diff --git a/.gitignore b/.gitignore index 83972fa..78b7a1f 100644 --- a/.gitignore +++ b/.gitignore @@ -149,6 +149,9 @@ activemq-data/ # Environments .env +.env.* +!.env.example +!.env.*.example .envrc .venv env/ @@ -216,3 +219,12 @@ __marimo__/ # Streamlit .streamlit/secrets.toml + +# Local scraped corpora and review handoff notes +/data/ +/LOCAL_REVIEW.md +/WORK_STATUS.md +/.codex/ + +# macOS metadata +.DS_Store diff --git a/AGENTS.md b/AGENTS.md index 230d1d9..6e8f67d 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -25,4 +25,6 @@ Use pytest. Name test files `test_*.py` under the owning component's `test/` dir ## Commit & Pull Request Guidelines +Use descriptive branch names without a `codex/` prefix, for example `cco-scraper-update`. + The existing history uses short, imperative, lowercase commit subjects such as `setup ruff linting and github ci`. Keep commits scoped to one logical change. Pull requests should describe the changed component, summarize behavior and tests run, link any issue or experiment context, and include screenshots only when user-facing rendered output changes. diff --git a/datasets/benchmarks/PDF_OCR_REVIEW.md b/datasets/benchmarks/PDF_OCR_REVIEW.md new file mode 100644 index 0000000..6a50c9e --- /dev/null +++ b/datasets/benchmarks/PDF_OCR_REVIEW.md @@ -0,0 +1,158 @@ +# PDF extraction and OCR review + +Local diagnostic, October 6, 2026. The Firecrawl versus FinePDFs comparison is +**not complete**. Firecrawl was skipped because `FIRECRAWL_API_KEY` is unset. +FinePDFs was inspected but its GPU OCR pipeline was not executed. “finepdf” is +provisionally interpreted as Hugging Face FinePDFs, pending user confirmation. + +## Measured sample and method + +Two committed English, born-digital WHO excerpts, six pages each: + +- Bloodstream infections, publication `9789240121805`, source PDF pages 30-35. +- HIV service delivery, publication `9789240124233`, source PDF pages 11-16. + +The fixtures, original URLs and reuse notes are documented in +[fixture attribution](../test/fixtures/pdf/README.md). Raw outputs are +research derivatives of the WHO sources and carry the source reuse conditions; +the repository's code licence does not replace those conditions. + +Hardware: Apple M4 MacBook Air, 16 GB RAM; macOS 26.2, Python 3.13.15. Docling +used its default automatic accelerator selection, not a forced CPU benchmark. +Its model weights were already cached by the integration tests. Each elapsed +time covers conversion/export; Tesseract times include 300-DPI rendering and +all six page OCR calls. These are single runs, not stable throughput estimates. + +Native extraction used PDFium's text layer. Tesseract received page PNGs only, +with English language data and automatic page segmentation (`--psm 3`). Docling +used OCR off and table structure/cell matching on. Rasterizing born-digital PDFs +tests an OCR path; it does not substitute for naturally scanned/degraded PDFs. + +Seven source-checked anchors per excerpt cover recommendation wording, +population, a negation, concentration/risk numbers, age-dependent certainty, +intervals, task sharing and a date-column footnote. Relevant source pages were +rendered and checked visually. The checks normalize whitespace and hyphenated +line breaks. Anchor presence is not a clinical accuracy, recall, CER or WER +score, and it does not establish table alignment or absence of hallucinations. + +## Results + +| Backend | Version/configuration | CVC anchors | HIV anchors | Seconds: CVC / HIV | +| --- | --- | --- | --- | --- | +| Native PDF text | pypdfium2 5.14.0 | 7/7 | 6/7 | 0.024 / 0.026 | +| Tesseract | 5.5.3, English, 300 DPI, PSM 3 | 7/7 | 7/7 | 17.029 / 13.848 | +| Docling before footnote fix | 2.134.0, default markdown export | 7/7 | 6/7 | 12.787 / 5.156 | +| AMFV Docling after footnote fix | 2.134.0, missing footnotes retained with page provenance | 7/7 | 7/7 | 11.282 / 5.035 | +| Firecrawl auto and forced OCR | Hosted API | skipped | skipped | not measured | +| FinePDFs | Published full pipeline | not run | not run | not measured | + +The native HIV miss is a text-layer artefact: `high-certainty` contains U+FFFE +in place of the hyphen. Both certainty populations remain readable in the raw +output, but exact matching fails. This is not an omitted recommendation. + +Docling preserved all 26 dated recommendation rows in the two-page HIV table, +including their date/status columns. Their order and the 2025 Updated/New +statuses were checked against the source rendering. Native output separates +status values into a block after the recommendation text, which weakens their +row association. Tesseract's anchor success alone does not prove a usable +three-column table: its output is unstructured text, groups statuses separately, +and omits the `New` status for the 2025 mental-health recommendation on page 2. + +The default Docling export omitted the footnote defining the date column. +The structured conversion still contained that footnote. AMFV now appends +detected footnotes missing from markdown under `PDF footnotes`, with PDF page +numbers. The HIV footnote retains its source wording and page 2 provenance. +The real-PDF test and refreshed reference output check this behavior. Table +captions, certainty phrases and recommendation rows remain present. + +Results and raw markdown: + +- [Initial comparison](results/2026-10-06/results.json) +- [Final Docling export](results/2026-10-06/footnotes-preserved/results.json) +- [Manifest and source hashes](pdf_ocr_manifest.json) +- [Reproducible runner](pdf_ocr.py) + +The five-document Docling/Marker/PyMuPDF4LLM results in the original +[WHO PR #6](https://github.com/MedARC-AI/amfv/pull/6) came from that author. +They were not reproduced in this update. + +## Firecrawl and FinePDFs status + +The runner supports Firecrawl API v2 `auto` and forced `ocr` modes, per its +[official API guide](https://github.com/firecrawl/firecrawl-docs/blob/84e2b419134e753e6f22ac1f3fcb713256154487/advanced-scraping-guide.mdx). +It verifies that the public fixture URL has exactly the local PDF's SHA-256 +before submitting it. Only the six-page public excerpts are configured. Missing +credentials produce an explicit skipped result. No Firecrawl request was sent. +Its hosted model revision is not pinned, so retain the API response metadata. + +[FinePDFs](https://github.com/huggingface/finepdfs/tree/15aa381fae828c8acd6efa058fa0b41d2c602ede) +is a corpus-processing pipeline, not a drop-in equivalent of the Firecrawl API. +Its published code uses modified vendored Docling for embedded text and +`reducto/RolmOCR` served by vLLM for OCR. Its embedded-text extractor disables +table structure, unlike this AMFV configuration. A stock Docling run therefore +cannot be reported as a FinePDFs result. Its top-level code is AGPL-3.0; this +update did not copy it into the AMFV package. + +No NVIDIA/vLLM inference host or remote OCR endpoint was provided for this Mac. +The next comparison needs the confirmed intended “finepdf” project, Firecrawl +credentials and actual FinePDFs outputs from the same hashed excerpts. A larger +test should also include naturally scanned pages and manually transcribed +ground truth, with table-cell/date/status associations and numeric/negation +errors graded separately. + +## Open OCR candidates + +These are candidates for the next controlled comparison, not winners established +by this two-excerpt diagnostic: + +| Candidate | Role | Evidence/status | +| --- | --- | --- | +| [Tesseract](https://github.com/tesseract-ocr/tesseract) | Local CPU baseline for scanned text | Tested here; text anchors survived, table structure needs a separate parser. | +| [RolmOCR](https://huggingface.co/reducto/RolmOCR) | First OCR model for a faithful FinePDFs comparison | FinePDFs' published OCR model; Apache-2.0 model card. Not run here. | +| [PaddleOCR / PP-StructureV3 / PaddleOCR-VL](https://github.com/PaddlePaddle/PaddleOCR) | Challenger for layout/table-rich guideline pages | Official toolkit provides structured Markdown/JSON and table-cell coordinates. Not run here. | +| [olmOCR](https://github.com/allenai/olmocr) | Independent document-OCR challenger | Apache-2.0; supports local NVIDIA inference or a remote server. Not run here. | + +Start with RolmOCR for FinePDFs fidelity, PaddleOCR's structure pipeline for +table quality, and Tesseract as the inexpensive baseline. Keep Firecrawl's +hosted parser as a separate backend rather than assuming its OCR model is +interchangeable with these candidates. Use native extraction for clean text +layers when its order and encoding survive source checks. + +## Running the diagnostic + +New local runs under `results/` are ignored by default. The reviewed +`results/2026-10-06/` evidence is deliberately retained in version control. +Use a new output directory when rerunning an experiment to preserve that evidence. + +```bash +uv sync --dev --group ocr +uv run --group ocr python datasets/benchmarks/pdf_ocr.py \ + --output-dir /tmp/amfv-pdf-results --backend native tesseract +``` + +The optional `docling` backend uses the AMFV exporter from the separate WHO +scraper update. It records a skip until that module and its PDF dependencies +are installed. After the WHO change is merged, reproduce the final export with: + +```bash +uv sync --dev --group pdf --group ocr +uv run --group pdf --group ocr python datasets/benchmarks/pdf_ocr.py \ + --output-dir /tmp/amfv-docling-results --backend docling +``` + +The initial default-export measurements are retained as historical evidence; +the runner uses the final AMFV exporter when that integration is available. + +Tesseract and its English language data must be installed separately. Tesseract +5.5.3 was installed with Homebrew for this run. PDF model weights were downloaded +to the normal cache, and FinePDFs was inspected in `/tmp/amfv-finepdfs-20261006`. + +With `FIRECRAWL_API_KEY` configured securely in the process environment: + +```bash +uv run --group ocr python datasets/benchmarks/pdf_ocr.py \ + --output-dir /tmp/amfv-firecrawl-results --backend firecrawl-auto firecrawl-ocr +``` + +The second command calls the paid hosted service when credentials are present. +The missing-key path was exercised here; live Firecrawl integration is untested. diff --git a/datasets/benchmarks/pdf_ocr.py b/datasets/benchmarks/pdf_ocr.py new file mode 100644 index 0000000..68f2224 --- /dev/null +++ b/datasets/benchmarks/pdf_ocr.py @@ -0,0 +1,230 @@ +"""Compare PDF text, Docling, raster OCR and Firecrawl on the same WHO excerpts. + +Run from the workspace root with `uv run --group ocr python +datasets/benchmarks/pdf_ocr.py --output-dir /tmp/amfv-pdf-results`. +This is a small diagnostic, not a character-error-rate or model leaderboard. +""" + +from __future__ import annotations + +import argparse +import hashlib +import json +import os +import platform +import re +import shutil +import subprocess +import tempfile +import time +from datetime import UTC, datetime +from importlib.metadata import version +from pathlib import Path +from typing import Any + +import httpx + +try: + import pypdfium2 as pdfium + + HAS_PDFIUM = True +except ImportError: + HAS_PDFIUM = False + +try: + from amfv_datasets.scraping.pdf import HAS_DOCLING as HAS_AMFV_DOCLING + from amfv_datasets.scraping.pdf import pdf_to_markdown + + HAS_AMFV_PDF = True +except ImportError: + HAS_AMFV_PDF = False + HAS_AMFV_DOCLING = False + +_FIXTURES = Path(__file__).resolve().parents[1] / "test" / "fixtures" / "pdf" +_MANIFEST = Path(__file__).with_name("pdf_ocr_manifest.json") +_BACKENDS = ("native", "tesseract", "docling", "firecrawl-auto", "firecrawl-ocr") + + +def _normalized(text: str) -> str: + text = text.translate(str.maketrans({"–": "-", "−": "-", "‑": "-", "\u00ad": ""})) + text = re.sub(r"-\s+", "-", text) + return " ".join(text.casefold().split()) + + +def _native_text(path: Path) -> str: + with pdfium.PdfDocument(path) as document: + pages: list[str] = [] + for page in document: + text_page = page.get_textpage() + try: + pages.append(text_page.get_text_range()) + finally: + text_page.close() + page.close() + return "\n\n".join(pages) + + +def _tesseract_text(path: Path, *, dpi: int, page_segmentation: int) -> str: + with tempfile.TemporaryDirectory(prefix="amfv-ocr-") as directory, pdfium.PdfDocument(path) as document: + pages: list[str] = [] + for index, page in enumerate(document, start=1): + image_path = Path(directory) / f"page-{index}.png" + bitmap = page.render(scale=dpi / 72) + try: + bitmap.to_pil().save(image_path) + finally: + bitmap.close() + page.close() + result = subprocess.run( + ["tesseract", str(image_path), "stdout", "-l", "eng", "--psm", str(page_segmentation)], + check=True, + capture_output=True, + text=True, + timeout=180, + ) + pages.append(result.stdout) + return "\n\n".join(pages) + + +def _firecrawl_text(document: dict[str, Any], *, mode: str, api_key: str) -> tuple[str, dict[str, Any]]: + """Check remote input identity before sending a public fixture URL for parsing.""" + url = document["public_fixture_url"] + with httpx.Client(timeout=240, follow_redirects=True) as client: + source = client.get(url) + source.raise_for_status() + actual_hash = hashlib.sha256(source.content).hexdigest() + if actual_hash != document["sha256"]: + raise ValueError(f"Remote fixture differs from the local PDF: {document['filename']}") + response = client.post( + "https://api.firecrawl.dev/v2/scrape", + headers={"Authorization": f"Bearer {api_key}"}, + json={ + "url": url, + "formats": ["markdown"], + "parsers": [{"type": "pdf", "mode": mode, "maxPages": document["pages"], "pages": True}], + "timeout": 180000, + "maxAge": 0, + "onlyMainContent": False, + }, + ) + response.raise_for_status() + payload = response.json() + if not payload.get("success"): + raise RuntimeError("Firecrawl returned an unsuccessful scrape; inspect the API dashboard") + data = payload.get("data", {}) + markdown = data.get("markdown", "") + if not markdown.strip(): + raise RuntimeError("Firecrawl returned no markdown; the PDF comparison is incomplete") + return markdown, {"mode": mode, "metadata": data.get("metadata"), "pages": data.get("pages")} + + +def main() -> int: + """Run selected backends and retain raw outputs, provenance and anchor checks.""" + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("--backend", choices=_BACKENDS, nargs="+", default=["native", "tesseract"]) + parser.add_argument("--output-dir", type=Path, required=True) + parser.add_argument("--dpi", type=int, default=300) + parser.add_argument("--psm", type=int, choices=range(14), default=3) + args = parser.parse_args() + if args.dpi < 72: + parser.error("--dpi must be at least 72") + + manifest = json.loads(_MANIFEST.read_text(encoding="utf-8")) + args.output_dir.mkdir(parents=True, exist_ok=True) + results: list[dict[str, Any]] = [] + report: dict[str, Any] = { + "created_at": datetime.now(UTC).isoformat(), + "platform": platform.platform(), + "machine": platform.machine(), + "python": platform.python_version(), + "methodology": manifest["methodology"], + "backends": args.backend, + "raster_dpi": args.dpi, + "tesseract_psm": args.psm, + "results": results, + } + report_path = args.output_dir / "results.json" + failed = False + for document in manifest["documents"]: + path = _FIXTURES / document["filename"] + actual_hash = hashlib.sha256(path.read_bytes()).hexdigest() + if actual_hash != document["sha256"]: + raise ValueError(f"Local fixture changed; review and update the manifest: {path}") + for backend in args.backend: + result: dict[str, Any] = { + "document": document["filename"], + "sha256": actual_hash, + "pages": document["pages"], + "backend": backend, + } + results.append(result) + api_key = os.environ.get("FIRECRAWL_API_KEY") + if backend.startswith("firecrawl-") and not api_key: + result.update(status="skipped", reason="FIRECRAWL_API_KEY is not configured") + elif backend == "tesseract" and shutil.which("tesseract") is None: + result.update(status="skipped", reason="Install the Tesseract executable and English language data") + elif backend in {"native", "tesseract"} and not HAS_PDFIUM: + result.update(status="skipped", reason="Install the raster dependencies with uv sync --group ocr") + elif backend == "docling" and not HAS_AMFV_PDF: + result.update(status="skipped", reason="The AMFV Docling exporter requires the WHO scraper PR") + elif backend == "docling" and not HAS_AMFV_DOCLING: + result.update(status="skipped", reason="Install WHO's PDF dependencies with uv sync --group pdf") + else: + start = time.perf_counter() + try: + if backend == "native": + text = _native_text(path) + result["version"] = version("pypdfium2") + elif backend == "tesseract": + text = _tesseract_text(path, dpi=args.dpi, page_segmentation=args.psm) + result["version"] = subprocess.check_output(["tesseract", "--version"], text=True).splitlines()[ + 0 + ] + elif backend == "docling": + text = pdf_to_markdown(path.read_bytes(), running_header=document["title"], name=path.name) + result["version"] = version("docling") + result["ocr"] = False + result["export"] = "AMFV markdown export, including footnotes" + result["accelerator"] = "Docling default auto" + result["table_structure"] = True + else: + text, details = _firecrawl_text( + document, mode=backend.removeprefix("firecrawl-"), api_key=api_key + ) + details_path = args.output_dir / f"{path.stem}.{backend}.response.json" + details_path.write_text(json.dumps(details, indent=2) + "\n", encoding="utf-8") + result["version"] = "hosted API v2; model version not pinned" + if not text.strip(): + raise ValueError("No readable output") + output_name = f"{path.stem}.{backend}.md" + (args.output_dir / output_name).write_text(text, encoding="utf-8") + normalized = _normalized(text) + checks = [ + { + "name": anchor["name"], + "text": anchor["text"], + "present": _normalized(anchor["text"]) in normalized, + } + for anchor in document["anchors"] + ] + result.update( + status="ok", + seconds=round(time.perf_counter() - start, 3), + characters=len(text), + output=output_name, + anchors=checks, + anchors_found=sum(check["present"] for check in checks), + anchors_total=len(checks), + ) + except Exception as exc: # noqa: BLE001 - keep completed backend evidence when another backend fails + failed = True + result.update(status="error", reason=str(exc), seconds=round(time.perf_counter() - start, 3)) + report_path.write_text(json.dumps(report, indent=2, ensure_ascii=False) + "\n", encoding="utf-8") + print(f"{path.name}: {backend}: {result['status']}", flush=True) + return int(failed) + + +__all__ = ["main"] + +if __name__ == "__main__": + raise SystemExit(main()) diff --git a/datasets/benchmarks/pdf_ocr_manifest.json b/datasets/benchmarks/pdf_ocr_manifest.json new file mode 100644 index 0000000..8cc74a4 --- /dev/null +++ b/datasets/benchmarks/pdf_ocr_manifest.json @@ -0,0 +1,81 @@ +{ + "methodology": "Two born-digital English WHO excerpts, six pages each. Native PDF text and stock Docling (OCR off) use the text layer; Tesseract receives only 300-DPI page rasters. Anchors were checked against rendered excerpt pages 1-2; checks normalize whitespace and hyphenated line breaks. Anchor presence is a diagnostic, not medical correctness, table alignment, recall, CER or WER. FinePDFs itself is not a backend in this runner.", + "documents": [ + { + "filename": "who_9789240121805_excerpt.pdf", + "title": "Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters: part 2: central venous catheters", + "pages": 6, + "sha256": "4795bf865b6d84dd4606bd94264962eea06b4a03d08188b39fb445e7006f6a66", + "publication_url": "https://www.who.int/publications/i/item/9789240121805", + "public_fixture_url": "https://raw.githubusercontent.com/conscioustahoe/amfv/ca7a330fc2e87022e24e6b1b72130b3ec5002408/datasets/test/fixtures/pdf/who_9789240121805_excerpt.pdf", + "anchors": [ + { + "name": "recommendation population", + "text": "daily in critically-ill adults, adolescents and children" + }, + { + "name": "recommendation certainty", + "text": "Conditional recommendation, low certainty of evidence" + }, + { + "name": "neonatal exclusion", + "text": "but not neonates" + }, + { + "name": "concentration range", + "text": "2% to 4%" + }, + { + "name": "risk estimate", + "text": "0.67; 95% confidence interval" + }, + { + "name": "confidence interval", + "text": "0.45 to 0.99" + }, + { + "name": "chemical burn rationale", + "text": "due to the risk of chemical burns" + } + ] + }, + { + "filename": "who_9789240124233_excerpt.pdf", + "title": "Consolidated HIV guidelines: service delivery", + "pages": 6, + "sha256": "ab24426012f0add3049b2e298e0ac73e702d7aa903fee5a24341fb3f8289bd0d", + "publication_url": "https://www.who.int/publications/i/item/9789240124233", + "public_fixture_url": "https://raw.githubusercontent.com/conscioustahoe/amfv/ca7a330fc2e87022e24e6b1b72130b3ec5002408/datasets/test/fixtures/pdf/who_9789240124233_excerpt.pdf", + "anchors": [ + { + "name": "conditional outside-facility initiation", + "text": "Antiretroviral therapy (ART) initiation may be offered outside the health facility" + }, + { + "name": "rapid initiation population", + "text": "Rapid ART initiation should be offered to all people living with HIV" + }, + { + "name": "adult/child certainty distinction", + "text": "high-certainty evidence for adults, low-certainty evidence for children" + }, + { + "name": "clinical visit interval", + "text": "clinical visits every 3-6 months" + }, + { + "name": "nonphysician task sharing", + "text": "Trained non-physician clinicians, midwives and nurses can initiate first-line ART" + }, + { + "name": "integrated diabetes care", + "text": "Diabetes and hypertension care should be integrated with HIV services" + }, + { + "name": "date-column footnote", + "text": "Date the recommendation/good practice statement was formulated" + } + ] + } + ] +} diff --git a/datasets/benchmarks/results/.gitattributes b/datasets/benchmarks/results/.gitattributes new file mode 100644 index 0000000..e8122e5 --- /dev/null +++ b/datasets/benchmarks/results/.gitattributes @@ -0,0 +1,2 @@ +# Preserve raw backend text, including native CRLF and trailing spaces. +**/*.md -text whitespace=-blank-at-eol,-blank-at-eof diff --git a/datasets/benchmarks/results/.gitignore b/datasets/benchmarks/results/.gitignore new file mode 100644 index 0000000..bf65807 --- /dev/null +++ b/datasets/benchmarks/results/.gitignore @@ -0,0 +1,5 @@ +# Keep reviewed evidence; ignore new local experiment runs by default. +/* +!/.gitignore +!/.gitattributes +!/2026-10-06/ diff --git a/datasets/benchmarks/results/2026-10-06/footnotes-preserved/results.json b/datasets/benchmarks/results/2026-10-06/footnotes-preserved/results.json new file mode 100644 index 0000000..4fc4e92 --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/footnotes-preserved/results.json @@ -0,0 +1,122 @@ +{ + "created_at": "2026-10-06T22:16:27.717736+00:00", + "platform": "macOS-26.2-arm64-arm-64bit-Mach-O", + "machine": "arm64", + "python": "3.13.15", + "methodology": "Two born-digital English WHO excerpts, six pages each. Native PDF text and stock Docling (OCR off) use the text layer; Tesseract receives only 300-DPI page rasters. Anchors were checked against rendered excerpt pages 1-2; checks normalize whitespace and hyphenated line breaks. Anchor presence is a diagnostic, not medical correctness, table alignment, recall, CER or WER. FinePDFs itself is not a backend in this runner.", + "backends": [ + "docling" + ], + "raster_dpi": 300, + "tesseract_psm": 3, + "results": [ + { + "document": "who_9789240121805_excerpt.pdf", + "sha256": "4795bf865b6d84dd4606bd94264962eea06b4a03d08188b39fb445e7006f6a66", + "pages": 6, + "backend": "docling", + "version": "2.134.0", + "ocr": false, + "export": "AMFV markdown export, including footnotes", + "accelerator": "Docling default auto", + "table_structure": true, + "status": "ok", + "seconds": 11.282, + "characters": 16672, + "output": "who_9789240121805_excerpt.docling.md", + "anchors": [ + { + "name": "recommendation population", + "text": "daily in critically-ill adults, adolescents and children", + "present": true + }, + { + "name": "recommendation certainty", + "text": "Conditional recommendation, low certainty of evidence", + "present": true + }, + { + "name": "neonatal exclusion", + "text": "but not neonates", + "present": true + }, + { + "name": "concentration range", + "text": "2% to 4%", + "present": true + }, + { + "name": "risk estimate", + "text": "0.67; 95% confidence interval", + "present": true + }, + { + "name": "confidence interval", + "text": "0.45 to 0.99", + "present": true + }, + { + "name": "chemical burn rationale", + "text": "due to the risk of chemical burns", + "present": true + } + ], + "anchors_found": 7, + "anchors_total": 7 + }, + { + "document": "who_9789240124233_excerpt.pdf", + "sha256": "ab24426012f0add3049b2e298e0ac73e702d7aa903fee5a24341fb3f8289bd0d", + "pages": 6, + "backend": "docling", + "version": "2.134.0", + "ocr": false, + "export": "AMFV markdown export, including footnotes", + "accelerator": "Docling default auto", + "table_structure": true, + "status": "ok", + "seconds": 5.035, + "characters": 23719, + "output": "who_9789240124233_excerpt.docling.md", + "anchors": [ + { + "name": "conditional outside-facility initiation", + "text": "Antiretroviral therapy (ART) initiation may be offered outside the health facility", + "present": true + }, + { + "name": "rapid initiation population", + "text": "Rapid ART initiation should be offered to all people living with HIV", + "present": true + }, + { + "name": "adult/child certainty distinction", + "text": "high-certainty evidence for adults, low-certainty evidence for children", + "present": true + }, + { + "name": "clinical visit interval", + "text": "clinical visits every 3-6 months", + "present": true + }, + { + "name": "nonphysician task sharing", + "text": "Trained non-physician clinicians, midwives and nurses can initiate first-line ART", + "present": true + }, + { + "name": "integrated diabetes care", + "text": "Diabetes and hypertension care should be integrated with HIV services", + "present": true + }, + { + "name": "date-column footnote", + "text": "Date the recommendation/good practice statement was formulated", + "present": true + } + ], + "anchors_found": 7, + "anchors_total": 7 + } + ] +} diff --git a/datasets/benchmarks/results/2026-10-06/footnotes-preserved/who_9789240121805_excerpt.docling.md b/datasets/benchmarks/results/2026-10-06/footnotes-preserved/who_9789240121805_excerpt.docling.md new file mode 100644 index 0000000..bd17e37 --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/footnotes-preserved/who_9789240121805_excerpt.docling.md @@ -0,0 +1,150 @@ +## 3.2 Pre-insertion + +## 3.2.1 Chlorhexidine-containing body wash + +- Recommendation 1. WHO suggests using a chlorhexidine-containing body wash (liquid or impregnated wipes) daily in critically-ill adults, adolescents and children. (Conditional recommendation, low certainly of evidence) + +## Remarks + +- The evidence did not differentiate between liquid and impregnated wipes and evaluated chlorhexidine concentrations ranging from 2% to 4%. It was noted that most of the studies provided evidence on the use of chlorhexidine, hence a decision was made to restrict the recommendation to chlorhexidine only and to adults, adolescents and children, but not neonates (due to the risk of chemical burns) (32). +- The GDG noted that the use of this intervention would be particularly beneficial in settings with high baseline CLABSI rates. However, the GDG highlighted the need to prioritize resources at the local level based on other competing needs in the health care facility. The GDG also expressed concerns on the potential development of resistance to chlorhexidine and on waste generation when using wipes. +- Chlorhexidine products should be used with caution in patients with burns and according to the manufacturer's instructions in order to ensure they are safe for use . + +## Rationale for the recommendation + +Overall, the GDG judged both the desirable and undesirable effects to be small, but with a balance of effects probably favouring the intervention, with a low certainty of evidence. Based on the evidence provided, the GDG opted for a conditional recommendation in favour of the intervention, in particular in the context of high CLABSI rates at baseline, where the overall benefit may be more noticeable. + +## Summary of the evidence + +The Systematic Reviews Group identified 31 studies (10 randomized controlled trials [RCTs] (33-42) and 21 non-randomized studies of interventions [NRSIs](43-63)) comparing daily antiseptic body washing with chlorhexidine to soap and water bathing in adults, adolescents and children. + +- There was low certainty of evidence that daily antiseptic body washing may result in fewer CLABSI than no routine antiseptic body washing (risk ratio [RR]: 0.67; 95% confidence interval [CI]: 0.45 to 0.99; 10 RCTs (33-42)). +- There was low certainty of evidence that routine antiseptic body washing may result in lower catheter-related mortality than no routine antiseptic body washing (RR: 0.20; 95% CI: 0.01 to 4.13; one RCT (41)). +- There was low certainty of evidence that routine antiseptic body washing may result in little to no difference in all-cause mortality when compared with no routine antiseptic body washing (odds ratio [OR]: 0.90; 95% CI: 0.76 to 1.07; five RCTs (33, 37-40)). +- There was very low certainty of evidence that routine antiseptic body washing may result in little to no difference in sepsis when compared with no routine antiseptic body washing (RR: 1.92; 95% CI: 0.66 to 5.60; one RCT (33)). + +- Subgroup analyses revealed no difference in the effect for CLABSI based on age (adults versus children) or setting (ICU versus general ward). + +## Evidence-to-decision + +A potentially reduced equity was noted due to the different availability of products around the world and their moderate costs, in particular when considering LMICs, but the intervention was considered probably cost-effective. No issues were identified in acceptability and feasibility. + +## Implementation considerations + +- The recommendation refers to using a chlorhexidine-containing body wash (liquid or impregnated wipes) daily and not immediately prior to line insertion. The GDG acknowledges that there are variations in the available protocols on body washing and directs health and care workers to follow their national and local policies. +- The recommendation refers to all critically-ill patients that require a CVC and not limited to the ICU. +- It is advisable to engage nursing and support staff as key stakeholders in the decision (that is, time required for bathing). +- It is also advisable to engage patients and their families, considering the potential adverse effects on the skin related to exposure to chlorhexidine, including asking for any known intolerance and rare allergic reactions to chlorhexidine. +- WASH services and standards (64, 65) are essential to enable critical IPC practices, including this recommendation on body wash. + +## Research needs + +- There is a significant variation in available protocols on how to perform patient bathing. The GDG identified this as an area of research interest to determine the most effective protocol and product to be used (that is, liquid wash versus wipes or other means, rinsing versus no rinsing, which solution to use). +- The GDG also identified the need for additional research on the use of antiseptic products other than chlorhexidine (that is, octenidine and others) using multicentre RCTs or other types of trials (that is, adaptive) and focusing on specific populations (such as haematology patients or the paediatric ICU). +- AMR and other wider implications (that is, impact on skin microbiome, cross-resistance to antibiotics and on the epidemiology of other infections caused by gram-positive and gramnegative bacteria) associated with the use of chlorhexidine and other body wash products were identified as important areas for research. +- Cost-benefit research studies including baseline surveillance data to define potential thresholds for universal use were also considered as other areas for further research. + +## 3.3 Insertion + +## 3.3.1 Insertion pack/kit + +- Good practice statement 6. WHO recommends that clinicians use a standardized insertion pack/kit when placing a CVC in adults, adolescents, children and neonates (Good practice statement) + +## 3.3.2 CVC insertion by an insertion team + +- WHO makes no recommendation regarding the use of a dedicated CVC insertion teams in adults, adolescents, children, or neonates, as the evidence remains inconclusive. + +## Remarks + +- A CVC insertion team is defined as consisting of individuals specifically trained and competent in CVC insertion who are dedicated to perform the insertion of CVCs. +- The search criteria used in the systematic review were inclusive as possible, including terms such as "insertion team", "vascular access team", intravenous therapy team". +- Establishing a dedicated CVC insertion team is an intervention that is different from ensuring specific training for all those tasked with CVC insertion, which is covered by good practice statements 1 to 3 in these guidelines. +- Consideration should also be given to specific populations (i.e., neonates, children, or patients with known difficult intravenous access) as the potential benefit in preventing CLABSI may be confined to these groups. +- The GDG members acknowledged the difficulties in designing a RCT to evaluate this topic and the very limited available evidence. + +## Rationale for the decision + +Based on the evidence provided (two uncontrolled before-after studies and no RCTs), the GDG considered both the desirable and undesirable effects to be small. The GDG judged that the crucial factor was the adequate training of the clinician inserting the CVC, rather than having a dedicated team itself. For this reason, the GDG considered that the balance of effects did not favour either the intervention (a formal insertion team) or the comparison (insertion by a clinician who is not part of a formal insertion team), provided that the clinician has undergone appropriate training and demonstrated competency in CVC insertion. Another concern expressed by the GDG was the potential loss of knowledge and skills in CVC insertion among non-team members (doctors and nurses), such that their insertion competency could decline. This may be especially impactful in emergency situations where they are required to urgently insert a CVC without the assistance of a formal insertion team, with potential implications affecting the wider hospital capacity and patient care. Overall, given the very low certainty of the evidence, no recommendation could be made. + +## Summary of the evidence + +- The Systematic Reviews Expert Group did not identify any RCTs on the topic, but only two uncontrolled before-after studies (66, 67) evaluating the use of a dedicated CVC insertion (for non-haemodialysis) team compared with not using a dedicated CVC insertion team. The risk of bias in both studies was rated as high due to their uncontrolled design (no formal assessment). +- There was very low certainty of evidence from one study in children and adults reporting a 58% reduction in the number of CLABSIs after the introduction of a highly specialized vascular access team (67). Similarly, there was very low certainty of evidence from another study in neonates reporting that overall CLABSI rates decreased by 65% after implementation of the line team (66). +- There were no studies identified reporting on catheter-related mortality, all-cause mortality, sepsis pneumothorax, local site infections, bleeding events that require action, cardiac tamponade, deep vein thrombosis or venous stenosis. + +## Evidence-to-decision + +Increased moderate costs (due to the resources required to maintain a dedicated team), the probably reduced equity, variable acceptability by health care workers and patients (hospital management might not be in favour due to the additional costs, but patient satisfaction is likely to be improved) and low feasibility (as having a dedicated team is not always possible due to staff shortages) were all considered, resulting in a decision to make no recommendation.. The overall cost-effectiveness judgement did not favour either the intervention or the comparison. + +## Research needs + +- Further research is needed to compare the potential cost-benefit of using a formal CVC insertion team versus an educational/behavioural programme for clinicians inserting CVC, including both ICU and non-ICU settings. + +## 3.3.3 Sterile technique + +- Good practice statement 7. WHO recommends using a sterile technique for the insertion of all CVCs in adults, adolescents, children and neonates. (Good practice statement) +- Good practice statement 8. WHO recommends that clinicians use sterile single-use gloves compared to non-sterile gloves when inserting a CVC in adults, adolescents, children and neonates. + +(Good practice statement) + +- Recommendation 2. WHO suggests using a sterile technique involving maximal sterile barrier precautions over a sterile technique not involving maximal sterile +- barrier precautions at the time of CVC insertion in adults, adolescents, children and neonates. + +(Conditional recommendation, very low certainty of evidence) + +## Remarks + +- Maximal sterile barrier precautions: sterile barrier precautions that include the use of a cap, mask, sterile gown, sterile gloves and a sterile full body drape (11). +- Sterile technique is defined as a set of specific practices and procedures performed to make equipment and areas free from all microorganisms and to maintain that sterility (11). +- GDG members acknowledged that the above definition of sterile technique as comparator also seems to be generic and it may not provide enough practical details. + +## Rationale for the recommendation + +There was an extensive discussion during which a number of GDG members expressed the opinion that a strong recommendation would have been more suitable; others suggested a conversion into a good practice statement. However, it was decided to assess the available evidence and to proceed with the evidence-to-decision framework. Following the assessment, the GDG judged the desirable and undesirable effects to be small and trivial, respectively, but with a balance of effects probably favouring the intervention, despite the very low certainty of evidence. In view of the data presented, the GDG members opted for a conditional recommendation in favour of maximal sterile barrier precautions, but acknowledging that the current evidence is weak and more studies are needed to assess if less barriers may be sufficient. + +## Summary of the evidence + +The Systematic Reviews Expert Group identified eight studies (two RCTs (68, 69) and six NRSIs (70-75)) comparing maximal sterile barrier precautions in adults, children and adolescents, and neonates. + +- There was very low certainty of evidence that maximal sterile barrier precautions may result in fewer CLABSIs than a lesser standard of precautions (RR: 0.47, 95% CI: 0.10 to 2.22; two RCTs (68, 69)). +- Two non-randomized studies also reported very low certainty of evidence that maximal sterile barrier precautions may result in fewer CLABSIs than a lesser standard of precautions (RR: 0.21; 95% CI: 0.05 to 0.94; two non-RCTs (70, 71)) +- There was very low certainty of evidence that maximal sterile barrier precautions may result in little to no difference in local insertion site infections when compared with a lesser standard of precautions (RR: 1.01; 95% CI: 0.06 to 16.03; one RCT (68)). +- There were no studies identified that reported on catheter-related mortality, all-cause mortality, sepsis and other complications. + +## Evidence-to-decision + +The intervention may increase costs due to the additional equipment required and it is probably accepted and feasible as it is already common practice. However, patients may have more discomfort with full body drapes. In addition, the GDG judged equity to be probably increased. + +## Implementation considerations + +- Ensure adequate training, understanding and competency in the application of maximal sterile barriers precautions. +- The GDG discussed the need for a dedicated procedure room for line insertion and noted that this is not practical for critically-ill patients and may be an issue in some health care facilities, particularly in LMICs. Furthermore, it is not part of the current definition of maximal sterile barrier precautions. + +## Research needs + +- More research studies are needed to assess if lesser sterile precautions are needed, with a trial design and platform (for example, adaptive (76, 77)) that would allow to clearly distinguish the impact of the different components within maximal sterile barrier precautions (that is, cap, mask, full body drape, the environment where the line is placed) and allow to answer multiple questions simultaneously. + +## 3.3.4 Skin disinfection + +- Good practice statement 9. WHO recommends that skin disinfection should always be used prior to the insertion of all CVCs in adults, adolescents, children and neonates. (Good practice statement) +- Recommendation 3. WHO suggests using an alcohol-based formulation over an +- aqueous-based antiseptic formulation for skin disinfection before CVC insertion in adults, adolescents, children, and neonates. (Conditional recommendation, low certainty of evidence) + +## Remarks + +- Alcohol-based antiseptics should be the preferred choice for skin disinfection before CVC insertion. +- It was acknowledged that alcohol-based formulations can cause skin irritation, dryness and a risk of pain in case of laceration and these potential side-effects should be taken into consideration when using these products. Some of these side-effects may be caused by either the alcohol or the antiseptic component. +- Alcohol-based antiseptics can cause skin lesions in preterm infants without intact skin. + +## Rationale for the recommendation + +The GDG considered alcohol-based formulations as the preferred option for skin disinfection over aqueous formulations, with a conditional recommendation due to the low certainty of evidence. The GDG considered the desirable effects of alcohol-based formulations for skin disinfection to be moderate and the undesirable effects trivial, with a balance of effects probably favouring the intervention. + +## Summary of the evidence + +The Systematic Reviews Expert Group found data from five RCTs (78-82). + +- There was low certainty of evidence that alcohol-based solutions may result in fewer CLABSIs than aqueous-based solutions for skin disinfection (RR: 0.59; 95% CI: 0.36 to 0.98; five RCTs (78-82)). +- There was very low certainty of evidence that alcohol-based solutions may result in less allcause mortality than aqueous-based solutions for skin disinfection (RR: 0.79; 95% CI: 0.48 to 1.30; two RCTs (78, 79)). +- There was very low certainty of evidence that alcohol-based solutions may result in fewer sepsis episodes when compared with aqueous-based solutions for skin disinfection (RR: 0.68; 95% CI: 0.35 to 1.30; one RCT (78)). +- There were no studies identified that reported on catheter-related mortality, pneumothorax, local insertion site infections and other complications. \ No newline at end of file diff --git a/datasets/benchmarks/results/2026-10-06/footnotes-preserved/who_9789240124233_excerpt.docling.md b/datasets/benchmarks/results/2026-10-06/footnotes-preserved/who_9789240124233_excerpt.docling.md new file mode 100644 index 0000000..5866d32 --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/footnotes-preserved/who_9789240124233_excerpt.docling.md @@ -0,0 +1,145 @@ +Table 1: HIV service delivery recommendations + +| Date* | Recommendations | Status | +|-------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------| +| Linkage and ART initiation | Linkage and ART initiation | Linkage and ART initiation | +| 2016 | Following an HIV diagnosis, a package of support interventions should be offered to ensure timely linkage to care for all people living with HIV (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2021 | Health systems should invest in people-centred practices and communication, including ongoing training, mentoring, supportive supervision and monitoring of health workers, to improve the relationships between clients and health-care providers (Good practice statement). | Revalidated | +| 2019 | ART initiation should follow the overarching principles of providing people-centred care. People-centred care should be focused and organized around the health needs, preferences and expectations of people and communities, upholding individual dignity and respect, especially for vulnerable populations (Good practice statement). | Revalidated | +| 2021 | Antiretroviral therapy (ART) initiation may be offered outside the health facility (Conditional recommendation, low- to moderate-certainty evidence). | Revalidated | +| 2017 | Rapid ART initiation should be offered to all people living with HIV following a confirmed HIV diagnosis and clinical assessment (Strong recommendation: high- certainty evidence for adults, low-certainty evidence for children) | Revalidated | +| 2021 | The offer of same-day ART initiation should include approaches to improve uptake, treatment adherence and retention such as tailored client education, counselling and support (Good practice statement). | Revalidated | +| Clinical visits and medication pick-up | Clinical visits and medication pick-up | Clinical visits and medication pick-up | +| 2021 | People established on ART should be offered clinical visits every 3-6 months, preferably every six months if feasible (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2021 | People established on ART should be offered refills of ART lasting 3-6 months, preferably six months if feasible (Strong recommendation, moderate-certainty evidence). | Revalidated | +| Adherence and engagement in care | Adherence and engagement in care | Adherence and engagement in care | +| 2025 | Adherence support interventions should be provided to people on ART (Strong recommendation, moderate-certainty evidence). | Updated | +| 2021 | Viral load for treatment monitoring should be complemented with non-judgemental, tailored approaches to assessing adherence (Good practice statement). | Revalidated | +| 2021 | HIV programmes should implement interventions to trace people who have disengaged from care and provide support for re-engagement (Strong recommendation, low-certainty evidence). | Revalidated | +| Services for adolescents and young people | Services for adolescents and young people | Services for adolescents and young people | +| 2021 | Psychosocial interventions should be provided to all adolescents and young adults living with HIV (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2013 | Adolescent-friendly services should be implemented in HIV services to ensure engagement and improved outcomes (Strong recommendation, moderate-certainty evidence). | Revalidated | + +Table 1: HIV service delivery recommendations (cont'd) + +| Date* | Recommendations | Status | +|-------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------| +| Services for adolescents and young people | Services for adolescents and young people | Services for adolescents and young people | +| 2013 | Adolescents should be counselled about the potential benefits and risks of disclosing their HIV status to others and empowered and supported to determine whether, when, how and to whom to disclose (Conditional recommendation, very-low -certainty evidence). | Revalidated | +| Task sharing | Task sharing | Task sharing | +| 2016 | Trained non-physician clinicians, midwives and nurses can initiate first-line ART (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2016 | Trained non-physician clinicians, midwives and nurses can maintain ART (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2016 | Trained and supervised community health workers can dispense ART between regular clinical visits (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2016 | Trained and supervised lay health-care providers can distribute ART (Strong recommendation, low-certainty evidence). | Revalidated | +| 2021 | Task sharing of specimen collection and point-of-care testing with non-laboratory personnel should be implemented when professional staffing capacity is limited (Strong recommendation, moderate-certainty evidence). | Revalidated | +| Service Delivery models and integration | Service Delivery models and integration | Service Delivery models and integration | +| 2013 | Decentralization of ART care should be considered as a way to increase access and improve retention in care (Strong recommendation, moderate-to-low-certainty evidence). | Revalidated | +| 2013 | In generalized epidemic settings, ART should be initiated and maintained in pregnant and postpartum women and in infants at maternal and child health-care settings, with linkage and referral to ongoing HIV care and ART, if appropriate (Strong recommendation, low-certainty evidence). | Revalidated | +| 2021 | Sexual and reproductive health services, including contraception, may be integrated with HIV services (Conditional recommendation, very-low-certainty evidence). | Revalidated | +| 2013 | In settings with a high burden of HIV and tuberculosis (TB), ART should be initiated in TB treatment settings, with linkage to ongoing HIV care and ART (Strong recommendation, very-low-certainty evidence). | Revalidated | +| 2013 | In settings with a high burden of HIV and TB, TB treatment may be provided for people living with HIV in HIV care settings if they have been diagnosed with TB (Strong recommendation, very-low-certainty evidence). | Revalidated | +| 2025 | Diabetes and hypertension care should be integrated with HIV services (Strong recommendation, moderate-to-very-low-certainty evidence). | Updated | +| 2025 | Mental health care for depression, anxiety and alcohol use disorders should be integrated with HIV services (Strong recommendation, moderate-to-very-low-certainty evidence). | New | + +## 1. Introduction + +HIV remains a major public health challenge and a leading cause of illness and death in many low- and middleincome countries, despite the wide availability of effective antiretroviral therapy and proven prevention tools. + +In many settings, evidence-informed interventions for HIV prevention, testing, treatment and long-term care are not fully implemented, delivered inconsistently or reach people too late, contributing to ongoing transmission, advanced disease and avoidable mortality. Persistent health system constraints - including workforce shortages, weak supply chains and limited integration with related services such as those for noncommunicable diseases and mental health care - further hinder effective service delivery. + +These HIV service delivery guidelines provide updated and consolidated recommendations to support countries, programme managers, health-care workers, communities and other stakeholders to organize services that are people-centred, integrated, equitable and grounded in the best available evidence. The guidelines aim to improve access to and continuity of high-quality HIV services, reduce inequities in outcomes and accelerate progress towards ending AIDS as a public health threat. + +## 2. Linkage and ART initiation + +## 2.1 Linkage from HIV testing to enrolment in care + +## Recommendation + +Following an HIV diagnosis, a package of support interventions should be offered to ensure timely linkage to care for all people living with HIV. + +(Strong recommendation, moderate-certainty evidence) + +The following interventions have demonstrated benefit in improving linkage to care following an HIV diagnosis: + +- streamlined interventions to reduce time between diagnosis and engagement in care, including (i) enhanced linkage with case management; (ii) support for HIV disclosure; (iii) tracing; (iv) training staff to provide multiple services; and (v) streamlined services (moderate-certainty evidence); +- peer support a and navigation approaches for linkage (moderate-certainty evidence); and +- quality improvement approaches using data to improve linkage (low-certainty evidence). + +a Includes peer counselling. + +## Good practice statements + +ART initiation should follow the overarching principles of providing people-centred care. People-centred care is care that is focused and organized around the health needs, preferences and expectations of people and communities, upholding individual dignity and respect, especially for vulnerable populations. It should promote the engagement and support of people and families to play an active role in their own care through informed decision-making. + +All people newly diagnosed with HIV should be retested to verify their HIV status before starting ART, using the same testing strategy and algorithm as the initial test. To minimize the risk of misdiagnosis, this approach should be maintained in settings in which rapid ART initiation is being implemented. + +Revalidated + + + +## Background + +Linkage to HIV prevention, treatment and care after testing remains a challenge, with disproportionate gaps for key populations, men, young people and previously diagnosed people living with HIV (not receiving ART, disengaged or lost to care). + +HIV testing service providers are primarily responsible for linkage to care. People testing HIV positive should be assessed for advanced HIV disease using CD4 cell count testing and offered ART, and those testing negative should be linked to prevention services. + +## Rationale + +Evidence-informed interventions to support linkage can be classified into three approaches (2): + +- streamlined services via enhanced linkage with case management, HIV disclosure support, client tracing for non-engagers and staff training for multiple services; +- peer support and navigation interventions, including home visits, system navigation and enhanced counselling; and +- quality improvement interventions enhance linkage to care (3). + +## Implementation considerations + +- Interventions may vary based on the context, including health-care delivery systems, geography and target population. +- A combination of interventions may be needed to improve linkage for specific underserved groups, especially key populations and men. +- Barriers to linkage to HIV care and ART initiation should be assessed and addressed, especially for populations facing multiple structural and individual barriers. +- Linkage for children can be improved using point-of-care early infant diagnosis, SMS/GSM/GPRS printers to accelerate access to laboratory results and family-centred models delivering care to mother, baby, and partner at the same point. +- Health facilities should have capacity to provide services for key populations. Community- and peer-based approaches can be especially effective in restrictive legal settings. + +WHO guidelines on HIV testing services (4) provide information on post-test services and linkage to care. + +## 2.2 People-centred care + +## Good practice statement + +Health systems should invest in people-centred practices and communication, including ongoing training, mentoring, supportive supervision and monitoring of health workers, to improve the relationships between clients and health-care providers. + +HIV programmes should: + +- provide people-centred care that is focused and organized around the health needs, preferences and expectations of people and communities, upholding individual dignity and respect, especially for underserved populations, and engage and support people and families to play an active role in their own care by informed decision-making; +- offer safe, acceptable and appropriate clinical and non-clinical services in a timely fashion, aiming to reduce morbidity and mortality associated with HIV infection and to improve health outcomes and quality of life in general; and +- promote the efficient and effective use of resources. + + + +## Background + +People-centred health services are an approach to care that consciously adopts the perspectives of individuals, families and communities and sees them as participants and beneficiaries of trusted health systems that respond to their needs and preferences in humane and holistic ways. This approach acknowledges the experiences and perspectives of health-care providers that may enable or prevent the delivery of high-quality people-centred care. + +## Rationale + +- A systematic review (5) identified several interventions to enhance people-centred care for people living with HIV. Overall, studies reported beneficial effects of these approaches across the HIV cascade outcomes, including improved ART uptake, adherence and viral load suppression. +- Studies have shown that people are willing to travel longer distances to be seen by a health-care provider with a respectful and caring attitude, and negative health-care worker attitudes contribute to loss to care and poor programme outcomes (6). + +Interventions found to improve relationships between clients and health-care providers can be classified into the following approaches: + +- providing friendly and welcoming services: training providers to make general HIV services more welcoming, including for clients who have disengaged from care; +- conducting sensitization training for clinical and non-clinical health-care providers to improve care for key populations both at the primary care and community levels; such training should include issues related to stigma and discrimination; +- offering individualized adherence counselling and person-centred communication such as shared decisionmaking and planning for ART initiation and adherence; +- facilitating client education in empowerment and communication skills; and +- providing feedback to health-care workers on client concerns and evaluation of service quality such as community score cards and client feedback mechanisms combined with quality improvement exercises. + +## Implementation considerations + +- Build capacity for ongoing training, mentoring and supportive supervision on people-centred communication, stigma reduction and welcoming services. +- Integrate sensitization training for all staff (clinical and non-clinical) on respectful attitudes, stigma and discrimination, re-engagement of clients who have disengaged, and support for women and other vulnerable groups experiencing violence or sexual health concerns. +- Establish routine client feedback mechanisms (such as satisfaction questionnaires, community score cards and suggestion boxes) and use the results in facility quality-improvement meetings. +- Involve clients and community representatives in designing and reviewing services. +- Review workload, staffing and client flow to allow time for person-centred interactions. + +## PDF footnotes + +- Page 2: * Date the recommendation/good practice statement was formulated \ No newline at end of file diff --git a/datasets/benchmarks/results/2026-10-06/results.json b/datasets/benchmarks/results/2026-10-06/results.json new file mode 100644 index 0000000..7741965 --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/results.json @@ -0,0 +1,352 @@ +{ + "created_at": "2026-10-06T22:09:29.141303+00:00", + "platform": "macOS-26.2-arm64-arm-64bit-Mach-O", + "machine": "arm64", + "python": "3.13.15", + "methodology": "Two born-digital English WHO excerpts, six pages each. Native PDF text and stock Docling (OCR off) use the text layer; Tesseract receives only 300-DPI page rasters. Anchors were checked against rendered excerpt pages 1-2; checks normalize whitespace and hyphenated line breaks. Anchor presence is a diagnostic, not medical correctness, table alignment, recall, CER or WER. FinePDFs itself is not a backend in this runner.", + "backends": [ + "native", + "tesseract", + "docling", + "firecrawl-auto", + "firecrawl-ocr" + ], + "raster_dpi": 300, + "tesseract_psm": 3, + "results": [ + { + "document": "who_9789240121805_excerpt.pdf", + "sha256": "4795bf865b6d84dd4606bd94264962eea06b4a03d08188b39fb445e7006f6a66", + "pages": 6, + "backend": "native", + "version": "5.14.0", + "status": "ok", + "seconds": 0.024, + "characters": 17869, + "output": "who_9789240121805_excerpt.native.md", + "anchors": [ + { + "name": "recommendation population", + "text": "daily in critically-ill adults, adolescents and children", + "present": true + }, + { + "name": "recommendation certainty", + "text": "Conditional recommendation, low certainty of evidence", + "present": true + }, + { + "name": "neonatal exclusion", + "text": "but not neonates", + "present": true + }, + { + "name": "concentration range", + "text": "2% to 4%", + "present": true + }, + { + "name": "risk estimate", + "text": "0.67; 95% confidence interval", + "present": true + }, + { + "name": "confidence interval", + "text": "0.45 to 0.99", + "present": true + }, + { + "name": "chemical burn rationale", + "text": "due to the risk of chemical burns", + "present": true + } + ], + "anchors_found": 7, + "anchors_total": 7 + }, + { + "document": "who_9789240121805_excerpt.pdf", + "sha256": "4795bf865b6d84dd4606bd94264962eea06b4a03d08188b39fb445e7006f6a66", + "pages": 6, + "backend": "tesseract", + "version": "tesseract 5.5.3", + "status": "ok", + "seconds": 17.029, + "characters": 17147, + "output": "who_9789240121805_excerpt.tesseract.md", + "anchors": [ + { + "name": "recommendation population", + "text": "daily in critically-ill adults, adolescents and children", + "present": true + }, + { + "name": "recommendation certainty", + "text": "Conditional recommendation, low certainty of evidence", + "present": true + }, + { + "name": "neonatal exclusion", + "text": "but not neonates", + "present": true + }, + { + "name": "concentration range", + "text": "2% to 4%", + "present": true + }, + { + "name": "risk estimate", + "text": "0.67; 95% confidence interval", + "present": true + }, + { + "name": "confidence interval", + "text": "0.45 to 0.99", + "present": true + }, + { + "name": "chemical burn rationale", + "text": "due to the risk of chemical burns", + "present": true + } + ], + "anchors_found": 7, + "anchors_total": 7 + }, + { + "document": "who_9789240121805_excerpt.pdf", + "sha256": "4795bf865b6d84dd4606bd94264962eea06b4a03d08188b39fb445e7006f6a66", + "pages": 6, + "backend": "docling", + "version": "2.134.0", + "ocr": false, + "status": "ok", + "seconds": 12.787, + "characters": 16672, + "output": "who_9789240121805_excerpt.docling.md", + "anchors": [ + { + "name": "recommendation population", + "text": "daily in critically-ill adults, adolescents and children", + "present": true + }, + { + "name": "recommendation certainty", + "text": "Conditional recommendation, low certainty of evidence", + "present": true + }, + { + "name": "neonatal exclusion", + "text": "but not neonates", + "present": true + }, + { + "name": "concentration range", + "text": "2% to 4%", + "present": true + }, + { + "name": "risk estimate", + "text": "0.67; 95% confidence interval", + "present": true + }, + { + "name": "confidence interval", + "text": "0.45 to 0.99", + "present": true + }, + { + "name": "chemical burn rationale", + "text": "due to the risk of chemical burns", + "present": true + } + ], + "anchors_found": 7, + "anchors_total": 7 + }, + { + "document": "who_9789240121805_excerpt.pdf", + "sha256": "4795bf865b6d84dd4606bd94264962eea06b4a03d08188b39fb445e7006f6a66", + "pages": 6, + "backend": "firecrawl-auto", + "status": "skipped", + "reason": "FIRECRAWL_API_KEY is not configured" + }, + { + "document": "who_9789240121805_excerpt.pdf", + "sha256": "4795bf865b6d84dd4606bd94264962eea06b4a03d08188b39fb445e7006f6a66", + "pages": 6, + "backend": "firecrawl-ocr", + "status": "skipped", + "reason": "FIRECRAWL_API_KEY is not configured" + }, + { + "document": "who_9789240124233_excerpt.pdf", + "sha256": "ab24426012f0add3049b2e298e0ac73e702d7aa903fee5a24341fb3f8289bd0d", + "pages": 6, + "backend": "native", + "version": "5.14.0", + "status": "ok", + "seconds": 0.026, + "characters": 14802, + "output": "who_9789240124233_excerpt.native.md", + "anchors": [ + { + "name": "conditional outside-facility initiation", + "text": "Antiretroviral therapy (ART) initiation may be offered outside the health facility", + "present": true + }, + { + "name": "rapid initiation population", + "text": "Rapid ART initiation should be offered to all people living with HIV", + "present": true + }, + { + "name": "adult/child certainty distinction", + "text": "high-certainty evidence for adults, low-certainty evidence for children", + "present": false + }, + { + "name": "clinical visit interval", + "text": "clinical visits every 3-6 months", + "present": true + }, + { + "name": "nonphysician task sharing", + "text": "Trained non-physician clinicians, midwives and nurses can initiate first-line ART", + "present": true + }, + { + "name": "integrated diabetes care", + "text": "Diabetes and hypertension care should be integrated with HIV services", + "present": true + }, + { + "name": "date-column footnote", + "text": "Date the recommendation/good practice statement was formulated", + "present": true + } + ], + "anchors_found": 6, + "anchors_total": 7 + }, + { + "document": "who_9789240124233_excerpt.pdf", + "sha256": "ab24426012f0add3049b2e298e0ac73e702d7aa903fee5a24341fb3f8289bd0d", + "pages": 6, + "backend": "tesseract", + "version": "tesseract 5.5.3", + "status": "ok", + "seconds": 13.848, + "characters": 14346, + "output": "who_9789240124233_excerpt.tesseract.md", + "anchors": [ + { + "name": "conditional outside-facility initiation", + "text": "Antiretroviral therapy (ART) initiation may be offered outside the health facility", + "present": true + }, + { + "name": "rapid initiation population", + "text": "Rapid ART initiation should be offered to all people living with HIV", + "present": true + }, + { + "name": "adult/child certainty distinction", + "text": "high-certainty evidence for adults, low-certainty evidence for children", + "present": true + }, + { + "name": "clinical visit interval", + "text": "clinical visits every 3-6 months", + "present": true + }, + { + "name": "nonphysician task sharing", + "text": "Trained non-physician clinicians, midwives and nurses can initiate first-line ART", + "present": true + }, + { + "name": "integrated diabetes care", + "text": "Diabetes and hypertension care should be integrated with HIV services", + "present": true + }, + { + "name": "date-column footnote", + "text": "Date the recommendation/good practice statement was formulated", + "present": true + } + ], + "anchors_found": 7, + "anchors_total": 7 + }, + { + "document": "who_9789240124233_excerpt.pdf", + "sha256": "ab24426012f0add3049b2e298e0ac73e702d7aa903fee5a24341fb3f8289bd0d", + "pages": 6, + "backend": "docling", + "version": "2.134.0", + "ocr": false, + "status": "ok", + "seconds": 5.156, + "characters": 23625, + "output": "who_9789240124233_excerpt.docling.md", + "anchors": [ + { + "name": "conditional outside-facility initiation", + "text": "Antiretroviral therapy (ART) initiation may be offered outside the health facility", + "present": true + }, + { + "name": "rapid initiation population", + "text": "Rapid ART initiation should be offered to all people living with HIV", + "present": true + }, + { + "name": "adult/child certainty distinction", + "text": "high-certainty evidence for adults, low-certainty evidence for children", + "present": true + }, + { + "name": "clinical visit interval", + "text": "clinical visits every 3-6 months", + "present": true + }, + { + "name": "nonphysician task sharing", + "text": "Trained non-physician clinicians, midwives and nurses can initiate first-line ART", + "present": true + }, + { + "name": "integrated diabetes care", + "text": "Diabetes and hypertension care should be integrated with HIV services", + "present": true + }, + { + "name": "date-column footnote", + "text": "Date the recommendation/good practice statement was formulated", + "present": false + } + ], + "anchors_found": 6, + "anchors_total": 7 + }, + { + "document": "who_9789240124233_excerpt.pdf", + "sha256": "ab24426012f0add3049b2e298e0ac73e702d7aa903fee5a24341fb3f8289bd0d", + "pages": 6, + "backend": "firecrawl-auto", + "status": "skipped", + "reason": "FIRECRAWL_API_KEY is not configured" + }, + { + "document": "who_9789240124233_excerpt.pdf", + "sha256": "ab24426012f0add3049b2e298e0ac73e702d7aa903fee5a24341fb3f8289bd0d", + "pages": 6, + "backend": "firecrawl-ocr", + "status": "skipped", + "reason": "FIRECRAWL_API_KEY is not configured" + } + ] +} diff --git a/datasets/benchmarks/results/2026-10-06/who_9789240121805_excerpt.docling.md b/datasets/benchmarks/results/2026-10-06/who_9789240121805_excerpt.docling.md new file mode 100644 index 0000000..bd17e37 --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/who_9789240121805_excerpt.docling.md @@ -0,0 +1,150 @@ +## 3.2 Pre-insertion + +## 3.2.1 Chlorhexidine-containing body wash + +- Recommendation 1. WHO suggests using a chlorhexidine-containing body wash (liquid or impregnated wipes) daily in critically-ill adults, adolescents and children. (Conditional recommendation, low certainly of evidence) + +## Remarks + +- The evidence did not differentiate between liquid and impregnated wipes and evaluated chlorhexidine concentrations ranging from 2% to 4%. It was noted that most of the studies provided evidence on the use of chlorhexidine, hence a decision was made to restrict the recommendation to chlorhexidine only and to adults, adolescents and children, but not neonates (due to the risk of chemical burns) (32). +- The GDG noted that the use of this intervention would be particularly beneficial in settings with high baseline CLABSI rates. However, the GDG highlighted the need to prioritize resources at the local level based on other competing needs in the health care facility. The GDG also expressed concerns on the potential development of resistance to chlorhexidine and on waste generation when using wipes. +- Chlorhexidine products should be used with caution in patients with burns and according to the manufacturer's instructions in order to ensure they are safe for use . + +## Rationale for the recommendation + +Overall, the GDG judged both the desirable and undesirable effects to be small, but with a balance of effects probably favouring the intervention, with a low certainty of evidence. Based on the evidence provided, the GDG opted for a conditional recommendation in favour of the intervention, in particular in the context of high CLABSI rates at baseline, where the overall benefit may be more noticeable. + +## Summary of the evidence + +The Systematic Reviews Group identified 31 studies (10 randomized controlled trials [RCTs] (33-42) and 21 non-randomized studies of interventions [NRSIs](43-63)) comparing daily antiseptic body washing with chlorhexidine to soap and water bathing in adults, adolescents and children. + +- There was low certainty of evidence that daily antiseptic body washing may result in fewer CLABSI than no routine antiseptic body washing (risk ratio [RR]: 0.67; 95% confidence interval [CI]: 0.45 to 0.99; 10 RCTs (33-42)). +- There was low certainty of evidence that routine antiseptic body washing may result in lower catheter-related mortality than no routine antiseptic body washing (RR: 0.20; 95% CI: 0.01 to 4.13; one RCT (41)). +- There was low certainty of evidence that routine antiseptic body washing may result in little to no difference in all-cause mortality when compared with no routine antiseptic body washing (odds ratio [OR]: 0.90; 95% CI: 0.76 to 1.07; five RCTs (33, 37-40)). +- There was very low certainty of evidence that routine antiseptic body washing may result in little to no difference in sepsis when compared with no routine antiseptic body washing (RR: 1.92; 95% CI: 0.66 to 5.60; one RCT (33)). + +- Subgroup analyses revealed no difference in the effect for CLABSI based on age (adults versus children) or setting (ICU versus general ward). + +## Evidence-to-decision + +A potentially reduced equity was noted due to the different availability of products around the world and their moderate costs, in particular when considering LMICs, but the intervention was considered probably cost-effective. No issues were identified in acceptability and feasibility. + +## Implementation considerations + +- The recommendation refers to using a chlorhexidine-containing body wash (liquid or impregnated wipes) daily and not immediately prior to line insertion. The GDG acknowledges that there are variations in the available protocols on body washing and directs health and care workers to follow their national and local policies. +- The recommendation refers to all critically-ill patients that require a CVC and not limited to the ICU. +- It is advisable to engage nursing and support staff as key stakeholders in the decision (that is, time required for bathing). +- It is also advisable to engage patients and their families, considering the potential adverse effects on the skin related to exposure to chlorhexidine, including asking for any known intolerance and rare allergic reactions to chlorhexidine. +- WASH services and standards (64, 65) are essential to enable critical IPC practices, including this recommendation on body wash. + +## Research needs + +- There is a significant variation in available protocols on how to perform patient bathing. The GDG identified this as an area of research interest to determine the most effective protocol and product to be used (that is, liquid wash versus wipes or other means, rinsing versus no rinsing, which solution to use). +- The GDG also identified the need for additional research on the use of antiseptic products other than chlorhexidine (that is, octenidine and others) using multicentre RCTs or other types of trials (that is, adaptive) and focusing on specific populations (such as haematology patients or the paediatric ICU). +- AMR and other wider implications (that is, impact on skin microbiome, cross-resistance to antibiotics and on the epidemiology of other infections caused by gram-positive and gramnegative bacteria) associated with the use of chlorhexidine and other body wash products were identified as important areas for research. +- Cost-benefit research studies including baseline surveillance data to define potential thresholds for universal use were also considered as other areas for further research. + +## 3.3 Insertion + +## 3.3.1 Insertion pack/kit + +- Good practice statement 6. WHO recommends that clinicians use a standardized insertion pack/kit when placing a CVC in adults, adolescents, children and neonates (Good practice statement) + +## 3.3.2 CVC insertion by an insertion team + +- WHO makes no recommendation regarding the use of a dedicated CVC insertion teams in adults, adolescents, children, or neonates, as the evidence remains inconclusive. + +## Remarks + +- A CVC insertion team is defined as consisting of individuals specifically trained and competent in CVC insertion who are dedicated to perform the insertion of CVCs. +- The search criteria used in the systematic review were inclusive as possible, including terms such as "insertion team", "vascular access team", intravenous therapy team". +- Establishing a dedicated CVC insertion team is an intervention that is different from ensuring specific training for all those tasked with CVC insertion, which is covered by good practice statements 1 to 3 in these guidelines. +- Consideration should also be given to specific populations (i.e., neonates, children, or patients with known difficult intravenous access) as the potential benefit in preventing CLABSI may be confined to these groups. +- The GDG members acknowledged the difficulties in designing a RCT to evaluate this topic and the very limited available evidence. + +## Rationale for the decision + +Based on the evidence provided (two uncontrolled before-after studies and no RCTs), the GDG considered both the desirable and undesirable effects to be small. The GDG judged that the crucial factor was the adequate training of the clinician inserting the CVC, rather than having a dedicated team itself. For this reason, the GDG considered that the balance of effects did not favour either the intervention (a formal insertion team) or the comparison (insertion by a clinician who is not part of a formal insertion team), provided that the clinician has undergone appropriate training and demonstrated competency in CVC insertion. Another concern expressed by the GDG was the potential loss of knowledge and skills in CVC insertion among non-team members (doctors and nurses), such that their insertion competency could decline. This may be especially impactful in emergency situations where they are required to urgently insert a CVC without the assistance of a formal insertion team, with potential implications affecting the wider hospital capacity and patient care. Overall, given the very low certainty of the evidence, no recommendation could be made. + +## Summary of the evidence + +- The Systematic Reviews Expert Group did not identify any RCTs on the topic, but only two uncontrolled before-after studies (66, 67) evaluating the use of a dedicated CVC insertion (for non-haemodialysis) team compared with not using a dedicated CVC insertion team. The risk of bias in both studies was rated as high due to their uncontrolled design (no formal assessment). +- There was very low certainty of evidence from one study in children and adults reporting a 58% reduction in the number of CLABSIs after the introduction of a highly specialized vascular access team (67). Similarly, there was very low certainty of evidence from another study in neonates reporting that overall CLABSI rates decreased by 65% after implementation of the line team (66). +- There were no studies identified reporting on catheter-related mortality, all-cause mortality, sepsis pneumothorax, local site infections, bleeding events that require action, cardiac tamponade, deep vein thrombosis or venous stenosis. + +## Evidence-to-decision + +Increased moderate costs (due to the resources required to maintain a dedicated team), the probably reduced equity, variable acceptability by health care workers and patients (hospital management might not be in favour due to the additional costs, but patient satisfaction is likely to be improved) and low feasibility (as having a dedicated team is not always possible due to staff shortages) were all considered, resulting in a decision to make no recommendation.. The overall cost-effectiveness judgement did not favour either the intervention or the comparison. + +## Research needs + +- Further research is needed to compare the potential cost-benefit of using a formal CVC insertion team versus an educational/behavioural programme for clinicians inserting CVC, including both ICU and non-ICU settings. + +## 3.3.3 Sterile technique + +- Good practice statement 7. WHO recommends using a sterile technique for the insertion of all CVCs in adults, adolescents, children and neonates. (Good practice statement) +- Good practice statement 8. WHO recommends that clinicians use sterile single-use gloves compared to non-sterile gloves when inserting a CVC in adults, adolescents, children and neonates. + +(Good practice statement) + +- Recommendation 2. WHO suggests using a sterile technique involving maximal sterile barrier precautions over a sterile technique not involving maximal sterile +- barrier precautions at the time of CVC insertion in adults, adolescents, children and neonates. + +(Conditional recommendation, very low certainty of evidence) + +## Remarks + +- Maximal sterile barrier precautions: sterile barrier precautions that include the use of a cap, mask, sterile gown, sterile gloves and a sterile full body drape (11). +- Sterile technique is defined as a set of specific practices and procedures performed to make equipment and areas free from all microorganisms and to maintain that sterility (11). +- GDG members acknowledged that the above definition of sterile technique as comparator also seems to be generic and it may not provide enough practical details. + +## Rationale for the recommendation + +There was an extensive discussion during which a number of GDG members expressed the opinion that a strong recommendation would have been more suitable; others suggested a conversion into a good practice statement. However, it was decided to assess the available evidence and to proceed with the evidence-to-decision framework. Following the assessment, the GDG judged the desirable and undesirable effects to be small and trivial, respectively, but with a balance of effects probably favouring the intervention, despite the very low certainty of evidence. In view of the data presented, the GDG members opted for a conditional recommendation in favour of maximal sterile barrier precautions, but acknowledging that the current evidence is weak and more studies are needed to assess if less barriers may be sufficient. + +## Summary of the evidence + +The Systematic Reviews Expert Group identified eight studies (two RCTs (68, 69) and six NRSIs (70-75)) comparing maximal sterile barrier precautions in adults, children and adolescents, and neonates. + +- There was very low certainty of evidence that maximal sterile barrier precautions may result in fewer CLABSIs than a lesser standard of precautions (RR: 0.47, 95% CI: 0.10 to 2.22; two RCTs (68, 69)). +- Two non-randomized studies also reported very low certainty of evidence that maximal sterile barrier precautions may result in fewer CLABSIs than a lesser standard of precautions (RR: 0.21; 95% CI: 0.05 to 0.94; two non-RCTs (70, 71)) +- There was very low certainty of evidence that maximal sterile barrier precautions may result in little to no difference in local insertion site infections when compared with a lesser standard of precautions (RR: 1.01; 95% CI: 0.06 to 16.03; one RCT (68)). +- There were no studies identified that reported on catheter-related mortality, all-cause mortality, sepsis and other complications. + +## Evidence-to-decision + +The intervention may increase costs due to the additional equipment required and it is probably accepted and feasible as it is already common practice. However, patients may have more discomfort with full body drapes. In addition, the GDG judged equity to be probably increased. + +## Implementation considerations + +- Ensure adequate training, understanding and competency in the application of maximal sterile barriers precautions. +- The GDG discussed the need for a dedicated procedure room for line insertion and noted that this is not practical for critically-ill patients and may be an issue in some health care facilities, particularly in LMICs. Furthermore, it is not part of the current definition of maximal sterile barrier precautions. + +## Research needs + +- More research studies are needed to assess if lesser sterile precautions are needed, with a trial design and platform (for example, adaptive (76, 77)) that would allow to clearly distinguish the impact of the different components within maximal sterile barrier precautions (that is, cap, mask, full body drape, the environment where the line is placed) and allow to answer multiple questions simultaneously. + +## 3.3.4 Skin disinfection + +- Good practice statement 9. WHO recommends that skin disinfection should always be used prior to the insertion of all CVCs in adults, adolescents, children and neonates. (Good practice statement) +- Recommendation 3. WHO suggests using an alcohol-based formulation over an +- aqueous-based antiseptic formulation for skin disinfection before CVC insertion in adults, adolescents, children, and neonates. (Conditional recommendation, low certainty of evidence) + +## Remarks + +- Alcohol-based antiseptics should be the preferred choice for skin disinfection before CVC insertion. +- It was acknowledged that alcohol-based formulations can cause skin irritation, dryness and a risk of pain in case of laceration and these potential side-effects should be taken into consideration when using these products. Some of these side-effects may be caused by either the alcohol or the antiseptic component. +- Alcohol-based antiseptics can cause skin lesions in preterm infants without intact skin. + +## Rationale for the recommendation + +The GDG considered alcohol-based formulations as the preferred option for skin disinfection over aqueous formulations, with a conditional recommendation due to the low certainty of evidence. The GDG considered the desirable effects of alcohol-based formulations for skin disinfection to be moderate and the undesirable effects trivial, with a balance of effects probably favouring the intervention. + +## Summary of the evidence + +The Systematic Reviews Expert Group found data from five RCTs (78-82). + +- There was low certainty of evidence that alcohol-based solutions may result in fewer CLABSIs than aqueous-based solutions for skin disinfection (RR: 0.59; 95% CI: 0.36 to 0.98; five RCTs (78-82)). +- There was very low certainty of evidence that alcohol-based solutions may result in less allcause mortality than aqueous-based solutions for skin disinfection (RR: 0.79; 95% CI: 0.48 to 1.30; two RCTs (78, 79)). +- There was very low certainty of evidence that alcohol-based solutions may result in fewer sepsis episodes when compared with aqueous-based solutions for skin disinfection (RR: 0.68; 95% CI: 0.35 to 1.30; one RCT (78)). +- There were no studies identified that reported on catheter-related mortality, pneumothorax, local insertion site infections and other complications. \ No newline at end of file diff --git a/datasets/benchmarks/results/2026-10-06/who_9789240121805_excerpt.native.md b/datasets/benchmarks/results/2026-10-06/who_9789240121805_excerpt.native.md new file mode 100644 index 0000000..df7876f --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/who_9789240121805_excerpt.native.md @@ -0,0 +1,244 @@ +3. Recommendations and Good Practice Statements +11 +3.2 Pre-insertion +3.2.1 Chlorhexidine-containing body wash +• Recommendation 1. WHO suggests using a chlorhexidine-containing body wash +(liquid or impregnated wipes) daily in critically-ill adults, adolescents and children. +(Conditional recommendation, low certainly of evidence) +Remarks +• The evidence did not differentiate between liquid and impregnated wipes and evaluated +chlorhexidine concentrations ranging from 2% to 4%. It was noted that most of the studies +provided evidence on the use of chlorhexidine, hence a decision was made to restrict the +recommendation to chlorhexidine only and to adults, adolescents and children, but not +neonates (due to the risk of chemical burns) (32). +• The GDG noted that the use of this intervention would be particularly beneficial in settings with +high baseline CLABSI rates. However, the GDG highlighted the need to prioritize resources at the +local level based on other competing needs in the health care facility. The GDG also expressed +concerns on the potential development of resistance to chlorhexidine and on waste generation +when using wipes. +• Chlorhexidine products should be used with caution in patients with burns and according to the +manufacturer’s instructions in order to ensure they are safe for use . +Rationale for the recommendation +Overall, the GDG judged both the desirable and undesirable effects to be small, but with a balance of +effects probably favouring the intervention, with a low certainty of evidence. Based on the evidence +provided, the GDG opted for a conditional recommendation in favour of the intervention, in particular in +the context of high CLABSI rates at baseline, where the overall benefit may be more noticeable. +Summary of the evidence +The Systematic Reviews Group identified 31 studies (10 randomized controlled trials [RCTs] (33-42) and +21 non-randomized studies of interventions [NRSIs](43-63)) comparing daily antiseptic body washing +with chlorhexidine to soap and water bathing in adults, adolescents and children. +• There was low certainty of evidence that daily antiseptic body washing may result in fewer +CLABSI than no routine antiseptic body washing (risk ratio [RR]: 0.67; 95% confidence interval +[CI]: 0.45 to 0.99; 10 RCTs (33-42)). +• There was low certainty of evidence that routine antiseptic body washing may result in lower +catheter-related mortality than no routine antiseptic body washing (RR: 0.20; 95% CI: 0.01 to +4.13; one RCT (41)). +• There was low certainty of evidence that routine antiseptic body washing may result in little to +no difference in all-cause mortality when compared with no routine antiseptic body washing +(odds ratio [OR]: 0.90; 95% CI: 0.76 to 1.07; five RCTs (33, 37-40)). +• There was very low certainty of evidence that routine antiseptic body washing may result in +little to no difference in sepsis when compared with no routine antiseptic body washing (RR: +1.92; 95% CI: 0.66 to 5.60; one RCT (33)). + +Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters +12 +Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters +• Subgroup analyses revealed no difference in the effect for CLABSI based on age (adults versus +children) or setting (ICU versus general ward). +Evidence-to-decision +A potentially reduced equity was noted due to the different availability of products around the world +and their moderate costs, in particular when considering LMICs, but the intervention was considered +probably cost-effective. No issues were identified in acceptability and feasibility. +Implementation considerations +• The recommendation refers to using a chlorhexidine-containing body wash (liquid or +impregnated wipes) daily and not immediately prior to line insertion. The GDG acknowledges +that there are variations in the available protocols on body washing and directs health and care +workers to follow their national and local policies. +• The recommendation refers to all critically-ill patients that require a CVC and not limited to the +ICU. +• It is advisable to engage nursing and support staff as key stakeholders in the decision (that is, +time required for bathing). +• It is also advisable to engage patients and their families, considering the potential adverse +effects on the skin related to exposure to chlorhexidine, including asking for any known +intolerance and rare allergic reactions to chlorhexidine. +• WASH services and standards (64, 65) are essential to enable critical IPC practices, including this +recommendation on body wash. +Research needs +• There is a significant variation in available protocols on how to perform patient bathing. The +GDG identified this as an area of research interest to determine the most effective protocol and +product to be used (that is, liquid wash versus wipes or other means, rinsing versus no rinsing, +which solution to use). +• The GDG also identified the need for additional research on the use of antiseptic products other +than chlorhexidine (that is, octenidine and others) using multicentre RCTs or other types of +trials (that is, adaptive) and focusing on specific populations (such as haematology patients or +the paediatric ICU). +• AMR and other wider implications (that is, impact on skin microbiome, cross-resistance to +antibiotics and on the epidemiology of other infections caused by gram-positive and gram￾negative bacteria) associated with the use of chlorhexidine and other body wash products were +identified as important areas for research. +• Cost-benefit research studies including baseline surveillance data to define potential thresholds +for universal use were also considered as other areas for further research. + +3. Recommendations and Good Practice Statements +13 +3.3 Insertion +3.3.1 Insertion pack/kit +• Good practice statement 6. WHO recommends that clinicians use a standardized +insertion pack/kit when placing a CVC in adults, adolescents, children and neonates +(Good practice statement) +3.3.2 CVC insertion by an insertion team +• WHO makes no recommendation regarding the use of a dedicated CVC insertion +teams in adults, adolescents, children, or neonates, as the evidence remains +inconclusive. +Remarks +• A CVC insertion team is defined as consisting of individuals specifically trained and competent +in CVC insertion who are dedicated to perform the insertion of CVCs. +• The search criteria used in the systematic review were inclusive as possible, including terms +such as “insertion team”, “vascular access team”, intravenous therapy team”. +• Establishing a dedicated CVC insertion team is an intervention that is different from ensuring +specific training for all those tasked with CVC insertion, which is covered by good practice +statements 1 to 3 in these guidelines. +• Consideration should also be given to specific populations (i.e., neonates, children, or patients +with known difficult intravenous access) as the potential benefit in preventing CLABSI may be +confined to these groups. +• The GDG members acknowledged the difficulties in designing a RCT to evaluate this topic and +the very limited available evidence. +Rationale for the decision +Based on the evidence provided (two uncontrolled before-after studies and no RCTs), the GDG +considered both the desirable and undesirable effects to be small. The GDG judged that the crucial +factor was the adequate training of the clinician inserting the CVC, rather than having a dedicated +team itself. For this reason, the GDG considered that the balance of effects did not favour either the +intervention (a formal insertion team) or the comparison (insertion by a clinician who is not part +of a formal insertion team), provided that the clinician has undergone appropriate training and +demonstrated competency in CVC insertion. Another concern expressed by the GDG was the potential +loss of knowledge and skills in CVC insertion among non-team members (doctors and nurses), such that +their insertion competency could decline. This may be especially impactful in emergency situations +where they are required to urgently insert a CVC without the assistance of a formal insertion team, with +potential implications affecting the wider hospital capacity and patient care. Overall, given the very low +certainty of the evidence, no recommendation could be made. + +Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters +14 +Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters +Summary of the evidence +• The Systematic Reviews Expert Group did not identify any RCTs on the topic, but only two +uncontrolled before-after studies (66, 67) evaluating the use of a dedicated CVC insertion (for +non-haemodialysis) team compared with not using a dedicated CVC insertion team. The risk of +bias in both studies was rated as high due to their uncontrolled design (no formal assessment). +• There was very low certainty of evidence from one study in children and adults reporting a +58% reduction in the number of CLABSIs after the introduction of a highly specialized vascular +access team (67). Similarly, there was very low certainty of evidence from another study in +neonates reporting that overall CLABSI rates decreased by 65% after implementation of the line +team (66). +• There were no studies identified reporting on catheter-related mortality, all-cause mortality, +sepsis pneumothorax, local site infections, bleeding events that require action, cardiac +tamponade, deep vein thrombosis or venous stenosis. +Evidence-to-decision +Increased moderate costs (due to the resources required to maintain a dedicated team), the probably +reduced equity, variable acceptability by health care workers and patients (hospital management might +not be in favour due to the additional costs, but patient satisfaction is likely to be improved) and low +feasibility (as having a dedicated team is not always possible due to staff shortages) were all considered, +resulting in a decision to make no recommendation.. The overall cost-effectiveness judgement did not +favour either the intervention or the comparison. +Research needs +• Further research is needed to compare the potential cost-benefit of using a formal CVC insertion +team versus an educational/behavioural programme for clinicians inserting CVC, including both +ICU and non-ICU settings. +3.3.3 Sterile technique +• Good practice statement 7. WHO recommends using a sterile technique for the +insertion of all CVCs in adults, adolescents, children and neonates. +(Good practice statement) +• Good practice statement 8. WHO recommends that clinicians use sterile single-use +gloves compared to non-sterile gloves when inserting a CVC in adults, adolescents, +children and neonates. +(Good practice statement) +• Recommendation 2. WHO suggests using a sterile technique involving maximal +sterile barrier precautions over a sterile technique not involving maximal sterile +barrier precautions at the time of CVC insertion in adults, adolescents, children and +neonates. +(Conditional recommendation, very low certainty of evidence) + +3. Recommendations and Good Practice Statements +15 +Remarks +• Maximal sterile barrier precautions: sterile barrier precautions that include the use of a cap, +mask, sterile gown, sterile gloves and a sterile full body drape (11). +• Sterile technique is defined as a set of specific practices and procedures performed to make +equipment and areas free from all microorganisms and to maintain that sterility (11). +• GDG members acknowledged that the above definition of sterile technique as comparator also +seems to be generic and it may not provide enough practical details. +Rationale for the recommendation +There was an extensive discussion during which a number of GDG members expressed the opinion +that a strong recommendation would have been more suitable; others suggested a conversion into a +good practice statement. However, it was decided to assess the available evidence and to proceed with +the evidence-to-decision framework. Following the assessment, the GDG judged the desirable and +undesirable effects to be small and trivial, respectively, but with a balance of effects probably favouring +the intervention, despite the very low certainty of evidence. In view of the data presented, the GDG +members opted for a conditional recommendation in favour of maximal sterile barrier precautions, but +acknowledging that the current evidence is weak and more studies are needed to assess if less barriers +may be sufficient. +Summary of the evidence +The Systematic Reviews Expert Group identified eight studies (two RCTs (68, 69) and six NRSIs (70-75)) +comparing maximal sterile barrier precautions in adults, children and adolescents, and neonates. +• There was very low certainty of evidence that maximal sterile barrier precautions may result +in fewer CLABSIs than a lesser standard of precautions (RR: 0.47, 95% CI: 0.10 to 2.22; two RCTs +(68, 69)). +• Two non-randomized studies also reported very low certainty of evidence that maximal sterile +barrier precautions may result in fewer CLABSIs than a lesser standard of precautions (RR: 0.21; +95% CI: 0.05 to 0.94; two non-RCTs (70, 71)) +• There was very low certainty of evidence that maximal sterile barrier precautions may result in +little to no difference in local insertion site infections when compared with a lesser standard of +precautions (RR: 1.01; 95% CI: 0.06 to 16.03; one RCT (68)). +• There were no studies identified that reported on catheter-related mortality, all-cause +mortality, sepsis and other complications. +Evidence-to-decision +The intervention may increase costs due to the additional equipment required and it is probably +accepted and feasible as it is already common practice. However, patients may have more discomfort +with full body drapes. In addition, the GDG judged equity to be probably increased. +Implementation considerations +• Ensure adequate training, understanding and competency in the application of maximal sterile +barriers precautions. +• The GDG discussed the need for a dedicated procedure room for line insertion and noted that +this is not practical for critically-ill patients and may be an issue in some health care facilities, +particularly in LMICs. Furthermore, it is not part of the current definition of maximal sterile +barrier precautions. + +Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters +16 +Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters +Research needs +• More research studies are needed to assess if lesser sterile precautions are needed, with a trial +design and platform (for example, adaptive (76, 77)) that would allow to clearly distinguish the +impact of the different components within maximal sterile barrier precautions (that is, cap, +mask, full body drape, the environment where the line is placed) and allow to answer multiple +questions simultaneously. +3.3.4 Skin disinfection +• Good practice statement 9. WHO recommends that skin disinfection should always be +used prior to the insertion of all CVCs in adults, adolescents, children and neonates. +(Good practice statement) +• Recommendation 3. WHO suggests using an alcohol-based formulation over an +aqueous-based antiseptic formulation for skin disinfection before CVC insertion in +adults, adolescents, children, and neonates. +(Conditional recommendation, low certainty of evidence) +Remarks +• Alcohol-based antiseptics should be the preferred choice for skin disinfection before CVC insertion. +• It was acknowledged that alcohol-based formulations can cause skin irritation, dryness +and a risk of pain in case of laceration and these potential side-effects should be taken into +consideration when using these products. Some of these side-effects may be caused by either +the alcohol or the antiseptic component. +• Alcohol-based antiseptics can cause skin lesions in preterm infants without intact skin. +Rationale for the recommendation +The GDG considered alcohol-based formulations as the preferred option for skin disinfection over +aqueous formulations, with a conditional recommendation due to the low certainty of evidence. The +GDG considered the desirable effects of alcohol-based formulations for skin disinfection to be moderate +and the undesirable effects trivial, with a balance of effects probably favouring the intervention. +Summary of the evidence +The Systematic Reviews Expert Group found data from five RCTs (78-82). +• There was low certainty of evidence that alcohol-based solutions may result in fewer CLABSIs than +aqueous-based solutions for skin disinfection (RR: 0.59; 95% CI: 0.36 to 0.98; five RCTs (78-82)). +• There was very low certainty of evidence that alcohol-based solutions may result in less all￾cause mortality than aqueous-based solutions for skin disinfection (RR: 0.79; 95% CI: 0.48 to +1.30; two RCTs (78, 79)). +• There was very low certainty of evidence that alcohol-based solutions may result in fewer sepsis +episodes when compared with aqueous-based solutions for skin disinfection (RR: 0.68; 95% CI: +0.35 to 1.30; one RCT (78)). +• There were no studies identified that reported on catheter-related mortality, pneumothorax, +local insertion site infections and other complications. \ No newline at end of file diff --git a/datasets/benchmarks/results/2026-10-06/who_9789240121805_excerpt.tesseract.md b/datasets/benchmarks/results/2026-10-06/who_9789240121805_excerpt.tesseract.md new file mode 100644 index 0000000..df7c335 --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/who_9789240121805_excerpt.tesseract.md @@ -0,0 +1,343 @@ +3. Recommendations and Good Practice Statements + +3.2 Pre-insertion + +3.2.1 Chlorhexidine-containing body wash + +© Recommendation 1. WHO suggests using a chlorhexidine-containing body wash +(liquid or impregnated wipes) daily in critically-ill adults, adolescents and children. +(Conditional recommendation, low certainly of evidence) + +Remarks + +e The evidence did not differentiate between liquid and impregnated wipes and evaluated +chlorhexidine concentrations ranging from 2% to 4%. It was noted that most of the studies +provided evidence on the use of chlorhexidine, hence a decision was made to restrict the +recommendation to chlorhexidine only and to adults, adolescents and children, but not +neonates (due to the risk of chemical burns) (32). + +e The GDG noted that the use of this intervention would be particularly beneficial in settings with +high baseline CLABSI rates. However, the GDG highlighted the need to prioritize resources at the +local level based on other competing needs in the health care facility. The GDG also expressed +concerns on the potential development of resistance to chlorhexidine and on waste generation +when using wipes. + +e Chlorhexidine products should be used with caution in patients with burns and according to the +manufacturer’s instructions in order to ensure they are safe for use. + +Rationale for the recommendation + +Overall, the GDG judged both the desirable and undesirable effects to be small, but with a balance of +effects probably favouring the intervention, with a low certainty of evidence. Based on the evidence +provided, the GDG opted for a conditional recommendation in favour of the intervention, in particular in +the context of high CLABSI rates at baseline, where the overall benefit may be more noticeable. + +Summary of the evidence + +The Systematic Reviews Group identified 31 studies (10 randomized controlled trials [RCTs] (33-42) and +21 non-randomized studies of interventions [NRSIs](43-63)) comparing daily antiseptic body washing +with chlorhexidine to soap and water bathing in adults, adolescents and children. + +e There was low certainty of evidence that daily antiseptic body washing may result in fewer +CLABSI than no routine antiseptic body washing (risk ratio [RR]: 0.67; 95% confidence interval +[Cl]: 0.45 to 0.99; 10 RCTs (33-42)). + +e There was low certainty of evidence that routine antiseptic body washing may result in lower +catheter-related mortality than no routine antiseptic body washing (RR: 0.20; 95% Cl: 0.01 to +4.13; one RCT (41)). + +e There was low certainty of evidence that routine antiseptic body washing may result in little to +no difference in all-cause mortality when compared with no routine antiseptic body washing +(odds ratio [OR]: 0.90; 95% Cl: 0.76 to 1.07; five RCTs (33, 37-40)). + +e There was very low certainty of evidence that routine antiseptic body washing may result in +little to no difference in sepsis when compared with no routine antiseptic body washing (RR: +1.92; 95% Cl: 0.66 to 5.60; one RCT (33)). 11 + + +12 + +Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters + +Subgroup analyses revealed no difference in the effect for CLABSI based on age (adults versus +children) or setting (ICU versus general ward). + +Evidence-to-decision + +A potentially reduced equity was noted due to the different availability of products around the world +and their moderate costs, in particular when considering LMICs, but the intervention was considered +probably cost-effective. No issues were identified in acceptability and feasibility. + +Implementation considerations + +The recommendation refers to using a chlorhexidine-containing body wash (liquid or +impregnated wipes) daily and not immediately prior to line insertion. The GDG acknowledges +that there are variations in the available protocols on body washing and directs health and care +workers to follow their national and local policies. + +The recommendation refers to all critically-ill patients that require a CVC and not limited to the +ICU. + +It is advisable to engage nursing and support staff as key stakeholders in the decision (that is, +time required for bathing). + +It is also advisable to engage patients and their families, considering the potential adverse +effects on the skin related to exposure to chlorhexidine, including asking for any known +intolerance and rare allergic reactions to chlorhexidine. + +WASH services and standards (64, 65) are essential to enable critical IPC practices, including this +recommendation on body wash. + +Research needs + +There is a significant variation in available protocols on how to perform patient bathing. The +GDG identified this as an area of research interest to determine the most effective protocol and +product to be used (that is, liquid wash versus wipes or other means, rinsing versus no rinsing, +which solution to use). + +The GDG also identified the need for additional research on the use of antiseptic products other +than chlorhexidine (that is, octenidine and others) using multicentre RCTs or other types of +trials (that is, adaptive) and focusing on specific populations (such as haematology patients or +the paediatric ICU). + +AMR and other wider implications (that is, impact on skin microbiome, cross-resistance to +antibiotics and on the epidemiology of other infections caused by gram-positive and gram- +negative bacteria) associated with the use of chlorhexidine and other body wash products were +identified as important areas for research. + +Cost-benefit research studies including baseline surveillance data to define potential thresholds +for universal use were also considered as other areas for further research. + + +3. Recommendations and Good Practice Statements + +3.3 Insertion + +3.3.1 Insertion pack/kit + +© Good practice statement 6. WHO recommends that clinicians use a standardized +insertion pack/kit when placing a CVC in adults, adolescents, children and neonates +(Good practice statement) + +3.3.2 CVC insertion by an insertion team + +© WHO makes no recommendation regarding the use of a dedicated CVC insertion +teams in adults, adolescents, children, or neonates, as the evidence remains +inconclusive. + +Remarks + +e ACVC insertion team is defined as consisting of individuals specifically trained and competent +in CVC insertion who are dedicated to perform the insertion of CVCs. +e The search criteria used in the systematic review were inclusive as possible, including terms + +6 + +such as “insertion team”, “vascular access team”, intravenous therapy team”. + +e Establishing a dedicated CVC insertion team is an intervention that is different from ensuring +specific training for all those tasked with CVC insertion, which is covered by good practice +statements 1 to 3 in these guidelines. + +e Consideration should also be given to specific populations (i.e., neonates, children, or patients +with known difficult intravenous access) as the potential benefit in preventing CLABSI may be +confined to these groups. + +e The GDG members acknowledged the difficulties in designing a RCT to evaluate this topic and +the very limited available evidence. + +Rationale for the decision + +Based on the evidence provided (two uncontrolled before-after studies and no RCTs), the GDG +considered both the desirable and undesirable effects to be small. The GDG judged that the crucial +factor was the adequate training of the clinician inserting the CVC, rather than having a dedicated + +team itself. For this reason, the GDG considered that the balance of effects did not favour either the +intervention (a formal insertion team) or the comparison (insertion by a clinician who is not part + +of a formal insertion team), provided that the clinician has undergone appropriate training and +demonstrated competency in CVC insertion. Another concern expressed by the GDG was the potential +loss of knowledge and skills in CVC insertion among non-team members (doctors and nurses), such that +their insertion competency could decline. This may be especially impactful in emergency situations +where they are required to urgently insert a CVC without the assistance of a formal insertion team, with +potential implications affecting the wider hospital capacity and patient care. Overall, given the very low +certainty of the evidence, no recommendation could be made. + +13 + + +14 + +Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters + +Summary of the evidence + +e The Systematic Reviews Expert Group did not identify any RCTs on the topic, but only two +uncontrolled before-after studies (66, 67) evaluating the use of a dedicated CVC insertion (for +non-haemodialysis) team compared with not using a dedicated CVC insertion team. The risk of +bias in both studies was rated as high due to their uncontrolled design (no formal assessment). + +e There was very low certainty of evidence from one study in children and adults reporting a +58% reduction in the number of CLABSIs after the introduction of a highly specialized vascular +access team (67). Similarly, there was very low certainty of evidence from another study in +neonates reporting that overall CLABSI rates decreased by 65% after implementation of the line +team (66). + +e There were no studies identified reporting on catheter-related mortality, all-cause mortality, +sepsis pneumothorax, local site infections, bleeding events that require action, cardiac +tamponade, deep vein thrombosis or venous stenosis. + +Evidence-to-decision + +Increased moderate costs (due to the resources required to maintain a dedicated team), the probably +reduced equity, variable acceptability by health care workers and patients (hospital management might +not be in favour due to the additional costs, but patient satisfaction is likely to be improved) and low +feasibility (as having a dedicated team is not always possible due to staff shortages) were all considered, +resulting in a decision to make no recommendation.. The overall cost-effectiveness judgement did not +favour either the intervention or the comparison. + +Research needs +e Further research is needed to compare the potential cost-benefit of using a formal CVC insertion +team versus an educational/behavioural programme for clinicians inserting CVC, including both +ICU and non-ICU settings. + +3.3.3 Sterile technique + +© Good practice statement 7. WHO recommends using a sterile technique for the +insertion of all CVCs in adults, adolescents, children and neonates. +(Good practice statement) + +© Good practice statement 8. WHO recommends that clinicians use sterile single-use +gloves compared to non-sterile gloves when inserting a CVC in adults, adolescents, +children and neonates. +(Good practice statement) + +© Recommendation 2. WHO suggests using a sterile technique involving maximal +sterile barrier precautions over a sterile technique not involving maximal sterile +barrier precautions at the time of CVC insertion in adults, adolescents, children and +neonates. +(Conditional recommendation, very low certainty of evidence) + + +3. Recommendations and Good Practice Statements + +Remarks + +e Maximal sterile barrier precautions: sterile barrier precautions that include the use of a cap, +mask, sterile gown, sterile gloves and a sterile full body drape (11). + +e Sterile technique is defined as a set of specific practices and procedures performed to make +equipment and areas free from all microorganisms and to maintain that sterility (11). + +e GDG members acknowledged that the above definition of sterile technique as comparator also +seems to be generic and it may not provide enough practical details. + +Rationale for the recommendation + +There was an extensive discussion during which a number of GDG members expressed the opinion + +that a strong recommendation would have been more suitable; others suggested a conversion into a +good practice statement. However, it was decided to assess the available evidence and to proceed with +the evidence-to-decision framework. Following the assessment, the GDG judged the desirable and +undesirable effects to be small and trivial, respectively, but with a balance of effects probably favouring +the intervention, despite the very low certainty of evidence. In view of the data presented, the GDG +members opted for a conditional recommendation in favour of maximal sterile barrier precautions, but +acknowledging that the current evidence is weak and more studies are needed to assess if less barriers +may be sufficient. + +Summary of the evidence + +The Systematic Reviews Expert Group identified eight studies (two RCTs (68, 69) and six NRSls (70-75)) +comparing maximal sterile barrier precautions in adults, children and adolescents, and neonates. + +e There was very low certainty of evidence that maximal sterile barrier precautions may result +in fewer CLABSIs than a lesser standard of precautions (RR: 0.47, 95% Cl: 0.10 to 2.22; two RCTs +(68, 69)). + +* Two non-randomized studies also reported very low certainty of evidence that maximal sterile +barrier precautions may result in fewer CLABSIs than a lesser standard of precautions (RR: 0.21; +95% Cl: 0.05 to 0.94; two non-RCTs (70, 71)) + +e There was very low certainty of evidence that maximal sterile barrier precautions may result in +little to no difference in local insertion site infections when compared with a lesser standard of +precautions (RR: 1.01; 95% Cl: 0.06 to 16.03; one RCT (68)). + +e There were no studies identified that reported on catheter-related mortality, all-cause +mortality, sepsis and other complications. + +Evidence-to-decision + +The intervention may increase costs due to the additional equipment required and it is probably +accepted and feasible as it is already common practice. However, patients may have more discomfort +with full body drapes. In addition, the GDG judged equity to be probably increased. + +Implementation considerations + +e Ensure adequate training, understanding and competency in the application of maximal sterile +barriers precautions. + +e The GDG discussed the need for a dedicated procedure room for line insertion and noted that +this is not practical for critically-ill patients and may be an issue in some health care facilities, +particularly in LMICs. Furthermore, it is not part of the current definition of maximal sterile +barrier precautions. + + +16 + +Guidelines for the prevention of bloodstream infections and other infections associated with the use of intravascular catheters + +Research needs + +More research studies are needed to assess if lesser sterile precautions are needed, with a trial +design and platform (for example, adaptive (76, 77)) that would allow to clearly distinguish the +impact of the different components within maximal sterile barrier precautions (that is, cap, +mask, full body drape, the environment where the line is placed) and allow to answer multiple +questions simultaneously. + +3.3.4 Skin disinfection + +© Good practice statement 9. WHO recommends that skin disinfection should always be +used prior to the insertion of all CVCs in adults, adolescents, children and neonates. +(Good practice statement) + +© Recommendation 3. WHO suggests using an alcohol-based formulation over an +aqueous-based antiseptic formulation for skin disinfection before CVC insertion in +adults, adolescents, children, and neonates. +(Conditional recommendation, low certainty of evidence) + +Remarks + +Alcohol-based antiseptics should be the preferred choice for skin disinfection before CVC insertion. + +It was acknowledged that alcohol-based formulations can cause skin irritation, dryness + +and a risk of pain in case of laceration and these potential side-effects should be taken into +consideration when using these products. Some of these side-effects may be caused by either +the alcohol or the antiseptic component. + +Alcohol-based antiseptics can cause skin lesions in preterm infants without intact skin. + +Rationale for the recommendation + +The GDG considered alcohol-based formulations as the preferred option for skin disinfection over +aqueous formulations, with a conditional recommendation due to the low certainty of evidence. The +GDG considered the desirable effects of alcohol-based formulations for skin disinfection to be moderate +and the undesirable effects trivial, with a balance of effects probably favouring the intervention. + +Summary of the evidence + +The Systematic Reviews Expert Group found data from five RCTs (78-82). + +There was low certainty of evidence that alcohol-based solutions may result in fewer CLABSIs than +aqueous-based solutions for skin disinfection (RR: 0.59; 95% Cl: 0.36 to 0.98; five RCTs (78-82)). + +There was very low certainty of evidence that alcohol-based solutions may result in less all- +cause mortality than aqueous-based solutions for skin disinfection (RR: 0.79; 95% Cl: 0.48 to +1.30; two RCTs (78, 79)). + +There was very low certainty of evidence that alcohol-based solutions may result in fewer sepsis +episodes when compared with aqueous-based solutions for skin disinfection (RR: 0.68; 95% Cl: +0.35 to 1.30; one RCT (78)). + +There were no studies identified that reported on catheter-related mortality, pneumothorax, +local insertion site infections and other complications. diff --git a/datasets/benchmarks/results/2026-10-06/who_9789240124233_excerpt.docling.md b/datasets/benchmarks/results/2026-10-06/who_9789240124233_excerpt.docling.md new file mode 100644 index 0000000..5607e1d --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/who_9789240124233_excerpt.docling.md @@ -0,0 +1,141 @@ +Table 1: HIV service delivery recommendations + +| Date* | Recommendations | Status | +|-------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------| +| Linkage and ART initiation | Linkage and ART initiation | Linkage and ART initiation | +| 2016 | Following an HIV diagnosis, a package of support interventions should be offered to ensure timely linkage to care for all people living with HIV (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2021 | Health systems should invest in people-centred practices and communication, including ongoing training, mentoring, supportive supervision and monitoring of health workers, to improve the relationships between clients and health-care providers (Good practice statement). | Revalidated | +| 2019 | ART initiation should follow the overarching principles of providing people-centred care. People-centred care should be focused and organized around the health needs, preferences and expectations of people and communities, upholding individual dignity and respect, especially for vulnerable populations (Good practice statement). | Revalidated | +| 2021 | Antiretroviral therapy (ART) initiation may be offered outside the health facility (Conditional recommendation, low- to moderate-certainty evidence). | Revalidated | +| 2017 | Rapid ART initiation should be offered to all people living with HIV following a confirmed HIV diagnosis and clinical assessment (Strong recommendation: high- certainty evidence for adults, low-certainty evidence for children) | Revalidated | +| 2021 | The offer of same-day ART initiation should include approaches to improve uptake, treatment adherence and retention such as tailored client education, counselling and support (Good practice statement). | Revalidated | +| Clinical visits and medication pick-up | Clinical visits and medication pick-up | Clinical visits and medication pick-up | +| 2021 | People established on ART should be offered clinical visits every 3-6 months, preferably every six months if feasible (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2021 | People established on ART should be offered refills of ART lasting 3-6 months, preferably six months if feasible (Strong recommendation, moderate-certainty evidence). | Revalidated | +| Adherence and engagement in care | Adherence and engagement in care | Adherence and engagement in care | +| 2025 | Adherence support interventions should be provided to people on ART (Strong recommendation, moderate-certainty evidence). | Updated | +| 2021 | Viral load for treatment monitoring should be complemented with non-judgemental, tailored approaches to assessing adherence (Good practice statement). | Revalidated | +| 2021 | HIV programmes should implement interventions to trace people who have disengaged from care and provide support for re-engagement (Strong recommendation, low-certainty evidence). | Revalidated | +| Services for adolescents and young people | Services for adolescents and young people | Services for adolescents and young people | +| 2021 | Psychosocial interventions should be provided to all adolescents and young adults living with HIV (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2013 | Adolescent-friendly services should be implemented in HIV services to ensure engagement and improved outcomes (Strong recommendation, moderate-certainty evidence). | Revalidated | + +Table 1: HIV service delivery recommendations (cont'd) + +| Date* | Recommendations | Status | +|-------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------| +| Services for adolescents and young people | Services for adolescents and young people | Services for adolescents and young people | +| 2013 | Adolescents should be counselled about the potential benefits and risks of disclosing their HIV status to others and empowered and supported to determine whether, when, how and to whom to disclose (Conditional recommendation, very-low -certainty evidence). | Revalidated | +| Task sharing | Task sharing | Task sharing | +| 2016 | Trained non-physician clinicians, midwives and nurses can initiate first-line ART (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2016 | Trained non-physician clinicians, midwives and nurses can maintain ART (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2016 | Trained and supervised community health workers can dispense ART between regular clinical visits (Strong recommendation, moderate-certainty evidence). | Revalidated | +| 2016 | Trained and supervised lay health-care providers can distribute ART (Strong recommendation, low-certainty evidence). | Revalidated | +| 2021 | Task sharing of specimen collection and point-of-care testing with non-laboratory personnel should be implemented when professional staffing capacity is limited (Strong recommendation, moderate-certainty evidence). | Revalidated | +| Service Delivery models and integration | Service Delivery models and integration | Service Delivery models and integration | +| 2013 | Decentralization of ART care should be considered as a way to increase access and improve retention in care (Strong recommendation, moderate-to-low-certainty evidence). | Revalidated | +| 2013 | In generalized epidemic settings, ART should be initiated and maintained in pregnant and postpartum women and in infants at maternal and child health-care settings, with linkage and referral to ongoing HIV care and ART, if appropriate (Strong recommendation, low-certainty evidence). | Revalidated | +| 2021 | Sexual and reproductive health services, including contraception, may be integrated with HIV services (Conditional recommendation, very-low-certainty evidence). | Revalidated | +| 2013 | In settings with a high burden of HIV and tuberculosis (TB), ART should be initiated in TB treatment settings, with linkage to ongoing HIV care and ART (Strong recommendation, very-low-certainty evidence). | Revalidated | +| 2013 | In settings with a high burden of HIV and TB, TB treatment may be provided for people living with HIV in HIV care settings if they have been diagnosed with TB (Strong recommendation, very-low-certainty evidence). | Revalidated | +| 2025 | Diabetes and hypertension care should be integrated with HIV services (Strong recommendation, moderate-to-very-low-certainty evidence). | Updated | +| 2025 | Mental health care for depression, anxiety and alcohol use disorders should be integrated with HIV services (Strong recommendation, moderate-to-very-low-certainty evidence). | New | + +## 1. Introduction + +HIV remains a major public health challenge and a leading cause of illness and death in many low- and middleincome countries, despite the wide availability of effective antiretroviral therapy and proven prevention tools. + +In many settings, evidence-informed interventions for HIV prevention, testing, treatment and long-term care are not fully implemented, delivered inconsistently or reach people too late, contributing to ongoing transmission, advanced disease and avoidable mortality. Persistent health system constraints - including workforce shortages, weak supply chains and limited integration with related services such as those for noncommunicable diseases and mental health care - further hinder effective service delivery. + +These HIV service delivery guidelines provide updated and consolidated recommendations to support countries, programme managers, health-care workers, communities and other stakeholders to organize services that are people-centred, integrated, equitable and grounded in the best available evidence. The guidelines aim to improve access to and continuity of high-quality HIV services, reduce inequities in outcomes and accelerate progress towards ending AIDS as a public health threat. + +## 2. Linkage and ART initiation + +## 2.1 Linkage from HIV testing to enrolment in care + +## Recommendation + +Following an HIV diagnosis, a package of support interventions should be offered to ensure timely linkage to care for all people living with HIV. + +(Strong recommendation, moderate-certainty evidence) + +The following interventions have demonstrated benefit in improving linkage to care following an HIV diagnosis: + +- streamlined interventions to reduce time between diagnosis and engagement in care, including (i) enhanced linkage with case management; (ii) support for HIV disclosure; (iii) tracing; (iv) training staff to provide multiple services; and (v) streamlined services (moderate-certainty evidence); +- peer support a and navigation approaches for linkage (moderate-certainty evidence); and +- quality improvement approaches using data to improve linkage (low-certainty evidence). + +a Includes peer counselling. + +## Good practice statements + +ART initiation should follow the overarching principles of providing people-centred care. People-centred care is care that is focused and organized around the health needs, preferences and expectations of people and communities, upholding individual dignity and respect, especially for vulnerable populations. It should promote the engagement and support of people and families to play an active role in their own care through informed decision-making. + +All people newly diagnosed with HIV should be retested to verify their HIV status before starting ART, using the same testing strategy and algorithm as the initial test. To minimize the risk of misdiagnosis, this approach should be maintained in settings in which rapid ART initiation is being implemented. + +Revalidated + + + +## Background + +Linkage to HIV prevention, treatment and care after testing remains a challenge, with disproportionate gaps for key populations, men, young people and previously diagnosed people living with HIV (not receiving ART, disengaged or lost to care). + +HIV testing service providers are primarily responsible for linkage to care. People testing HIV positive should be assessed for advanced HIV disease using CD4 cell count testing and offered ART, and those testing negative should be linked to prevention services. + +## Rationale + +Evidence-informed interventions to support linkage can be classified into three approaches (2): + +- streamlined services via enhanced linkage with case management, HIV disclosure support, client tracing for non-engagers and staff training for multiple services; +- peer support and navigation interventions, including home visits, system navigation and enhanced counselling; and +- quality improvement interventions enhance linkage to care (3). + +## Implementation considerations + +- Interventions may vary based on the context, including health-care delivery systems, geography and target population. +- A combination of interventions may be needed to improve linkage for specific underserved groups, especially key populations and men. +- Barriers to linkage to HIV care and ART initiation should be assessed and addressed, especially for populations facing multiple structural and individual barriers. +- Linkage for children can be improved using point-of-care early infant diagnosis, SMS/GSM/GPRS printers to accelerate access to laboratory results and family-centred models delivering care to mother, baby, and partner at the same point. +- Health facilities should have capacity to provide services for key populations. Community- and peer-based approaches can be especially effective in restrictive legal settings. + +WHO guidelines on HIV testing services (4) provide information on post-test services and linkage to care. + +## 2.2 People-centred care + +## Good practice statement + +Health systems should invest in people-centred practices and communication, including ongoing training, mentoring, supportive supervision and monitoring of health workers, to improve the relationships between clients and health-care providers. + +HIV programmes should: + +- provide people-centred care that is focused and organized around the health needs, preferences and expectations of people and communities, upholding individual dignity and respect, especially for underserved populations, and engage and support people and families to play an active role in their own care by informed decision-making; +- offer safe, acceptable and appropriate clinical and non-clinical services in a timely fashion, aiming to reduce morbidity and mortality associated with HIV infection and to improve health outcomes and quality of life in general; and +- promote the efficient and effective use of resources. + + + +## Background + +People-centred health services are an approach to care that consciously adopts the perspectives of individuals, families and communities and sees them as participants and beneficiaries of trusted health systems that respond to their needs and preferences in humane and holistic ways. This approach acknowledges the experiences and perspectives of health-care providers that may enable or prevent the delivery of high-quality people-centred care. + +## Rationale + +- A systematic review (5) identified several interventions to enhance people-centred care for people living with HIV. Overall, studies reported beneficial effects of these approaches across the HIV cascade outcomes, including improved ART uptake, adherence and viral load suppression. +- Studies have shown that people are willing to travel longer distances to be seen by a health-care provider with a respectful and caring attitude, and negative health-care worker attitudes contribute to loss to care and poor programme outcomes (6). + +Interventions found to improve relationships between clients and health-care providers can be classified into the following approaches: + +- providing friendly and welcoming services: training providers to make general HIV services more welcoming, including for clients who have disengaged from care; +- conducting sensitization training for clinical and non-clinical health-care providers to improve care for key populations both at the primary care and community levels; such training should include issues related to stigma and discrimination; +- offering individualized adherence counselling and person-centred communication such as shared decisionmaking and planning for ART initiation and adherence; +- facilitating client education in empowerment and communication skills; and +- providing feedback to health-care workers on client concerns and evaluation of service quality such as community score cards and client feedback mechanisms combined with quality improvement exercises. + +## Implementation considerations + +- Build capacity for ongoing training, mentoring and supportive supervision on people-centred communication, stigma reduction and welcoming services. +- Integrate sensitization training for all staff (clinical and non-clinical) on respectful attitudes, stigma and discrimination, re-engagement of clients who have disengaged, and support for women and other vulnerable groups experiencing violence or sexual health concerns. +- Establish routine client feedback mechanisms (such as satisfaction questionnaires, community score cards and suggestion boxes) and use the results in facility quality-improvement meetings. +- Involve clients and community representatives in designing and reviewing services. +- Review workload, staffing and client flow to allow time for person-centred interactions. \ No newline at end of file diff --git a/datasets/benchmarks/results/2026-10-06/who_9789240124233_excerpt.native.md b/datasets/benchmarks/results/2026-10-06/who_9789240124233_excerpt.native.md new file mode 100644 index 0000000..b8a771d --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/who_9789240124233_excerpt.native.md @@ -0,0 +1,227 @@ +ix Consolidated HIV guidelines +Date* Recommendations Status +Linkage and ART initiation +2016 Following an HIV diagnosis, a package of support interventions should be offered to +ensure timely linkage to care for all people living with HIV (Strong recommendation, +moderate-certainty evidence). +2021 Health systems should invest in people-centred practices and communication, +including ongoing training, mentoring, supportive supervision and monitoring +of health workers, to improve the relationships between clients and health-care +providers (Good practice statement). +2019 ART initiation should follow the overarching principles of providing people-centred +care. People-centred care should be focused and organized around the health needs, +preferences and expectations of people and communities, upholding individual +dignity and respect, especially for vulnerable populations (Good practice statement). +2021 Antiretroviral therapy (ART) initiation may be offered outside the health facility +(Conditional recommendation, low- to moderate-certainty evidence). +2017 Rapid ART initiation should be offered to all people living with HIV following a +confirmed HIV diagnosis and clinical assessment (Strong recommendation: high￾certainty evidence for adults, low-certainty evidence for children) +2021 The offer of same-day ART initiation should include approaches to improve uptake, +treatment adherence and retention such as tailored client education, counselling and +support (Good practice statement). +Clinical visits and medication pick-up +2021 People established on ART should be offered clinical visits every 3–6 months, preferably +every six months if feasible (Strong recommendation, moderate-certainty evidence). +2021 People established on ART should be offered refills of ART lasting 3–6 months, +preferably six months if feasible (Strong recommendation, moderate-certainty evidence). +Adherence and engagement in care +2025 Adherence support interventions should be provided to people on ART (Strong +recommendation, moderate-certainty evidence). +2021 Viral load for treatment monitoring should be complemented with non-judgemental, +tailored approaches to assessing adherence (Good practice statement). +2021 HIV programmes should implement interventions to trace people who +have disengaged from care and provide support for re-engagement (Strong +recommendation, low-certainty evidence). +Services for adolescents and young people +2021 Psychosocial interventions should be provided to all adolescents and young adults +living with HIV (Strong recommendation, moderate-certainty evidence). +2013 Adolescent-friendly services should be implemented in HIV services to ensure engagement +and improved outcomes (Strong recommendation, moderate-certainty evidence). +Table 1: HIV service delivery recommendations +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Updated +Revalidated +Revalidated +Revalidated +Revalidated + +Executive summary x +Date* Recommendations Status +Services for adolescents and young people +2013 Adolescents should be counselled about the potential benefits and risks of disclosing +their HIV status to others and empowered and supported to determine whether, when, +how and to whom to disclose (Conditional recommendation, very-low -certainty evidence). +Task sharing +2016 Trained non-physician clinicians, midwives and nurses can initiate first-line ART +(Strong recommendation, moderate-certainty evidence). +2016 Trained non-physician clinicians, midwives and nurses can maintain ART (Strong +recommendation, moderate-certainty evidence). +2016 Trained and supervised community health workers can dispense ART between regular +clinical visits (Strong recommendation, moderate-certainty evidence). +2016 Trained and supervised lay health-care providers can distribute ART (Strong +recommendation, low-certainty evidence). +2021 Task sharing of specimen collection and point-of-care testing with non-laboratory +personnel should be implemented when professional staffing capacity is limited +(Strong recommendation, moderate-certainty evidence). +Service Delivery models and integration +2013 Decentralization of ART care should be considered as a way to increase access and +improve retention in care (Strong recommendation, moderate-to-low-certainty evidence). +2013 In generalized epidemic settings, ART should be initiated and maintained in +pregnant and postpartum women and in infants at maternal and child health-care +settings, with linkage and referral to ongoing HIV care and ART, if appropriate (Strong +recommendation, low-certainty evidence). +2021 Sexual and reproductive health services, including contraception, may be integrated +with HIV services (Conditional recommendation, very-low-certainty evidence). +2013 In settings with a high burden of HIV and tuberculosis (TB), ART should be initiated +in TB treatment settings, with linkage to ongoing HIV care and ART (Strong +recommendation, very-low-certainty evidence). +2013 In settings with a high burden of HIV and TB, TB treatment may be provided for +people living with HIV in HIV care settings if they have been diagnosed with TB (Strong +recommendation, very-low-certainty evidence). +2025 Diabetes and hypertension care should be integrated with HIV services (Strong +recommendation, moderate-to-very-low-certainty evidence). +2025 Mental health care for depression, anxiety and alcohol use disorders should be integrated +with HIV services (Strong recommendation, moderate-to-very-low-certainty evidence). +Table 1: HIV service delivery recommendations (cont’d) +* Date the recommendation/good practice statement was formulated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Revalidated +Updated +New + +1 +HIV remains a major public health challenge and a leading cause of illness and death in many low- and middle￾income countries, despite the wide availability of effective antiretroviral therapy and proven prevention tools. +In many settings, evidence-informed interventions for HIV prevention, testing, treatment and long-term care are +not fully implemented, delivered inconsistently or reach people too late, contributing to ongoing transmission, +advanced disease and avoidable mortality. Persistent health system constraints – including workforce shortages, +weak supply chains and limited integration with related services such as those for noncommunicable diseases and +mental health care – further hinder effective service delivery. +These HIV service delivery guidelines provide updated and consolidated recommendations to support countries, +programme managers, health-care workers, communities and other stakeholders to organize services that are +people-centred, integrated, equitable and grounded in the best available evidence. The guidelines aim to improve +access to and continuity of high-quality HIV services, reduce inequities in outcomes and accelerate progress towards +ending AIDS as a public health threat. +1. Introduction + +2 +2. Linkage and ART initiation +2.1 Linkage from HIV testing to enrolment in care +Recommendation +Following an HIV diagnosis, a package of support interventions should be offered to +ensure timely linkage to care for all people living with HIV. +(Strong recommendation, moderate-certainty evidence) +The following interventions have demonstrated benefit in improving linkage to care following an +HIV diagnosis: +• streamlined interventions to reduce time between diagnosis and engagement in care, including +(i) enhanced linkage with case management; (ii) support for HIV disclosure; (iii) tracing; (iv) training +staff to provide multiple services; and (v) streamlined services (moderate-certainty evidence); +• peer supporta and navigation approaches for linkage (moderate-certainty evidence); and +• quality improvement approaches using data to improve linkage (low-certainty evidence). +a Includes peer counselling. +Good practice statements +ART initiation should follow the overarching principles of providing people-centred care. +People-centred care is care that is focused and organized around the health needs, preferences and +expectations of people and communities, upholding individual dignity and respect, especially for +vulnerable populations. It should promote the engagement and support of people and families to play an +active role in their own care through informed decision-making. +All people newly diagnosed with HIV should be retested to verify their HIV status before starting ART, +using the same testing strategy and algorithm as the initial test. To minimize the risk of misdiagnosis, +this approach should be maintained in settings in which rapid ART initiation is being implemented. +Background +Linkage to HIV prevention, treatment and care after testing remains a challenge, with disproportionate gaps for key +populations, men, young people and previously diagnosed people living with HIV (not receiving ART, disengaged or +lost to care). +HIV testing service providers are primarily responsible for linkage to care. People testing HIV positive should be +assessed for advanced HIV disease using CD4 cell count testing and offered ART, and those testing negative should +be linked to prevention services. +Revalidated +Revalidated + +3 Consolidated HIV guidelines +Rationale +Evidence-informed interventions to support linkage can be classified into three approaches (2): +• streamlined services via enhanced linkage with case management, HIV disclosure support, client tracing for +non-engagers and staff training for multiple services; +• peer support and navigation interventions, including home visits, system navigation and enhanced +counselling; and +• quality improvement interventions enhance linkage to care (3). +Implementation considerations +• Interventions may vary based on the context, including health-care delivery systems, geography and target +population. +• A combination of interventions may be needed to improve linkage for specific underserved groups, especially +key populations and men. +• Barriers to linkage to HIV care and ART initiation should be assessed and addressed, especially for populations +facing multiple structural and individual barriers. +• Linkage for children can be improved using point-of-care early infant diagnosis, SMS/GSM/GPRS printers to +accelerate access to laboratory results and family-centred models delivering care to mother, baby, and partner +at the same point. +• Health facilities should have capacity to provide services for key populations. Community- and peer-based +approaches can be especially effective in restrictive legal settings. +WHO guidelines on HIV testing services (4) provide information on post-test services and linkage to care. +2.2 People-centred care +Good practice statement +Health systems should invest in people-centred practices and communication, including ongoing training, +mentoring, supportive supervision and monitoring of health workers, to improve the relationships +between clients and health-care providers. +HIV programmes should: +• provide people-centred care that is focused and organized around the health needs, preferences and +expectations of people and communities, upholding individual dignity and respect, especially for +underserved populations, and engage and support people and families to play an active role in their +own care by informed decision-making; +• offer safe, acceptable and appropriate clinical and non-clinical services in a timely fashion, aiming to +reduce morbidity and mortality associated with HIV infection and to improve health outcomes and +quality of life in general; and +• promote the efficient and effective use of resources. +Revalidated + +4 +Background +People-centred health services are an approach to care that consciously adopts the perspectives of individuals, +families and communities and sees them as participants and beneficiaries of trusted health systems that respond +to their needs and preferences in humane and holistic ways. This approach acknowledges the experiences and +perspectives of health-care providers that may enable or prevent the delivery of high-quality people-centred care. +Rationale +• A systematic review (5) identified several interventions to enhance people-centred care for people living +with HIV. Overall, studies reported beneficial effects of these approaches across the HIV cascade outcomes, +including improved ART uptake, adherence and viral load suppression. +• Studies have shown that people are willing to travel longer distances to be seen by a health-care provider with +a respectful and caring attitude, and negative health-care worker attitudes contribute to loss to care and poor +programme outcomes (6). +Interventions found to improve relationships between clients and health-care providers can be classified into the +following approaches: +• providing friendly and welcoming services: training providers to make general HIV services more welcoming, +including for clients who have disengaged from care; +• conducting sensitization training for clinical and non-clinical health-care providers to improve care for key +populations both at the primary care and community levels; such training should include issues related to +stigma and discrimination; +• offering individualized adherence counselling and person-centred communication such as shared decision￾making and planning for ART initiation and adherence; +• facilitating client education in empowerment and communication skills; and +• providing feedback to health-care workers on client concerns and evaluation of service quality such as +community score cards and client feedback mechanisms combined with quality improvement exercises. +Implementation considerations +• Build capacity for ongoing training, mentoring and supportive supervision on people-centred communication, +stigma reduction and welcoming services. +• Integrate sensitization training for all staff (clinical and non-clinical) on respectful attitudes, stigma and +discrimination, re-engagement of clients who have disengaged, and support for women and other vulnerable +groups experiencing violence or sexual health concerns. +• Establish routine client feedback mechanisms (such as satisfaction questionnaires, community score cards +and suggestion boxes) and use the results in facility quality-improvement meetings. +• Involve clients and community representatives in designing and reviewing services. +• Review workload, staffing and client flow to allow time for person-centred interactions. +Linkage and ART initiation \ No newline at end of file diff --git a/datasets/benchmarks/results/2026-10-06/who_9789240124233_excerpt.tesseract.md b/datasets/benchmarks/results/2026-10-06/who_9789240124233_excerpt.tesseract.md new file mode 100644 index 0000000..c5a623d --- /dev/null +++ b/datasets/benchmarks/results/2026-10-06/who_9789240124233_excerpt.tesseract.md @@ -0,0 +1,349 @@ +Consolidated HIV guidelines + +Table 1: HIV service delivery recommendations + +Linkage and ART initiation + +2016 _— Following an HIV diagnosis, a package of support interventions should be offered to +ensure timely linkage to care for all people living with HIV (Strong recommendation, +moderate-certainty evidence). + +2021 Health systems should invest in people-centred practices and communication, +including ongoing training, mentoring, supportive supervision and monitoring +of health workers, to improve the relationships between clients and health-care +providers (Good practice statement). + +2019 — ART initiation should follow the overarching principles of providing people-centred +care. People-centred care should be focused and organized around the health needs, +preferences and expectations of people and communities, upholding individual +dignity and respect, especially for vulnerable populations (Good practice statement). + +2021 Antiretroviral therapy (ART) initiation may be offered outside the health facility +(Conditional recommendation, low- to moderate-certainty evidence). + +2017 Rapid ART initiation should be offered to all people living with HIV following a +confirmed HIV diagnosis and clinical assessment (Strong recommendation: high- +certainty evidence for adults, low-certainty evidence for children) + +2021 ~The offer of same-day ART initiation should include approaches to improve uptake, +treatment adherence and retention such as tailored client education, counselling and +support (Good practice statement). + +Clinical visits and medication pick-up + +2021 People established on ART should be offered clinical visits every 3-6 months, preferably +every six months if feasible (Strong recommendation, moderate-certainty evidence). + +2021 — People established on ART should be offered refills of ART lasting 3-6 months, +preferably six months if feasible (Strong recommendation, moderate-certainty evidence). + +Adherence and engagement in care + +2025 _— Adherence support interventions should be provided to people on ART (Strong +recommendation, moderate-certainty evidence). + +2021 Viral load for treatment monitoring should be complemented with non-judgemental, +tailored approaches to assessing adherence (Good practice statement). + +2021 HIV programmes should implement interventions to trace people who + +have disengaged from care and provide support for re-engagement (Strong +recommendation, low-certainty evidence). + +Services for adolescents and young people + +2021 Psychosocial interventions should be provided to all adolescents and young adults +living with HIV (Strong recommendation, moderate-certainty evidence). + +2013 Adolescent-friendly services should be implemented in HIV services to ensure engagement +and improved outcomes (Strong recommendation, moderate-certainty evidence). + + +Table 1: HIV service delivery recommendations (cont’d) + +Date* + +2013 + +Recommendations + +Services for adolescents and young people + +Adolescents should be counselled about the potential benefits and risks of disclosing +their HIV status to others and empowered and supported to determine whether, when, +how and to whom to disclose (Conditional recommendation, very-low -certainty evidence). + +Task sharing + +Executive summary + +Revalidated + +2016 + +2016 + +2016 + +2016 + +2021 + +Trained non-physician clinicians, midwives and nurses can initiate first-line ART +(Strong recommendation, moderate-certainty evidence). + +Trained non-physician clinicians, midwives and nurses can maintain ART (Strong +recommendation, moderate-certainty evidence). + +Trained and supervised community health workers can dispense ART between regular +clinical visits (Strong recommendation, moderate-certainty evidence). + +Trained and supervised lay health-care providers can distribute ART (Strong +recommendation, low-certainty evidence). + +Task sharing of specimen collection and point-of-care testing with non-laboratory +personnel should be implemented when professional staffing capacity is limited +(Strong recommendation, moderate-certainty evidence). + +Revalidated + +Revalidated +Revalidated + +Revalidated + +Revalidated + +2013 + +2013 + +2021 + +2013 + +2013 + +2025 + +2025 + +Service Delivery models and integration + +Decentralization of ART care should be considered as a way to increase access and +improve retention in care (Strong recommendation, moderate-to-low-certainty evidence). + +In generalized epidemic settings, ART should be initiated and maintained in +pregnant and postpartum women and in infants at maternal and child health-care +settings, with linkage and referral to ongoing HIV care and ART, if appropriate (Strong +recommendation, low-certainty evidence). + +Sexual and reproductive health services, including contraception, may be integrated +with HIV services (Conditional recommendation, very-low-certainty evidence). + +In settings with a high burden of HIV and tuberculosis (TB), ART should be initiated +in TB treatment settings, with linkage to ongoing HIV care and ART (Strong +recommendation, very-low-certainty evidence). + +In settings with a high burden of HIV and TB, TB treatment may be provided for +people living with HIV in HIV care settings if they have been diagnosed with TB (Strong +recommendation, very-low-certainty evidence). + +Diabetes and hypertension care should be integrated with HIV services (Strong +recommendation, moderate-to-very-low-certainty evidence). + +Mental health care for depression, anxiety and alcohol use disorders should be integrated +with HIV services (Strong recommendation, moderate-to-very-low-certainty evidence). + +* Date the recommendation/good practice statement was formulated + +Revalidated + +Revalidated + +Revalidated + +Revalidated + +Revalidated + +Updated + + + +1. Introduction + +HIV remains a major public health challenge and a leading cause of illness and death in many low- and middle- +income countries, despite the wide availability of effective antiretroviral therapy and proven prevention tools. + +In many settings, evidence-informed interventions for HIV prevention, testing, treatment and long-term care are +not fully implemented, delivered inconsistently or reach people too late, contributing to ongoing transmission, +advanced disease and avoidable mortality. Persistent health system constraints - including workforce shortages, +weak supply chains and limited integration with related services such as those for noncommunicable diseases and +mental health care - further hinder effective service delivery. + +These HIV service delivery guidelines provide updated and consolidated recommendations to support countries, +programme managers, health-care workers, communities and other stakeholders to organize services that are +people-centred, integrated, equitable and grounded in the best available evidence. The guidelines aim to improve +access to and continuity of high-quality HIV services, reduce inequities in outcomes and accelerate progress towards +ending AIDS as a public health threat. + + + +2.1 Linkage from HIV testing to enrolment in care + +Following an HIV diagnosis, a package of support interventions should be offered to +ensure timely linkage to care for all people living with HIV. + +(Strong recommendation, moderate-certainty evidence) + +The following interventions have demonstrated benefit in improving linkage to care following an +HIV diagnosis: + +¢ streamlined interventions to reduce time between diagnosis and engagement in care, including +(i) enhanced linkage with case management; (ii) support for HIV disclosure; (iii) tracing; (iv) training +staff to provide multiple services; and (v) streamlined services (moderate-certainty evidence); + +peer support? and navigation approaches for linkage (moderate-certainty evidence); and + +quality improvement approaches using data to improve linkage (low-certainty evidence). + +a Includes peer counselling. -Revalidated - + +Good practice statements + +ART initiation should follow the overarching principles of providing people-centred care. + +People-centred care is care that is focused and organized around the health needs, preferences and +expectations of people and communities, upholding individual dignity and respect, especially for +vulnerable populations. It should promote the engagement and support of people and families to play an +active role in their own care through informed decision-making. + +All people newly diagnosed with HIV should be retested to verify their HIV status before starting ART, +using the same testing strategy and algorithm as the initial test. To minimize the risk of misdiagnosis, +this approach should be maintained in settings in which rapid ART initiation is being implemented. + +Background + +Linkage to HIV prevention, treatment and care after testing remains a challenge, with disproportionate gaps for key +populations, men, young people and previously diagnosed people living with HIV (not receiving ART, disengaged or +lost to care). + +HIV testing service providers are primarily responsible for linkage to care. People testing HIV positive should be +assessed for advanced HIV disease using CD4 cell count testing and offered ART, and those testing negative should +be linked to prevention services. + + + +Consolidated HIV guidelines + +Rationale + +Evidence-informed interventions to support linkage can be classified into three approaches (2): + +streamlined services via enhanced linkage with case management, HIV disclosure support, client tracing for +non-engagers and staff training for multiple services; + +peer support and navigation interventions, including home visits, system navigation and enhanced +counselling; and + +quality improvement interventions enhance linkage to care (3). + +Implementation considerations + +Interventions may vary based on the context, including health-care delivery systems, geography and target +population. + +A combination of interventions may be needed to improve linkage for specific underserved groups, especially +key populations and men. + +Barriers to linkage to HIV care and ART initiation should be assessed and addressed, especially for populations +facing multiple structural and individual barriers. + +Linkage for children can be improved using point-of-care early infant diagnosis, SMS/GSM/GPRS printers to +accelerate access to laboratory results and family-centred models delivering care to mother, baby, and partner +at the same point. + +Health facilities should have capacity to provide services for key populations. Community- and peer-based +approaches can be especially effective in restrictive legal settings. + +WHO guidelines on HIV testing services (4) provide information on post-test services and linkage to care. + +2. + +2 People-centred care + +Good practice statement + +Health systems should invest in people-centred practices and communication, including ongoing training, +mentoring, supportive supervision and monitoring of health workers, to improve the relationships +between clients and health-care providers. + +HIV programmes should: + +e provide people-centred care that is focused and organized around the health needs, preferences and +expectations of people and communities, upholding individual dignity and respect, especially for +underserved populations, and engage and support people and families to play an active role in their +own care by informed decision-making; + +e offer safe, acceptable and appropriate clinical and non-clinical services in a timely fashion, aiming to +reduce morbidity and mortality associated with HIV infection and to improve health outcomes and +quality of life in general; and + +e promote the efficient and effective use of resources. + +Revalidated + + +Linkage and ART initiation + +Background + +People-centred health services are an approach to care that consciously adopts the perspectives of individuals, +families and communities and sees them as participants and beneficiaries of trusted health systems that respond +to their needs and preferences in humane and holistic ways. This approach acknowledges the experiences and +perspectives of health-care providers that may enable or prevent the delivery of high-quality people-centred care. + +Rationale + +« Asystematic review (5) identified several interventions to enhance people-centred care for people living +with HIV. Overall, studies reported beneficial effects of these approaches across the HIV cascade outcomes, +including improved ART uptake, adherence and viral load suppression. + +e Studies have shown that people are willing to travel longer distances to be seen by a health-care provider with +a respectful and caring attitude, and negative health-care worker attitudes contribute to loss to care and poor +programme outcomes (6). + +Interventions found to improve relationships between clients and health-care providers can be classified into the +following approaches: + +¢ providing friendly and welcoming services: training providers to make general HIV services more welcoming, +including for clients who have disengaged from care; + +* conducting sensitization training for clinical and non-clinical health-care providers to improve care for key +populations both at the primary care and community levels; such training should include issues related to +stigma and discrimination; + +¢ offering individualized adherence counselling and person-centred communication such as shared decision- +making and planning for ART initiation and adherence; + +° facilitating client education in empowerment and communication skills; and + +* providing feedback to health-care workers on client concerns and evaluation of service quality such as +community score cards and client feedback mechanisms combined with quality improvement exercises. + +Implementation considerations + +¢ Build capacity for ongoing training, mentoring and supportive supervision on people-centred communication, +stigma reduction and welcoming services. + +« Integrate sensitization training for all staff (clinical and non-clinical) on respectful attitudes, stigma and +discrimination, re-engagement of clients who have disengaged, and support for women and other vulnerable +groups experiencing violence or sexual health concerns. + +« Establish routine client feedback mechanisms (such as satisfaction questionnaires, community score cards +and suggestion boxes) and use the results in facility quality-improvement meetings. + +¢ Involve clients and community representatives in designing and reviewing services. + +* Review workload, staffing and client flow to allow time for person-centred interactions. diff --git a/datasets/test/fixtures/pdf/README.md b/datasets/test/fixtures/pdf/README.md new file mode 100644 index 0000000..698e7dc --- /dev/null +++ b/datasets/test/fixtures/pdf/README.md @@ -0,0 +1,15 @@ +# WHO PDF excerpts + +These fixtures were contributed by Ashu in [AMFV PR #6](https://github.com/MedARC-AI/amfv/pull/6), +commit `ca7a330fc2e87022e24e6b1b72130b3ec5002408`. +The PDFs are unchanged apart from page selection. + +| Fixture | WHO publication | Source PDF pages | +| --- | --- | --- | +| `who_9789240121805_excerpt.pdf` | [Bloodstream infections: central venous catheters](https://www.who.int/publications/i/item/9789240121805) | 30–35 | +| `who_9789240124233_excerpt.pdf` | [Consolidated HIV guidelines: service delivery](https://www.who.int/publications/i/item/9789240124233) | 11–16 | + +© World Health Organization. Licensed under [CC BY-NC-SA 3.0 IGO](https://creativecommons.org/licenses/by-nc-sa/3.0/igo/). +The repository's Apache-2.0 code licence does not cover these PDFs or their +extracted text. Retain attribution and the source licence when redistributing +the fixtures or derivative benchmark outputs. diff --git a/datasets/test/fixtures/pdf/who_9789240121805_excerpt.pdf b/datasets/test/fixtures/pdf/who_9789240121805_excerpt.pdf new file mode 100644 index 0000000..649cb7f Binary files /dev/null and b/datasets/test/fixtures/pdf/who_9789240121805_excerpt.pdf differ diff --git a/datasets/test/fixtures/pdf/who_9789240124233_excerpt.pdf b/datasets/test/fixtures/pdf/who_9789240124233_excerpt.pdf new file mode 100644 index 0000000..bf25b8b Binary files /dev/null and b/datasets/test/fixtures/pdf/who_9789240124233_excerpt.pdf differ diff --git a/datasets/test/test_pdf_ocr.py b/datasets/test/test_pdf_ocr.py new file mode 100644 index 0000000..39439e7 --- /dev/null +++ b/datasets/test/test_pdf_ocr.py @@ -0,0 +1,76 @@ +"""Exercise the standalone PDF/OCR diagnostic's public command.""" + +import hashlib +import json +import os +import subprocess +import sys +from pathlib import Path + +import pytest + +_DATASETS = Path(__file__).resolve().parents[1] +_RUNNER = _DATASETS / "benchmarks" / "pdf_ocr.py" +_PDF_FIXTURES = _DATASETS / "test" / "fixtures" / "pdf" + + +def test_firecrawl_without_credentials_records_skips(tmp_path: Path) -> None: + """A missing API key records every requested skip without calling Firecrawl.""" + environment = os.environ.copy() + environment.pop("FIRECRAWL_API_KEY", None) + result = subprocess.run( + [sys.executable, str(_RUNNER), "--output-dir", str(tmp_path), "--backend", "firecrawl-auto", "firecrawl-ocr"], + env=environment, + capture_output=True, + text=True, + timeout=60, + check=False, + ) + + assert result.returncode == 0, result.stderr + report = json.loads((tmp_path / "results.json").read_text()) + assert len(report["results"]) == 4 + assert {entry["backend"] for entry in report["results"]} == {"firecrawl-auto", "firecrawl-ocr"} + assert all(entry["status"] == "skipped" for entry in report["results"]) + assert all("FIRECRAWL_API_KEY" in entry["reason"] for entry in report["results"]) + assert not list(tmp_path.glob("*.md")) + + +def test_native_backend_retains_pdf_provenance_and_raw_text(tmp_path: Path) -> None: + """Native extraction retains the source hashes and clinically relevant wording.""" + pytest.importorskip("pypdfium2", reason="requires the root ocr dependency group") + result = subprocess.run( + [sys.executable, str(_RUNNER), "--output-dir", str(tmp_path), "--backend", "native"], + capture_output=True, + text=True, + timeout=60, + check=False, + ) + + assert result.returncode == 0, result.stderr + report = json.loads((tmp_path / "results.json").read_text()) + assert len(report["results"]) == 2 + for entry in report["results"]: + assert entry["status"] == "ok" + assert entry["sha256"] == hashlib.sha256((_PDF_FIXTURES / entry["document"]).read_bytes()).hexdigest() + markdown = (tmp_path / entry["output"]).read_text() + assert markdown.strip() + assert "certainty" in markdown + cvc = next(entry for entry in report["results"] if "1805" in entry["document"]) + assert cvc["anchors_found"] == 7 + + +def test_invalid_raster_resolution_is_rejected_before_output(tmp_path: Path) -> None: + """An invalid DPI fails before creating an experiment directory.""" + output = tmp_path / "invalid" + result = subprocess.run( + [sys.executable, str(_RUNNER), "--output-dir", str(output), "--dpi", "0"], + capture_output=True, + text=True, + timeout=60, + check=False, + ) + + assert result.returncode == 2 + assert "--dpi must be at least 72" in result.stderr + assert not output.exists() diff --git a/pyproject.toml b/pyproject.toml index ba89b1e..8d532e7 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -16,6 +16,7 @@ dev = [ "pytest>=8.0", "ruff>=0.11", ] +ocr = ["pypdfium2>=5.14.0,<6.0.0", "pillow>=11.0.0,<13.0.0"] [tool.uv] package = false diff --git a/uv.lock b/uv.lock index 0bfc4aa..3e78de1 100644 --- a/uv.lock +++ b/uv.lock @@ -137,6 +137,10 @@ dev = [ { name = "pytest" }, { name = "ruff" }, ] +ocr = [ + { name = "pillow" }, + { name = "pypdfium2" }, +] [package.metadata] @@ -150,6 +154,10 @@ dev = [ { name = "pytest", specifier = ">=8.0" }, { name = "ruff", specifier = ">=0.11" }, ] +ocr = [ + { name = "pillow", specifier = ">=11.0.0,<13.0.0" }, + { name = "pypdfium2", specifier = ">=5.14.0,<6.0.0" }, +] [[package]] name = "amfv-datasets" @@ -869,6 +877,68 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/0f/54/68a0978d1ef8502b8492099beaa6e7a0c1b32e3b5d4f677f5810cb08711c/pandas-3.0.3-cp314-cp314t-win_arm64.whl", hash = "sha256:b2c95f8bfc1ee412bf482605d7bfd30c12d1d26bd59fdd91efeef1d4718decb1", size = 9466464, upload-time = "2026-05-11T18:54:22.754Z" }, ] +[[package]] +name = "pillow" +version = "12.3.0" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/1c/3d/bb7fca845737cf9d7dbde16ed1843984665ff2e0a518f5db43e77ec540b9/pillow-12.3.0.tar.gz", hash = "sha256:3b8182a766685eaa002637e28b4ec8d6b18819a0c71f579bf0dbaa5830297cce", size = 47025035, upload-time = "2026-07-01T11:56:38.965Z" } +wheels = [ + { url = "https://files.pythonhosted.org/packages/9d/ac/31fb64e1e7efb5a4b50cd3d92049ba89ac6e4d8d3bb6a74e15048ca3353e/pillow-12.3.0-cp313-cp313-ios_13_0_arm64_iphoneos.whl", hash = "sha256:21900ce7ba264168cd50defae43cd75d25c833ad4ad6e73ffc5596d12e25ac89", size = 4161684, upload-time = "2026-07-01T11:54:25.934Z" }, + { url = "https://files.pythonhosted.org/packages/87/b4/9805e23d2b4d77842b468513841fda254ee42f0289d25088340e4ff46e2d/pillow-12.3.0-cp313-cp313-ios_13_0_arm64_iphonesimulator.whl", hash = "sha256:4e8c2a84d977f50b9daed6eeaf3baef67d00d5d74d932288f02cb94518ee3ace", size = 4255487, upload-time = "2026-07-01T11:54:27.935Z" }, + { url = "https://files.pythonhosted.org/packages/df/39/ecf519435a200c693fe053a6ee4d835b41cf963a4dfc2551c4e637cb2a71/pillow-12.3.0-cp313-cp313-ios_13_0_x86_64_iphonesimulator.whl", hash = "sha256:ae26d61dfa7a47befdc7572b521024e8745f3d809bd95ca9505a7bba9ef849ec", size = 3696433, upload-time = "2026-07-01T11:54:29.813Z" }, + { url = "https://files.pythonhosted.org/packages/42/92/2fc3ffad878ae8dd5469ec1bc8eb83b71f48e13efdf68f02709003982a32/pillow-12.3.0-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:7a743ff716f746fc19a9557f60dab1600d4613255f8a7aeb3cdde4db7eb15a66", size = 5345889, upload-time = "2026-07-01T11:54:31.97Z" }, + { url = "https://files.pythonhosted.org/packages/10/76/8803c13605b763d33d156c4678fc77f8443389c0c51c8aef707bb02015f4/pillow-12.3.0-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:d69141514cc30b774ceea5e3ed3a6635c8d8a96edf664689b890f4089111fb35", size = 4780109, upload-time = "2026-07-01T11:54:34.026Z" }, + { url = "https://files.pythonhosted.org/packages/1f/01/e18aff37cb0b4aac47ac90f016d347a49aca667ef97f190b06ac2aabc928/pillow-12.3.0-cp313-cp313-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:f7401aebd7f581d7f83a439d87d474999317ee099218e5ad25d125290990ba65", size = 6263736, upload-time = "2026-07-01T11:54:36.131Z" }, + { url = "https://files.pythonhosted.org/packages/f7/62/de5bdd77d935331f4f802edc11e4d82950f642caad6cb2f949837b8560e2/pillow-12.3.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:0847a763afefb695bc912d7c131e7e0632d4edc1d8698f58ddabec8e46b8b6d3", size = 6937129, upload-time = "2026-07-01T11:54:38.216Z" }, + { url = "https://files.pythonhosted.org/packages/70/4d/105627a13300c5e0df1d174230b32fd1273062c96f7745fd552b945d1e1d/pillow-12.3.0-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:571b9fcb07b97ef3a492028fb3d2dc0993ca23a06138b0315286566d29ef718a", size = 6339562, upload-time = "2026-07-01T11:54:40.354Z" }, + { url = "https://files.pythonhosted.org/packages/6b/1d/f13de01a553988ab895ba1c722e06cf3144d4f57656fd5b81b6d881f1179/pillow-12.3.0-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:756c768d0c9c2955feb7a56c37ea24aea2e369f8d36a88da270b6a9f19e62b5e", size = 7049439, upload-time = "2026-07-01T11:54:42.489Z" }, + { url = "https://files.pythonhosted.org/packages/c9/f9/066794cca041b969964f779ee5fa66a9498bbf34248ac39c5d7954e4198f/pillow-12.3.0-cp313-cp313-win32.whl", hash = "sha256:a876864214e136f0eb367788dbd7df045f4806801518e2cfe9e13229cfe06d8f", size = 6473287, upload-time = "2026-07-01T11:54:44.9Z" }, + { url = "https://files.pythonhosted.org/packages/a6/9b/7a58e61d62be561da3a356fe2384d4059a6345fc130e23ef1c36a5b81d24/pillow-12.3.0-cp313-cp313-win_amd64.whl", hash = "sha256:1cca606cd25738df4ed873d5ad46bbdb3d83b5cbca291f6b4ff13a4df6b0bbe8", size = 7239691, upload-time = "2026-07-01T11:54:47.141Z" }, + { url = "https://files.pythonhosted.org/packages/aa/b0/c4ed4f0ef8f8fa5ee8351537db6650bb8189f7e118842978dd6589065692/pillow-12.3.0-cp313-cp313-win_arm64.whl", hash = "sha256:b629de27fda84b42cde7edef0d85f13b958b47f6e9bbcbba9b673c562a89bd8b", size = 2568185, upload-time = "2026-07-01T11:54:49.137Z" }, + { url = "https://files.pythonhosted.org/packages/dc/01/001f65b68192f0228cc1dbbc8d2530ab5d58b61037ba0587f946fea607cd/pillow-12.3.0-cp314-cp314-ios_13_0_arm64_iphoneos.whl", hash = "sha256:9cf95fe4d0f84c82d282745d9bb08ad9f926efa00be4697e767b814ce40d4330", size = 4161736, upload-time = "2026-07-01T11:54:51.156Z" }, + { url = "https://files.pythonhosted.org/packages/1a/d2/0219746d0fd16fc8a84498e79452375be3797d3ce4044596ce565164b84f/pillow-12.3.0-cp314-cp314-ios_13_0_arm64_iphonesimulator.whl", hash = "sha256:8728f216dcdb6e6d555cf971cb34076139ad74b31fc2c14da4fafc741c5f6217", size = 4255435, upload-time = "2026-07-01T11:54:53.414Z" }, + { url = "https://files.pythonhosted.org/packages/c8/02/8d0bc62ef0302318c46ff2a512822d2610e81c7aa46c9b3abe6cbaca5ad0/pillow-12.3.0-cp314-cp314-ios_13_0_x86_64_iphonesimulator.whl", hash = "sha256:a45650e8ce7fafffd731db8550230db6b0d306d181a90b67d3e6bca2f1990930", size = 3696262, upload-time = "2026-07-01T11:54:55.739Z" }, + { url = "https://files.pythonhosted.org/packages/85/e2/73c77d218410b14f5f2d565e8a998d5317b7b9c75368d29985139f7a46f0/pillow-12.3.0-cp314-cp314-macosx_10_15_x86_64.whl", hash = "sha256:ba54cfebe86920a559a7c4d6b9050791c20513650a1952ebe3368c7dc70306f8", size = 5350344, upload-time = "2026-07-01T11:54:57.657Z" }, + { url = "https://files.pythonhosted.org/packages/c7/da/32c752228ae345f489e3a42499d817b6c3996da7e8a3bc7a04fc806b243b/pillow-12.3.0-cp314-cp314-macosx_11_0_arm64.whl", hash = "sha256:e158cb00350dc278f3b91551101aa7d12415a66ebf2c91d8d5ac14e56ddd3ad0", size = 4780131, upload-time = "2026-07-01T11:54:59.713Z" }, + { url = "https://files.pythonhosted.org/packages/b1/9d/8b2c807dbef61a5197c047afe99823787eb66f63daf9fb2432f91d6f0462/pillow-12.3.0-cp314-cp314-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:e9aeb04d6aef139de265b29683e119b638208f88cf73cdd1658aa07221165321", size = 6263757, upload-time = "2026-07-01T11:55:01.778Z" }, + { url = "https://files.pythonhosted.org/packages/5c/44/c85361f65dbe00eea8576ee467c768d25129989efb76e94f205e9ca9bb46/pillow-12.3.0-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:251bf95b67017e27b13d82f5b326234ca62d70f9cf4c2b9032de2358a3b12c7b", size = 6936962, upload-time = "2026-07-01T11:55:03.93Z" }, + { url = "https://files.pythonhosted.org/packages/18/7e/e483414b35800b86b6f08dbbc7803fb5cd52c4d6f897f47d53ea2c7e6f65/pillow-12.3.0-cp314-cp314-musllinux_1_2_aarch64.whl", hash = "sha256:fe3cca2e4e8a592be0f269a1ca4835c25199d9f3ce815c8491048f785b0a0198", size = 6339171, upload-time = "2026-07-01T11:55:05.989Z" }, + { url = "https://files.pythonhosted.org/packages/f0/f4/68c491844841ede6bed70189546b3ee9731cf9f2cbad396faff5e1ccba45/pillow-12.3.0-cp314-cp314-musllinux_1_2_x86_64.whl", hash = "sha256:23aceaa007d6172b02c277f0cd359c79492bbb14f7072b4ede9fbcaf20648130", size = 7048116, upload-time = "2026-07-01T11:55:08.131Z" }, + { url = "https://files.pythonhosted.org/packages/a3/34/77f3f793fed8efc7d243f21b33c5a3f0d1c97ee70346d3db855587e155ff/pillow-12.3.0-cp314-cp314-win32.whl", hash = "sha256:af8d94b0db561cf68b88a267c5c44b49e134f525d0dc2cb7ed413a66bc23559a", size = 6467209, upload-time = "2026-07-01T11:55:10.408Z" }, + { url = "https://files.pythonhosted.org/packages/f1/e0/492879f69d94f91f60fc8cd05ba03650e9520afebb2fb7aa12777d7c7f38/pillow-12.3.0-cp314-cp314-win_amd64.whl", hash = "sha256:fdafc9cce40277e0f7a0feabce0ee50dd2fa1800f3b38015e51296b5e814048d", size = 7237707, upload-time = "2026-07-01T11:55:12.745Z" }, + { url = "https://files.pythonhosted.org/packages/c9/ac/6b11f2875f1c2ac040d84e1bbf9cf22a88038f901ca1037898b280b38365/pillow-12.3.0-cp314-cp314-win_arm64.whl", hash = "sha256:e91206ee562682b51b98ef4b26a6ef48fd84e15fd4c4bc5ec768eb641d206838", size = 2565995, upload-time = "2026-07-01T11:55:14.736Z" }, + { url = "https://files.pythonhosted.org/packages/52/69/c2208e56af9bfc1913afb24020297a691eb1d4ef688474c8a04913f65e04/pillow-12.3.0-cp314-cp314t-macosx_10_15_x86_64.whl", hash = "sha256:164b31cd1a0490ab6efae01aa5df49da7061be0af1b30e035b6e9a1bfe34ee6e", size = 5352503, upload-time = "2026-07-01T11:55:17.076Z" }, + { url = "https://files.pythonhosted.org/packages/07/70/e5686d753e898a45d778ff1718dba8516ead6ab6b95d85fc8c4b70650cf2/pillow-12.3.0-cp314-cp314t-macosx_11_0_arm64.whl", hash = "sha256:5afb51d599ea772b8365ae807ae557f18bccfe46ab261fd1c2a9ed700fc6eb17", size = 4782956, upload-time = "2026-07-01T11:55:19.448Z" }, + { url = "https://files.pythonhosted.org/packages/d5/37/25c6692f06927ee973ff18c8d9ee98ad0b4d84ee67a09610c2dd1447958e/pillow-12.3.0-cp314-cp314t-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:3edce1d53195db527e0191f84b71d02022de0540bf43a16ed734ed7537b07385", size = 6322855, upload-time = "2026-07-01T11:55:21.613Z" }, + { url = "https://files.pythonhosted.org/packages/cc/91/420637fcb8f1bc11029e403b4538e6694744428d8246118e45719f944556/pillow-12.3.0-cp314-cp314t-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:bf16ba1b4d0b6b7c8e534936632270cf70eb00dbe09005bc345b2677b726855c", size = 6989642, upload-time = "2026-07-01T11:55:24.006Z" }, + { url = "https://files.pythonhosted.org/packages/10/08/b94d7811281ccf0d143a1cf768d1c49e1e54af63e7b708ab2ee3eb87face/pillow-12.3.0-cp314-cp314t-musllinux_1_2_aarch64.whl", hash = "sha256:24870b09b224f7ae3c39ed07d10e819d06f8720bc551847b1d623832b5b0e28d", size = 6391281, upload-time = "2026-07-01T11:55:26.252Z" }, + { url = "https://files.pythonhosted.org/packages/d2/87/24233f785f55474dc02ce3e739c5528a77e3a862e9333d1dd7a25cc31f70/pillow-12.3.0-cp314-cp314t-musllinux_1_2_x86_64.whl", hash = "sha256:30f2aa603c41533cc25c05acd0da21636e84a315768feb631c937177db558931", size = 7096716, upload-time = "2026-07-01T11:55:28.318Z" }, + { url = "https://files.pythonhosted.org/packages/23/26/fcb2f6e37175b04f53570b59937867e2b80ee1685e744023153028fc14f9/pillow-12.3.0-cp314-cp314t-win32.whl", hash = "sha256:4b0a7fe987b14c31ebda6083f74f22b561fd3739bc0ac51e019622e3d72668c7", size = 6474125, upload-time = "2026-07-01T11:55:30.956Z" }, + { url = "https://files.pythonhosted.org/packages/90/de/3634abee5f1c9e13c56787b7d5517b0ba8d6de51700b95578cf338349c9f/pillow-12.3.0-cp314-cp314t-win_amd64.whl", hash = "sha256:962864dc93511324d51ddbb5b9f8731bf71675b93ca612a07441896f4688fb8c", size = 7242939, upload-time = "2026-07-01T11:55:34.044Z" }, + { url = "https://files.pythonhosted.org/packages/ce/2a/fd13f8eb24de5714a6eb444a3d67e2842c6c576e159a43793adf23051351/pillow-12.3.0-cp314-cp314t-win_arm64.whl", hash = "sha256:0740a512dc522224c77d9aa5a8d70d8b7d73fb91f2c21125d8d025d3b8990e45", size = 2567506, upload-time = "2026-07-01T11:55:35.988Z" }, + { url = "https://files.pythonhosted.org/packages/5d/dc/8fdce34ec725a33c81c6ba122b904d6b9024e50ea9ac7bede62fab54506c/pillow-12.3.0-cp315-cp315-ios_13_0_arm64_iphoneos.whl", hash = "sha256:0feb2e9d6ad6c9e3c06effe9d00f3f1e618a6643273576b016f591e9315a7139", size = 4162063, upload-time = "2026-07-01T11:55:37.941Z" }, + { url = "https://files.pythonhosted.org/packages/76/66/2044b9a63d3b84ff048228dfcb7cd9bf0df983e8470971bf7d4c57b693de/pillow-12.3.0-cp315-cp315-ios_13_0_arm64_iphonesimulator.whl", hash = "sha256:9e881fca225083806662a5c43d627d215f258ff43c890f831966c7d7ba9c7402", size = 4255549, upload-time = "2026-07-01T11:55:40.022Z" }, + { url = "https://files.pythonhosted.org/packages/52/7e/1f67e6f4ece6b582ee4b539decbcc9f848dc245a93ed8cd7338bafef72f1/pillow-12.3.0-cp315-cp315-ios_13_0_x86_64_iphonesimulator.whl", hash = "sha256:4998562bf62a445225f22e07c896bb04b35b1b1f2eb6d760584c9c51d7a5f78c", size = 3696331, upload-time = "2026-07-01T11:55:41.98Z" }, + { url = "https://files.pythonhosted.org/packages/12/40/d306fc2c8e4d45d7f175c77edca7063be7b86fe7fe6e68f4353bf71d808c/pillow-12.3.0-cp315-cp315-macosx_10_15_x86_64.whl", hash = "sha256:dc624f6bc473dacdf7ef7eb8678d0d08edf15cd94fad6ae5c7d6cc67a4e4902f", size = 5350370, upload-time = "2026-07-01T11:55:44.028Z" }, + { url = "https://files.pythonhosted.org/packages/dd/44/668fb1437e8ce420f62d6106eb66e44a5971602a4d794615bdf79315d82d/pillow-12.3.0-cp315-cp315-macosx_11_0_arm64.whl", hash = "sha256:71d6097b330eea8fd15097780c8e89cb1a8ce7838669f48c5bacd6f663dd4701", size = 4780147, upload-time = "2026-07-01T11:55:46.073Z" }, + { url = "https://files.pythonhosted.org/packages/0c/08/93fa2e70e30a2d81547e481b6ee2bb9522117221fb1e0ce4b5df70967677/pillow-12.3.0-cp315-cp315-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:28ce87c5ab450a9dd970b52e5aca5fe63ed432d18a2eaddd1979a00a1ba24ace", size = 6273659, upload-time = "2026-07-01T11:55:48.264Z" }, + { url = "https://files.pythonhosted.org/packages/f8/6d/043e96ff814fc31a33077e4cba86082167db520c93632afdf2042febbb0c/pillow-12.3.0-cp315-cp315-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:6b02afb9b97f65fbca5f31db6a2a3ba21aa93030225f150fa3f249717e938fb4", size = 6947439, upload-time = "2026-07-01T11:55:50.503Z" }, + { url = "https://files.pythonhosted.org/packages/af/92/ba71d2ee2ac0edf3fa33bd9d5ee9ee080da70b1766f3ca3934f9938ddac9/pillow-12.3.0-cp315-cp315-musllinux_1_2_aarch64.whl", hash = "sha256:1182d52bc2d5e5d7d0949503aa7e36d12f42205dc287e4883f407b1988820d39", size = 6353577, upload-time = "2026-07-01T11:55:52.697Z" }, + { url = "https://files.pythonhosted.org/packages/0f/ce/e63064e2122923ff687c8ad792d0d736a7b3920a56a46982e81a7fdd25d6/pillow-12.3.0-cp315-cp315-musllinux_1_2_x86_64.whl", hash = "sha256:e795b7eb908249c4e43c7c99fac7c2c75dab0c43566e37db472a355f63693d71", size = 7060394, upload-time = "2026-07-01T11:55:55.149Z" }, + { url = "https://files.pythonhosted.org/packages/54/76/a09cc3ccc8d773a7283d34c38bec1708f9e3cc932093cbc4c5e71ac4060b/pillow-12.3.0-cp315-cp315-win32.whl", hash = "sha256:57b3d78c95ba9059768b10e28b813002261d3f3dfc55cc48b0c988f625175827", size = 6467375, upload-time = "2026-07-01T11:55:57.769Z" }, + { url = "https://files.pythonhosted.org/packages/3e/03/1846c49ba3b1d5550392a4bbd06d6fb4578e1cd91a803198b5c90f5f7d53/pillow-12.3.0-cp315-cp315-win_amd64.whl", hash = "sha256:fa4ecea169a355be7a3ade2c783e2ed12f0e40d2c5621cda8b3297faf7fbb9f5", size = 7237048, upload-time = "2026-07-01T11:55:59.975Z" }, + { url = "https://files.pythonhosted.org/packages/fb/bb/89f35dcc79610423f9f195504d7def7f0d1416a711541b42867e25fe3412/pillow-12.3.0-cp315-cp315-win_arm64.whl", hash = "sha256:877c3f311ff35410f690861c4409e7ccbf0cd2f878e50628a28e5a0bb689e658", size = 2566006, upload-time = "2026-07-01T11:56:02.143Z" }, + { url = "https://files.pythonhosted.org/packages/30/88/707027ba09942dfa2c28759b5c222d769290a41c6d20ea60ec250801941f/pillow-12.3.0-cp315-cp315t-macosx_10_15_x86_64.whl", hash = "sha256:e9871b1ffbfa9656b60aeee92ed5136a5742696006fa322b29ea3d8da0ecc9cf", size = 5352509, upload-time = "2026-07-01T11:56:04.2Z" }, + { url = "https://files.pythonhosted.org/packages/b0/6d/00352fa25332c2569cd387851f568cc5a4b75a9adbfb37ac4fbce4c02eec/pillow-12.3.0-cp315-cp315t-macosx_11_0_arm64.whl", hash = "sha256:53aa02d20d10c3d814d536aa4e5ac9b84ca0ff5a88377963b085ad6822f93e64", size = 4783167, upload-time = "2026-07-01T11:56:06.631Z" }, + { url = "https://files.pythonhosted.org/packages/13/4f/9e049dfa21af7c22427275720e2490267ba8138120add5c4c574deb69782/pillow-12.3.0-cp315-cp315t-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:446c34dcc4324b084a53b705127dc15717b22c5e140ae0a3c38349d4efec071e", size = 6329237, upload-time = "2026-07-01T11:56:08.868Z" }, + { url = "https://files.pythonhosted.org/packages/36/16/cf6eeaae8d0fce8dd390a33437cf68c5d5bd73834a2bc6e2f14efda0ab45/pillow-12.3.0-cp315-cp315t-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:cf1845d02ad822a369a49f2bb9345b1614744267682e7a03527dc3bf6eea1777", size = 6997047, upload-time = "2026-07-01T11:56:11.379Z" }, + { url = "https://files.pythonhosted.org/packages/1e/69/dbf769bdd55f48bf5733cac28edc6364ffaa072ec9ba336266e4fe66be55/pillow-12.3.0-cp315-cp315t-musllinux_1_2_aarch64.whl", hash = "sha256:186941b6aef820ad110fb01fb06eb925374dc3a21b17e37ec9a53b250c6fe2d1", size = 6400440, upload-time = "2026-07-01T11:56:13.908Z" }, + { url = "https://files.pythonhosted.org/packages/a0/e1/ffc9cfc2eea0d178da8018e18e959301ad9d6bc9f3edb7181e748a474b97/pillow-12.3.0-cp315-cp315t-musllinux_1_2_x86_64.whl", hash = "sha256:f13c32a3abd6079a66d9526e18dad9b6d280384d49d7c54040cd57b6424041d9", size = 7105895, upload-time = "2026-07-01T11:56:16.575Z" }, + { url = "https://files.pythonhosted.org/packages/18/f0/a5595c1e8c3ae44b9828cb2f0fa8155e5095ef04d6327b8f61cf44a3df85/pillow-12.3.0-cp315-cp315t-win32.whl", hash = "sha256:1657923d2d45afb66526e5b933e5b3052e6bdea196c90d3abb2424e18c77dae8", size = 6474384, upload-time = "2026-07-01T11:56:18.855Z" }, + { url = "https://files.pythonhosted.org/packages/e4/04/62bcd9f844984c5938d3b05264a61d797a29d3e0812341a8204af70bbdee/pillow-12.3.0-cp315-cp315t-win_amd64.whl", hash = "sha256:8cd2f7bdda092d99c9fc2fb7391354f306d01443d22785d0cbfafa2e2c8bb418", size = 7243537, upload-time = "2026-07-01T11:56:21.214Z" }, + { url = "https://files.pythonhosted.org/packages/3d/68/1f3066acedf37673694a7141381d8f811ae97f30d34413d236abe7d489f1/pillow-12.3.0-cp315-cp315t-win_arm64.whl", hash = "sha256:06ff022112bc9cbf83b60f8e028d94ad87b60621706487e65f673de61610ab59", size = 2567491, upload-time = "2026-07-01T11:56:23.506Z" }, +] + [[package]] name = "pluggy" version = "1.6.0" @@ -1000,6 +1070,36 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/f4/7e/a72dd26f3b0f4f2bf1dd8923c85f7ceb43172af56d63c7383eb62b332364/pygments-2.20.0-py3-none-any.whl", hash = "sha256:81a9e26dd42fd28a23a2d169d86d7ac03b46e2f8b59ed4698fb4785f946d0176", size = 1231151, upload-time = "2026-03-29T13:29:30.038Z" }, ] +[[package]] +name = "pypdfium2" +version = "5.14.0" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/95/d0/c81d3a7c2a9af37b817ace1de0acd40cf44d15f12407c5e86b3668364a5c/pypdfium2-5.14.0.tar.gz", hash = "sha256:c5f009b3157f10e97dceb55963f5910eff92feb00587ba10a76f12b87ce1a4b6", size = 376498, upload-time = "2026-10-04T15:19:19.835Z" } +wheels = [ + { url = "https://files.pythonhosted.org/packages/91/03/79e89eac9d811e83d606342e129f5f39e168442ddf23b024fea4a7ee4762/pypdfium2-5.14.0-py3-none-android_23_arm64_v8a.whl", hash = "sha256:bed597b2cea3990164e43f9003f71db18959d0abd5d73adc9c176e7be2d84b98", size = 3453370, upload-time = "2026-10-04T15:18:40.79Z" }, + { url = "https://files.pythonhosted.org/packages/cc/68/369b80e408017b18eaecaa3c730bded07d90bfb65562215df200b56fb8e2/pypdfium2-5.14.0-py3-none-android_23_armeabi_v7a.whl", hash = "sha256:1951f0aed469150b13c62eabd501a9839e608ab9983ca8579be9eb73213b72b6", size = 2889924, upload-time = "2026-10-04T15:18:42.825Z" }, + { url = "https://files.pythonhosted.org/packages/d1/ea/14673bc9d8b7beeaa1eb46e9951b22543edaf2a4676c586e3b1e032ff6ee/pypdfium2-5.14.0-py3-none-macosx_13_0_arm64.whl", hash = "sha256:2de384df66ba55fcaab0775f30f28ec1090af3dfa60276a07821efc96d993118", size = 3542294, upload-time = "2026-10-04T15:18:44.345Z" }, + { url = "https://files.pythonhosted.org/packages/a6/11/b720097b01fa0874854f2f6669cbea4e4ea4e075769687714fac64d68964/pypdfium2-5.14.0-py3-none-macosx_13_0_x86_64.whl", hash = "sha256:e4e203ea9710fd00e5448edb6f1615dc8587035357f75f40b432dde0c33e8da1", size = 3735845, upload-time = "2026-10-04T15:18:45.975Z" }, + { url = "https://files.pythonhosted.org/packages/92/b4/0c31aa51887cd6cd032191dfe010a6d01ed43cf03204cfbd2184ebe4b715/pypdfium2-5.14.0-py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:f1b696e6901e16f114a2ec6332e5e3f8f5033a901614ead28499ab18ca6024f5", size = 3719672, upload-time = "2026-10-04T15:18:47.455Z" }, + { url = "https://files.pythonhosted.org/packages/93/a8/ae6ef96bf66559328d07b9e402ea704352ea00c49b6a73573da57e1fb378/pypdfium2-5.14.0-py3-none-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:593f2c952ae3ffdca0efcbb3d9464fbccb876254386114ff900cabef21157c3f", size = 3435593, upload-time = "2026-10-04T15:18:49.131Z" }, + { url = "https://files.pythonhosted.org/packages/59/ff/a78405fab4c8bad0ec25b49c5efba2c85ed14609ec73645f95220560bd81/pypdfium2-5.14.0-py3-none-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:d436ee9e024f981e68f5775f5a9d115f93ea14ee6c2c6efd35dd17d83edf4942", size = 3868604, upload-time = "2026-10-04T15:18:51.304Z" }, + { url = "https://files.pythonhosted.org/packages/5d/6e/09e9b62ab66c9acef5ad14f8a8c0d7b4d8d6ea6492e4e65b612ef146d373/pypdfium2-5.14.0-py3-none-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:f6f13bbcc5f4adabc2676e52f662c6cb375de86b314790b0ae08f3ab62eb116a", size = 4279333, upload-time = "2026-10-04T15:18:52.948Z" }, + { url = "https://files.pythonhosted.org/packages/4f/a3/c9cc797fc8bdfb8f37b9b0f8b9d02a5fc196b2015f408d53624cab5b0519/pypdfium2-5.14.0-py3-none-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:11f281613fa22313d9c7ab89947665e84eccf8ebe40e1198a84a88352305648d", size = 3799581, upload-time = "2026-10-04T15:18:54.913Z" }, + { url = "https://files.pythonhosted.org/packages/b9/76/54355a4bbd88bdd5ed3f4405bdc345eb593df9995daf90d285cbdf5c1410/pypdfium2-5.14.0-py3-none-manylinux_2_27_s390x.manylinux_2_28_s390x.whl", hash = "sha256:51d9e9b64ebc34effaf57f9b6d4511b3f66ad3744bd1690d2cc6700853173dcf", size = 4113022, upload-time = "2026-10-04T15:18:56.774Z" }, + { url = "https://files.pythonhosted.org/packages/7d/bc/ea461961ed0e0c4866df7a5610e76f769ef468bff28cd007e2aeecc8b882/pypdfium2-5.14.0-py3-none-manylinux_2_34_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:605ab9d0d4c5e223599c9065b88d16b2c1f131c807c80dea8adbb16f1433e95b", size = 4062832, upload-time = "2026-10-04T15:18:58.471Z" }, + { url = "https://files.pythonhosted.org/packages/32/30/dde99bc8cb3f8ace1d856095c2b4a29c80eecf9089b186a3b0845d0abc69/pypdfium2-5.14.0-py3-none-musllinux_1_2_aarch64.whl", hash = "sha256:382de7fe20d32c42993a274d7b6c555a5623a97570dfc1d2f5e0a16fe0d5d482", size = 5058436, upload-time = "2026-10-04T15:18:59.993Z" }, + { url = "https://files.pythonhosted.org/packages/ec/16/5314182dda2695fdf5bd414a450ee866087068cca4725703932770d4be04/pypdfium2-5.14.0-py3-none-musllinux_1_2_armv7l.whl", hash = "sha256:dbfd6deff68cc46b134acd6be380d98d694a9f018fbb622c07229225c85db389", size = 4595505, upload-time = "2026-10-04T15:19:01.835Z" }, + { url = "https://files.pythonhosted.org/packages/63/3f/474c42e726f0020095c7d5f3fb88cfd4e5d39c1361105a72899ada0ecd1b/pypdfium2-5.14.0-py3-none-musllinux_1_2_i686.whl", hash = "sha256:9f4d77db5232826dd03a63481f32164331b96c21fd68f0667b2e43dbae141a93", size = 5309775, upload-time = "2026-10-04T15:19:03.564Z" }, + { url = "https://files.pythonhosted.org/packages/6b/0c/723a6cf11cff00f125310d8c2c08362dc6c100d05fff8f92285a4df1bd41/pypdfium2-5.14.0-py3-none-musllinux_1_2_ppc64le.whl", hash = "sha256:b40a0913196a1483f0fdc22a53f8719c3aef87f1c4d8d9c38d2ad4e207500fdf", size = 5224565, upload-time = "2026-10-04T15:19:05.264Z" }, + { url = "https://files.pythonhosted.org/packages/5c/c5/86ab02a41e77a7aa962af6545a406815aeb9abaecd9f25dec34dbc336b72/pypdfium2-5.14.0-py3-none-musllinux_1_2_riscv64.whl", hash = "sha256:790e2cac1641a65912b73bd7243f45195d36f1663c85a3e1a126a8f5867c82a3", size = 4704416, upload-time = "2026-10-04T15:19:07.05Z" }, + { url = "https://files.pythonhosted.org/packages/ac/de/fb75013f924c5a4dde4a4a41ec13e7495f9b80022bf35dd51baa54e05910/pypdfium2-5.14.0-py3-none-musllinux_1_2_s390x.whl", hash = "sha256:09b99c8f0cb427eb17fec13c0862ed598bba34b4843df153f70fff806a2820bc", size = 5163621, upload-time = "2026-10-04T15:19:09.021Z" }, + { url = "https://files.pythonhosted.org/packages/cd/77/e59c814f10b533bc4565abe90ccef888ba29be45ada4627ebbf710961f0d/pypdfium2-5.14.0-py3-none-musllinux_1_2_x86_64.whl", hash = "sha256:e70d87cb0577eab38f2106f9c9606b458930beef612a1b5f298772ed259f5ec0", size = 5121606, upload-time = "2026-10-04T15:19:10.609Z" }, + { url = "https://files.pythonhosted.org/packages/21/25/e067396b4bdd26c19f0997bfa3422d3975a49ceec2c59668e7599f2adcba/pypdfium2-5.14.0-py3-none-pyemscripten_2026_0_wasm32.whl", hash = "sha256:c73be14076bedebd9bcaf9b062579c95c668580043bccd29eb0db502101d5716", size = 2675501, upload-time = "2026-10-04T15:19:12.588Z" }, + { url = "https://files.pythonhosted.org/packages/7f/0c/6c21f68a57d0c4c506b9e5f72506ba91d8dde47eef699f3fd9561f7bff0e/pypdfium2-5.14.0-py3-none-win32.whl", hash = "sha256:9fd5cc94a389d50298e4d8cb79af6b9b8e0d785606e2a937725dc6e271c9c6e6", size = 3805374, upload-time = "2026-10-04T15:19:14.357Z" }, + { url = "https://files.pythonhosted.org/packages/00/dc/ca7874924c9cfd701ad53f89529968523790e70473e0b71e834668316148/pypdfium2-5.14.0-py3-none-win_amd64.whl", hash = "sha256:149fd5c6397b8df8bf7911a93506eff0be874f877afe7ac936cf5d37d21a6a06", size = 3947280, upload-time = "2026-10-04T15:19:16.302Z" }, + { url = "https://files.pythonhosted.org/packages/46/ab/35f2276deeeebb781925e2647dd88a39f8ea1a910104a0dbb28218473502/pypdfium2-5.14.0-py3-none-win_arm64.whl", hash = "sha256:eb8aeca157808f323e39ea298cc6d6c8e080c192ea2efb1ca81daa0f0ff4d095", size = 3745021, upload-time = "2026-10-04T15:19:18.276Z" }, +] + [[package]] name = "pytest" version = "9.0.3"