A common challenge in systematic reviews is maintaining a clear, auditable trail from literature discovery through to final inclusion. While ResearchRabbit excels at surfacing relevant papers via its similarity and citation networks, its suggestion lists aren't directly compatible with the PRISMA flow diagram required for publication. This post details a methodical workflow for integrating ResearchRabbit's output into a PRISMA-compliant screening process.
My approach uses a dedicated reference manager (Zotero, in this case) as the system of record, with careful tagging to mirror PRISMA stages. Here is the step-by-step integration:
* **Establish a Dedicated Collection:** Create a new Zotero collection for your review project. All literature will pass through here.
* **Initial Import from Databases (PRISMA's "Identification"):** Import all records from your primary database searches (e.g., Scopus, Web of Science) into this collection. Apply the tag **"01_identified"**.
* **Processing ResearchRabbit Suggestions:** When you explore a seed paper in ResearchRabbit and generate a list of "Similar Work" or "Later Citations":
* Manually export the suggestion list (CSV/BibTeX) from ResearchRabbit.
* Import this file into Zotero, placing items into your project collection.
* Immediately apply a distinct tag: **"02_rabbit_suggestions"**. This is crucial for tracking this unique source.
* **Deduplication:** Use Zotero's duplicate detection function across the entire collection. Merged records should retain all relevant source tags (`01_identified`, `02_rabbit_suggestions`).
* **Screening & Tagging:** Proceed with title/abstract screening. As you exclude records, tag them with **"excluded_[reason]"**. For those advancing to full-text retrieval, use **"03_screened"**. After full-text assessment, tag with **"included"** or **"excluded_ft_[reason]"**.
This tagging structure allows you to generate accurate PRISMA numbers. You can report that *n* additional records were identified via ResearchRabbit's snowballing (count of unique items with the `02_rabbit_suggestions` tag). Your final flow diagram will accurately reflect both traditional search results and AI-assisted discovery without compromising the rigor of the screening process. The key is treating ResearchRabbit as a distinct, trackable source within your reference manager's workflow.
Your bill is too high.