Silvi vs Covidence

1. Ease of Use

Covidence

Covidence is optimized for systematic reviews in life sciences where researchers conduct systematic reviews regularly.

When you start a new review in Covidence, the first thing you are asked is if you are going to conduct a Cochrane review or not.

For a new user intending to conduct a systematic review in social sciences or for someone not familiar with Cochrane reviews, this may cause confusion.

Covidence also asks you to specify the type of question you are going to answer through your systematic review. This means you would need to have sufficient familiarity with the systematic review process before starting a project in Covidence.

Prior knowledge and experience with the systematic review process makes it difficult for students and new researchers to get started with Covidence.

Silvi

Compared to Covidence, Silvi is very easy to get started especially if you are a student or starting your research career. Starting a project in Silvi doesn’t require any familiarity with the systematic review process. Silvi has been designed to make it easy for researchers to get started on a systematic review.

To start a project in Silvi, simply click on “Create new project” in the top-right corner of your screen, type in the title of your project, and your project will be created.

2. Database Integrations

Covidence

Once you have started a project you will need to import studies into it. Covidence does not provide you with direction integrations with databases.

To import studies into a Covidence project, you will need a file in any of the following formats: EndNote XML, PubMed, or RIS. This means you will first have to download an XML, PubMed, or RIS file from a database and then upload it in Covidence. This makes the process lengthier and cumbersome.

Silvi

Silvi offers direct integrations with three databases: PubMed, ERIC (Education Resources Information Center), and OpenAlex. This feature lets you import studies directly from a given database. You won’t have to download an XML or a PubMed file from a database and then upload it to Silvi.

To import studies from PubMed, simply run a search and copy the search string by clicking on “Advanced” under the search bar. Paste the search string in Silvi and it will automatically fetch all the studies related to the search string.

For ERIC and Open Alex, you can simply copy the URL of your search results and paste it in Silvi, and it will retrieve all the relevant studies for you.

If you want to import studies from any other database like Web of Science or Scopus, you import them using an RIS file like Covidence.

3. Bulk Screening

Covidence

In Covidence you screen titles and abstracts one by one. Covidence gives you the titles and abstracts of studies in the form of a long list. Screening studies one by one is tedious and Covidence feels like an “old school” app since this is how systematic literature reviews have been done traditionally.

Silvi

When doing literature reviews, you often come across a group of studies that can be excluded. Here Silvi lets you search for these studies and select them all at once to give you an overview. You can then remove these studies in Bulk instead of going through them one by one. This can really speed up the screening process.

That said, if you need to be thorough, it is possible in Silvi to screen studies one-by-one as well. This is your choice and depends on the type of review you are conducting.

4. Screening with AI

Covidence

While screening titles and abstracts, Covidence offers researchers a “Most relevant” feature, which uses a machine learning model to identify trends in your past screening behavior to determine which studies are likely to be included first. Since this machine learning model uses your screening pattern, it won’t give you relevancy scores until you have not screened 25 studies with at least 2 studies included and 2 excluded.

Silvi

Unlike Covidence, Silvi’s AI screening feature asks you explicitly about studies that are eligible for inclusion or exclusion instead of relying on your earlier screening patterns.

Silvi AI screening in contrast to Covidence offers you to be explicit towards the AI about what studies are eligible instead of relying on AI to pick up what you are looking for purely based on your screening pattern

Silvi’s AI feature doesn’t require you to screen 25 studies like Covidence. Instead, Silvi asks you to specify your inclusion/exclusion criteria, which is very similar to the way you would specify the criteria to your team members. Silvi’s AI feature requires you to include and exclude only one study each although if you screen more studies, it would certainly be helpful.

Based on the specified inclusion and exclusion criteria, Silvi will give you suggestions to include or exclude a given study. AI suggestions will appear in black color in the bottom-left corner of the abstract.

5. Finding PDFs

Covidence

After you are done with the title and abstract screening, you will move on to the full text screening phase. In Covidence, you will have upload PDFs for each paper yourself. Covidence doesn’t retrieve PDFs of open access papers automatically. This requires you to look for PDFs of studies, download them, and then upload them to Covidence, which makes the whole process lengthy and cumbersome.

Silvi

When you reach the full text phase in Silvi, it will automatically retrieve PDFs of open access papers for you. You will only have to upload PDFs of papers that are behind paywalls. This will help you save valuable time.

6. Screening full-texts

Covidence

When you move on to screening full texts, you will have to click on the PDF you have uploaded to open it in a new window. Covidence does not display the list of articles along with the full text of the article that you want to screen. This makes the process of screening full texts cumbersome because you have to keep moving from one screen to another.

Silvi

Silvi displays the full text of an article right next to the list of articles thereby making it very convenient to access other articles. In Silvi, you can screen full texts within the app and look at the overview of all the studies together with the PDF in a single view. This doesn’t disrupt your flow.

7. Data Extraction with AI

Covidence

Covidence gives you editable templates for extracting data from full text of papers. Templates that Covidence gives you at this stage include Identification, Methods, Population, Interventions, Outcomes, and Results data.

Furthermore, templates in Covidence are optimized for biomedical research. If you are conducting a systematic review in a field like education you will have to design your own templates, which can become a cumbersome process.

Silvi

In Silvi, you can, of course, extract data manually but it also gives you an AI-powered feature that lets you extract data based on certain tags that you create. You can create three types of tags: text, numeric, and category. While creating these tags, you can also tell Silvi briefly about the kind of data you want it to extract from the full text of papers.

For example, you can create a category tag for all the countries from where the participants in a particular study came from. While creating the tag, you can specify that Silvi should not consider the countries of authors mentioned in the paper and only focus on the countries of the participants.

8. Transparent Data Extraction

Covidence

Once you have extracted data from the full text of a paper, you cannot link the extracted data with the relevant part in the text of the paper. It can make verifying the origin of the extracted data very difficult when coming back to the review.

Silvi

Silvi will extract relevant data, and every bit of extracted data will be automatically linked to the relevant part in the text so you can see for yourself the place where it extracted the data from. This ensures transparency in the data extraction process.

Even though Silvi’s AI feature extract data for you, it gives you full control and decision-making authority. Once it has extracted data, you approve it if you agree with Silvi’s suggestion. Otherwise, you can easily delete these suggestions.

9. Tables and Plots

Covidence

After you are done with data extraction, the only chart Covidence helps you create is PRISMA. You can’t generate any other plots or tables in Covidence using your extracted data. You will have to download the extracted data as separate files, which you will then have to upload to other tools like RevMan or Excel for analysis.

Silvi

In Silvi, you can easily generate various plots and tables using your extracted data in addition to a PRISMA chart. Simply select the kind of data you want to create a table or a plot with and Silvi will give you the relevant table or plot.

Once have done the hard work of extracting data from full texts, you want to look at the data in an organized manner. In Silvi, you can look at the data directly within the app. You can create a table with all your extracted data or a certain part of it in a sub-table.

This enables you to go back and forth between full text and analysis to quickly extract more data and see the results immediately without having to deal with organising and downloading files.