Estimated filtration: a number she was never measured by
Educational use only. Not clinically reviewed81-year-old woman for Review of what the reported estimate claims, review of what it was generated from, a differently generated marker, reassessment, and handoff of an owned medicine decision with the uncertainty explicit. About 60 simulated minutes.
What you will practise
- Read the reported estimate as an estimate, with the width it actually carries.
- Ask what the estimate was built from before trusting it for this patient.
- Request a differently generated marker, and read disagreement as information.
- Distinguish a repeated marker from a narrower answer.
- Transfer the uncertainty explicitly, with the medicine decision owned.
Review and sources
Not clinically reviewed. No clinician has signed this scenario.
- Inker LA, Eneanya ND, Coresh J, Tighiouart H, et al. New creatinine- and cystatin C-based equations to estimate GFR without race. The New England Journal of Medicine. 2021;385:1737–1749. doi:10.1056/NEJMoa2102953. The accuracy quoted here is a population statistic: 85% within 30% of measured also means up to one value in seven falls outside even that band, and median biases of a few mL/min/1.73 m² in either direction are reported. Accuracy in development and validation cohorts is not a guarantee for an individual.
- Iversen E, Bengaard AK, Leegaard Andersen A, Tavenier J, et al. Performance of panel-estimated GFR among hospitalized older adults. American Journal of Kidney Diseases. 2023;82:715–724. doi:10.1053/j.ajkd.2023.05.004. One hundred and six patients at a single Danish hospital, a substudy of a trial rather than a designed accuracy study. The authors state that only White patients were included and that only a subset underwent GFR measurement, so the small biases it reports come from a narrow sample.
- Ibe Y, Ishigo T, Aigami T, Fujii S, et al. Effect of discrepancy in estimated renal function on vancomycin area under the blood concentration-time curve. International Journal of Clinical Pharmacy. 2025;47:1458–1466. doi:10.1007/s11096-025-01960-w. Retrospective, one cohort, one drug, and association rather than causation: the patients whose estimates disagreed were also older and longer in hospital, which are the same features that make both markers harder to interpret. It does not establish that either marker is correct.
Open Sim Lab is an educational simulator, not for clinical use. It is not a clinical decision-support tool, not a dosing calculator, and is not validated for any decision affecting a real patient.