diagnostic-study-quality-assessment-quadas-2

QUADAS-2 assessment evaluates four domains: Patient Selection, Index Test, Reference Standard, and Flow and Timing. Each domain uses signaling questions answered Yes, No, or Unclear, with every judgment supported by a direct quote from the paper.

Use the bundled pdf_extractor.py script with PyPDF2 installed. Run python scripts/pdf_extractor.py paper.pdf to extract full text, or pass start and end page numbers to extract a specific page range into a UTF-8 text file.

The four domains are Patient Selection, Index Test, Reference Standard, and Flow and Timing. Each domain has signaling questions covering issues like consecutive sampling, blinding, pre-specified thresholds, and whether all patients received the same reference standard.

When information is missing, the rating must be marked Unclear rather than guessed. The skill enforces strictness by requiring a direct quote for every judgment and defaulting to Unclear when no supporting evidence exists.

PyPDF2 extraction fails on scanned PDFs that lack a text layer, since they contain only images. The script also cannot guarantee correct Methods section detection if the paper uses nonstandard section headings.