Data Extraction From Unstructured
Documents for Living Reviews
Automate every stage of your literature review to produce evidence-based research faster and more accurately.
Living reviews are continually updated by incorporating relevant new evidence as it becomes available. Data extraction involves the retrieval or collection of data from different sources, both physical and digital, for use in the data analysis stage of the systematic review process. These data sources can be classified into structured and unstructured data. Structured data appears in predefined models, usually organized into templates and spreadsheets that are easy to analyze.
Unstructured data, on the other hand, is not in a predefined model—it must be collected from different pages or documents and combined into one model in a spreadsheet. This article will show you how to conduct data extraction from unstructured documents for living reviews.
The Process of Data Extraction
The data extraction process begins after identifying the individual studies that match the research question of the living review. There are many data extraction methods in systematic reviews that you can apply to gather data for analysis.
1. Generate Data Extraction Forms
The first step is to create a data extraction form to help you collect useful data from unstructured documents; many examples of data extraction forms for systematic reviews are available to guide you in generating the right one for the process. Given the importance of data extraction forms, ample time should be invested in their design. Use existing systematic reviews on your topic to guide you on the type of information to collect.
The methods of data extraction you use to generate the data extraction forms can include systematic review software, spreadsheet software, and Google Docs.
2. Train Your Team on Extraction Categories
Training your team helps reduce discrepancies that may occur when you begin the data extraction process. Show them the type of data categories to collect and establish a standard. This way, other researchers taking part in the review process will collect relevant data.
3. Perform a Trial of the Extraction Form
Conducting the trial helps you catch any errors that may occur in the data extraction process. You want to avoid researcher bias and have every researcher present similar data for synthesis.
4. Discuss Inconsistencies in the Process
After a successful trial, conduct the data extraction process. Address any inconsistencies if researchers on your team collect irrelevant data. This way, you’ll improve the validity and reliability of the living review once you’re done updating it.
Meta-analysis data extraction involves collecting data from multiple studies to find common results and synthesize them so as to identify overall trends. Meta-analysis can be part of a systematic review process to establish statistical significance with studies that have conflicting results. It can also provide a more complex analysis of harms, safety data, and benefits.
Learn More About DistillerSR
(Article continues below)
Living reviews are continually updated when new evidence related to the research question becomes available. Data extraction is a process using systematic review software, spreadsheets, and optical recognition software through which researchers collect data that can be analyzed and synthesized.