Dr Kathryn Griffith explains how the chronic kidney disease indicators in the quality and outcomes framework have evolved, and how they fit in with NICE guidance

The identification of people with early chronic kidney disease (CKD) is vital to offering them treatment that can both delay the progression of the disease and also the complications associated with the advanced stages of this condition. As normal blood creatinine levels range from 60–120 mmol/l, a normal reading in a patient could conceal an almost 50% loss of kidney function.1 The opportunity to intervene in the early phases of kidney disease is lost if testing for creatinine alone is used to identify patients at risk.

The Kidney Disease Outcomes and Quality Initiative (KDOQI) classification of CKD has only existed for 6 years.2 This categorises the stages of CKD by evidence of kidney damage and a fall in the glomerular filtration rate (GFR). The KDOQI staging of CKD has been modified by NICE and is shown in Table 1. As the GFR is difficult to measure directly, an estimated GFR (eGFR) is used, as derived from a formula developed in the Modification of Diet in Renal Disease Study.4

Table 1: Stages of chronic kidney disease as modified by NICE3

Stagea GFR (ml/min/1.73m2) Description
1 ?90 Normal or increased GFR, with other evidence of kidney damage
2 60-89 Slight decrease in GFR, with other evidence of kidney damage
3A 45-59 Moderate decrease in GFR, with or without other evidence of kidney damage
3B 30-44
4 15-29 Severe decrease in GFR, with or without other evidence of kidney damage
5 <15 Established renal failure
aUse the suffix (p) to denote the presence of proteinuria when staging CKD
GFR=glomerular filtration rate; CKD=chronic kidney disease

National Institute for Health and Care Excellence. Chronic kidney disease: Early identification and management of the chronic kidney disease in adults in primary and secondary care. Clinical Guideline 73. London: NICE, 2008. Reproduced with kind permission. Available at: www.nice.org.uk/guidance/CG73

Identification of patients with CKD

The National service framework for renal services, published in 2005, recommended the use of eGFR along with serum creatinine results for the identification of CKD.5 An eGFR of <60 ml/min/m2 on at least two occasions over a period of ?3 months indicates a diagnosis of CKD stages 3–5. The majority of these patients will have CKD stage 3 that will not progress further and so will be managed entirely in primary care. To facilitate medical management, systems needed to be established, both to optimise patient care and also to identify those with more rapid progression or significant kidney disease. For these reasons it was important that CKD was included as a clinical domain in the quality and outcomes framework (QOF).

Introduction of CKD indicators: GMS contract 2006/07

Indicators for CKD first appeared in the QOF in the 2006/07 revisions.6 Primary care took up the challenge despite the lack of read codes, little education, or understanding, and within one year had identified almost 3% of the population as having CKD!7

However, the initial CKD indicators led to some confusion regarding blood pressure medications. It is clear that for a patient with early CKD, effective control of blood pressure reduces progression of kidney disease and also the most significant outcome: the occurrence of cardiovascular (CV) events. This was indicated by the audit standard for blood pressure, which was set lower for CKD compared with any other clinical domain. Indicator CKD 4 in the GMS contract 2006/07 suggested that the drugs of choice for all patients with CKD were the A drugs—angiotensin-converting enzyme (ACE) inhibitors or angiotensin-receptor blockers (ARBs). This resulted in some practitioners switching patients with well-controlled blood pressure onto ACE inhibitors and ARBs unnecessarily; for example, those patients with CKD without either diabetes or proteinuria.

Revised CKD domain: GMS contract 2008/09

The situation became clearer with the publication of the revisions to the GMS contract in 2008.8 Indicator CKD 4 was relabelled CKD 5 and amended to clarify that it should be patients with CKD, hypertension, and proteinuria who are treated with an ACE inhibitor or an ARB. This was an improvement, but it was disappointing for two reasons: it was not made clear that patients with CKD should have regular urine testing for proteinuria; and the method of testing for proteinuria was not made apparent. Fortunately, the NICE guideline on Early identification and management of chronic kidney disease in adults in primary and secondary care was published in 2008, and contained recommendations for best practice.3

Testing for proteinuria

The NICE guideline on the management of CKD emphasised that standard reagent strips are not recommended to test for proteinuria.3 They are not a quantitative test and do not detect low levels of protein in the urine. However, there are some strips that measure albumin at low concentrations and these could be used if quantitative testing is not available.

The test of preference for proteinuria is an albumin:creatinine ratio (ACR). This is calculated based on a spot urine sample and is the most specific test to detect low levels of albumin in the urine, and thus the early stages of damage to the basement membrane in the renal glomerulus. It also has the advantage of being the same test used during the review of diabetes, which is one of the commonest causes of CKD.

Values of ACR that are considered significant for patients with CKD without diabetes are higher than for those with CKD and diabetes. In the latter group, it is important to consider albuminuria (microalbuminuria) significant if the ACR is >2.5 mg/mmol in men and >3.5 mg/mmol in women.

In people without diabetes, proteinuria is defined in the NICE guideline as a level of ACR ?30 mg/mmol.3 It is important to confirm one abnormal value with at least one more sample and preferably one taken in the early morning to reduce the effects of the upright posture. Values of ?70 mg/mmol do not need repeating as they are considered highly significant and warrant referral for specialist assessment.

The protein:creatinine ratio (PCR) may be used as an alternative to ACR if there are high levels of proteinuria or if ACR testing is not available. The PCR is a cheaper test than ACR and has been used to monitor significant levels of protein by nephrologists for many years. The test for PCR also measures other proteins apart from albumin and can be used to determine comparative levels of ACR:3

  • ACR of 30 mg/mmol ?
    PCR of 50 mg/mmol
  • ACR of 70 mg/mmol ?
    PCR of 100 mg/mmol.

At high levels of proteinuria, a PCR test is considered adequate.3

Revised CKD domain: GMS contract 2009/10

The QOF clinical indicators for 2009/10 for the CKD domain brings all of these elements together (see Table 2).9 Indicator CKD 5 clearly shows that ACE inhibitors and ARBs are used for patients with CKD, hypertension, and proteinuria. Moreover, there is a new indicator—CKD 6—for recording urine ACR or PCR in the past 15 months. There are additional points both for urine testing (6 points—CKD 6) and achieving tighter blood pressure control (an additional 5 points to CKD 5, making 9 in total for this indicator). It will be important to ensure that urine is tested regularly and that positive proteinuria results are coded to include people with proteinuria in the denominator for CKD 5. When proteinuria is detected, the suffix (p) can be added to the staging, for example, stage 4p, and there are now codes for this.

Table 2: QOF 2009/10 clinical indicators for chronic kidney disease9

No. Indecator Points Payment stages
CKD 1 The practice can produce a register of patients aged 18 years and over with CKD (US National Kidney Foundation: stage 3 to 5 CKD) 6  
CKD 2 The percentage of patients on the CKD register whose notes have a record of blood pressure in the previous 15 months 6 40–90%
CKD 3 The percentage of patients on the CKD register in whom the last blood pressure reading (measured in the previous 15 months) is ?140/85 mmHg 11 40–70%
CKD 5 The percentage of patients on the CKD register with hypertension and proteinuria who are treated with an ACE inhibitor or ARB (unless a contraindication or side-effects are recorded) 9 40–80%
CKD 6 The percentage of patients on the CKD register whose notes have a record of a urine albumin:creatinine ratio (or protein:creatinine ratio) test in the previous 15 months 6 40–80%
  Total points 38  
CKD=chronic kidney disease; ACE=angiotensin-converting enzyme; ARB=angiotensin-receptor blocker

Challenges for primary care

Case finding
Studies suggest that in populations with a normal age distribution, the prevalence of CKD should be around 10%.10 Regular searches of eGFR results for levels <60 ml/min/1.73m2; should reveal new cases of CKD, thereby updating disease registers. All people with cardiovascular disease (CVD), diabetes, hypertension, structural renal disease, and multi-system diseases (e.g. lupus) or who are taking nephrotoxic drugs should have an annual test for kidney function.3 This will also identify new cases and reduce the effects of low prevalence due to loss of the cut-off and square root formula.11 It is important to remember that statistically, when we are looking at differences from the mean some practices will inevitably fall short of the average and will thus receive reduced payments.

Clinical review
It will be important to build into place regular testing for blood and urine samples along with measurements for ACR. These are routine in our practice vascular clinic, as is a measure of blood glucose. The tight audit standard for blood pressure reflects the importance of control; however, controlling blood pressure in patients with CKD can often be very difficult and usually requires three drugs,12 one of which is likely to be an ACE inhibitor or ARB. In order to achieve good blood pressure control it is important to remember that the actual targets for patients with CKD range between 120–139 mmHg (systolic) and below 90 mmHg (diastolic).3 This needs to be highlighted to all members of the clinical team because of the confusion with all the different audit standards for blood pressure within the QOF.

It will be necessary for patients with CKD to have more frequent follow up while titrating up medication, and blood tests both before and 1–2 weeks after making changes to ACE inhibitor or ARB dose.3 This should improve patient outcomes as well as help to achieve QOF points.

Age and CKD

Deciding whether elderly patients have CKD or if the decline in eGFR is a direct result of age is a common dilemma. Many practice registers will have large numbers of elderly women with an eGFR of around 60 ml/min/1.73m². The practice computer database will show if these patients have a stable renal function or if levels are declining at a slow rate (1 ml/year) that is compatible with increasing age. In our practice, if there is an obvious decline then the patient is coded as CKD because there are benefits to this (e.g. better control of blood pressure). It is also important to review their other medication, and in particular, the regular use of non-steroidal anti-inflammatory drugs, which should be discontinued if possible.

Improvement to the CKD domain

There are two important issues not covered in the QOF. The first is the concept of CKD and CV risk. When eGFR falls below 45 ml/min/1.73m², the risk of CV mortality and morbidity increases dramatically, and other CV risk factors should be aggressively managed.13

The second area is the detection and management of renal anaemia. If eGFR drops below 45 ml/min/1.73m2; anaemia becomes more common and is in itself responsible for increasing the risk of CVD. Anaemia is more frequent in patients with CKD and diabetes14 and the annual review for patients with both CKD and diabetes should include an estimation of haemoglobin. It was the identification of increasing CVD risk when eGFR falls below 45 ml/min/1.73m2 that led to the division of stage 3 CKD into stage 3A and 3B (see Table 1).2

With the move of patient services from hospital to the community we can look to the development of services managing renal anaemia, with erythropoetin and iron infusions funded either through a local QOF or locally enhanced service.

Possible future changes

A simplification of the blood pressure audit standards in QOF would be useful as at present there are three different levels, which many healthcare professionals find confusing. I would suggest the following: a level for primary prevention; another for secondary prevention, stroke and CVD; and a third for those at the highest risk (for example, patients with diabetes and microalbuminuria, and CKD and albuminuria). Now that NICE is overseeing the QOF review process, it is unlikely that points will be earned for measuring blood pressure without achieving standards, and duplicate payments for treating blood pressure for several domains are also likely to be reduced. I also believe that payments should be increased for the management of high-risk patients, such as those with CKD.

It is possible that the method used to calculate eGFR will change from using the formula from the Modification of Diet in Renal Disease Study to another that has a different correction for age and sex; this may exclude some elderly women at low risk of progression from being diagnosed as having CKD.

Summary

I believe that the level of achievement of primary care in this clinical area is a testament to the high standards of practice teams. This third incarnation of the clinical domain in the GMS contract 2009/10 now fits in clearly with the NICE guideline and targets medication use in those with proteinuria and CKD. The management of these patients may be very challenging however, because they will often have several long-term conditions and are on drug therapy that will impact on their renal function. With the focus on reduction of CV risk in CKD patients, primary care is best placed to optimise the care of these patients with early CKD.

I am keen to make contact with other GPs who, like myself, seek to bridge the gap between primary and secondary care renal services: kathryn.griffith@york.nhs.uk
  • The NICE guideline on the diagnosis and management of CKD is likely to increase workload in primary care and referrals to secondary care
  • It is likely to create immediate referral and prescribing cost pressures, with potential savings and health benefits realised sometime in the future
  • Many of the suggested interventions can be provided in primary care or the community
  • A constructive and pragmatic dialogue between hospital and primary care clinicians is likely to craft local care pathways that can best meet the demands of this guideline
  • CKD specialist nurses could be employed in primary care to avoid paying an expensive tariff charge when working for hospital trusts
  • The majority of people with CKD will be elderly, often with complex
    co-morbidities. Sensitive and personalised interpretation of the NICE guideline in discussion with the patient is likely to be necessary
  • Tariff costs: general medical outpatient = £189 (new), £91 (follow up)a
  1. Renal Unit of the Royal Infirmary of Edinburgh website. Kidney tests. renux.dmed.ed.ac.uk/EdREN/EdRenINFObits/kidneytests.html (accessed 25 June 2009).
  2. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification and stratification. Am J Kidney Dis 2002; 39 (2 suppl 1): S1–266. Available at: www.kdigo.org/guidelinescompare/kdoqi.html
  3. National Institute for Health and Care Excellence. Chronic kidney disease: Early identification and management of chronic kidney disease in adults in primary and secondary care. Clinical Guideline 73. London: NICE, 2008. Available at: www.nice.org.uk/guidance/CG73
  4. Levey S, Bosch J, Lewis J et al. A more accurate method to estimate glomerular filtration rate from serum creatinine; a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999; 130 (6): 461–470.
  5. Department of Health. National service framework for renal services. Part two: chronic kidney disease, acute renal failure and end of life care. London: DH, 2005. Available at: www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/Browsable/DH_4102941
  6. British Medical Association. Revisions to the GMS contract, 2006/07. Delivering Investment in General Practice. London: BMA, 2006.
  7. The Information Centre. National quality and outcomes framework statistics for England 2006/07. The Information Centre, Prescribing Support Unit, 2007. Available at: www.ic.nhs.uk/statistics-and-data-collections/supporting-information/audits-and-performance/the-quality-and-outcomes-framework/qof-2006/07/qof-20
  8. British Medical Association. Quality and outcomes framework guidance for GMS contract 2008/09. London: BMA, 2008. Available at: www.bma.org.uk/employmentandcontracts/independent_contractors/quality_outcomes_framework/qof06.jsp
  9. British Medical Association. Quality and outcomes framework guidance for GMS contract, 2009/10. Delivering investment in general practice. London: BMA, 2009. Available at: www.bma.org.uk/employmentandcontracts/independent_contractors/quality_outcomes_framework/
  10. Stevens P, O’Donoghue D, de Lusignan S et al. Chronic kidney disease management in the United Kindgom: NEOERICA project results. Kidney Int 2007: 72 (1): 92–99.
  11. British Medical Association website. QOF introduction of raw prevalence on income. www.bma.org.uk/employmentandcontracts/independent_contractors/quality_outcomes_framework/QOFPrevalence.jsp (accessed 23 June 2009)
  12. Sarnak M, Greene T, Wang X et al. The effect of a lower target blood pressure on the progression of kidney disease: long-term follow-up of the modification of diet in renal disease study. Arch Intern Med 2005; 142 (5): 342–351.
  13. Go A, Chertow G, Fan D et al. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalisation. N Eng J Med 2004; 351 (13): 1296–1305.
  14. Thomas M, MacIsaac R, Tsalamandris C et al.
    Unrecognized anemia in patients with diabetes: a cross-sectional survey. Diabetes Care 2003; 26 (4): 1164–1169. G