Evolution of Chronic Kidney Disease: Defining a Model of Care
World Congress of Nephrology (ASN/ISN), San Francisco, CA, October, 2001
Prevalence and Current Treatment of Chronic Kidney Disease:
Paul Kimmel, M.D.:
It's a great personal pleasure to introduce a good friend of mine, Dr. William McClellan, who is Professor of Clinical Medicine at Emory University, and I would say, one of the leading lights of epidemiology of chronic kidney disease in the United States. We're very glad that he has consented to speak to us today about the "Prevalence and Current Treatment of Chronic Kidney Disease: The Background." Bill.
Dr. William McClellan:
Thank you Paul. It's a delight to be here and a real privilege to address this audience on epidemiology and pattern of care of people with chronic kidney disease.
I will talk about what we can learn about the epidemic of chronic kidney disease or end-stage renal disease, and what observational studies have to teach us about the care of these patients.
Trends in Incidence and Prevalence of ESRD
This slide, I think, is familiar to all of us. It's from the USRDS, and it shows trends in the incidence and prevalence of end-stage renal disease. When Fritz Port first showed me this slide in a much earlier version a number of years ago, I was impressed that the rate of increase of incident patients in the United States back four or five years ago was still increasing monotonically and had not reached its inflection point where it would begin to curve off. As you can see, that pattern is continuing during the last few years. It's projected to continue to increase unless we do something about the epidemic.
Unlike epidemics of infectious disease, chronic disease epidemics occur over long periods of time. The actual chronic end-stage disease -- and this is an estimate in 1999 for individuals on end-stage renal disease -- reflects the slow progression of people with chronic kidney disease over a prolonged period of time.
Prevalence of CRI in the US Population
These are data from the National Health and Nutrition Examination Survey that randomly sampled individuals throughout the population of the United States, and among other things, allowed Camille Jones to estimate the level of renal function in these individuals. This is actually serum creatinine of 1.5, 1.7 and 2.0 mg per deciliter or greater. What you can see is there are substantial proportions of the US population, nearly 11 million individuals, with serum creatinines in excess of 1.5 mg per deciliter and as many as 800,000 in excess of 2.
Recent estimates using Cockcroft-Gault estimating formulates actually increase this number of at-risk individuals. The problem with a slide like this is it's very disconnected, if you will, from the realities of everyday practice. So what I am going to talk to you about is how to translate those observational studies from NHANES and the USRDS into some detection strategies that will allow us to identify in familiar settings individuals at risk for chronic kidney disease.
I'm going to show you data from three groups, high-risk groups that are constituted by family members of end-stage renal disease patients. I'm going to talk to you a little bit about the prevalence of chronic kidney disease and physician practices, and then finally the prevalence of chronic kidney disease in patients who are hospitalized for cardiovascular problems.
Prevalence of Anemia among High Risk Families
These studies are currently ongoing in our shop at Emory and the Georgia Medical Care Foundation. These data were shown from the National Kidney Association of Georgia KEEP screening initiative, which was funded, in part, by our host tonight, Ortho Biotech Products, L.P., at the session earlier today.
Frequency of CKD among Family Members of ESRD Patients (n=184)
Basically, what we did was we went to dialysis facilities throughout the state of Georgia, asked the dialysis staff to recruit members of families to come into an off-center screening site. At that site, we measured their blood pressure, we checked their blood sugar. We got a urinalysis and a serum creatinine and some anthropometric information as well as information about their medical history.
Reported History of Kidney Disease
184 people, at the time we constructed this slide, had participated in that screening activity, who told us at the beginning of the visit that they had a family member who was an ESRD patient. This shows you the Cockcroft-Gault creatinine clearance for those 184 individuals. 14% of them walking in off the street simply because the family member said, "You really need to get checked," had a creatinine clearance of less than 60 mL per minute.
Blood Pressure Status among Family Members of ESRD Patients
There was a substantial proportion of those individuals that had either blood pressure that was elevated and they were unaware of it, or that was elevated and they were aware of it and not on treatment, or they were on treatment, but their blood pressure was not controlled. That's these three slices of the pie right here. Nearly 30% of these family members had uncontrolled high blood pressure.
Diabetes Mellitus Status among Family Members of ESRD Patients
Similarly, 10% had blood sugar on a random finger stick of greater than 200 mg per deciliter or had an elevated blood sugar with a previous history of diabetes mellitus. So walking off the street, about a third of these family members had uncontrolled high blood pressure, and 10% of them had evidence for hyperglycemia.
Frequency of Anemia among Family Members of ESRD Patients
We also measured the hemoglobin on them and defined the individuals with a hemoglobin of greater than 14 as having no evidence of anemia. Men with a hemoglobin between 12 and 14 and women between 10 and 12 as having mild anemia, and those with less than 10 grams per deciliter as having severe anemia. What you can see is by level or stratum of renal function, greater than or equal to 90, 60-90 and less than 60 mL per minute, the presence of severe anemia was elevated two-fold in the individuals with moderately severe anemia.
Another way of saying that -- and I have regretted not having this slide constructed in a different way -- had we shown you the prevalence of anemia by level of GFR, you would have seen that anemia increases as GFR -- estimated Cockcroft-Gault GFR -- falls.
So what we know is that family members of patients that we deal with on a day-to-day basis have a high prevalence of chronic kidney disease as manifested by a diminished GFR. Also, a third of them had 1+ or greater proteinuria. I didn't show you those data.
They have poorly controlled hypertension or diabetes and they frequently are anemic. The anemia tends to increase as their level of renal function deteriorates. Again, I didn't show you these data, but many of these patients are unaware. In fact, 86% of the individuals who had a creatinine clearance of less than 60 were unaware of their impaired renal function.
Prevalence of Anemia in Early Renal Insufficiency Study: PAERI
The next study I would like to share with you is also a study that is being supported by Ortho Biotech Products, L.P. This is a national, multicenter study that has engaged 250 primary care physicians, nephrologists and endocrinologists across the country. The purpose of this study is to define the prevalence of chronic kidney disease in these practices, and within the patients with chronic kidney disease, the prevalence of anemia. That's why it's called the Prevalence of Anemia in Early Renal Insufficiency study.
To get into the study you have to have a serum creatinine between 2.0 and 6.0 if you're a male, 1.5 and 6.0 if you're a female. At the time that these data were pulled together, we had enrolled about a third of the total 5,000 patients that will be enrolled in the study at study conclusion.
We had screened 120,000 charts and had found about 2.5% of the patients whose charts were coming across the desk and who had not been previously identified as having chronic kidney disease, fell into those inclusion criteria. This translates into a detection rate of about one patient every other week in these primary care practices. Not a tremendously large volume, but not nearly as unusual as one might suspect. These patients were older, tended to be female, they were disproportionately African-American, and half of them had renal failure attributed to diabetes mellitus.
They had a lot of comorbid conditions, and I want you to keep some of these prevalences in mind. A fifth of them had had a history of heart attack. Nearly a fifth had angina pectoris. A quarter of them had heart failure. 17% of them had some mention of ventricular hypertrophy in their chart, and the overwhelming majority, nearly 90% of them, were hypertensive.
Prevalence of Anemia according to Cockcroft-Gault
In this population too, there was evidence of symptomatic renal disease. So as the Cockcroft-Gault estimated creatinine clearance fell from greater than 90 at risk -- and the way you could get a greater than 90 creatinine clearance in this population, of course, is to have a relatively low creatinine and a large body mass. But in any event, a few of them, about 5%, fell in this stratum. As you can see, as the creatinine clearance declined from 90-60 range, 60-30 range, 30-15 and less than 15 mL per minute, the prevalence of severe anemia -- defined as a hemoglobin less than 10 grams per deciliter, moderate anemia, 10-12 grams per deciliter, and any anemia increase to the point where just at the point of severe chronic renal insufficiency as defined by the new NKF-K/DOQI guidelines, nearly half of the individuals had some degree of anemia. About 15% were severely anemic.
None of these patients who were anemic were receiving any therapy for their anemia at the time of enrollment in the study.
So the PAERI study tells us that chronic kidney disease is prevalent among the patients that we see in our day-to-day practices as primary care physicians, that anemia increases in prevalence in these patients as their renal function deteriorates, and cardiovascular comorbidity is common in these patients.
That's probably the way that patients are coming through our practices. They are not coming to us as patients with chronic kidney disease, but they are being seen by our colleagues as a patient with heart failure or ischemic heart disease with some mild and probably insignificant degree of anemia and some milder and probably insignificant degree of chronic kidney disease.
I hope the data from this final study that I am going to share with you will dissuade you from that description of mild anemia or mild renal insufficiency.
Anemia and Renal Insufficiency in Medicare Patients with Heart Failure
These are data that we developed in part of our outcome studies with the Medicare population in Georgia.
Characteristics of Patients
This is a retrospective cohort study wherein we randomly sampled Medicare beneficiaries who had been discharged from Georgia hospitals during 1998. This was a random probability sample, so this represents what you and I are doing in Georgia anyway, on a day-to-day basis. There were 645 patients sampled. They had a final diagnosis at the time of discharge of congestive heart failure. They were alive at discharge. Their mean age was 76 years. 60% of them were female, and 71% were white. Again, a disproportionate number of blacks, 29%. For the Georgia population, it's closer to 18%.
In their medical histories, there was also a very frequent pattern of cardiovascular comorbidity. As you would expect in patients who are discharged with heart failure, two-thirds were hypertensive, 44% were diabetic, a fifth had had a stroke, a quarter of them had had a myocardial infarction. Half of them had a history of coronary artery disease, and 15% had angina pectoris. These are not mutually exclusive diagnoses. But the point is that this is a population that has a lot of cardiovascular disease, and probably would be labeled as individuals with heart disease.
Prevalence of Renal Insufficiency among Heart Failure Patients
When we look at their serum creatinines, just under two-thirds, 62% had a serum creatinine of less than 1.5 mg per deciliter. 38% of them had some degree of renal insufficiency, either a serum creatinine between 1.5 and 2 or in excess of 2.0 mg per deciliter. It would have been nice to convert these to Cockcroft-Gault creatinine clearances, but unfortunately the primary study did not entail a need for weights. So we were unable to do that. I think even just using serum creatinines, this, in an elderly population, is unequivocal chronic kidney disease.
Anemia among Patients with Heart Failure and CKD
This is the pattern of anemia among these individuals with heart failure and chronic kidney disease. So what we have done here, again, is here are the strata of renal function less than 1.5, 1.5-2.0 and greater than 2.0 mg per deciliter of serum creatinine. These stacked bar graphs show you the proportion of individuals who had a hematocrit -- again, there were no hemoglobin data for this particular observational group -- who had a hematocrit greater than 40, 36-40, 30-36 and less than 30%.
What you can see is that as much as we observed in the PAERI cohort, as renal function declines, the proportion of individuals who have severe anemia, defined as less than 30%, increases dramatically in this group of patients. Similarly, the group of individuals who, by any definition, would be non-anemic, falls to really a trivial percent, if you think about these individuals as being heart failure patients and not as patients with complex cardiorenal problems.
This is a Kaplan-Meier plot showing you the survival over follow-up for these individuals by hemoglobin stratum. So here is those individuals at discharge who had a hematocrit greater than 40, 36-40, 30-36 and less than 30. What you can see is that survival in this population of individuals with heart failure who were anemic began to diverse almost immediately after discharge. The presence of anemia of any degree, red, green or blue, conferred substantial increased risk of morbidity and mortality following discharge and continuing throughout the end of follow-up.
I think I have misplaced a slide here, but I need to tell you now that 38% of these patients also had chronic kidney disease. When we entered the presence of chronic kidney disease into multivariate models with the level of hemoglobin and controlled for other factors including age, race, sex and other comorbidities, we found that both chronic kidney disease and the presence of anemia had an independent effect on survival. If you had chronic kidney disease, defined as a creatinine of 1.5 or greater, you had a 28% increased risk of mortality during the first year of follow-up.
If you had anemia, defined as a hematocrit of 30% or less, you had a 26% risk of death during the first year. If you had the combined presence of anemia and chronic kidney disease -- as did nearly 80% of the anemic patients -- you had a twofold increase of risk in death, compared to the nonanemic patients.
ACEI Use at Discharge in Heart Failure Patients
These patients weren't particularly well served by the medical care at discharge. ACE use in these patients, particularly those who had left ventricular systolic dysfunction, declined with increasing serum creatinine. Even in those who had reportedly normal serum creatinines of less than 1.4 mg per deciliter, only 58% of them were sent home on an ACE inhibitor. That fell to 42% of those individuals with serum creatinines greater than 2.
It might be argued that this is rational therapy, but there is substantial evidence to support the use of ACE inhibitors in patients with heart failure, even in the presence of mild to moderate elevations in serum creatinine.
So in summary, end-stage renal disease in the United States is epidemic. The epidemic reflects the slow progression of patients with chronic kidney disease through a prolonged illness. These individuals belong to groups that can be readily identified and accessed through our dialysis clinics, our practices and our hospitals. When we access, screen and identify these individuals and look at their management, what we are likely to find is that there are substantial opportunities to improve the blood pressure control of these individuals, that many of them will be diabetic, but poorly treated, that they will frequently not be benefited by the presence of an angiotensin-converting enzyme inhibitor, even when it's indicated for conditions like left ventricular systolic dysfunction, and they will frequently be anemic and potential candidates for correction of their anemia because of their chronic kidney disease, not because of the cardiovascular comorbidity. With that, I will close.
1. USRDA Annual Data Report, 2000
This educational activity is supported by an educational grant from Ortho Biotech Products, L.P.
This activity has been planned and produced in accordance with the ACCME Essential Areas and Policies.
This educational activity is based on an ASN/ISN evening symposium which was
planned by the World Congress of Nephrology program committee.