A Survey of Characteristics and Potential Contribution of Registered Studies for 2019 novel coronavirus disease (COVID-19)


 Background

The World Health Organization characterized the 2019 novel coronavirus disease (COVID-19) as a pandemic on March 11. Many clinical trials on COVID-19 have been registered, and we aim to review the characteristics of the trials and provide guidance for future trials to avoid duplicated effort.
Methods

All the studies on COVID-19 registered before Mar 3, 2020 on eight registry platforms worldwide were searched and the data of design, participants, interventions, and outcomes were extracted and analyzed. The most promising trials were screened based on study design, rationale, and resource availability.
Results

393 studies registered were identified until Mar 3 2020 and 380 (96.7%) studies were from mainland China, while 3 in Japan, 3 in France, 2 in the US, and 3 were international collaborative studies. 363 studies (92.4%) recruited participants from hospitals and 266 studies (67.7%) aimed at therapeutic effect, others were for prevention, diagnosis, prognosis, etc. 202 studies (51.4%) were randomized controlled trials (RCTs). The average sample size was 1061 and ranged from 8 to 150,000 per study. 177 out of 266 therapeutic studies (66.5% ) tested Western medicines including antiviral drugs (17.7%), stem cell and cord blood therapy (10.2%), chloroquine and derivatives (8.3%), 16 (6.0%) on Chinese medicines, and 73 (27.4%) on integrated therapy of Western and Chinese medicines. 14 Chinese medicines had its clear rationale for evaluation of therapeutic effects. 31 studies among 266 therapeutic studies (11.7%) used mortality as primary outcome, while the most designed secondary outcomes were symptoms and signs (47.0%). 106 studies (27.0%) were funded by the government, and 268 (68.2%) demonstrated ethical approval. 45.5% studies (179 out of 266) had not started recruiting till Mar 3. Eight RCTs were evaluated as the most promising trials.
Conclusions

Majority of the studies focused on assessing therapeutics for COVID-19 but inappropriate outcome setting, delayed recruitment and insufficient numbers of new cases in China implied many studies may fail to complete. Strategies and protocols of the studies with robust and rapid data sharing from international collaboration are warranted for emergency public health events, helping to accelerate priority setting for timely evidence-based decision-making.

priority setting for timely evidence-based decision-making.

Background
Coronaviruses are a large family of viruses that cause illness ranging from the common cold to more severe diseases such as Middle East respiratory syndrome (MERS) and Severe Acute Respiratory Syndrome (SARS) [1]. 2019 novel coronavirus disease (COVID-19) occurred in December 2019 and the first case was reported in Wuhan, China [2]. On January 31, 2020, the World Health Organization (WHO) announced that the new coronavirus epidemic constituted a public health emergency of international concern and characterized COVID-19 as a pandemic on March 11 [3]. As of April 22, 2,565,879 people were confirmed infected with COVID-19 worldwide in 210 countries, areas or territories with cases [4]. While carrying out public health control and clinical management, Chinese government also encouraged speeding up clinical trials of new drugs, and it was necessary to promptly launch them into the frontline of treatment, improve cure and reduce death [5]. As there is no specific treatment for COVID-19, and the main management is for symptomatic treatment and supportive care, clinical evidence is urgently needed to support clinical decision-making. Researchers in China reacted quickly, and the first clinical trial was registered in the China clinical trials registry on Jan 23 2020. Following a short period, more than 200 clinical trials have been registered, involving a variety of therapeutic approaches [6]. Facing the increasing ongoing trials, it would be important to review the research questions and characteristics of these studies to inform clinical practice for the prevention and treatment of COVID-19. These questions need to be answered for our increased understanding of the most promising trials. Therefore, our aim is to investigate the current status and characteristics of the registered studies, providing guidance for future trials and avoiding duplicated effort worldwide.

Search strategy
All the clinical studies on COVID-19 registered before Mar 3 2020 on the trial registry platforms were

Data extraction
Nine authors (MX, YJ, YZ, YYZ, YXS, ZYT, XYJ, QBJ, MY) abstracted data extraction in parallel, including registration number and date, title, e-mail, leading institutions, country and province, setting, ethic information, funding, design, study objectives, anticipated start date, interventions and control, population, sample size, recruiting status and outcomes. We also checked the numbers of confirmed cases in China from the official website of the National Health Commission of the People's Republic of China [7]. All abstracted data were entered into a pre-defined data extraction sheet.

Screen for promising trials
We designed the criteria for the most promising trials as follow: Those with a rigorous design (randomized controlled trials, RCTs) for therapeutic effects; Those testing a treatment with a good rationale for safety and effectiveness, with existing evidence of human studies on COVID-19, pneumonia, SARS, Ebola, MERS, influenza, human immunodeficiency virus (HIV) and other viral diseases; Treatments which are widely available (e.g. Chinese herbal medicines); Those measuring important outcomes (mortality or exacerbation as primary outcomes); Those with a reasonable sample size (estimated based on hypothesis, and a reasonable chance of completing recruitment).
The trials were screened with the data extraction sheet, and the supporting evidence was searched on PubMed, characteristics of the eligible RCTs were extracted.

Basic information of registered studies
After searching on the registries, 406 records were retrieved and after removing duplicates, 393 were eligible and included in the analysis. 321 studies were from ChiCTR, 69 from the Clinicaltrials.gov, and 3 from Japanese Registry. No records were found from other registries.
Among those 380 studies in China, 36 studies were supposed to be conducted in more than 2 provinces, and 328 studies each in one province (26 provinces in total) except 16 studies with no information about the place. One study was an online survey investigating quality of life of Chinese residents during or after the outbreak of new coronavirus pneumonia .

Western medicine
The therapeutic clinical trials contributed the largest proportion in the registrations with the number of 266 studies (Table 2). Among the tested Western medicine, 47 studies tested on antivirals including arbidol, lopinavir-ritonavir, darunavir, interferon, ribavirin, and danoprevir. Apart from these antivirals in China, five RCTs registered for remdesivir, including two phase Ⅲ randomized, doubleblind, placebo-controlled multicenter study for mild/moderate and severe patients, two phase Ⅲ RCTs comparing different duration with standard therapy, and one phaseⅡmulticenter, adaptive, randomized blinded controlled trial. Other trials tested antivirals not marketed in China including fapilavir, xofluza, azvudine, triazavirin, and ASC09F.
Other tested Western treatments included stem cell and cord blood (27 studies

Outcomes of the registered studies
In Table 4, only 17 RCTs, 2 controlled clinical trials, and 12 observational studies (31 studies totally, 11.7%) used mortality as a primary outcome. Other clinical important outcomes such as exacerbation rate/time (26 studies, 9.8%) and length of intensive care unit (ICU) stay (2 studies, 0.8%) were also seldom used as primary outcomes. Symptoms and signs were listed in 105 studies (39.5%) as primary outcomes, while 125 studies (47.0%) as secondary outcomes. Common laboratory parameters were used in 40 studies (15.0%) as primary outcomes, while 109 studies (41.0%) as secondary outcomes.
Trials identified as the most promising clinical trials Totally 8 RCTs were judged as the most promising trials for the treatment of COVID-19, which were RCTs for therapeutic effects with clear and available intervention, primary outcomes with mortality or exacerbation, and potential human evidence for the treatment of COVID-19 ( Table 5). Four of them had direct evidence on the treatment of COVID-19 patients and others had evidence on H7N9, severe community-acquired pneumonia, and sanitation workers.

Discussion
This study systematically reviewed available registered studies for COVID-19 with the analyses of their distributions and characteristics. 393 studies were registered in eight registries, aiming at the prevention (16 studies), treatment (266), diagnosis (25) and prognosis (19) of COVID-19. Majority of the studies were randomized trials, followed by observational studies testing different interventions such as antiviral drugs, Chinese medicine, and integrated therapies. Except for 50 studies, clinical important outcomes such as mortality and exacerbation rate/time were not set as primary outcomes in majority trials. 179 studies had not started recruiting and would hardly be able to carry on in China due to insufficient patients.
As a new communicable disease, direct evidence for the prevention of COVID-19 is not available. We found insufficient evidence to support the rationale for tested Western medicines, while based on historical records and human evidence of SARS and H1N1 influenza prevention, Chinese herbal formula is considered as an alternative approach for prevention of COVID-19 in high-risk population [29]. The therapeutic clinical studies made up the largest proportion of the registrations. Antivirals, the most promising category of Western medicine, accounted for 17.7% out of the therapeutic studies. Eight RCTs were identified as promising trials under the criteria of study design and related human evidence. In terms of Chinese medicines, 14 had clinical or laboratory evidence, showing the potential therapeutic effects on COVID-19 patients.
Flaws in study design, such as the setting of the control and outcomes, and the lack of coordination were discovered from the registrations. Even in an outbreak, investigational products should be evaluated in scientifically and ethically sound studies [30]. Do no harm is always the first rule for all human studies. Methodologically, double-blind randomized, placebo controlled trials are considered to be the gold standard for therapeutic clinical trials [31]. However, considering the emergency and the practical issues of ethics and informed consent, the implementation of RCT faces more challenges. In the context of COVID-19 pandemic, the control interventions should be supportive care.
As statistics shows, the mortality of COVID-19 was 4.3% in Wuhan, China, indicating severe lifethreatening disease [32]. New studies on clinical characteristics of COVID-19 also reported outcomes on exacerbation, such as the median time from first symptom to dyspnea, acute respiratory distress syndrome (ARDS), transfer to the intensive care unit (ICU) due to complications and death of multiple organ failure [32,33], and other symptoms and laboratory findings for example neutrophilia, organ and coagulation dysfunction, which were potential risk factors for ARDS [34] and elevated d-dimer as risk factors for mortality [35]. On the contrary, clinical important outcomes such as mortality and exacerbation were only used as primary outcomes in 21.5% analyzed registrations, and the observational measures in clinical practice such as symptoms, signs, common laboratory tests and viral nucleic acid/viral load were used more frequently. What's more, the most used primary and secondary outcomes were similar and clear measurements and time points were seldom available.
The design of more than three primary outcomes in one trial may bring problems in the interpretation of research results [6].
Although the number of registered trials is increasing, only carefully conducted trials can show which measures work [36]. Without a coordination of the research teams in the whole country, potential participants could be scattered in numerous small studies, resulting in less powerful results or incomplete trials. There were only 3 international and 36 domestic collaboration studies, suggesting a low level of cooperation. In fact, the registrations showed that nearly half of the studies had not started recruiting by Mar 3, while the new cases in China were sharply reducing. Besides, a few of the sponsors had withdrawn their studies due to lack of patients. In fact, we could learn from the experience on study design of Ebola virus disease according to WHO documents [31, 37]. For example, the adaptive trial designs were used in the Ebola epidemic, it has the capacity to yield meaningful and interpretable data quickly, while more complex to coordinate among different sites.
The key points of study design in these documents may also helpful for the design and implementation of COVID-19 clinical trials.
There are several limitations in our research. First, the registrations provide limited information on the trials, and our analysis is based on the registered information but not the full protocols. Second, the required information of the registrations are not unique across different registries, and the information could be revised by sponsors after the search and analysis, so the results may not include the whole registered information of the COVID-19 studies. Third, the criteria of the most promising trials were based on the authors' consensus but not a widely used standard, we still look forward to more encouraging results of the registered trials not limited to the identified ones.
More international collaborations, rapid data sharing, and strengthened coordination are needed in the searching for effective therapy. As WHO suggested, enhancing global coordination of all relevant stakeholders, a clear and transparent global research and innovation priority setting and common platforms for standardized process are needed in the research during the outbreak [38]. In addition, WHO and partners are launching SOLIDARITY trial and aims to generate robust data for the most effective treatment for COVID-19 [3]. In addition, rapid data sharing is warranted once they are adequately quality controlled for release [39]. To response the outbreak of COVID-19, a quick upload of data is recommended when registered trials initiated so for immediate analysis and inform upcoming trials. Thirdly, the coordination of the trials are urgently needed. More rigorous regulations by the National Health Commission in China have been delivered for the clinical studies on COVID-19 recently, aim to strengthen overall coordination, promote data integration, and improve research efficiency [40]. With the statistics of registered information, we will trace the trials for the update status regularly. Further research could be conducted to investigate the impact factors of a successful trial in the emergency of public events, and summarize valuable experience for the protocols of

Ethics approval and consent to participate
Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
All data generated and analysed during this study are included in this article.  Other comparisons: different dosage or duration of the tested intervention, "historical comparison" (without details), γ-Globulin, bag-valve mask oxygenation Assisted tracheal intubation, psychological intervention (without details), Chinese medicine.  Table 3 is only available as a download in the supplemental files section