Legacy Python analysis for a COVID-19 goals-of-care cohort from a state veterans home outbreak. The workflow summarizes admission characteristics, advance-care-planning variables, patient versus surrogate decision-maker groups, code-status transitions, and hospitalization outcomes.
This is an abstract-only legacy analysis repository. A verifiable abstract DOI or indexed abstract record was not recovered in the current public repository, CV, or website materials during this cleanup pass. Until that record is supplied, cite this repository as research software using CITATION.cff.
The patient-level source workbook is restricted clinical data and is not part of this public repository.
Maintainer: Brian W. Locke, MD, MSCI (@reblocke, ORCID 0000-0002-3588-5238).
Additional abstract authors, affiliations, funding, and conflicts of interest were not verified from current public repository materials. Add them here and in CITATION.cff if a citable abstract record is recovered.
Expected local-only input:
data/private/WorkingDb.xls
The workbook appears to contain row-level clinical and goals-of-care data. Do not commit source data, derived row-level exports, PHI, private workbooks, or local collaborator files. The public repository documents the expected schema and legacy workflow only.
python -m pip install -r requirements.txt
python SVH.py --input data/private/WorkingDb.xls --output-dir outputs/legacy-pythonGenerated outputs are written under the selected output directory:
output.xlsx: cleaned analysis tabletables.xlsx: legacy aggregate summary tablesstatistical_tests.txt: printed statistical test outputcode_status_alluvial.html: manually specified legacy alluvial/Sankey diagramfigures/*.png: generated distribution plots
| Path | Purpose |
|---|---|
SVH.py |
Legacy Python workflow for loading, cleaning, summarizing, testing, and plotting the cohort. |
data_dictionary.md, data_dictionary.csv |
Human-readable and machine-readable documentation of expected workbook columns, derived fields, and outputs. |
REPRODUCING.md |
Reproduction guide for running the workflow with a compatible local workbook. |
llms.txt |
Concise machine-readable index for LLM and search-agent use. |
CITATION.cff |
Structured repository-software citation metadata. |
tests/ |
No-PHI synthetic-workbook smoke tests. |
Historical aggregate outputs that were previously tracked in the repository root and dist figs/ are archived in the GitHub release legacy-covid-goc-outputs-2026-06-04.
| Dependency | Use |
|---|---|
| Python 3.10+ | Runtime used for the cleaned legacy workflow. |
| pandas, openpyxl, xlrd | Workbook loading and Excel export. |
| matplotlib, seaborn | Static distribution plots. |
| scipy | Legacy t tests and Fisher exact tests. |
| plotly | Code-status alluvial/Sankey HTML output. |
| pytest | No-PHI smoke tests. |
Use CITATION.cff for the repository citation. If a citable abstract record is later supplied, add it as preferred-citation and update this README and llms.txt.
Repository code is released under the MIT License. Restricted clinical data, third-party content, private drafts, and publisher-formatted text are excluded.
Use GitHub issues or pull requests for repository-specific questions.