Cockcroft-Gault calculator for creatinine clearance

Cockcroft and Gault published the equation in 1976 (Nephron 1976;16:31-41) from hospital inpatients, as a way to estimate creatinine clearance from a single serum creatinine without a timed urine collection. It is still the equation behind many drug labels and trial protocols, because the dosing studies that supported those drugs used it. The form is simple: the term (140 − age) captures the fall in clearance with age, weight stands in for muscle mass and body size, and the factor 72 converts the units so that the answer comes out in millilitres per minute. Creatinine must be in mg/dL. Women get a multiplier of 0.85 because they have, on average, less muscle mass per kilogram.

Take a 60-year-old man who weighs 80 kg and has a serum creatinine of 1.0 mg/dL. Then CrCl = (140 − 60) × 80 / (72 × 1.0) = 6,400 / 72 = 88.9, reported as 89 mL/min. For a 60-year-old woman with the same weight and creatinine the answer is 88.9 × 0.85 = 75.6, reported as 76 mL/min. The sex difference of 13 mL/min comes entirely from the 0.85 factor, which is worth knowing when you test a spreadsheet or an eCRF formula: if the male and female results are identical, the factor has been dropped.

Many central laboratories report creatinine in µmol/L. Divide by 88.4 to get mg/dL: 88 µmol/L is 1.0 mg/dL, 106 µmol/L is 1.2 mg/dL and 115 µmol/L is 1.3 mg/dL. The lab unit converter lists the other conversions you will meet. Unit slips are the commonest source of absurd values: entering 88 as if it were mg/dL gives a clearance near 1 mL/min, and entering 1.0 as if it were µmol/L gives a number far above any physiological value. Collect creatinine in one labelled unit per field and put a plausibility range on it.

The CKD-EPI 2021 equation estimates glomerular filtration rate, indexed to 1.73 m² of body surface area, from creatinine, age and sex alone. Cockcroft-Gault estimates creatinine clearance in absolute mL/min and uses weight. They are not interchangeable even though both are loosely called renal function. The differences show most in the extremes of body size and age. A small older woman, 78 years old and 52 kg with a creatinine of 1.3 mg/dL, has a Cockcroft-Gault clearance of 29 mL/min and a CKD-EPI eGFR of 42. With an eligibility cut-off of 30, the first excludes her and the second does not. A tall, heavy young man shows the opposite pattern, where Cockcroft-Gault with actual weight gives a much higher number than CKD-EPI (178 against 133 in the table above).

For the person writing a protocol, the practical rule is to name the equation, the weight and the unit, and never write only “renal function” or “eGFR”. Choose Cockcroft-Gault when the investigational drug, or the comparator, has its dosing or its renal restrictions defined by creatinine clearance, which is true of a long list of older drugs. Choose CKD-EPI when the criterion is about kidney disease staging or a cohort definition. Whichever you pick, apply it identically at every site and for every visit. Some protocols deliberately report both, using one for eligibility and the other for a safety listing, and say so.

Creatinine clearance appears in clinical trial protocols in two ways. As an eligibility criterion, for example “CrCl of at least 60 mL/min by Cockcroft-Gault at screening”, it protects participants whose kidneys may not clear the study drug and keeps the population consistent. As a dose-adjustment or hold rule, the protocol states bands of clearance and the action that goes with each, which the investigator then applies. This page does not tell you what any band should be or how to dose any drug: those rules come from the investigator’s brochure, the label and the protocol, and they are clinical decisions. What the data team can do is make sure the number that triggers the rule is calculated the same way every time.

A usable criterion therefore names the equation, the weight, the threshold and its direction, the sample (local or central laboratory, and how many days before randomisation), whether the result is rounded before comparison, and what happens at a borderline value. Because creatinine moves a few tenths of a mg/dL from day to day, a participant at 59.6 mL/min one week can be at 61 the next. Say whether a repeat inside the screening window is allowed and which result counts. The eligibility screening eCRF template shows how each criterion and its source value can be recorded.

For a trial with repeated dosing, the same calculation recurs at later visits as a safety check, so the formula and the weight convention must be fixed in the data management plan, not decided by whoever is filling in the form. A calculated field applies one formula to every record, removes hand calculation and the queries that come with it, and leaves the lab report as the source for the creatinine itself.

This is a reference calculation, not clinical advice. It reproduces a published equation so that a protocol writer, data manager or coordinator can check a number. It is not a medical device, it has not been validated as clinical software, and it must not be used to decide a treatment, a dose, or whether a real person may enter a trial. Those decisions belong to the protocol and the investigator, using the laboratory’s reported value and local procedures.

Like every creatinine-based estimate, Cockcroft-Gault assumes a steady state and is unreliable when creatinine is changing quickly, as in acute kidney injury. It was derived in adults and is not used for children. Creatinine reflects muscle mass, diet and a few drugs that block its tubular secretion, and weight is a poor proxy for muscle in people who are very obese, oedematous, amputees or very muscular. A cystatin C equation, a measured clearance or a timed urine collection is the better route where kidney function decides a dose, and the protocol will say so. Capture also has the lab data and reference range management approach for recording which laboratory produced each creatinine.