Creatinine Clearance for Clinical Trial Eligibility

Many clinical trial protocols set a minimum kidney function for enrollment, and creatinine clearance from the Cockcroft-Gault equation is the measure they name most often. This page explains how study staff usually calculate and record that number. It does not decide eligibility: the protocol, the sponsor and the investigator do.

Why Protocols Set a Kidney Function Cutoff

A drug that leaves the body through the kidneys can build up when kidney function is low. Until a new drug has been studied in people with reduced kidney function, protocols often exclude them or limit their dose. A review of 55 anticancer drugs approved in the United States from 2015 to 2019 found that 95% of their 74 pivotal trials had kidney-related eligibility criteria. Among trials with a kidney function cutoff, the median lower limit was 45 mL/min (or mL/min/1.73 m²), with a range from 20 to 60.

So the cutoff is usually written in creatinine clearance terms, and 45 mL/min is a common value, but it is not universal. The number that applies is always the one in your protocol.

CrCl, eGFR or Serum Creatinine: Use the Method the Protocol Names

Kidney criteria can be written as a serum creatinine limit, a creatinine clearance, an eGFR, or a combination. In the same review, a combined serum creatinine and GFR criterion was the most common form (40% of trials), followed by serum creatinine only (27%) and GFR only (27%). Use the exact method the protocol names, because the methods are not interchangeable:

• Cockcroft-Gault CrCl is reported in mL/min and uses age, sex, body weight and serum creatinine.
• eGFR is reported in mL/min/1.73 m², scaled to a standard body surface area. FDA guidance notes that drug clearance follows the person's own GFR in mL/min, and converts eGFR by multiplying by the person's body surface area and dividing by 1.73.
• Measured CrCl from a timed urine collection is a separate test, not a calculation.

For dedicated studies of how kidney impairment changes a drug's pharmacokinetics, the FDA's 2024 guidance recommends eGFR over estimated creatinine clearance, while calling measured GFR, measured creatinine clearance or another widely accepted equation reasonable. It also says participants should be enrolled using one common estimation method across the control and renal impairment groups. That advice covers that type of study; your protocol text still governs. For how the two measures differ, see CrCl vs eGFR, and for the bands the guidance uses, see the FDA renal impairment categories.

Which Body Weight the Protocol Means

Cockcroft-Gault multiplies by body weight, so the weight definition alone can move a result across a cutoff. Protocols may specify actual (total) body weight, ideal body weight, adjusted body weight, or the lower of actual and ideal weight. Follow the protocol wording exactly.

When the protocol is silent, the evidence gives background, not a rule for the trial. In 3,678 patients with stable kidney function, Winter and colleagues found that Cockcroft-Gault was unbiased with actual weight in underweight patients and with ideal body weight in patients of normal weight. In overweight, obese and morbidly obese patients, adjusted body weight with a 0.4 factor was the least biased and most accurate. The FDA guidance likewise says that in overweight or obese people, ideal or adjusted body weight is likely to be more accurate than actual weight. The National Kidney Foundation notes that weight choices are applied inconsistently across clinicians and institutions, which produces variable results.

That variability is why an unclear protocol is a question for the sponsor's medical monitor, not a judgment call at the site. Ask in writing and file the answer with the study records, so every participant is handled the same way.

Units, Creatinine Floors and Weight Caps

Cockcroft-Gault expects serum creatinine in mg/dL and weight in kilograms. If the lab reports creatinine in µmol/L, divide by 88.4 to get mg/dL (the creatinine unit converter does this). To convert pounds to kilograms, divide by 2.20462. Record both the reported value and the converted value, and decide how conversions are rounded before you calculate.

Some research groups add their own rules. SWOG Policy 38, for example, uses Cockcroft-Gault with weight capped at 140% of ideal body weight, a minimum serum creatinine of 0.7 mg/dL, and a maximum CrCl of 125 mL/min for drug dosing. SWOG states that the creatinine minimum was chosen for consistency with other professional guidelines. Winter's study, by contrast, found that rounding low creatinine values up did not improve accuracy or bias. Other groups and sponsors use different rules or none, so apply a floor or cap only when your protocol or a policy it references requires it.

Borderline Results and Rounding

Take a protocol that requires CrCl of at least 45 mL/min by Cockcroft-Gault, using the lower of actual and ideal body weight. A woman aged 67 is 5 ft 4 in (64 in) tall, weighs 52 kg and has a serum creatinine of 1.0 mg/dL.

1. Ideal body weight (Devine, female): 45.5 + 2.3 × (64 - 60) = 54.7 kg. Actual weight is lower, so 52 kg is used.
2. CrCl = (140 - 67) × 52 × 0.85 / (72 × 1.0) = 3,226.6 / 72 = 44.8 mL/min.

Is 44.8 at least 45? That depends on the protocol's rounding rule. If results are rounded to whole numbers before comparison, 44.8 becomes 45. If the unrounded value is compared, 44.8 is below the cutoff. Neither the calculator nor this page can settle that. The same example shows how sensitive a borderline result is:

• Had the protocol named ideal body weight, the result would be 73 × 54.7 × 0.85 / 72 = 47.1 mL/min.
• Had the lab reported 88 µmol/L, the unrounded conversion is 0.9955 mg/dL, and the result would be 45.0 mL/min.

Which Creatinine Result to Use

If the protocol sets a time window for screening labs, use a serum creatinine from a specimen drawn inside that window, and check the specimen collection date rather than the date the report was printed. Use the lab source the protocol specifies. Do not average several results or pick the most favorable one unless the protocol tells you how to choose.

Cockcroft-Gault turns a single creatinine into a clearance estimate, so it is meaningful only when creatinine is stable. When kidney function is changing, as in acute kidney injury, the estimate lags behind the true value. The FDA guidance makes the same point for renal impairment studies: participants should have stable renal function, and patients with acute changes should not be included. If a screening creatinine differs sharply from earlier values, raise it with the investigator before relying on it.

Recording the Calculation

Under 21 CFR 312.62(b), an investigator must “prepare and maintain adequate and accurate case histories that record all observations and other data pertinent to the investigation” for each participant. Sites meet this in different ways, but a creatinine clearance worksheet or source note usually shows:

• The serum creatinine value, unit, lab and specimen date, plus any converted value.
• Age, sex, height and weight, with units and the date each was measured.
• The weight type used (actual, ideal, adjusted, or lower of actual and ideal) and the protocol section that requires it.
• The formula: Cockcroft-Gault, with the 0.85 factor for women.
• The result with units, and the rounding rule applied.
• Any sponsor clarification relied on, and who calculated and checked the value, and when.

The creatinine clearance calculator on the homepage lets you set the body-weight method and shows the weight used and the result. Its Copy summary button copies the inputs and result as text for a note, and its Print button produces a printable record. Check that the weight method matches the protocol before using its number. The eligibility decision itself stays with the investigator, under the protocol and the sponsor's guidance.