Full text retrieval
Full texts, supplements, code and data
Every candidate record is checked against Europe PMC, PubMed Central and Unpaywall. Open texts and supplements are downloaded automatically; the rest are retrieved through institutional access. Files are kept privately and never published: only the manifest of what was retrieved, with checksums and licenses, is public.
Availability by retrieval tier
| Tier | Open, Europe PMC | Open, PDF | Free to read | Paywalled | No DOI | Total |
|---|---|---|---|---|---|---|
| 1. Reports of registered randomized trials | 595 | 194 | 126 | 437 | 15 | 1,367 |
| 2. Other randomized trial reports | 1,222 | 1,043 | 587 | 2,007 | 357 | 5,216 |
| 3. Systematic reviews | 1,190 | 585 | 351 | 1,476 | 96 | 3,698 |
| 4. Nonrandomized studies | 2,735 | 1,733 | 960 | 3,641 | 297 | 9,366 |
| 5. Protocols and other trials | 1,027 | 580 | 350 | 1,385 | 211 | 3,553 |
Automatic downloads so far: fulltext pdf failed: 1,906; fulltext pdf failed_not_pdf: 217; fulltext pdf no_text_layer: 1; fulltext pdf retrieved: 1,121; fulltext pdf wrong_file: 5; fulltext xml failed: 360; fulltext xml retrieved: 5,732; supplement zip failed: 1,210; supplement zip none_available: 910; supplement zip retrieved: 4,044.
Manual retrieval queue
Items no open route reaches, in priority order. To help with retrieval through institutional access, save each PDF under the name in save_as into data/fulltext/inbox/ of a local clone and run ingest_inbox(); see the instructions.
Loading data…