eGFR Batch
Batch-compute eGFR with the published CKD-EPI 2021 equations from a CSV of age, sex and creatinine, with the creatinine unit (mg/dL or umol/L) declared explicitly — the tool refuses to guess it. Optionally adds cystatin-C eGFR, Cockcroft-Gault and de-indexed values. Use when asked to calculate eGFR for a cohort or CSV file, run one CKD-EPI calculation, or convert creatinine units for eGFR. Not for general lab-unit conversion (use labunits), child growth z-scores (use growthz), de-identification checks (use safeharbor-check), or randomisation lists (use blockrandom). Research/educational only; not a medical device.
egfr-batch
What this tool does
egfr-batch computes eGFR for a CSV of people with the published CKD-EPI 2021 equations (creatinine, and creatinine–cystatin C where cystatin C is given), returning the same CSV with the results, the converted creatinine actually used, and a per-row column naming the exact equation behind each number. It refuses to run unless the creatinine unit is declared: mg/dL and µmol/L differ by a factor of 88.4, and a guessed unit silently corrupts every row. It does arithmetic only — no diagnosis, no staging, no interpretation.
Install
pip install git+https://github.com/RakTikoo/egfr-batch
This provides the egfr-batch command. Verify an install with
egfr-batch check (recomputes a published worked example; prints PASS).
Run
egfr-batch convert patients.csv --scr-unit mg/dL --scys-unit mg/L --cockcroft-gault --cg-weight actual --deindex --out patients_egfr.csv
egfr-batch single --age 60 --sex F --scr 106 --scr-unit umol/L --scys 1.2 --scys-unit mg/L
- Input columns are
age,sex(F/M),creatinine, plus optionalcystatin_c,height_cm,weight_kg. --scr-unitis required:mg/dLorumol/L. µmol/L values are converted with the NIDDK factor (÷ 88.4); the value used is written to every row asscr_used_mg_dl.- If any row has cystatin C,
--scys-unit mg/Lis required. --cockcroft-gaultrequires an explicit--cg-weightofactual,idealoradjusted; the weight used is recorded per row.--deindexadds eGFR in mL/min using Mosteller BSA; it needsheight_cmandweight_kgon every row.
Reading the results
Exit code 0 means the run computed. The summary states the declared
unit, which equations ran on how many rows, the Cockcroft-Gault weight
choice, and the output file. Output values are rounded to 1 decimal
place (scr_used_mg_dl keeps 4, so the conversion is auditable), and
each result column is paired with a provenance column naming its
equation and version.
Exit code 2 means the tool refused and no output file was written:
a missing --scr-unit, cystatin C without --scys-unit, an age under
18 or over 130, a missing, non-numeric or non-positive value, or
Cockcroft-Gault without a declared weight choice. One bad row refuses
the whole run, with the row number in the error — there is no partial
output to misread.
Limits
- Research and educational tool only. Not a medical device. This tool does published-formula arithmetic only — it does not diagnose, stage, or interpret anything. Verify any result before clinical use.
- Arithmetic only, by design: no CKD staging, no diagnosis, no reference-range flagging, no interpretation. What a number means for a person is a clinical judgement this software does not make.
- Adults only: ages under 18 are refused. The CKD-EPI 2021 equations are adult equations; there is no paediatric equation in this tool.
- Assay assumptions are yours: the equations assume standardised (IDMS-traceable) creatinine and reference-standardised cystatin C; the tool cannot tell whether laboratory values meet that.
- eGFR is an estimate, not a measurement; known non-GFR determinants of creatinine and cystatin C (muscle mass, diet, some medicines, illness) affect the inputs and therefore the outputs.
- The published equations define female/male coefficients only; sex
values outside
F/Mare refused rather than mapped to something the publication does not define.
Related skills
- Downloads
- 73All time
- Security audit
- Pass
- Last updated
- 1d ago
- Current version
- v0.1.0
- License
- MIT-0
