Characteristics of Registered Systematic Reviews on Treatment for COVID-19 in Prospero Platform

Background: Characteristics and research collaboration of registered systematic reviews (SRs) on treatment for coronavirus disease 2019 (COVID-19) remain unclear. This study aimed to analyze research collaboration, interventions, and outcome measures in registered SRs of treatment associated with COVID-19 and point out the problem. Methods: PROSPERO was searched for SRs of treatment related to COVID-19 as of June 2, 2020. Excel 2016 was used for descriptive statistical analysis of the extracted information. VOSviewer 1.6.14 software was used to generate network maps for collaborations between countries and institutions. Results: A total of 189 SRs were included, which were registered by 301 institutions from 39 countries. China (69, 36.50%) was the country with the highest output. Cooperation between countries was not close enough. Chengdu university of traditional Chinese medicine (7, 3.70%) was the institutions with the highest output. There is close cooperation between institutions. Interventions included antiviral therapy (81, 42.86%), respiratory support (16, 8.47%), circulation support (11, 5.82%), plasma therapy for convalescent patients (11, 5.82%), immunotherapy (9, 4.76%), TCM Treatment (9, 4.76%), rehabilitation treatment (5, 2.65%), anti-inammatory treatment (16, 8.47%), and other treatments (31, 16.40%). In antiviral therapy (81, 42.86%), the most commonly used drugs were chloroquine/hydroxychloroquine (26, 13.76%), followed by remdesivir (12, 6.35%), lobinavir/ritonavir(11, 5.82%), favipiravir (5, 2.65%), ribavirin (5, 2.65%), interferon (5, 2.65%), abiron (4, 2.12%), abidor (4, 2.12%), but the description was brief, and no speci�c implementation plan was provided. The most frequently used primary outcome was mortality rate (92, 48.68%), and the most frequently used secondary outcome was length of hospital stay (48, 25.40%). The expression of the outcomes was not standardized. Conclusions


Introduction
In late December 2019, an outbreak of pneumonia of unknown origin characterized by strong interpersonal transmission.2][3][4][5] This is a coronavirus of the genus beta.It is enveloped with round or oval particles.The main clinical manifestations of COVID-19 are fever, dry cough, and fatigue.Pathological changes of lung, spleen, hilar lymph nodes, bone marrow, heart, and blood vessels are observed.The virus is generally susceptible to human infection. 6Within three months, it has affected six continents. 7,8 s of August 20, 2020, 22817751 cases have been reported, including 793379 deaths. 9ROSPERO is an international database of prospectively registered systematic reviews in health and social care, welfare, public health, education, crime, justice, and international development.After the outbreak, there was no clear and effective treatment and no speci c medicine, medical workers and scienti c researchers actively carry out research and have registered numerous COVID-19 treatment systematic reviews (SRs).However, no research has focused on the characteristics of these registered SRs.This study was designed to evaluate the cooperation between countries and institutions and the distribution of outcome measures in registered COVID-19 treatment SRs, to provide a reference for future researchers to register and carry out COVID-19 SRs.

Data sources
We systematically searched the PROSPERO registration platform (https://www.crd.york.ac.uk/prospero) to identify all registered COVID-19 treatment SRs.The deadline for retrieval is June 2, 2020.

Inclusion and exclusion criteria
The type of included study was registered SRs on PROSPERO.The study population was patients diagnosed with COVID-19, and there were no restrictions on age, gender, race, and course of disease.
Intervention was arbitrary treatment.We excluded basic science, diagnostic tests, empirical studies, and health services.Duplicate records were also excluded.

Study selection and data extraction
Two researchers independently reviewed and screened and the retrieved records based on pre-determined inclusion and exclusion criteria, and then exchanged checks with each other.Finally, through communication with the third researcher, the dispute was resolved.
One researcher used a prede ned form to extract detailed data from included registrations, while another reviewer veri ed the accuracy of the extracted data.Speci c data include: subject, author, language, research type, discipline type, interventions, control measures, country, institution, primary and secondary outcomes, registration time and expected completion time.

Data management and analysis
We preprocess the extracted data and standardize institutions, interventions, and outcomes with different expressions.Microsoft Excel was used for descriptive analysis of the extracted data.Then, VOSviewer 1.6.141][12][13] In the obtained network diagram, nodes represent the elements of analysis (countries and institutions), node size reveals frequency, color of nodes indicate different clusters, and lines represent the cooperation between different nodes.The parameters of VOSviewer are as follows: counting method (fractional counting), ignoring documents with multiple authors (maximum number of authors per document is 25).

General characteristics of registered SRs
By June 2, 2020, 205 SRs were retrieved, of which 189 met the inclusion criteria.By August 3, 2020, 122 SRs had reached the expected completion time, of which 111 SRs were still under review ongoing (90.98%), and only 11 SRs (9.02%) have been completed but have not yet been published.All details are shown in Table 1.
In terms of research funding sources, 136 SRs were not funded, accounting for 71.96%, and 53 SRs were funded, accounting for 28.04%.The most frequent source of nancial assistance was from the National Natural Science Foundation of China for 11 SRs, accounting for 20.75% of all nancial assistance.All details are shown in Table 1.
Among the SRs included, March, April, May, and June were registered 25, 109, 50 and 5 SRs, respectively, accounting for 13.13%, 57.67%, 26.46% and 2.65% respectively.According to the speci c date, the number of registered SRs on April 20, 2020 was the largest, with 14 SRs.The relationship between the speci c date and the number of registered SRs is shown in Fig. 1.

Discussion
The PROSPERO platform opened the COVID-19 retrieval channel, carried out reasonable and meticulous classi cation management of the literature, and the retrieved literature met the inclusion standards.By August 3, 2020, 122 of the 189 SRs included should have reached the expected completion time, but in fact, only 11 SRs were completed, with the completion rate of only 9.02%.It can be explained that the reasons for the low completion rate may be that the data resources obtained in the early stage are insu cient, the data acquisition method is di cult, the expected completion time is too short, and the di culty of the research is estimated incorrectly.Therefore, in the future research, we should try our best to avoid it, and we should make a correct evaluation of the research, whether from the feasibility or the time.To achieve a reasonable planning progress, as far as possible in the scheduled time to complete.
Through the reports on the retrieval database, it is not di cult to see that among the 188 SRs, except for one which is a single database, the remaining 187 SRs are all joint retrieval of multiple databases to obtain the corresponding data.).This shows that there are a lot of repeated studies and there may be a waste of scienti c research resources.The description of intervention measures was not standard enough.The included literature did not describe the treatment method, dosage and time in detail.Therefore, in the future research, researchers should be more careful to check the registered projects to avoid repeated research.
Interventions used should be described in more detail and in a more standardized way.
Among all the outcome indicators, the most important outcome indicators were mortality rate (92, 48.66%), then adverse events (28, 14.81%) and the time of becoming negative for the coronavirus (22, 11.64%); the most frequently used secondary outcome was length of hospital stay (48, 25.40%), and then adverse events (43, 22.75%), ICU length of stay (30, 15.87%) and mechanical cultivation (23, 12.17%).By comparing the two types of outcome indicators, the common items were length of hospital stay, adverse events, ICU length of stay, the time of becoming negative for the coronavirus, and monitoring of mechanical ventilation.However, there are some differences between the two.The main outcome indicators are monitoring clinical symptoms and signs, while the secondary outcome indicators tend to be prognostic indicators, laboratory test data (IL-6, C-reactive protein, ALT), and the adverse reactions mainly focus on adverse drug reactions.Therefore, we can see that the determination of outcome indicators should be complementary as far as possible, so as to obtain more detailed and perfect information of outcome indicators, and avoid excessive repetition.

Conclusions
A social network analysis of countries A social network analysis of institutions

Figure 1 Registration
Figure 1
Each SR included has multiple outcome indicators.The main outcome indicators are related to a series of symptoms, signs, examination, prognosis and so on.The most common outcome measure was mortality rate (92, 48.68%), followed by adverse events (28, 14.81%), time of becoming negative for the coronavirus(22, 11.

Table 7
The top 16 secondary outcome measures in terms of frequency [N (%)] Among them, PubMed / MEDLINE combined with EMBASE (152, 80.85%) was used most.At the same time, 136 (72.34%)SRs retrieved only English databases, and the remaining 52 (27.66%) were combined with Chinese and English databases.In this view, most of the databases are limited to English, and their data are not representative enough to a certain extent.The above shows that in the use of databases, it is advisable to search multi language databases at the same time to make the data more representative.164 (86.77%)SRs were completed by one country independently, and 25 (13.23%)SRs were completed by cooperation.69 (36.51%) of the SRs were undertaken by China, 22 (11.64%) by UK and 19 (10.05%) by Brazil.A total of 39 countries have participated in the registered SRs and 27 (69.23%)countries have formed cooperative relations.UK has the largest cooperation intensity, followed by China, Canada, Egypt, and Italy.A total of 301 institutions contributed to the registered SRs and 232 (77.08%) had cooperative relations.Chengdu University of traditional Chinese medicine has undertaken 7 (3.7%)projects, Liaoning treatment (81, 42.86%), of which 11 SRs did not specify the speci c drugs.The most common used drugs were chloroquine / hydroquinone (26, 13.76%), followed by remdesivir (12, 6.35%), lobinavir / ritonavir (11, 5.82%), favipiravir (5, 2.65%), ribavirin (5, 2.65%), interference (5, 2.65%), abiron (4, 2.12%), abidor (4, 2.12%