Thread by @medrants

1/ #UncleBob presents #5goodminutes on estimated GFR - eGFR
Before estimating equations, we measured 24 hr urine creatinine. GFR = (Ucr * V/1440 minutes)/ Scr

Collecting 24 hour urine is fraught with difficulty.
@ABsteward @CPSolvers @kidney_boy @uabimres @TMcCarty2010

2/ The first estimation equation appeared in 1976 - pubmed.ncbi.nlm.nih.gov/1244564/ Cockcroft-Gault

They developed an equation for estimating the numerator. The idea - weight in kg and age (and gender) gives a reasonable estimate of the creatinine produced each day.

3/ Consider the numerator - Ucr * V/1400 min - this equals the amount of creatinine created & excreted in 24 hours. Creatinine production is a function of muscle mass. Muscle mass decreases with age. Their equation for creatinine production each day - (28-age)mg/cc* wt in kg

4/ Now our equation looks like (28-age)mg/cc * wt in kg/ (Scr *1440)
Multiply the numerator and denominator by 5, adjust the variables and you get (140-age)* wt in kg / Scr*72
subtract 15% for women.
There are adjustments for patients too thin or too large.

5/ There are some problems w/ this formula - especially with morbid obesity. The next formula was MDRD (from Modifying Diet in Renal Disease study. Reference - kidney.org/content/mdrd-s…
This formula has some advantages, but assumes a normal muscularity for age, gender and race.

6/ The newest equation is CKD-epi - Reference kidney.org/content/ckd-ep…

7/ All 3 equations have potential inaccuracies and strengths. The key to all equations is that they depend on accurate estimation of muscle mass. Also all equations assume a stable serum creatinine. You cannot use any equation with changing creatinines

8/ Some patients have muscle mass significantly greater than average for their age and gender.

9/ I had a 3rd year student who was a serious muscle builder. As a 2nd year student he worked out daily. Creatinine = 1.3. As a 3rd year student, he could not work out as much - his creatinine = 1. His GFR had not really changed but his eGFR improved dramatically.

10/ Recently I have noticed several patients with eGFR over 300 - this would not happen with Cockcroft-Gault but current basic metabolic panels use MDRD or CKD-epi
Obviously these patients had greatly reduced muscle mass

11/ You will miss CKD in patients with anorexia, major amputations, spinal cord injury, cancer cachexia. In such patients Cockcroft-Gault will clearly give a better estimate, because when you have major weight loss you will lose significant muscle

12/ You may have noticed that PharmDs prefer Cockcroft-Gault for calculating drug doses. I think this is because it does not assume muscle mass.

13/ The bottom line - I wrote this because understanding that eGFR is not a "real" GFR and has many potential errors may help you recognize those potential errors. In general eGFR is helpful at clueing us to CKD, but like any lab test, we should become aware of its assumptions.

Actually there is a typo here

The formula for 24 hr urine creatinine is (28-0.2Age) * wt in kg

You can use this to assess the accuracy of 24 hr urine collection

confidence intervals appoximatly + or - 25%