CME Slides Forum - A.S. Levey

CHRONIC KIDNEY DISEASE: A NEW PUBLIC HEALTH PROBLEM

Andrew S. Levey, Boston, USA

Chair: Joan Fort, Barcelona, Spain

Carmine Zoccali, Reggio Calabria, Italy

levey

Prof. Andrew S. Levey
Division of Nephrology
Tufts University School of Medicine
Boston, USA

Good morning, members and guests of the society, Doctor Zoccali thank you very much for inviting me to speak before you toady. It’s an honour and a privilege to address you on this topic.

The objectives for my talk today are to give some background on why chronic kidney disease is recognised as a public health problem. We are principally nephrologists so we take care of many patients with chronic kidney disease. We recognise for them and for those of us with kidney disease in our families that chronic kidney disease is a personal health problem but it is also a public health problem.
I will spend some time to give a few thoughts on the definition of chronic kidney disease to reflect back on some of the things that we have learned over the past 5 years since introducing this definition. I will provide an update on some changes in the incidence and prevalence of chronic kidney disease. I will be focusing principally on results from the United States but I think that you will see that these results are reflected throughout the world and I know there have been many presentations at the meeting and before the meeting about chronic disease around the world.
I will outline briefly an example of a public health approach that is developing in the United States and I will look back on our role as nephrologists in the public health approach.

I want to begin with acknowledging the contributions of many others to the work that I’ll be presenting here and to my own personal work throughout the years, particularly those involved in the MDRD study both during the initial clinical trial and in the epidemiologic results that have been published since then. The people who have been instrumental in working with me in the K/DOQI and the KDIGO guidelines particularly Joe Coresh in the original CKD work group and as well the current leadership of KDIGO and the evidence review team at New England Medical Centre. I want to acknowledge the work of those in the National Kidney Disease Education Program in the United States who have been responsible for the efforts to calibrate the serum creatinine assay throughout the US and soon throughout the world and my own research team, CKD-EPI that is working to improve the estimating equations for the GFR.

I want to begin by reiterating the simple message from World Kidney day in 2000 that chronic kidney disease is common, it is harmful and it is treatable but for us here attending the EDTA and for us as kidney specialists we should change the last line to; we have treatment to emphasise our unique role in leading the way towards a public health approach for this disease.

Now this slide shows the trends in the death rates for the 6 leading causes of death in the United States over a span of years of more than 30 years. If we simply focus from about 1986 to the year 2000, we see the continuing dramatic decrease in death from heart disease, as well as from stroke.

By contrast, when we look from the same period of time from the mid 80s to about 2000 in the United States, we see a doubling in the incidence and the prevalence of kidney failure treated by dialysis and transplantation and therefore, registered with the United States Renal Data System. So, that in the year 2000 there were more deaths from kidney failure in the United States, 99.000 than there were from colon cancer, breast cancer or prostate cancer, diseases that are far more recognisable to pose a public health problem than kidney disease has been recognised.

During the same period of time from the early 90s through to 2000 costs in the United States to Medicare which is the insurance system for the elderly have risen to 3 and a half billion dollars now consuming 7%. The total cost for kidney disease is 3 and a half billion dollars out of a total of 7 and a half billion for all of Medicare and representing this percentage of total health expenditures in the United States. The same thing is being seen from employer group health plans, private insurance for younger people. So kidney failure in the United States is characterised by a rising prevalence, poor outcomes and high costs.

Now, studying kidney failure is a little bit like in cardiovascular disease studying people who have survived their first heart attack. If we wish to do something to reduce the incidence in prevalence of kidney disease, we clearly have to begin earlier. This slide shows the conceptual model for chronic kidney disease that we have developed through the K/DOQI guidelines recognising that kidney failure is simply the end stage of a disease that must begin before and by definition a disease that culminates in kidney failure must be detectable beforehand as a reduction in kidney function and we use the GFR as the overall index of kidney function and that prior to a chronic decrease in GFR there must be some sort of kidney damage that can in principal be detected from markers of damage before the GFR declines. Prior to that there must exist an increased risk state characterised by the presence of risk factors which lead to an increased risk of developing kidney damage and later stages of kidney disease. Usually, these risk factors are a variation of the normal characteristics of the population and therefore, we have a continuum of states on the way to the progression towards kidney failure. The green circles represent chronic kidney disease, the blue represent antecedents of kidney disease and that at any point along the way there can be complications of chronic kidney disease not just of the end stage that by themselves can lead to death. We even show a slide from complications from normo because we have to recognise that once we begin to intervene and look for kidney disease, in fact there can be complications to normal people from our interventions to try to detect or prevent the disease.

Now, there’s been a great deal of work with this conceptual model and we now have a clinical action plan for all the different stages of CKD that focuses on what clinicians should do at these different stages to try and reduce the risk for progression to later stages. We recognise that these bold horizontal arrows represent risk factors for progression and potential targets for intervention. This is now the clinical action plan for CKD that many of you have implemented so successfully in your own practices, in your own countries. We are woefully inadequate in implementing all of the guidelines that exist for the different stages of kidney disease and this remains a major challenge for us as kidney specialists. This is not the same as public health plan. This is the clinical action plan for the care of patients. Embedded within this conceptual diagram and conceptual model are the seeds of a public health plan.

Now, you know very well that the definition for these different stages was developed first in 2002 by the CKD K/DOQI team modified later by KDIGO as kidney damage or GFR less than 60 for 3 months or more and the modification by KDIGO to include patients who have had a kidney transplant irrespective of whether there is manifest kidney damage or a reduction in GFR.

You’re also aware that the stages have been stratified according to the level of GFR into these 5 stages given these titles and in the United States at least have now been embedded in the administrative coding that makes possible claims-based analysis, at least in the US and there is a move to adopt these codes throughout all countries and to have them updated into subsequent versions of the international classification of disease.

Now, we’ve been able to use the same model to make operational definitions of CKD for epidemiological studies and I would propose that the same definitions can be used in public health programs. Kidney damage has been defined in large detection or screening programs as an elevated spot albumin to creatinine ratio greater than 30mg/g, decreased GFR as a decreased estimated GFR less than 60, kidney failure as an estimated GFR less than 15 and these have been applied in large national surveys in the United States.

The first survey these were applied was the NHANES III survey from 1988-1994 and the base population for that NHANES survey was 177 million. We found that 6.3% of the population had kidney damage another 4.5% had a decreased estimated GFR, that about 0.1% had kidney failure for a total prevalence of chronic kidney disease of 19.5 million or 10.9% of the population. This clearly led us to recognise that the earlier stages of kidney disease are more prevalent than diabetes, more prevalent than the forms of cancer that I showed you on the earlier slide, slightly less prevalent than hypertension but a major antecedent to kidney failure.

The outcomes of chronic kidney disease by this diagram include more than kidney failure, they include the states of low GFR and kidney damage and complications that can exist from them even for patients that do not reach kidney failure. By complications of CKD we generally refer to complications of decreased GFR such as hypertension, anaemia, malnutrition, bone and mineral disorders and we now recognise that chronic kidney disease is a risk factor for cardiovascular disease and I’ll just show you a couple of the slides that I think have become quite well know outlining this.

These are analyses from NHANES showing with a reduction in estimated GFR the increasing proportion of the population that suffers from these comorbid conditions or I should say complications of kidney disease, hypertension, a low serum calcium, a low haemoglobin, a low albumin, an elevated phosphorous and a marker of functional limitation unable to walk one quarter mile.

You’re well familiar with all of these slides showing chronic kidney disease as a risk factor for cardiovascular disease. Doctor Sarnacks analysis of cardiovascular disease mortality in dialysis patients compared to the general population. Doctor Go’s analysis published in the New England Journal showing the rate of cardiovascular events according to the estimated GFR in more than 1 million people receiving care at the Keiser Permanente Medical Foundation, the WHO international compilation of data showing proteinuria as a risk factor for death in type 1 and type 2 diabetes. A more recent analysis based on urinary albumin concentration from the Groeningen cohort showing that in the range except that there’s microalbuminuria, there’s a substantial increase in cardiovascular disease mortality. We now recognise that chronic kidney disease is an independent risk factor for cardiovascular disease although the reasons for this are not well known and remain an active area of research in our field.

To me the most striking analysis however, was this one and this was the first I was aware of from Keith et al published in Archives of Internal Medicine in 2005 who compared the outcomes among a large group of patients in managed care organisation in the United States and the patients were stratified according to their level of estimated GFR. Here 60-89 and no urine protein, a group felt to be free of kidney disease compared to patients with stage 2 , stage 3 or stage 4 CKD. First we have to look at the number of people. This is a fairly large group of patients particularly in this stage 3. The mean age at the time of detection for the stage 1 and the stage 2 patients is close to 60 but for stage 3 and stage 4 the mean age is over 70. The amount of follow up was 4-5 years in these two groups. The striking finding to me shown in blue is that only 1 % with stage 3 CKD developed kidney failure 1% and that in stage 4 people who we feel are close to dialysis, in fact only 17% of the population develop kidney failure. Very few receive the transplant before developing kidney failure, some disenrolled which is always a problem with research in manage care populations but a large fraction of the patients died.

So even amongst stage 4 CKD, the risk of death was 3 times greater than the risk of kidney failure and in stage 3 CKD the risk of death is 25 times greater than the risk of kidney failure.

We now know that among the outcomes of chronic kidney disease complications of decreased GFR and death are far more important, as far as the number of patients affected than kidney failure. The same epidemiologic model can be used to look at risk factors. Now we begin to look at the identification of risk factors as a way to intervene to prevent not just kidney failure but to prevent all the outcomes and we now need to split this into risk factors for developing chronic kidney disease compared to risk factors for progression and complications of chronic kidney disease. Those for developing kidney disease can be thought of as two types, susceptibility to develop kidney disease or exposure to an initiation factor. I’ve highlighted here examples of susceptibility and initiation factors. You’re well aware that diabetes high blood pressure are accepted factors that literally cause chronic kidney disease and it’s likely that obesity and dyslipidemia maybe initiation factors. Factors that increase susceptibility when exposed to an initiation factor likely include older age, a family history of chronic kidney disease, and in the United States being a member of a racial or ethnic minority group. I’ll deal with some of these as we go on in the talk. I think these are the prime targets for detecting CKD early.

Risk factors for progression and complications are those which worsen kidney damage or accelerate the GFR decline such as a higher level of proteinuria, or a higher level of blood pressure, factors that increase the risk for complications of decreased GFR and pathophysiology we’ve been studying this for years, factors related to hypertension, anaemia, malnutrition, bone disease and then cardiovascular disease risk factors either traditional or non-traditional risk factors. Of course, we have been studying factors related to adverse outcomes of kidney failure.

Now most important for a public health program is to recognise that we have treatment. It’s very important to appreciate non-treatable diseases but that’s not the situation for kidney disease at this point. We have treatment to reduce morbidity and mortality of kidney failure. We have treatment to slow the progression of kidney disease, to delay and prevent the complications of decreased GFR and to reduce cardiovascular disease risk. I would submit that we even have treatments to prevent kidney disease.

Before moving on to recent trends in incidence and prevalence of kidney disease in the United States, let me reflect back on some thoughts on the definition. We have used albuminuria and I would in fact, like to get away from the term microalbuminuria because it’s simply confusing to our non-nephrology colleagues. We’ve used the level of albuminuria, a ratio greater than 30 in order to indicate chronic kidney disease. Now, some have suggested sex and race specific cut-offs because of the variation in creatinine excretion. So far we have opted for a simpler definition simply to facilitate implementation. Now, this is 2-3 times greater than the normal value and these are the NHANES data for albumin: creatinine ratio showing that this level identifies these people which is less than about 10% of the population, that it is well recognised to be the earliest marker of kidney damage due to diabetes, glomerular disease and the best marker we have for hypertension although not sensitive enough and by itself it’s associated with adverse outcomes.

Now, the importance of albuminuria goes far past it’s being a marker of kidney damage. You’re well aware of this but these things are often confused in discussions. It’s certainly a clue to the type of CKD whether it’s a glomerular disease or a non-glomerular disease albuminuria appears to be the strongest risk factor for progression of kidney diseases and therefore, we detect a progression factor as well as the disease. It’s an effect modifier for ACE inhibitor therapy in non-diabetic kidney disease. People with more proteinuria respond better to an ACE inhibitor. It is also hypothesised to be a marker of vascular permeability and many have referred to the presence of albuminuria as indicating generalized endothelial dysfunction and at levels far less than 30 it’s a risk factor for cardiovascular disease. And such has been hypothesised to be a surrogate outcome for kidney disease progression and CVD risk reduction and we don’t yet have clinical trials to show that therapies targeted at lowering the albuminuria per se are more effective than therapies that do not target the albuminuria.

It’s very difficult to design these trials but if so this would open up the field, at least in the United States, for the food and drug administration to license products, pharmaceutical products that target albuminuria per se.

I do want to point out one study that I think has been overlooked. This is the study by Mann et al looking at microalbuminuria, the outcomes of microalbuminuria in the HOPE study in particular microalbuminuria as identifying people at risk of developing clinical proteinuria. Of course, we know this in diabetes but people have not recognised that this in fact is a reasonable definition in non-diabetic kidney disease well.
In all of HOPE of the patients who had microalbuminuria 16% developed clinical proteinuria compared to 0.8% without microalbuminuria giving an odds ratio for microalbuminuria as a risk factor for clinical proteinuria of 26 which is still substantial despite adjustment for most other risk factors for developing clinical proteinuria.
In patients with diabetes the odds were higher for developing clinical proteinuria 22.8% and they were also higher in those without microalbuminuria 1.5% . The risk ratio or the odds ratio were very similar. In those without diabetes the absolute risk was much lower 6.6% and the absolute risk in those without microalbuminuria was also much lower 0.3% but the odds ratio which is the best expression of this as a risk factor was the same in non-diabetics as well as in diabetics and survived a multivariable adjustment just as in diabetes. So when we talk about microalbuminuria being a marker of kidney damage in the earlier stages of kidney disease, it conveys the same diagnostic importance in diabetics as it does in non-diabetics. The risk of course is much lower but its risk factor relationship as far as the odds ratio adjusted is exactly the same.

Let’s spend a moment or two talking about the definition of chronic kidney disease as a reduction in GFR. The rational for choosing 60 as this is substantially above the level associated with kidney failure. This slide goes down to 40, shows the inulin clearance results in a compilation of data from Wesson in his textbook in 1973 and this is substantially above the kidney failure level. It’s less than the normal range. This is the normal range in men, young men about 120-140. Young women, 120-130 and the CKD level is shown by the red line, it’s very infrequent in people less than 40 but you’ll clearly see that we have identified a few people in these older studies at ages greater than 70 who were thought to be normal as having CKD.

Now, this slide is a compilation of studies in normos from 1973. The structure of the elderly population in 2000 is far different from what it was in 1973 and we have many, many more people alive greater than 70 then there used to be and we are undoubtedly identifying a large number of people with a GFR under 60 as having chronic kidney disease and.
The question has arisen are these people normal or do they have something that we need to be concerned with? Now, there are a number of age related changes in the human kidney. The structure shows a decreased size and cortical thickness. It shows substantial glomerulosclerosis which is quite uncommon in younger people. It shows tubular atrophy and fibrosis and vascular sclerosis. The function is also markedly different. It shows a decreased GFR as shown before, a decrease in renal plasma flow, an increase in filtration fraction, a decrease in maximal urinary concentration and a decrease in maximal acid excretion.
These studies don’t tell us whether this is normal or not. They tell us that the kidney is changing with age. There are many features in this which are quite characteristic of disease. But we also know that patients with chronic kidney disease are the elderly are the fastest growing population developing kidney failure and that a low GFR is associated with tremendous risk particularly in the elderly. So I think it should be an open question for research as to the meaning of a low estimated GFR in an elderly patient whether it’s normo or not I think we can be informed by some of the previous epidemiological studies in hypertension.

Now I picked hypertension because hypertension is pretty much thought to be the major risk factor for cardiovascular disease that has come under control in the last 30 years in the slides that I showed you at the very beginning. If we look back at the story for hypertension, we find that in the 40s it was recognised that hypertension as a disease that could kill. It became quite well popularised with the death of Franklin Roosevelt in the United States. We then recognised that even moderate hypertension was associated with substantial risk of cardiovascular disease mortality. We recognised that blood pressure rises with age and it had long been thought to be a normal phenomenon of aging that blood pressure was elevated and initially there was a different definition of hypertension in the elderly than there was in young people. It was a systolic greater than 160 rather than 140.
We then recognised that even moderate hypertension in the elderly was associated with substantial risk and we recognised through clinical trials that there was a benefit of antihypertensive therapy in moderate hypertension particularly in the elderly, particularly in the range that had been considered normal and has now led to a single definition of hypertension for all ages although most elderly people are recognised as having hypertension.

So, I think there’s strong precedent for us to maintain the definition of chronic kidney disease as it is. It’s objective, it’s based on laboratory measures, it’s independent of the cause of CKD and it’s irrespective of age. The laboratory measures are imperfect but the measures do identify people at substantial increased risk for adverse outcomes.
Cardiovascular disease and complications of reduced GFR are far more common than kidney failure in chronic kidney disease and there is a very large population affected especially among the elderly. The implications for us kidney specialists are that we have to change the way we think about CKD. We have to recognise that much more research is necessary, particularly in the elderly to understand the underpinnings of the decrease in GFR and the decrease in function and the susceptibility to these adverse outcomes.

I want to spend a few minutes talking about trends in incidence and prevalence in the United States. This slide shows the adjusted incident rate for new cases of kidney failure. For some years since the 80s there’s been noticed that the rate of rise has been slowing. You can see this slowing in the rate of rise or this can be plotted on this axis as a percent increase from the previous years. The percent increase has been going down.

But you know our practice as nephrologists in the United States changed during this period. We simply started to offer the treatment to everyone who had the disease and that by itself can make this curve look this way. The incident population is not growing because we’re treating the same number of incident patients with the same proportion, we’re treating them all. We didn’t use to treat them all, so as we treated more the incidence went up. That will not explain this last year 2004 and 2003 when the incidence went down, you can barely see it but this curve is starting to go down in the United States. This suggests something different than just treating the population affected, this suggests that maybe we’re getting somewhere with the treatment of the disease.

Now, this is the most clearly shown in my mind if we isolate those people with kidney failure due to diabetes and at a young age, aged 20-39, so now we’re really talking about kidney failure due to type 1 diabetes. Among whites aged 20-29 this line or aged 30-39 this line there’s been a clear decrease in the incidence of new cases of kidney failure due to type 1 diabetes in the United States. I think those of us in the audience would immediately think that this is related to treatment. The pivotal clinical trials of ACE inhibitors and type 1 diabetics were conducted in the late 80s; the results were disseminated in the early 90s. The drugs were approved by the FDA shortly thereafter and there has been widespread implementation in the United States of these drugs and throughout the world. They, I think, maybe responsible of this decline in diabetes but this is just in whites. Look at the results in blacks. Look at the results in Native Americans and Asians and look at the results in Hispanics in the United States. This is still a public health problem even if we are successful in the white population, we are woefully inadequate in the non-white populations and in the Hispanic ethnicity. This is still a public health problem, even if the incidence of kidney failure is coming under control. This is still a public health problem.

If we look at new cases adjusted by age, these are the people over 65-75 and these are the people aged 75 plus. So, although we maybe having an effect in the young, the disease is still a tremendous problem in the elderly. We maybe having an effect at this point in the elderly as well but look at this disparity between rates in the elderly versus rates in the young. I would not say that having a low GFR is innocuous in the elderly; this is the group at greatest risk of getting kidney failure.

Now, I want to turn to trends in the earlier stages of chronic kidney disease. As you may remember, I first showed you data from NHANES 1988-1994. These data are reproduced in this column, although I have to tell you that there has been further work on recalibration of the serum creatinine measurement for use in the MDRD study equation and these results are very sensitive to creatinine calibration and with the most recent calibration to the new national standard the rates during this period are 10% rather than 10.9%. It seems like a small difference but this difference across a big population is a lot of people. This is our best estimate as of today and these analyses have just been done by the CKD-EPI group with Joe Coresh leading the team.

Now, there have been 3 additional NHANES surveys since 1994 which we have lumped together because they include a substantial number of people and there have been 3 separate calibrations corresponding to these new NHANES surveys of the serum creatinine and again, the results are very sensitive to creatinine calibration but our best estimate is that in this 5 year interval or this 6 year interval, that the prevalence is increased to 13.1%.

The confidence interval is fairly tight and the prevalence ratio is 1.3. You can think of this as a 30% odds. For an increase in the prevalence since this earlier interval and the best number for the number affected in the United States in the year 2000 would be 26 million. So, this is a major point. We are just seeing a decrease in the incidence of kidney failure but what we should be seeing looming ahead for us is a tremendous problem that we have a big increase in the prevalence of earlier stages of kidney disease.

Now, in Doctor Coresh’s analysis when we look at risk factors for having a lower estimated GFR, remember the prevalence ratio of 1.3 we should keep that in mind, when we adjust this for age, sex and race it’s actually higher for the low eGFR 1.5.

When this is adjusted for the known diagnosis of diabetes and hypertension, the odds ratio is 1.4% and when this is adjusted for the known increase in body mass index in the population, the odds ratio remains higher than 1.

When this is adjusted for potential lack of residual areas in calibrating the serum creatinine assay, this appears to be robust. Although we do get very close to 1.0 here with a P value of only 0.3. So we think this is a robust analysis. When we look at albuminuria, in fact the increase in albuminuria does seem to be accountable for by the aging of the population, the increase in prevalence of diabetes and hypertension and obesity. But there maybe something going on in addition with the eGFR but even if there was nothing special with the eGFR and it’s just due to these factors, the wave of diabetes and hypertension that is destined to affect our population because of the epidemic of obesity will drive these numbers even further up.

Who is most affected by this? It’s clearly the elderly population. Age greater than 70. According to these analyses we have 45% of the population meeting the criteria for chronic kidney disease and at age 60-69 it’s about 20%.

When we try and next go on to say well what are the public health strategies, we can divide this into three types according to the classical paradigm; primary prevention would be preventing the development of CKD, secondary prevention would be preventing the progression or complications of CKD and tertiary prevention would be treating kidney failure which we think we’re already doing.

Now, the centres for disease control and prevention in the United States convened an expert panel which met in March 2007 and came up with draft recommendations which are still under consideration and identified a 10 point plan for implementing a public health strategy for chronic kidney disease.

On the left are recommendations and on the right are actions what I’ve not shown are further tables on the right that say what should the coordination be with other public health agencies, what’s the time table and what should be the measurable outcomes?

I believe this could be standardised throughout the world. This, of course, will be specific to the country of origin and although I list 10, I want to return to number 5 for the conclusion of the talk but these ten are to define national goals for improving the outcomes, to implement existing quality improvement measures, to define new measures, to implement a surveillance for CKD and its risk factors. In the United States we have the USRDS and we have NHANES but we don’t have much more. To detect CKD in the at risk populations, for research to identify and close gaps in knowledge about CKD primary and secondary prevention. To develop tools to predict progression to kidney failure. So we can see which of those elderly patients who are affected are more likely to get kidney failure. To strengthen and provide better education and to improve laboratory procedures for CKD reporting, these are targeted at providers and hospitals and clinical labs and then for the general public to increase public awareness.
CKD detection is really a two --- effort o promote CKD testing for early detection and to identify patients at increased risk and I would say that we should be focusing on people with cardiovascular disease and CVD risk factor conditions, especially among the elderly.

For CKD testing, of course, you know there are many tests but the one that is easiest to implementing a public health program is proteinuria testing, the spot albumin: creatinine ratio I think is pretty reasonable although I must say that urine albumin assay is not standardised and many have suggested that we could do better with sex specific cut-offs. I think a single cut-off is reasonable for public health program. Laboratory reporting needs to be standardised and we need to appreciate the difference between the level that we detect as a marker of kidney damage versus a predictor of CVD and of course, we need better markers for other types of kidney disease.

For the GFR this is the performance of the MDRD study equation. This is the measured GFR and this is the estimated GFR in the MDRD study population and as you all know, it’s a pretty good fit with an R² of about 0.9 and 90% of the predictions being within 30% of the measurement.

Now the problems with the MDRD study population is that most people had a GFR less than 60 and still the points are not that tight.

The problem with creatinine calibration, at least in the United States, appears to have been solved by moving from the Beckman which was the MDRD study lab which had a bias of about 1.3.

Against the base creatinine of 0.9 to an assay system at least in CKD-EPI research hands which is the Roche enzymatic which is right on target with the new national standard.

With this we can then apply the MDRD study equation refitted for the Roche enzymatic to a large population of data where we have standards available.

The CKD-EPI database includes 5000 people now compared to 1600 from MDRD. It includes 30% blacks, 30% diabetics, 5% transplant recipients. The mean GFR was 40. Here’s the mean standardised creatinine compared to the uncalibrated creatinine.

Here’s the performance of the MDRD study equation in the CKD-EPI population. The P30 is 83%, not as good as in the original population meaning there’s more spread.

There’s a tremendous spread above 60. As you can see these patients were not in the original MDRD study population, below 60 there’s very little bias.

Therefore, the recommendation has been to use the MDRD study equation particularly and it’s best if it’s standardised, again standardised creatinine in people with a GFR less than 60.

To automatically report eGFR whenever creatinine is measured and to teach others that despite uncertainty this is a strong predictor of adverse outcomes, particularly cardiovascular disease. We clearly need new research to identify filtration markers that maybe better especially in the elderly such as cystatin C and we need a very cautious interpretation in the elderly that recognises their low risk of kidney failure and their high risk of other conditions that are treatable.

We clearly need educational campaigns. I’ve been involved in the campaigns with some clinical laboratories to explain GFR estimates through the national kidney foundation’s effort and through NKEDP and many of you have been involved in local efforts.

This is clearly central. Without this sort of educational campaign, physicians don’t understand the meaning of the eGFR reports they’re getting back.

Finally, I just want to mention who is the population that I think should be targeted for screening?

I think it should be people with cardiovascular disease and those with CVD risk factors. CVD is the most common cause of death in the United States and the most common cause of death in people with CKD. So efforts to prevent CVD are paramount. CKD is common in people with CVD risk factors. CKD multiplies the risk for CVD. The risk factor levels are higher. The risk factor control is more difficult, the outcomes are worse and the cost of care is higher. Most important, the risk factor management differs in patients with CKD compared to those without CKD. Finally, treatments are available and effective for all the complications.

For those of you who need to teach your colleagues as to why they need to identify CKD in people with CVD risk factors the goals are different. In hypertension we say there is a lower goal and we think ACE or ARBs are indicated rather than diuretics if there is proteinuria. There is a need for an increased frequency of monitoring. In diabetics glipizide is preferred as a --- urea of preference and metformin needs to be avoided with low GFR. There are newer antidiabetic agents which need to be adjusted based on the level of kidney function. Among those with dyslipidemia there’s currently a debate as to whether CKD is a CVD equivalent and whether the LDL levels need to be lower than 100 as they are for other CVD risk factors compared simply to 130.

The dose of fibrates has to be reduced and there’s an increased risk of side effects from combination therapy.

So, in summary we have a conceptual model, we have a clinical action plan, we still need a public health approach. We need better implementation and we need more research. Our role as nephrologists is that we need to change the way we think about CKD, we need to maintain our traditional strengths in patient care especially for CKD stage 4-5, we need more research and we need education, we need to educate our colleagues and our students about the earlier stages of CKD and we need to work with public health officials.

So the message from World Kidney Day was that CKD is common, it’s 13% in US adults with a higher prevalence in the elderly and in patients with CVD risk factors. It’s harmful, there’s an increased risk for heart disease, complications of decreased function and progression to kidney failure. It’s treatable and more important we have the treatments.

Next year on World Kidney Day I hope that we can go even further than what we did in 2007 with a public health approach across the world. Thank you very much.