Rationale and impact
- Creatinine-based estimate of glomerular filtration rate (GFR)
- Investigations for proteinuria
- Reagent strips for proteinuria and haematuria
- Who should be tested for CKD
- Frequency of monitoring
- Risk assessment, referral criteria and shared care
- Pharmacotherapy for blood pressure control
- Pharmacotherapy for proteinuria and choice of antihypertensive agent
- Diagnostic role of glomerular filtration rate
- Optimal Hb levels
- Correcting iron deficiency
- Hyperphosphataemia in people with CKD stage 4 or 5
These sections briefly explain why the committee made the recommendations and how they might affect practice.
Evidence on the specific eGFR equations or ethnicity adjustments seen by the committee was not from UK studies so may not be applicable to UK black, Asian and minority ethnic groups. None of the studies included children and young people. The committee was also concerned about the value of P30 as a measure of accuracy (P30 is the probability that the measured value is within 30% of the true value), the broad range of P30 values found across equations and the relative value or accuracy of ethnicity adjustments to eGFR equations in different ethnic groups. The committee agreed that adding an ethnicity adjustment to eGFR equations for different ethnicities may not be valid or accurate. Categorisations based on ethnicity lump together people with a diverse range of family backgrounds and differences in eGFR across ethnicities are likely to at least partly arise because of differences in average muscle mass between ethnic groups. However, muscle mass also differs from person to person within the same ethnicity and so making an adjustment based on ethnicity may be inaccurate for some people. Therefore, the committee agreed to remove the 2014 recommendation on how to adjust the CKD-EPI creatinine equation for adults of African-Caribbean or African family origin. The committee highlighted the 2008 recommendation, which states that caution should be used when interpreting eGFR and in adults with extremes of muscle mass and on those who consume protein supplements (this was added to recommendation 1.1.1). The committee made recommendations for research on appropriate eGFR equations for black, Asian and minority ethnic groups (adults, children and young people) in the UK (see research recommendations 1 and 2). They agreed that factors other than ethnicity should also be explored as biomarkers.
The committee agreed that in the absence of good evidence for their accuracy, the 2014 recommendations that cystatin-c equations should be considered during diagnosis in certain circumstances, should be removed. In particular, they noted that although using cystatin-c equations may reduce false-positive results, it is likely to also increase false-negative results. This will avoid potentially misleading tests being conducted and the costs associated with these. They made a research recommendation for a large study using UK data to evaluate the accuracy of cystatin-c equations (see other recommendations for research).
There will be an impact on practice, as the adjustment of the CKD-EPI creatinine equation for adults of African-Caribbean or African family origin has been removed from the guideline. Only a small number of centres in the UK currently use cystatin-c equations regularly, so most should not be affected by the removal of the cystatin-c recommendations.
For children and young people with CKD, there was no evidence for the accuracy of measuring albumin:creatinine ratio (ACR) compared with protein:creatinine ratio (PCR) to quantify proteinuria. The committee discussed the recommendations for adults and agreed that, overall, these fit well with current practice and can be recommended for children and young people as well.
The committee discussed the eGFR threshold recommended for quantifying urinary albumin or urinary protein loss in adults without diabetes. They agreed that this threshold is not appropriate for children and young people because any reduction in GFR in this population would prompt measuring proteinuria. Therefore, for children and young people they set the threshold for creatinine as above the upper limit of the age-appropriate reference range.
The committee agreed to make a research recommendation to identify the effect of measuring proteinuria with ACR compared with PCR on the timing of treatment changes in children and young people with CKD and the consequences of the delay in treatment changes on different levels of proteinuria. (See other recommendations for research.)
The recommendations are in line with current practice, so no additional resources should be needed.
The evidence showed that reagent strips were less useful to rule out than to rule in proteinuria. The committee highlighted that ruling out proteinuria with confidence was the main goal when using reagent strips. Therefore, they agreed that reagent strips should not be used to identify proteinuria in children and young people. The evidence was not reviewed for adults and so the committee agreed to retain the 2014 recommendation not to use reagent strips to identify proteinuria in adults unless the strips are capable of specifically measuring albumin at low concentrations and expressing the result as an ACR. The committee also highlighted that these tests are commonly used in clinical practice and agreed to make a further recommendation for further investigations in adults, children and young people with an incidental finding of unexplained proteinuria on reagent strips. Further testing is needed to confirm CKD by identifying other markers of kidney damage (such as ACR or glomerular filtration rate).
There was limited evidence on the accuracy of reagent strips for albuminuria, so the committee did not feel able to make recommendations. There were only 2 studies, and only 1 showed that reagent strips could be useful.
There was no evidence on the accuracy of reagent strips for haematuria in children and young people. The 2014 guideline (which did not cover children and young people) recommended reagent strips for detecting haematuria in adults. The committee agreed to extend this recommendation to children and young people, because the evidence for adults is likely to be applicable to this population.
The recommendations are in line with current practice, so no additional resources should be needed. The committee noted that if all dipstick tests are confirmed by laboratory testing anyway, there would be extra costs attached to using dipsticks as a first step, which were not justified by the benefits.
For children and young people, the evidence showed that acute kidney injury and solitary functioning kidney were clinically significant risk factors for developing CKD. The committee highlighted that solitary functioning kidney was not due to kidney donation but to nephrectomy secondary to congenital anomalies of the kidney and urinary tract or to a lack of a kidney at birth or a non-functioning kidney.
The committee highlighted that there were other important risk factors for developing CKD in children and young people, but that no evidence was found for these. Based on their clinical knowledge and experience, they added 'gout' as a risk factor for adults and 'low birth weight' as a risk factor for children and young people.
The committee agreed that the frequency of monitoring (for developing CKD or progression) should be individualised for adults, children and young people. This is to address the different characteristics and risks that each person will have.
The committee agreed that more research on risk factors for developing CKD in children and young people would help to strengthen current guidance, so they made a research recommendation. (See other recommendations for research.)
The recommendations are in line with current practice, so no additional resources should be needed.
Most of the evidence showed that with eGFR decline, the risk of kidney disease progression and mortality increases, and this risk increases with the rate of eGFR decline. The committee agreed this is observed in clinical practice and any person presenting with an increase in eGFR decline would be monitored more frequently. The committee reviewed the recommendations and agreed that they are consistent with the evidence and clinical practice. They agreed to clarify monitoring by stating that repeat assessment is to be agreed with each person with or at risk of CKD.
The committee agreed that the frequency of monitoring they recommended was a minimum level and that more frequent monitoring would be appropriate for some patients. This should also be guided by rate of change in eGFR or ACR and specific comorbidities, including diabetes. ACR monitoring should be individualised. For example, ACR might be monitored more frequently in people with high ACR (categories A2 or A3), or if a change in ACR would affect management.
The committee made a research recommendation to identify the optimal frequency of ACR monitoring in adults, children and young people with CKD. (See other recommendations for research.)
The committee discussed whether specific recommendations are needed for children and young people with CKD and decline in eGFR, but agreed that this population would be referred to specialist care.
The committee noted that no changes had been made to the previous suggested monitoring schedule, and they believed it was relatively well implemented in clinical practice. Therefore, they were confident there should not be a substantial impact on practice from the new recommendations.
Results from a meta-analysis (including the SPRINT trial) showed no meaningful difference between standard and more intensive blood pressure targets for adults with CKD. The 2014 guideline recommended maintaining systolic blood pressure below 140 mmHg and diastolic blood pressure below 90 mmHg. This is consistent with clinical practice and with the NICE guideline on managing hypertension. The committee noted that although there is limited evidence on blood pressure targets in people with CKD and proteinuria, it is important to maintain a systolic blood pressure below 130 mmHg and a diastolic pressure below 80 mmHg.
The committee agreed that none of the evidence they had seen warranted changing the recommendations. They also noted that intensive blood pressure targets only result in a marginal reduction in stroke and kidney failure, but put a large burden on patients in terms of polypharmacy and associated risks and side effects (such as falls).
The committee agreed that a useful target for blood pressure in children and young people with CKD and proteinuria is a systolic blood pressure below the 50th percentile for height.
The committee agreed that particular care had to be taken with people who were frail or who had multiple morbidities. However, the NICE guideline on hypertension already covers this group, so the committee did not make new recommendations.
The recommendations for adults are consistent with current practice and should not have an impact on resources. The recommendation for blood pressure targets in children and young people may have some cost implications, although the committee did not think they would be significant.
The interventions recommended are intended to improve a range of outcomes, including rates of progression to end-stage renal disease. There was evidence for adults, but not for children and young people. Paediatric experts on the committee agreed that the evidence for adults was also applicable to children and young people. Therefore, the committee did not make separate recommendations for different age groups.
The evidence for adults covered people with proteinuria or albuminuria, and included people with diabetes. This allowed the committee to make separate recommendations for people with and without diabetes. In the committee's experience, many people with diabetes and CKD are frail, or are taking a lot of medicines, so they made a recommendation to address this.
The evidence showed that, compared with placebo, ACE inhibitors reduced the risk of end-stage renal disease in people without diabetes. ARBs did not show the same effect. However, the committee did not believe the evidence was sufficiently robust to show that ACE inhibitors were better than ARBs. In addition, for people with type 2 diabetes, ARBs did reduce the risk of end-stage renal disease and heart failure. Based on the limitations of the evidence and the evidence available for people with type 2 diabetes, the committee recommended both ACE inhibitors and ARBs.
For people with type 2 diabetes, ARBs reduced the risk of end-stage renal disease and heart failure. The committee also recommended ACE inhibitors because the evidence did not show a clear difference between ACE inhibitors and ARBs on the following outcomes:
reduction of proteinuria
end-stage renal disease
non-fatal cardiovascular events
adverse events (hypotension)
There was no evidence comparing ACE inhibitors with placebo in people with type 2 diabetes. The evidence for people without diabetes did show that ACE inhibitors reduced the risk of end-stage renal disease, compared with placebo. The committee used this evidence to make the recommendation for people with diabetes.
The committee also noted evidence that SGLT2 inhibitors reduced the risk of end-stage renal disease, mortality and hospitalisation in adults with type 2 diabetes, and made a recommendation for this. The committee agreed that this was probably a class effect rather than a single medicine. When making the recommendation, the committee discussed the different proteinuria thresholds used in the studies to recruit participants. They agreed that an ACR of 30 mg/mmol was appropriate and consistent with the inclusion criteria of all the studies they had considered. The committee cautioned that these medicines are not suitable for everyone and should only be used within their marketing authorisation. The committee highlighted that people taking these medicines should also have monitoring, in line with the advice in the BNF. Volume depletion was a particular issue of concern for the committee, as this may cause falls for people with frailty.
The committee noted the limited cost-effectiveness evidence, and the potential impact the cost of these medicines may have. However, they noted the medicines did have positive NICE technology appraisal guidance for use in people with diabetes without CKD, and agreed that with the additional renal benefits in people with CKD, they were likely to be more cost effective in this population. Specifically, they noted NICE technology appraisal guidance that SGLT2 inhibitors are cost effective as first-line treatment in people with diabetes for whom metformin, sulfonylurea and pioglitazone are not appropriate, or as part of dual and triple therapy when earlier lines of therapy are not sufficient. These technology appraisals were mostly conducted before the publication of recent large trials directly looking at cardiovascular events and mortality, and were therefore based on extrapolations from intermediate endpoints (in particular HbA1c). The committee noted that their recommendations would bring forward the use of these medicines for some people (those who develop CKD and proteinuria before they meet the criteria for an SGLT2 inhibitor based solely on their diabetes).
The committee noted that the doses of SGLT2 inhibitors used in people with diabetes and CKD are lower than in people without renal impairment, and the committee would not expect the medicines to have the same benefits for blood glucose control as in people with diabetes but without CKD. However, with benefits on proteinuria and overall kidney outcomes, they were confident that the overall clinical benefit in people with diabetic kidney disease would be as large as the benefits estimated in the technology appraisals for people with diabetes but not CKD. The committee were therefore confident that, with a at least as large a benefit for a similar cost, it is appropriate that these medicines are available earlier for people with diabetes and CKD, and that this would represent a cost-effective use of NHS resources.
The committee agreed that this was a fast-moving area and that studies were being done to assess the usefulness of SGLT2 inhibitors in people with CKD who do not have diabetes, but they agreed the evidence was not yet strong enough to make a recommendation, even though it looked promising.
The recommendations on ARBs and ACE inhibitors reflect current practice, so no additional resources should be needed.
The recommendation for SGLT2 inhibitors might result in a significant change in practice, because it will mean these medicines are prescribed more widely, and this would come with a substantial cost impact. The committee noted, however, that this was likely to represent a cost-effective use of resources, with these medicines providing additional benefits on renal outcomes, as well as the benefits they provide for diabetes management.
There was limited evidence showing that eGFR thresholds below 60 ml/min/1.73 m2 could be used to identify anaemia as being due to CKD. The committee questioned the applicability of this evidence because the studies did not rule out other causes of anaemia (which is usually done in practice).
The limited evidence meant that the committee was unable to recommend specific thresholds or probabilities. Instead, they used the available evidence and their expertise to specify ranges of GFR indicating that anaemia is more or less likely to be caused by CKD.
When anaemia may have other causes (such as gastrointestinal bleeding and certain cancers), investigating further will increase the chance of the real cause being identified and treated.
Clinical judgement is needed on how extensively to look for other causes when eGFR is between 30 and 60 ml/min/1.73 m2. Healthcare professionals will need to balance the risks of:
putting people through extensive and unnecessary investigations when their anaemia is caused by CKD
missing the real cause of their anaemia by assuming it is caused by CKD.
The committee agreed that when eGFR is below 30 ml/min/1.73 m2, anaemia is more likely to be caused by CKD. However, healthcare professionals should still use their clinical judgement and think about people's circumstances when deciding whether further assessment is needed.
Only 1 study included people with diabetes, and no studies included children and young people. However, the recommendations still apply to these populations, because other causes of anaemia would be ruled out before attributing the anaemia to CKD.
The committee noted a need for further research on the diagnostic test accuracy of different eGFR thresholds, particularly for eGFR thresholds of 30 and 60 ml/min/1.73 m2. They highlighted that in clinical practice, an eGFR threshold of 45 ml/min/1.73 m2 can also trigger investigation into anaemia due to CKD, but limited evidence was identified for the diagnostic accuracy of this threshold. The committee made a research recommendation on the diagnostic accuracy of these specific eGFR thresholds for determining the likelihood of anaemia being CKD related. (See other recommendations for research.)
These recommendations should not increase the cost to primary care, because they reflect current practice and act as cautions for healthcare professionals to explore the cause of anaemia. They may reduce costs by ensuring that the correct cause of anaemia is identified more quickly with appropriate investigations.
In the 2015 guideline, an aspirational Hb range between 100 and 120 g/litre was recommended for adults, young people and children aged 2 years and over. For children under 2 years, the Hb range was between 95 and 115 g/litre. These were based on evidence for adults. In 2020, the committee reviewed the evidence specifically for children and young people. The only evidence for this population came from a single small low-quality study, comparing the effects of a high and low Hb target on left ventricular mass index. No difference in effect was found. Given the lack of evidence, the committee agreed that the recommendations made in 2015 should not be changed.
The 2015 guideline recommended using the same target Hb range as adults for children and young people over 2 years, and a slightly lower level in children under 2. However, children and young people have different coagulation risks than adults, and are more prone to reductions in Hb from blood loss in haemodialysis circuits. In practice, higher Hb targets (up to 130 g/litre) are often used for children and young people. Because of the lack of evidence in this age group, the committee agreed that research is needed to inform future guidance (see research recommendation 4).
For people with stage 5 CKD who are on in-centre haemodialysis, the evidence showed that high-dose intravenous iron was better than a low-dose regimen at increasing levels of serum ferritin and haemoglobin as well as increasing the haematocrit. The committee agreed that the type of intravenous iron was not relevant and that there was no reason to recommend a specific preparation. They also highlighted that there are differences between iron preparations that affect their bioequivalence. Therefore, pharmacist advice is likely to be needed when choosing iron preparations. An example regimen for adults using iron sucrose was taken from the evidence to help guide practice. Ultimately, the choice of preparation should be based on local availability and policies. The committee agreed that children and young people should be given a high dose as set out in the BNFc, although they noted that use of intravenous iron preparations in children under 14 years was off label.
The committee was aware of a MHRA alert on intravenous iron and serious hypersensitivity reactions. The alert states that 'intravenous iron products should only be administered when staff trained to evaluate and manage anaphylactic or anaphylactoid reactions – as well as resuscitation facilities – are immediately available.' The committee agreed that intravenous iron should not be administered at home but recognised that this has a significant impact on people on home dialysis.
Most of the evidence was from studies with participants on haemodialysis. The committee agreed that more research would help to inform future guidance on intravenous iron for people with stage 5 CKD who are on peritoneal dialysis.
The recommendations are unlikely to lead to a substantial change in costs, as intravenous iron is relatively inexpensive, and there was evidence found in adults that use of high-dose iron leads to lower doses of erythropoiesis-stimulating agents being used, thereby offsetting any extra costs.
There was a significant amount of evidence (of varying quality) for adults with stage 5 CKD who are having dialysis. However, evidence was limited for adults not on dialysis, and for children and young people. The committee agreed to extrapolate from the evidence for adults with stage 5 CKD on dialysis, so they could make recommendations for the other groups.
People's preferences need to be taken into account when offering phosphate binders, because this could have an impact on adherence. The differences in phosphate binder formulations (for example, chewable and non-chewable) and the effect this has on how they are taken (before, with or after food) mean that people will often prefer one phosphate binder over the others. Oral phosphate binders are also unpleasant to take, and this might affect adherence as well. It is important to involve people in the choice of phosphate binder as far as possible, to ensure they are prescribed one they are happy with and can take as recommended.
The committee highlighted several factors that renal physicians assess at clinical reviews for people who are taking phosphate binders (including parathyroid hormone, vitamin D and serum calcium).
The committee reviewed the recommendations from the 2013 guideline in the light of limited new evidence. For children and young people with high serum calcium, they agreed to recommend sevelamer carbonate instead of sevelamer hydrochloride. This is because sevelamer carbonate offers a better balance of benefits and costs. The committee highlighted that in growing children and young people, calcium is often maintained close to, but not above the upper limit of the age-related reference range. Calcium is essential for bone development in children.
The committee reviewed the evidence for phosphate binders both in adults on dialysis and adults not having dialysis. Although the evidence for those not on dialysis was limited, it did reflect the evidence for adults on dialysis in every area apart from sucroferric oxyhydroxide. As there was no evidence on sucroferric oxyhydroxide in adults not on dialysis, the committee did not recommend it for this group.
The evidence showed that the most cost-effective treatment strategy is to start with calcium acetate, and switch to sevelamer carbonate if the person gets hypercalcaemia. This is because:
calcium acetate as a first-line treatment provides the best balance of benefits, harms and costs
calcium carbonate is cheaper than calcium acetate, but is more likely to cause high serum calcium levels and associated adverse outcomes
sevelamer carbonate and sevelamer hydrochloride are more expensive than calcium acetate, and do not provide enough benefit as a first-line treatment to justify the extra expense
when people have high serum calcium levels and cannot take calcium acetate, sevelamer carbonate is the best alternative; it is cheaper than sevelamer hydrochloride, and provides similar benefits, however, it still costs more than calcium acetate and, for first-line treatment, it does not provide enough benefit to justify this extra expense
sucroferric oxyhydroxide is not cost effective as a first-line treatment, but is a reasonable choice for people who cannot take calcium acetate or sevelamer carbonate
lanthanum carbonate is much more expensive than calcium acetate and sevelamer carbonate and may provide less benefit than other non-calcium-based phosphate binders.
Based on this evidence, the committee recommended a treatment sequence and alternatives for different situations.
The committee also agreed that diet and dialysis (when appropriate) had a large impact on serum phosphate levels. Therefore, before offering phosphate binders it is important to provide dietary advice and ensure people are on the dialysis regime that works best for them.
The committee made research recommendation 5 to address the lack of evidence in adults not on dialysis.
Replacing sevelamer hydrochloride with sevelamer carbonate may result in lower resource use, because there is a cheap generic version of sevelamer carbonate available.
There is currently variation across the UK in use of sucroferric oxyhydroxide. The recommendation on this phosphate binder may increase costs. However, this increase is unlikely to be substantial, because sucroferric oxyhydroxide is only recommended as a third-line option.