam). Both groups received standard-of-care consisting of antipyretics, analgesic drugs, nutrition supplementation and fluid replacement. Acetaminophen, the non-steroidal anti-inflammatory drug, could be applied for ameliorating fever if the temperature reached 38.5 degrees or higher. Antivirals or medications with core components of LHQW were prohibited. Sites could follow local guidelines and protocols in their countries and regions. Patients attended four in-hospital (in mainland China) or out-patient (in Thailand, Vietnam and the Philippines) visits (days 3, 7, 10 and an end-of-study visit, typically scheduled at day 14). Patients who prematurely discontinued treatment due to accelerated symptom recovery or other reasons could attend all planned visits. During the study, patients were requested to fill out the diary card twice daily to evaluate the changes in symptoms.


The primary endpoint was evaluated at day 14 - the median time to sustained clinical improvement or resolution of the nine above-mentioned major symptoms, rated as being less than or equal to mild (scored 1 or 0) and remained stable for >24 hours (see Supplementary File for the symptom diary cards).

Pre-specified secondary endpoints included the proportion of patients with sustained improvement or resolution of nine major symptoms at day 14, the median time to sustained improvement or resolution of each of these individual symptoms, the median time to onset of antipyretic effect and return to normal temperature (axillary temperature ≤ 37.0°C or oral temperature ≤ 37.3°C for >24 hours), the median time to sustained improvement or resolution of gastrointestinal symptoms, anosmia and ageusia, the proportion of patients with sustained improvement or resolution of all symptoms, the time to negative conversion of NAAT findings, and the rate of NAAT negative conversion (days 0, 7, 10, 14), the proportion of patients with major improvement in chest imaging, the incidence of COVID-19-related severe/critical disease, COVID-19-related and all-cause mortality within day 14. A designated experienced radiologist (blinded to study allocation) reviewed chest X-ray or computed tomography (CT) images and rated the outcomes. An improvement in chest radiology denoted a decreased area of infiltration, a decreased area of any radiologic abnormality, or decreased density of ground-glass opacity or nodules [15].

Safety endpoints were evaluated from the first dosing to the end of follow-up, including vital signs, physical examination, major changes in laboratory test, abnormal twelve-lead electrocardiogram findings, and the adverse event (AE) and serious adverse event (SAE). See Online Supplement for details.

Statistical analysis

Assuming that the median time to sustained improvement or resolution was 12 days in control group and 9 days in LHQW group, 652 patients would be randomized to LHQW or placebo group (1:1) with a 95% power with a two-sided significance of 0.05 according to PASS software. In practice, patients were enrolled while taking into account RAT findings, and 344 patients per group would be needed when assuming that 95% of patients with positive RAT findings would yield positive NAAT findings. Recruitment of 860 patients would be needed while considering a 20% dropout rate.

We conducted statistical analyses with SAS 9.4 software (SAS Institute, Cary, North Carolina). All patients who had been randomized and taken at least one dose of study medication and had a confirmed diagnosis of COVID-19 based on NAAT were included in the full-analysis set. Patients who fully complied with the protocol (adherence: 80% or greater) were included in per-protocol set. We prioritized data presentation of the full-analysis set. The primary endpoint was analyzed by using the Log-rank test and displayed with Kaplan-Meier curve. The time to events was presented as the median duration and 95% confidence interval (95%CI). The hazards ratio (HR) of clinical events was demonstrated. We analyzed the following endpoints with chi-square test or Fisher’s exact probability model, including the proportion of patients with alleviation of symptoms, reduction in viral shedding (censored at day 14), major improvement in radiology, severe and critical diseases, death and all-cause death. We also analyzed the median time to sustained alleviation of single symptom, the alleviation of fever, digestive symptoms, ageusia or anosmia and all clinical symptoms, and the duration of viral shedding with the same analytical strategy with the primary endpoint. We conducted post-hoc subgroup analysis of the primary endpoint according to the strata of nationality, sex, age, vaccination status, concomitant antiviral drugs or other Traditional Chinese Medicine compounds, and the duration of symptom onset.

The FLOSAN trial was registered with Chinese Clinical Trial Registry (No.: ChiCTR2200056727). The CONSORT checklist can be found in the supplemental file.

Role of funding source

The sponsor participated in the study design along with the principal investigators, study medication provision and data collection. An independent third party participated in data analysis. The first and corresponding authors had full access to the data and the corresponding author had the final decision to submit the manuscript for publication.

Source link