How to Use MetaCensus

MetaCensus has numerous functions. At its core, it is a tool for systematic reviews, meta-analyses, and consensus statements that reflect the current overall interpretation of the evidence, its limitations, and what investigation needs to be prioritized next to elucidate further what is known on a scientific topic.

Step 1:
Ask a scientific question, develop a systematic review and meta-analysis protocol

All meta-analyses begin by identifying the correct question to which they seek an answer. From there, a protocol will be developed to search for and analyze evidence related to the question systematically. MetaCensus strongly advocates using Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). MetaCensus has a protocol developer tool (insert link to the protocol page) that helps structure data fields to be abstracted and structured for further analysis. Select all relevant fields the review team will need to perform a robust systematic review and meta-analysis during this step.

Although MetaCensus initially focused on medical literature, it aims to expand to all relevant scientific domains that leverage scientific literature in future efforts.

Step 2:
Perform a literature search

Currently, MetaCensus has developed a literature search that allows a participant to structure a scientific question with relevant Boolean operators on the MetaCensus literature search page (link to the search page) that queries PubMed's Advanced Search Builder without having to leave MetaCensus's web app. The search will return search results, including publication title, DOI, authors, publish date, and abstract. From there, a reviewer can decide which search results are relevant for the review group to further investigate and analyze for relevance to the evidence related to their scientific question. Important note- investigating teams will likely need to search more scientific literature repositories than PubMed. While MetaCensus has yet to build integrated queries to other literature repositories, we strongly advocate for teams to search other repositories. As MetaCensus seeks further development support, we intend to add other repositories to our integrated literature search within MetaCensus. If you are someone who would like to support building such features, please message us here (link to the search page).

Step 3:
Full text upload

Once all relevant abstracts have been selected for further review, collaborators can begin pulling full texts of the publications for data abstraction. Copyright infringement warning*** Uploading full texts of scientific literature must comply with each user's licensing agreement through which they access the full text of any publication. MetaCensus does not condone the dissemination, use, or misuse of literature that negates or infringes on the copyright laws and licensing agreements through which any publication is accessed or utilized. Individual users are solely responsible for how they choose to use their judgment and agency related to compliance with copyright laws and regulations through which they access scientific literature. Meta-analysis groups are often permitted to collaborate on reviewing and analyzing publications being deliberated over for inclusion in a meta-analysis. Papers can be uploaded for individual use or shared with collaborators on a meta-analysis, depending on what is appropriate and in compliance with copyright rules, regulations, and laws.

Step 4:
Data abstraction

Once full-text PDFs have been uploaded, collaborators can begin extracting relevant data from the texts using the data abstraction tool (link to the tool page). MetaCensus has found one risk of compromised data integrity can be seen in transcription errors- it is easy to mistype a number, put a comma somewhere incorrectly, misspell a word, or misrepresent what was initially published through some other error. To reduce the risk of such errors (and to save collaborators time during data abstraction), MetaCensus has developed a feature where the abstractor can highlight a word, number, or character that they want to be transferred over to the text field and then hit the transfer button for the text or number to be copied into the field. As the data fields are abstracted, they will be encoded as JSON formatted data to be transported for structured data that can be used in tables and calculations as data are added and amassed for future use and analysis in the meta-analysis. It is strongly recommended that data be independently abstracted by no less than two abstractors that can then be compared against each other to ensure data integrity and accuracy.

Step 4:
PRISMA protocol assessment of each paper for inclusion, evidence weighting, and resulting analysis

Once data are abstracted and ready for the review team to assess the evidence using guidance from Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA, papers can be scored and weighed for the substance of their evidence in aggregate. Voting on inclusion and weighing of evidence in each paper as is conventionally done in meta-analyses. Voting members can each leverage PRISMA protocol recommendations to determine how to assess and score evidence being considered for each meta-analysis. MetaCensus has developed some preliminary web application tools to generate forest plots that allow basic odds ratios to be calculated and resultant forest plots to be generated. We seek biostatisticians, data scientists, and web developers to strengthen our analytic offerings and further empower the community to perform better science more quickly. Please get in touch with us with any recommendations for features you think might be beneficial to develop, and better yet, reach out to us if you are someone who can help build such features (link to the contact page).