Kinetic GFR: The Evidence Behind the Calculator

A 58-year-old man from Mangaluru is admitted with pneumonia and septic shock. He weighs 70 kg. His creatinine was 1.0 mg/dL on admission. Twenty-four hours later it is 2.2 mg/dL. The intern has meropenem and vancomycin to write up for the night, and she does what she was taught: she puts today’s creatinine into Cockcroft-Gault. (140 minus 58) times 70, divided by (72 times 2.2), is about 36 mL/min. She doses for a clearance of 36. This is an illustrative case, not a real patient.

The arithmetic is right. The answer is wrong, and it is wrong in the direction that harms him. By the end of this article you will know why a moving creatinine cannot be read like a stable one. You will know what the kinetic equation does instead, and which way the error runs at the drug chart.

1. A rising creatinine is reporting yesterday’s kidney

Every equation you know, Cockcroft-Gault, MDRD and CKD-EPI, assumes a steady state. The muscle makes a fixed amount of creatinine each day, the kidney clears the same amount, and so one value tells you the clearance. When the kidney fails suddenly, that balance breaks. Creatinine builds up in the body water, and the blood level climbs for two to three days before it settles at a new plateau. On the first morning of an injury, the static equation is reading the clearance of the day before yesterday.

“But surely a creatinine of 2.2 means the kidney is badly hurt?” It does. The trouble is that it does not tell you how badly, or how much worse it will look tomorrow. Waikar and Bonventre modelled it directly. Twenty-four hours after clearance fell by 90 per cent, creatinine had risen by about 1.8 to 2.0 mg/dL whatever the starting kidney function, but the proportional rise was 246 per cent with normal baseline function and only 47 per cent in stage 4 chronic kidney disease. The same injury looks entirely different depending on where the patient started, which is why the KDIGO creatinine criteria carry both an absolute and a relative arm.

Put simply: the same injury moves the creatinine by about the same amount in mg/dL, whoever the patient is. In someone who started normal, that is a huge percentage jump. In someone who started with advanced kidney disease, it is a modest one. That is why you cannot judge an acute injury by the creatinine alone, and why the staging criteria look at both the absolute and the relative rise.

2. What the kinetic equation does

Creatinine bookkeeping, not a new biomarker. Over the interval between two samples the muscle contributed a known quantity and the body retained a quantity readable from the rise in concentration multiplied by the volume it is dissolved in. The difference is what the kidney removed; divided by the average concentration, that is clearance.

“Bookkeeping?” Think of a bank account. You know the salary that comes in each month (the creatinine the muscle makes). You can see how much the balance rose (the rise in concentration, multiplied by the body water it is dissolved in). Whatever came in but did not stay in the account must have been spent. What the kidney “spent” is what it cleared.

Chen published the equation in this form in 2013. It starts from the clearance the patient had at baseline, and scales it by the baseline creatinine over the average creatinine during the interval. Then it corrects for how fast the creatinine is rising, compared with the fastest it could rise if the kidney had stopped altogether.

For our man, worked by hand with round figures. His baseline creatinine is 1.0, and the MDRD estimate at that value for a man of 58 is about 77 mL/min. His average creatinine over the day is 1.6. So the first part is 1.0 times 77, divided by 1.6, which is about 48 mL/min. Now the correction. His creatinine rose 1.2 mg/dL in 24 hours. If the fastest possible rise is 3 mg/dL a day, he has used 1.2 of 3, or 40 per cent, of the maximum, and 60 per cent of the 48 remains: about 29 mL/min. That is already lower than the 36 she wrote up.

3. The one number that changes everything

That “fastest possible rise” is the most influential number in the equation, and calculators disagree about it. QxMD and Medscape fix it at 1.5 mg/dL per day for everybody. Chen’s own group, in the 2019 dosing study, worked it out for each patient as the creatinine production divided by the volume it dissolves in, and the tool does the same. A heavily muscled man retains creatinine far faster than a frail woman.

Run our man again with the fixed 1.5. His rise of 1.2 is now 80 per cent of the maximum, only 20 per cent of the 48 remains, and the answer is about 10 mL/min. The same blood gives 29 by one method and 10 by the other. The tool prints both, and tells you that the lower one is the conservative choice for a drug with a narrow margin.

“So which is true?” Neither exactly. Observed rises in truly anuric patients are usually 1 to 2 mg/dL a day, below both, because a critically ill patient makes less creatinine: he is immobile, eating little and losing muscle. What both agree on is that his clearance is well below the 36 the static equation suggested.

4. Which way the error runs at the drug chart

This is the part to remember on a night round. The static equation errs in a predictable direction.

The rising case is the one residents know. The falling case is the one they miss, and it is easy to see why: once a dose has been reduced for the kidney, nobody feels urgency about putting it back. In the second week of an intensive care stay, the under-dosed recovering patient is the commoner error.

5. Does it change anything in practice?

It does, and an Indian study shows how often.

• 107 critically ill patients, Bangalore tertiary centre, prospective. Using kinetic GFR rather than the CKD-EPI estimate changed the dose of at least one drug in 65 per cent of patients, most often vancomycin, acyclovir and meropenem.
• 946 critically ill patients, ARDS Network fluid and catheter trial. Substituting the kinetic figure moved the Cockcroft-Gault dosing category in 33.5 per cent of those with acute kidney injury, and the effect was largest at the lowest clearances, where the dosing bands are narrowest.

In plain words: in about two of every three critically ill patients in Bengaluru, using the kinetic figure would have changed at least one dose. And the change was largest where it matters most, at the lowest clearances, where one band of a dosing table sits close to the next. In an Indian intensive care unit, where a vancomycin level is rarely back on the day you need it, this arithmetic does part of the job the level would do.

Where does the guidance stand? The KDIGO 2012 guideline remains the published standard. A draft KDIGO guideline for acute kidney injury and acute kidney disease, released for public review in March 2026, suggests calculating a kinetic GFR when a non-steady-state estimate will affect a decision. It is a draft, and the tool marks it as one.

Back to our patient

His clearance tonight is somewhere between 10 and 29 mL/min, not 36. For meropenem and vancomycin that is a different band, and for vancomycin the level should decide as soon as it is available. Tomorrow, if his creatinine starts to fall, the error flips. The static equation will read his recovering kidney as worse than it is, so the reduced doses should be revisited rather than left to run.

Try this

Day 7 in the same intensive care unit. A 45-year-old woman’s creatinine has fallen from 3.4 to 2.6 mg/dL in 24 hours as her acute kidney injury recovers. She is on a renally reduced dose of piperacillin-tazobactam, chosen on a Cockcroft-Gault from 2.6. Is her dose more likely to be too high or too low? Illustrative case.

Answer. Too low. A falling creatinine still reflects yesterday’s injury, so the static equation underestimates her clearance, and a beta-lactam may be under-dosed. Put both values and the interval into the tool. Where her baseline creatinine is known, the tool will not let the working figure rise above her pre-illness clearance, and prints the calculated figure beside it.

Take-home points

• A static equation on a rising creatinine overestimates clearance; on a falling one it underestimates it.
• The kinetic equation needs two creatinines, the hours between them and a baseline; it accounts for creatinine still building up.
• The maximum daily rise drives the answer: a fixed 1.5 mg/dL per day gives a lower figure than one derived from the patient.
• For a drug with a narrow margin, dose on the lower figure, and let a drug level decide where one is available.
• Revisit renally reduced doses as the creatinine falls, before the recovering patient is under-treated.

Declaration: the author built and maintains the tool. It is free to use. The figures in this article are those the tool carried in its Evidence tab on 5 November 2026, each from the source the tool names. The illustrative cases are worked by hand with round figures; check the tool’s result for a real patient.

References

1. Chen S. Retooling the creatinine clearance equation to estimate kinetic GFR when the plasma creatinine is changing acutely. J Am Soc Nephrol. 2013;24(6):877-888.
2. Kwong YD, Chen S, Bouajram R, et al. The value of kinetic glomerular filtration rate estimation on medication dosing in acute kidney injury. PLoS One. 2019;14(11):e0225601.
3. Waikar SS, Bonventre JV. Creatinine kinetics and the definition of acute kidney injury. J Am Soc Nephrol. 2009;20(3):672-679.
4. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl. 2012;2(1):1-138.
5. Kidney Disease: Improving Global Outcomes (KDIGO). KDIGO 2026 clinical practice guideline for acute kidney injury and acute kidney disease. Public review draft, March 2026.
6. Dinakar D, Chandan GS, Sreedhara R, et al. Kinetic estimated glomerular filtration rate and drug dosing in critically ill patients with acute kidney injury: a prospective observational study. Sci Prog. 2025;108(1):00368504251315806.
7. Bairy M, Khoo B, Tan SH, et al. Using KeGFR for vancomycin dosing when renal clearance is acutely changing: a simulation study in a retrospective cohort. Kidney Med. 2025;7(4):100970.
8. Mishra RC, Sodhi K, Prakash KC, et al. ISCCM guidelines on acute kidney injury and renal replacement therapy. Indian J Crit Care Med. 2022;26(suppl 2):S13-S42.
9. Vairakkani R, Fernando ME, Sujith S, et al. Acute kidney injury in a tertiary care center of South India. Indian J Nephrol. 2022;32(3):206-215.
10. Jelliffe RW, Jelliffe SM. A computer program for estimation of creatinine clearance from unstable serum creatinine levels, age, sex, and weight. Math Biosci. 1972;14(1-2):17-24.
11. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16(1):31-41.
12. Levey AS, Coresh J, Greene T, et al. Using standardized serum creatinine values in the Modification of Diet in Renal Disease Study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145(4):247-254.