I hate to write this, but it appears that the New York Times may be more up to speed than your primary care doctor or even your nephrologist when it comes to kidney disease in the elderly. For Older Adults, Questioning a Diagnosis of Chronic Kidney Disease accurately describes a problem which has not received the attention from the medical community that it should have. The problem is that we’re over diagnosing kidney disease in elderly patients because of a combination of measurements and extrapolations biased against old age mixed with a failure to account for the effect of normal aging on a normal kidney.
First some background. We have known for more than half a century that kidney function declines with age. The best way to quantify kidney function is by measuring the glomerular filtration rate (GFR). It is the amount of blood that is filtered by the kidney per unit time. I’ve left out a few details, but this is all you need to know to understand what follows. A normal GFR is about 120 ml/minute. We will assume that anything above 90 is normal. Thus, if we follow a subject, Mr Nephron, from age 20 to 80, we will find a 50% decline in GFR. Say it was 110 when he was 20, it will obviously be about 55 when he reaches 80. We will further stipulate that except for a few aches and pains Mr Nephron remains healthy as he ages. Most importantly he does not have hypertension, diabetes, or urinary tract obstruction, and has a normal urinalysis. He’s healthy, but old. The reason for his decline in GFR is aging not disease. Like a car with 200,000 miles on it, his kidneys are not sick – they’re just a bit worn down.
GFR is not routinely measured; rather it is estimated by extrapolating from a commonly measured blood test – usually the blood creatinine concentration. The National Kidney Foundation’s GFR calculator is here. Assume that Mr Nephron’s creatinine is 1.2 mg/dl and that he is not African-American. Don’t worry about the IDMS button on the calculator leave it checked yes. Be sure to enter his age. You will find that his estimated GFR is 57 or 58 depending on what equation is used to make the estimate. Also note that the units for GFR are mL/min/1.73 m2. Meters squared (that’s m2) is the value of body surface area. We correct for body surface area to account for the difference in kidney function due to differences in size. To know the body surface area you must know the subject’s height and weight. But we haven’t told the calculator anything about Mr Nephron’s body habitus. So unless our subject has a body surface area that is close to 1.73 our calculated value is way off.
Let me show how far from reality our calculated value can be. Suppose Mr Nephron is 5′-3″ and weighs 125 pounds (he’s a retired jockey). His body surface area is 1.62 m2. If his measured GFR, as opposed to estimated, is 60 ml/min, when corrected for 1.73 m2 it is 64 ml/min. But suppose Mr Nephron is a former linebacker and is 6′-6″ and weighs 250. His body surface area is 2.51. His GFR corrects to 41. But the NKF calculator gives us the same estimated GFR when we plug in an 80 year old white man with a creatinine of 1.2.
The classification of kidney disease in common clinical use is below. It’s an example of medicine’s desire to make distinctions which separate the clinician from useful information. Instead of saying here’s a patient with a GFR of 57, we fell more scientific by saying this patient has Stage 3 kidney disease while hoping we can remember all the cutoff values that separate one stage from another.
By using the estimated GFR that is supplied every time a doctor orders a panel of routine blood chemistries we are at risk of making two errors. First we will over diagnose kidney disease in elderly patients. Doing so may send them on a diagnostic voyage which is costly and unnecessary. Second, we may confuse normal aging with a disease. Perhaps half or more of elderly patients in whom a blood creatinine is measured will be told they have Stage 3 kidney damage, and be scared half to death, when there is nothing wrong with them other than the weight of their years. Again, if a patient has a normal blood pressure, a normal urinalysis, is not diabetic, and doesn’t have urinary tract obstruction, a doctor doesn’t need a blood test to know he doesn’t have kidney disease. A physician can usually tell if his patient has most forms of urinary tract obstruction by talking to him, though the requirements of our modern medical system limit this sort of interaction.
The short message, ask your doctor to read the NY Times article cited above or have him read the August 11, 2015 issue of the Journal of the American Medical Association for two learned views of this subject. Primary care doctors need to get up to speed on this issue and refer less patients to nephrologists solely because of an isolated lab result misinterpreted by a machine. A little common sense (aka good judgment) can solve this problem.