Number of COVID-19 clinical trials registered on ChiCTR and ClinicalTrials.gov
In total, 744 COVID-19 clinical trials were registered as of March 28 (534 in ChiCTR and 210 in ClinicalTrials.gov). We excluded 67 clinical trials for the following reasons: 10 trials with large amounts of missing data or misinformation, 57 other trials for diagnosis, epidemiological, prevention, basic science, health service, prognosis, and treatment studies. After a scrupulous check on the title and the text, 677 clinical trials were eventually included in our review (Figure 1).
Figure 1 Flowchart of screening process and search results
Trends in the number of registrations over time
As shown in Figure 2, the pandemic was still in its early stages in January with the lowest number (7 of 677, [1.03%]) of registrations. The number of related registrations grew rapidly in February due to the outbreak of COVID-19 in China. The trend curve fluctuates slightly in March. The stagnant growth in registrations from March 20 to 24 may be due to a significant decline in the number of confirmed cases in China at the time. Considering the global spread of COVID-19 and the prevalence of unlisted drugs and vaccines, registrations will continue to increase over time.
Figure 2. The monthly trend of COVID-19 clinical trial registrations (N=677)
An analysis on the trend of overall registrations (Figure 3) demonstrates that the number of interventional studies was significantly higher than the number of observational studies in the early stages of COVID-19. In March, enrollment in observational studies increased significantly. Although the number of interventional trials decreased slightly in mid-March compared to that in February, the trend went upward since March 24. These results were generally correlated to the stage of development of the epidemic, but it was also important to consider the potential influence of other factors (e.g. time differences, unstable website systems, etc.).
Figure 3. Statistics on single-day clinical trial registrations of COVID-19 (N=677)
The sponsor and period of trials
As shown in Table 1. Of the 677 studies, 522 (77.10%) trials were sponsored by hospitals, 74 (10.93%) by colleges and universities, 30 (4.43%) by companies and 27 (3.99%) by institutes. There were also 15 (2.22%) for individual initiators and 9 (1.33%) for others. COVID-19 is an emerging infectious disease that is highly contagious and destructive. To find effective methods of prevention and treatment as soon as possible, trials period was concentrated in an one-year period. 144 (21.27%) Trials period range from 0 to 2 months, 217 (32.05%) from 3 to 6 months and 173 (25.22%) from 7 to 12 months.97(14.33%) Trials period range from 13 to 24 months and 26(3.84%)over 24 months. In addition, 20 (2.95%) trials did not provide a clear period.
Table 1 Information of the sponsor and the period of trials (N=677)
Category
|
Information
|
Number of trials
|
Percentage of total
|
Sponsor
|
Hospital
|
522
|
77.10%
|
|
Colleges and universities
|
74
|
10.93%
|
|
Company
|
30
|
4.43%
|
|
Institute
|
27
|
3.99%
|
|
Personal
|
15
|
2.22%
|
|
Other
|
9
|
1.33%
|
period
|
0~2 months
|
144
|
21.27%
|
|
3~6 months
|
217
|
32.05%
|
|
7~12 months
|
173
|
25.55%
|
|
13~24 months
|
97
|
14.33%
|
|
>24 months
|
26
|
3.84%
|
|
Not provided
|
20
|
2.95%
|
Status of recruitment and approval by the Ethics Committee
In the 677 studies, 332 (49.04%) clinical trials recruited volunteers while 292 (43.13%) did not; 397 (58.64%) clinical trials were approved by ethics committees while 83(12.26%) were not (Table 2). Notably, 24 of 332 (7.23%) trials in recruitment and 3 of 25 (12.00%) completed trials have not been approved by the Ethics Committee. Approval by the Ethics Committee is a prerequisite for clinical trial registration,5 but the urgency to suppress further exacerbation of the COVID-19 outbreak has resulted in some unregistered trials.
Table 2 Status of recruitment and approval by the Ethics Committee (N=677)
Recruiting status
|
Approved by the Ethics Committee
|
Total
|
Yes
|
No
|
Not provided
|
Active, not recruiting
|
0
|
0
|
5
|
5
|
Completed
|
13
|
3
|
9
|
25
|
Enrolling by invitation
|
0
|
0
|
3
|
3
|
Not yet recruiting
|
150
|
54
|
88
|
292
|
Recruiting
|
223
|
24
|
85
|
332
|
Withdrawn, Suspended or Terminated
|
11
|
2
|
7
|
20
|
Total
|
397
|
83
|
197
|
677
|
Characteristics of subjects
As shown in Table 3, basic characteristics of subjects were classified according to the 'inclusion criteria' of 677 clinical trials. The main types of subjects included confirmed cases (77.25%), unconfirmed cases (13.00%), suspected cases (6.50%) and rehabilitation patients (2.36%). Since most trials received cases of different grades for COVID-19, the total number of different cases was higher than the total number of confirmed cases. In studies with confirmed cases, severe cases were selected more frequently (441 of 532 [84.32%]) as a condition for inclusion. In studies of unconfirmed cases, there were more (38 of 88 [43.18%]) trials with close contacts as the main subjects. The age demographics of the subjects was also accounted in this study. There were 375(55.39%) trials that included both adults and elders, and 233(34.42%) trials included children, adults and elders at the same time. Fewer (59 of 677, [8.71%]) trials were conducted for a single age group.
Table 3 Characteristics of subjects(N=677)
*All age groups were defined by ClinicalTrials.gov: age ≤ 17 is 'Child'; 18 to 64 is 'Adult' and ≥ 65 is 'Older Adult'.
Category
|
Information
|
Number of trials
|
Percentage of total
|
Type of subject
|
Confirmed Cases
|
523
|
77.25%
|
|
Mild Cases
|
316
|
60.42% (N=523)
|
|
Moderate Cases
|
390
|
74.57% (N=523)
|
|
Severe Cases
|
441
|
84.32% (N=523)
|
|
Critical Cases
|
374
|
71.51% (N=523)
|
|
Unconfirmed cases
|
88
|
13.00%
|
|
Close Contact
|
38
|
43.18% (N=88)
|
|
Healthy Volunteers
|
29
|
32.95% (N=88)
|
|
Patients with other diseases
|
21
|
23.86% (N=88)
|
|
Suspected Cases
|
44
|
6.50%
|
|
Rehabilitation Patients
|
16
|
2.36%
|
|
Not Provided
|
6
|
0.89%
|
Age group
|
Adult, Older Adult
|
375
|
55.39%
|
|
Child, Adult, Older Adult
|
233
|
34.42%
|
|
Adult
|
46
|
6.79%
|
|
Child, Adult
|
10
|
1.48%
|
|
Child
|
10
|
1.48%
|
|
Older Adult
|
3
|
0.44%
|
Design of intervention model and blinding in interventional clinical trials
A summary of the statistics for the interventional study types and blinding is shown in Table 4. The main methods of interventional study are randomized parallel assignments (70.27%), the non-randomized parallel assignments (9.58%) and single group assignments (12.29%). Control and experimental groups in randomized parallel trials are observed experimentally simultaneously, which helps eliminate errors caused by factors such as time and experimental conditions. Since there is not any specific drug for treating COVID-19 and its survival period is longer than that of other epidemics, most of interventional studies were designed by a randomized parallel assignment.
Blinding is an important tool for reducing subjects' perceptions of treatment allocation schemes between groups, eliminating bias and improving the scientific and validity of trials.6 Interventional studies with blinding are listed as the following: double- blind accounted for 11.79%; single-blind for 5.41%, open label for 44.23% and not provided for 38.57%.
Table 4 Design of intervention model and blinding in interventional clinical trials (N=407)
Intervention Model
|
Blinding
|
Number of trials
[n (%)]
|
Open Label [n (%)]
|
Single
[n (%)]
|
Double
[n (%)]
|
Not provided [n (%)]
|
Randomized Parallel
|
132 (32.43)
|
20 (4.91)
|
47 (11.55)
|
87 (21.38)
|
286 (70.27)
|
Non-randomized Parallel
|
9 (2.21)
|
0
|
1(0.25)
|
29 (7.13)
|
39 (9.58)
|
Single Group
|
30 (7.37)
|
0
|
0
|
20 (4.91)
|
50 (12.29)
|
Sequential
|
7 (1.72)
|
1 (0.25)
|
0
|
10 (2.46)
|
18 (4.42)
|
Crossover
|
1 (0.25)
|
1 (0.25)
|
0
|
1 (0.25)
|
3 (0.74)
|
Other
|
1 (0.25)
|
0
|
0
|
10 (2.46)
|
11 (2.70)
|
Total
|
180 (44.23)
|
22 (5.41)
|
48 (11.79)
|
157 (38.57)
|
407 (100)
|
Groups
The number of groups involved in COVID-19 clinical trials were counted (Figure 4). 265 trials involved a single group (39.14%), 323 trials involved two groups (47.71%), 55 trials involved three groups (8.12%), and 34 trials involved at least four groups (5.03%). During COVID-19, most trials were divided into two groups, one being controlled studies. This categorization allows differences across groups to be efficiently assessed, benefiting the discovery of 'potent drugs' that can counter the virus.
Figure 4 Groups for all clinical trial registrations (N=677)
Sample size
Directly related to the reliability, reproducibility and efficiency of clinical trials, , the sample size of a clinical trial is critical and demands deliberate consideration. It was noted that the sample size of clinical trials was mainly within the range of 0-299 (Figure 5). The number of participants in interventional and observational trials peaked in the sample range of 100-299. Studies with more than 300 participants accounted for a smaller portion (189 of 677 [27.92%]). There were 29 (4.28%) and 50 (7.93%) observational studies in the sample interval of 500 to 999 and >1000, respectively. There were more observational studies with >500 participants than interventional studies.
Figure 5. Relationship between different study (interventional and observational study) types and sample size (N=677)
Primary endpoint of interventional clinical trials
When designing a clinical trial, one of the most challenging and critical issues is how to select the primary endpoint used to assess efficacy. Given that a clinical trial should provide reliable evidence for its benefits and risks, the primary endpoint is preferably an outcome measure that clearly informs the patients of the benefits.7 Among 407 interventional studies, the required time for clinical recovery constituted 13.51%, clinical parameters and laboratory index for 11.55%, the change of pneumonia severity for 11.06%, questionnaire or scales for 9.83%, virus negative conversion rate of time for 8.11 %, the cure rate for 7.37%, and the mortality rate for 6.14% (Figure 6).
Figure 6. Statistics of primary endpoint (N=407)
Interventions
Interventions from 407 interventional trials were discussed in this section. Table 5 presents the details. Chemical or biological drugs accounted for 43.49%, of which 14.50% were antivirals, 8.85% were antimalarial drugs, and 7.86% were antineoplastics and immunomodulators. Clinical trials involving Traditional Chinese Medicine (TCM) or integrated medicine accounted for 24.08%, and cell therapy, behavioral intervention and medical instruments respectively accounted for 8.60%, 7.13% and 4.91%. In addition, five vaccine clinical trials for COVID-19 are in progress and their registration numbers are NCT04299724, NCT04276896, NCT04283461, NCT04313127 and NCT04324606.
Table 5 Statistics of interventions (N=407)
Category
|
Number of trials
|
Percentage of total
|
Chemical or Biological
|
177
|
43.49%
|
Antivirals
|
59
|
14.50%
|
Antimalarials
|
36
|
8.85%
|
Antineoplastics and Immunomodulators
|
32
|
7.86%
|
Antipyretic Analgesics
|
19
|
4.67%
|
Glucocorticoids
|
9
|
2.21%
|
Mucolytics
|
7
|
1.72%
|
Other
|
16
|
3.93%
|
TCM or Integrated Medicine
|
98
|
24.08%
|
Cell Therapy
|
35
|
8.60%
|
Plasma
|
15
|
3.69%
|
Vaccine
|
5
|
1.23%
|
Medical Instruments
|
20
|
4.91%
|
Behavioral Intervention
|
29
|
7.13%
|
Psychological intervention
|
4
|
0.98%
|
Other
|
23
|
5.65%
|
Combination of antiviral drugs and other drugs
This section further studied the interventional clinical trials of the 59 antiviral drugs aforementioned, and analyzed the combined use of antivirals with other drugs (Figure 7). The most commonly used antivirals were Lopinavir/ritonavir (LPV/r), Arbidol and Ribavirin. The drug with which they were most frequently combined was interferon, followed by Favipiravir with Tocilzumab and Darunavir with Cobicistat.
Figure 7. Relationship between antivirals and other drugs
* The size of the node in the figure corresponds to how many times it is used.
* The thickness of the connection between the drugs indicates the closeness of the joint use between the two.