New challenges in CKD management
GPs now increasingly manage chronic kidney disease in its early stages. Dr Kathryn Griffith advises on dealing with the issues this raises
For many years we have had very poor indicators of renal function in primary care. Renal function has to deteriorate by about 50 per cent before creatinine levels go up. Creatinine assays themselves were thought to be inaccurate and there was a lot of variation between different labs depending on the method used. Even when levels rose we were led to believe that it was all part of the ageing process and we could expect deteriorating renal function with a loss of glomerular filtration of 10ml/minute each decade after 40 years of age.
While there was a lack of interest in renal disease in primary care, renal physicians were busy managing patients with end-stage renal disease (ESRD) and starting them on dialysis at ever-increasing ages. Some patients arrived so late that they needed to immediately start dialysis (so-called crash-landers), sometimes with disastrous results. These patients have a particularly poor prognosis related to poor vascular access, infection of lines and metabolic complications. There was a clear need to identify patients at high risk of developing ESRD.In 2003 the US National Kidney Foundation produced practice guidelines for the evaluation and classification of chronic kidney disease (CKD). They used estimated glomerular filtration rates (eGFR), derived from the Modification of Diet in Renal Disease (MDRD) equation to produce five stages of CKD (table 1). UK laboratories now report eGFR to primary care and CKD is in the QOF. The rest, as they say, is history.
Where does the MDRD equation come from?
The gold standard test to assess kidney function is inulin clearance. This involves radioisotopes, is time consuming and costly and so is not suitable for all. It is used in secondary care when the real and not estimated glomerular filtration rate is needed, for example in a live donor prior to transplantation. The MDRD study gave researchers the opportunity to devise a short-cut method of estimating GFR using age, sex, creatinine and ethnic group, (those of African ethnic groups having larger muscle mass and thus serum creatinine levels). They produced a very complicated equation giving a result in ml/min/1.73m= thus correcting for standard body surface area.
This is not the same as the Cockroft-Gault equation. This does not calculate GFR but estimates creatinine clearance. It corrects for weight and is still used in therapeutics to calculate drug doses. It is not interchangeable with MDRD eGFR and cannot be used to categorise patients for QOF.The MDRD equation has a major benefit in that it can be easily calculated by the laboratories using the standard data on the request form. Since 2005 laboratories have been recommended by the renal NSF part 2 to routinely calculate eGFR from U&E results. Despite this, because of problems related to the creatinine assays, some areas have only just been routinely getting the results. This has caused significant coding problems for QOF. If the value is not routinely available it is easy to use the calculator on the Renal Association website, www.renal.org (I can highly recommend this as a source of information for both professionals and patients) .The value you calculate using the online calculator will not be the same as the one provided by your laboratory. This is because your lab should also use a correction factor to allow for the type of assay they use and to standardise to national values.The benefit of the online calculator is that it will give you a standard error. This should remind you that you are calculating an estimated value. The MDRD study involved patients with known chronic renal disease. The calculation is most accurate when renal function is worse and has most error nearer to normal values between 60 and 90, which is where we have the most problems in primary care. Laboratories are being discouraged in national guidance from calculating eGFR where creatinine levels are normal as the equation is not valid at these levels. Despite this some labs currently continue to send eGFR values when creatinine levels are normal. It is recommended not to include these patients in QOF registers.The test is not appropriate in under 18s, pregnant women, where renal function is rapidly changing, and when body habitus is abnormal for example body-builders or amputees.
Why is it important to identify patients with early kidney disease?
Patients with CKD are at increased risk of premature death but this is related to cardiovascular events, not renal failure. They are more likely to die of a heart attack or stroke than need dialysis.
Management is targeted at slowing the deterioration of their renal function and reducing cardiovascular risk. Renal patients have not only the traditional risk factors of age, hypertension, diabetes and smoking but also additional ones related to calcium metabolism, anaemia, electrolyte and fluid balance, uraemia and oxidative stress.CKD is a sign of target organ damage and management should include treatment of hyperlipidaemia, aspirin and of course blood pressure.Patients should only be coded as having CKD if the changes are persistent (over at least three months). Patients over 80 can be expected to have a GFR of 60 or less because of age-related deterioration (10 per cent per decade after 40), and although this will be 'normal for age', it will be associated with increased risk and problems with some medications, particularly NSAIDs.
What clinical assessment should patients with early CKD have?
The majority will have CKD related to hypertension or diabetes and obviously fasting glucose is an important investigation in those not known to have diabetes. It is also important to rule out the most important reversible cause of CKD – bladder outlet obstruction. All guidelines stress the importance of asking for symptoms of bladder outlet obstruction. I think most men over 65 will have these symptoms and so assessment should include examination of the abdomen to assess the presence of a palpable bladder and urological referral if found.
Urine should also be stick-tested for both blood and protein. Haematuria is most commonly associated with a renal stone, urological tumour or glomerular disease and is also likely to warrant an initial urological, rather than renal, referral.It is important to review the patients' medication and look for nephrotoxic drugs, particularly NSAIDs, which should be withdrawn if possible.
The importance of blood pressure control and proteinuria
The most important intervention to reduce the progression of CKD is to treat high blood pressure to optimum levels and not just what the British Hypertension Society (BHS) guidelines refer to as the 'minimally acceptable' level, which has been used for QOF. This means that we should treat to <130 0mmhg="" not="" just="" 140/85mmhg="" and="" even="" lower="" levels="" are="" suggested="" in="" bhs="" guidelines="" with="" the="" presence="" of="" proteinuria.="" there="" is="" actually="" very="" little="" clinical="" evidence="" for="" these="" targets="" and="" patients="" are="" likely="" to="" need="" multiple="" drug="" therapy="" and="" may="" not="" be="">130>
There is much debate about the importance of blood pressure lowering per se, or the class of the drugs used.Patients with significant CKD were excluded from traditional trials and the evidence for CKD patients is largely those with proteinuria and/or diabetes and funded by the manufacturers of the newer classes of therapy, in other words ACE inhibitors or angiotensin receptor blockers (ARBs). There is little evidence for the use of these drugs in elderly patients with early CKD and no proteinuria.A recent consensus meeting at the College of Physicians of Edinburgh produced a very helpful document that can guide primary care in the management of these patients, and can be easily accessed from its website www.rcpe.ac.ukPersistent proteinuria is the most important indicator of patients likely to have deterioration of renal function and cardiovascular complications. All patients with CKD should have stick testing of urine to check for blood and protein and Protein Creatinine Ratio (PCR) measurement if stick testing is persistently positive. Early morning urine sampling reduces the effect of orthostatic proteinuria or exercise, and thus false positive results. Patients with a PCR of >100 should be referred to a renal physician, and should receive an ACE inhibitor or ARB.It is still recommended that patients with diabetes should have microalbumin assessment if protein dipstick is negative. These patients are at highest risk and this method will detect the earliest signs of renal damage, and they should obviously then receive an ACE inhibitor or ARB.The consensus document recommends that in patients without diabetes or proteinuria, the management of high blood pressure should follow national guidelines and may include thiazide diuretics and calcium channel blockers in the elderly. It concludes that where blood pressure is controlled there is no indication to change to an ACE inhibitor or ARB and that these patients can be exempted from this section of QOF.
A new coding for stage 3?
The prevalence of CKD 3 is about 4 per cent of the population. Not all these patients have the same risk of progression and the renal consensus meeting suggested we could use the level of eGFR and the presence of proteinuria to identify those at highest risk who will need more monitoring, intervention and referral. The stage was divided by level of GFR and the presence of proteinuria. GFR levels of 45-59 were 3A and 30-44 3B, presence of proteinuria was identified by the suffix p so 3Bp have the highest risk.
Is there a role for the management of stage 4 patients in primary care?
Guidelines recommend that all patients with an eGFR of <30 should="" be="" referred="" to="" a="" renal="" physician.="" patients="" with="" deteriorating="" renal="" function="" should="" obviously="" be="" referred;="" however="" this="" could="" include="" a="" large="" number="" of="" elderly="" patients="" who="" have="" had="" stable="" creatinine="" levels="" monitored="" by="" the="" practice="" cardiovascular="" clinics="" for="" many="" years.="" many="" renal="" teams="" are="" offering="" the="" option="" of="" telephone="" or="" email="" advice="" to="" continue="" the="" management="" of="" these="" patients="" in="" primary="" care,="" and="" indeed="" are="" discharging="" stable="" patients="" from="" long-term="" follow-up="" back="" to="" primary="">30>
Metabolic complications are more likely to occur when eGFR falls below 30 (or 45 in patients with diabetes), and so the management of patients with CKD 4 should involve more frequent monitoring (3-4 monthly), including haemoglobin and calcium levels. Despite earlier advice it is not helpful to check PTH levels if calcium levels are normal. Management of renal anaemia is now being directed towards primary care. Initiation of erythropoesis stimulating agents (ESA) usually involves a specialist, however administration and follow-up commonly occurs 'closer to home' in the GP surgery. I now have several patients in their 90s with CKD 4 and 5 and symptoms of anaemia or heart failure having erythropoietin although they would not be suitable for renal replacement therapy. It is also important to ensure that these patients are iron replete and because ferritin can be raised as an acute phase protein reaction, levels under 100µg/l are considered abnormal and transferrin saturation levels are used as a better indicator of iron stores.All patients with CKD 3-5 should have influenza and pneumococcal vaccination. Those with deteriorating function who may require renal replacement therapy should also have hepatitis B vaccination. If the initiation of this is delayed until they have CKD 5 then the chances of them sero-converting become less and so the sooner they get started the better. There is therefore a role for general practice initiating vaccination where renal function is deteriorating at 5ml/min/year or more.
What do you tell your patients?
The identification of a new 'disease' from a blood test result which the patient was previously told was 'normal' is fraught with difficulties. Even when I explain the benefits of closer monitoring and my particular interest in their problem I am subjected to anger or tears.
Until recently there have been very few helpful information leaflets that don't imply that it is likely that CKD leads to renal failure and a kidney machine. There are helpful leaflets from the Renal Association and the Edinburgh renal unit, and a new leaflet from the RCGP to match the one sent to general practice last year.The important message is to translate eGFR to percentage renal function and explain that a fit healthy person can donate a kidney to have 50 per cent function and a normal life. It is not until function deteriorates to <15 per="" cent="" that="" people="" feel="" unwell="" and="">15><10 per="" cent="" before="" dialysis="" may="" be="" required.="" we="" do="" know="" that="" the="" majority="" of="" patients="" with="" ckd="" will="" not="" require="" renal="" replacement="" therapy="" and="" that="" there="" are="" treatments="" that="" we="" can="" give="" to="" slow="" the="" deterioration="" of="" the="">10>
What can patients do to help themselves?
The most important change involves strict salt restriction and means avoiding convenience foods and ready meals. Patients with CKD are very salt sensitive and benefit from improved blood pressure control. Weight loss also has benefits for blood pressure, but obesity may be a contraindication to renal transplant if required. Smoking cessation advice is obviously particularly important.
Kathryn Griffith is a GP in York and GP lead for York and Harrogate Renal Local Implementation Group
Competing interests Dr Griffith has received payment from Takeda to lecture on CKD and an unrestricted grant from Sanofi-Aventis to set up a renal clinic in her practice
Take-home points• The MDRD equation to estimate eGFR is only valid if the creatinine is abnormal• GFR calculated in this way is an estimate and has most error when values are between 60 and 90• Patients over 80 will have a GFR of =60 because of age-related deterioration – 10 per cent per decade over 40• Patients with early CKD are more likely to have a cardiovascular event than develop end-stage renal disease• Patients with highest risk complications are those with proteinuria; if PCR >100, refer and start an ACE• BP target for treatment of those with proteinuria is <130 0mmhg•="" ckd="" patients="" are="" likely="" to="" need="" more="" than="" one="" drug="" to="" treat="" high="" blood="">130>