Consultant nephrologist Dr Colin Jones answers questions on reviewing CKD patients during the pandemic, overdiagnosis and eGFR measurement
What is the bare minimum GPs should do in terms of reviewing CKD patients, bearing in mind that these consultations may be remote and we may have limited access to investigations and specialist help?
The Covid pandemic has imposed limitations on the services that can be provided in both primary and secondary care. Assuming there is an established diagnosis of CKD and that the underlying cause is irreversible, three essential elements need to be addressed in all cases and an additional element in those with more advanced CKD.
First, is everything possible being done to minimise the risk of progression and to manage cardiovascular risk factors? This will largely focus on blood pressure control and statin therapy.
Home blood pressure monitoring is useful in self-management of all people with CKD and has really come to the fore during the pandemic. Blood pressure targets are lower in people with proteinuric kidney disease (under 130/80mmHg) than those without proteinuria (under 140/90mmHg).1 Atorvastatin should be offered to manage hypercholesterolaemia.1
Second, are any complications developing because of reduced kidney function? These include uraemia, hypertension, oedema, anaemia, hypocalcaemia (manifest by proximal muscle weakness), acidosis and hyperkalaemia. A number of these can only be detected with a blood test.
Third, medication should be reviewed concentrating on two questions. Is any prescribed or OTC medication nephrotoxic and contributing to CKD? Is any prescribed or OTC medication excreted by the kidneys? In either case, medication withdrawal may be appropriate. In the latter case, dose adjustment or a change of medication to a drug that is not renally excreted may be required.
Finally, in people with advanced CKD (eGFR less than 20ml/min/1.73m2), is there a future care plan? People with CKD often have multimorbidity and there should be a clear plan with regards to the possibility of future renal replacement therapy (RRT) or end-of-life care, recognising that the majority of people with advanced CKD will die of comorbid disease and not of end-stage kidney disease (ESKD).
We may be unable to admit patients when we would like to, and we may be managing very sick patients discharged prematurely from hospital because of the pandemic. How can we minimise the risk of acute kidney injury (AKI), and what drugs should we be wary of using?
People with CKD are at increased risk of acute on chronic kidney injury. The commonest reason for a sudden drop in GFR is pre-renal failure caused by hypovolaemia, often in the setting of gastrointestinal fluid losses or a new or increased dose of diuretic medication. Inappropriate initiation of prescribed or OTC NSAIDs is also common. The risk of AKI can be reduced by following sick-day rules (tinyurl.com/ckd-leaflet), recognising the potential harm of injudicious medication withdrawal (tinyurl.com/ckd-sick-day) especially in people with heart disease.
In general, most people with AKI require hospital admission, both to make a diagnosis and to treat reversible factors (for instance, correct volume depletion with intravenous fluid; treat specific primary renal disease; relieve obstruction in post renal failure). These admissions should be accepted by secondary care.
People may be discharged earlier than previously. Attention should be paid to adequate hydration and restarting critical medications, including drugs that block the angiotensin system, as soon as possible. Withholding these drugs may increase mortality.8 Hydration status is best assessed by daily weight, which can be self-managed by the patient. A daily weight change of greater than 0.5kg suggests either fluid retention or worsening volume depletion.
CKD as a label causes a lot of anxiety in patients. Are we ‘overdiagnosing’ CKD in elderly patients?
A diagnosis of CKD may allow clinicians and patients to try to reduce cardiovascular risk and rate of loss of kidney function.1 If loss of kidney function cannot be prevented, those likely to need RRT – dialysis or a transplant – or to die of ESKD can be counselled. The diagnosis may also highlight those at risk of drug toxicity,2 either through the toxic effects of a medication on the kidney (such as NSAIDs) or accumulation of medication excreted by the kidneys (such as aciclovir toxicity if the dose is not corrected for GFR).
However, we need to balance the potential benefits of labelling an asymptomatic person as having CKD based on a laboratory test against the psychological impact of the diagnosis. Many people are anxious about the perceived ‘threat’ of ESKD and RRT.
When explaining to the patient, it may help to put things into some sort of perspective. Normal GFR in young adults is 80-120ml/min/1.73m2. It may help patients if they understand that symptoms of kidney failure typically develop at less than 15ml/min/1.73m2 and dialysis is not usually needed until GFR is 5-7 ml/ min/1.73m2 in adults. People choosing supportive care of ESKD typically die of comorbid disease or when GFR is 3-5ml/min/1.73m2. eGFR should also be interpreted in the context of the expected range for the patient’s age. Identifying a stable eGFR of 50ml/min/1.73m2 in a 90-year-old may be useful in reducing the risk of drug toxicity, but is unlikely to have any relevance in terms of risk of ESKD.
One of our local nephrologists has advised that in patients with eGFR below 60 the CKD ‘rules’ should not be invoked if serum creatinine is normal. In that circumstance, eGFR estimate is too inaccurate. How accurate is eGFR measurement? What factors lead to over- or underestimation?
Serum creatinine is dependent on kidney function, muscle mass and meat intake. Creatinine will be higher in people with an atypical muscle mass and following a meat-rich meal. Estimated GFR (eGFR) is based on regression equations derived from relatively small populations, many of whom had abnormal kidney function. The estimate of GFR becomes less reliable the closer the serum creatinine is to the normal range. There are a number of regression equations available and none has been shown to have a significant advantage.3
There is an argument that eGFR should not be reported if the serum creatinine is within the normal range, both because a ‘normal’ serum creatinine demonstrates that CKD is not present and because the formulae used to derive eGFR are not validated for a creatinine in the normal range. If creatinine is normal, kidney disease should only be suspected if there is other evidence of kidney damage (such as haematuria or proteinuria, abnormal renal imaging or a family history of genetic renal disease).
NICE guidance states that eGFR values greater than or equal to 60ml/min/1.73m2 should be interpreted with caution and that clinicians should consider measuring a cystatinC-based eGFR in people with a creatinine based eGFR of 45-59ml/min/1.73m2 when there is no albuminuria (ACR less than 3mg/mmol) and no other evidence of kidney disease before making a diagnosis of CKD.1 CKD should not be diagnosed in this group if the cystatinC-based eGFR is higher than 60ml/min/1.73m2. CystatinC measurements may not be universally available, but given the implications of a misdiagnosis and overtreatment, the onus is on clinicians to ensure that the label of CKD is applied correctly.
The same nephrologist feels that good blood pressure control in patients with CKD is more important than which agent is used, and says there is a good case for leaving the well controlled patient on existing blood pressure medications rather than changing to an ACE inhibitor or ARB. Would you agree?
There is good evidence that blood pressure control is pivotal in delaying progressive loss of kidney function, as well as reducing the risk of cardiovascular morbidity and mortality. Evidence for the superiority of drugs that block the renin-angiotensin-aldosterone system (RAAS) on renal outcomes is largely derived from studies of populations with diabetes and proteinuric renal disease.
Benefits are greater with higher levels of proteinuria. There is therefore a rationale to support good blood pressure control following NICE hypertension guidelines,4 rather than prioritising a particular agent in all people. However RAAS inhibitors should be the first choice in people with diabetes or proteinuria.
At what level should we consider investigating and managing anaemia in people with CKD? Is treatment with erythropoietin something we should leave to secondary care physicians?
Anaemia should always be investigated, focusing on diagnosing and treating reversible causes including iron, folic acid and vitamin B12 deficiency. Anaemia of CKD remains a diagnosis of exclusion and is unlikely to be due to CKD if eGFR is above 60 mlmin/1.73 m2.5 Treatment with erythrocyte stimulating agents (ESA) should only be considered if iron levels are adequate and is associated with harm as well as benefits. No trial has ever shown a survival benefit from treatment with an ESA, but quality of life and symptom burden can be improved.
NICE recommends investigation when haemoglobin is under 110g/l,5 although a lower threshold for treatment is supported by trial data.6 Target haemoglobin is 100-120g/l. ESA therapy is most successfully delivered through algorithms managed by an MDT with a high volume of activity to minimise risk and maximise cost-effectiveness.
Which patients with CKD should be referred for a renal opinion?
Referral would depend on the underlying diagnosis. CKD is not a diagnosis; it is a label indicating an irreversible reduction in GFR. Referral also depends on the presence or absence of blood or protein in the urine, the presence of metabolic complications or renal anaemia, the rate of decline in GFR and the likelihood that RRT would be considered by the patient or their medical team. Referral is appropriate if there is diagnostic uncertainty, a disease where specific therapy would alter the natural history of the disease, metabolic complications that could be managed by a renal MDT, or RRT needs to be discussed. Even if ESKD is inevitable, there may be opportunities for delaying disease progression. Planning for the appropriate management of ESKD, including consideration of pre-emptive live or deceased donor renal transplant, haemo- or peritoneal dialysis or symptomatic management of terminal uraemia requires time. Referral before the eGFR is below 20ml/min/1.73m2 or more than one year before the anticipated need for transplant or dialysis gives the renal MDT the opportunity to achieve the best RRT management plan.
Why do we send urine for protein quantification instead of dipsticking the urine in our practices? And we sometimes get confused with the urine ACR in diabetes and CKD. Why is proteinuria in non-diabetic patients not clinically significant until the ACR is 30mgmmol or more, but clinically significant at a much lower figure in diabetic patients (over 2.5mg/mmol in males and 3.5mg/mmol in females)?
Urine dipsticks were designed to ignore low-level proteinuria at a time when the medical focus was on glomerular disease and the clinical significance of low-level proteinuria had not been recognised.
Albuminuria is both an indicator of microvascular injury and of glomerular disease. In the general population, increasing albuminuria predicts the risk of cardiovascular death, even within the normal range. Albuminuria is an indicator of microvascular injury in people with hypertension4 and is the first indicator of glomerular injury in patients with diabetic nephropathy, a disease that is classically described as progressing from hyperfiltration to, albuminuria,, to dipstick positive proteinuria, to progressive kidney failure. However, albuminuria remains a stronger predictor of cardiovascular death in people with type 2 diabetes than it does of progression to ESKD. The injury to the glomerular filtration barrier that occurs in glomerulonephritis is characterised by higher levels of proteinuria, which often includes high molecular weight proteins as well as albumin. Many nephrologists prefer to use protein:creatinine ratio rather than ACR for patients with glomerular disease.
How important is it to advise patients on ACE inhibitors or ARBs to temporarily stop their treatment when they get diarrhoea or vomiting? Does this apply to all patients or just those at high risk – and if the latter, how may they be identified?
This is an area of controversy. Discontinuing medication that blocks the RAAS may reduce the risk of developing acute on chronic kidney injury or hyperkalaemia during an episode of intercurrent illness. However, there is increasing evidence that discontinuing these drugs in this situation is associated with an increased risk of death.8 The benefits and risks of stopping medication must be carefully assessed in each person. Those with heart disease, especially heart failure, may be better continuing their medication. If these drugs are stopped, they should be restarted as soon as possible either prior to discharge from secondary care or in primary care and with monitoring of kidney function and serum potassium. The RCGP offers excellent advice on AKI and the management of medication that blocks the RAAS in this context.9
NICE says: ‘stop metformin if the serum creatinine exceeds 150μmol/l or the eGFR is below 30.’ Given how useful metformin is in the treatment of diabetes, what would be your pragmatic advice for primary care on the prescribing of metformin and serum creatinine levels?
Metformin is associated with a decreased risk of death in people with type 2 diabetes. However, the risk of lactic acidosis, a rare but life-threatening complication of metformin treatment, increases as GFR falls. This complication is very rare at GFR greater than or equal to 30ml/min/1.73m2 and remains rare at GFR below <30ml/min/1.73m2, but has a high mortality rate. National and international guidelines therefore recommend discontinuing metformin at GFR below 30ml/min/1.73m2. Increasing evidence supports the use of SGLT-2 inhibitors because of their benefits on cardiovascular and renal outcomes in people with type 2 diabetes,7 but current trials are also limited to those with GFR above or equal to 30ml/min/1.73m2.
To what extent should we be looking for a specific ‘cause’ when we make a diagnosis of CKD? To this end, what assessment and which investigations should we perform?
As emphasised previously, CKD is not a diagnosis. Kidney failure may be pre-renal (usually acute or acute on chronic), intra-renal or post-renal. Hypertension, diabetes and vascular disease are the commonest causes of intra-renal CKD in an aging population with multimorbidity. There may be an obvious explanation for impaired kidney function based on the history, clinical examination and medication list.
A urine dipstick test for blood and laboratory testing for ACR should be performed in all patients and those with positive findings investigated for an intra-renal cause. Patients with systemic symptoms, an elevated CRP or disproportionate anaemia should be investigated for intra-renal causes, such as vasculitis or myeloma. Medication lists should be carefully scrutinised for potential nephrotoxins.
There may be few clues to a drug-induced interstitial nephritis, Proton pump inhibitors and NSAIDs are now the most commonly implicated medications in my practice. Obstructive uropathy (post-renal) can be caused by benign or malignant disease and may be reversible with prompt diagnosis and management. A renal tract ultrasound will exclude obstruction with a high degree of confidence. If there is diagnostic uncertainty and a diagnosis appears to be important, for example in someone who is likely to require RRT including renal transplant, referral to a renal physician for full investigation, which may include a renal biopsy, should be considered.
In our annual testing of patients with CKD, the calcium quite often comes back a little on the low side. What would be a sensible strategy to deal with this?
NICE guidance states that calcium, phosphate and PTH should only be measured when eGFR is below 30ml/min/1.73/m2. However, calcium is often available on routine biochemical profiles. Hypocalcaemia in the context of CKD suggests either vitamin D deficiency (prevalent in the general population) or secondary hyperparathyroidism due to the inability of the failing kidney to convert 25-hydroxyvitamin D3 into the more active 1,25-dihydroxyvitamin D3. In primary care, a low serum calcium should prompt a measurement of vitamin D. Insufficiency or deficiency should be managed according to local protocols. The management of secondary hyperparathyroidism and hyperphosphataemia requires the input of a renal MDT.
Key learning points
• The bare minimum CKD review should include management of CV risk factors, a check for complications and a review of medication
• Ensure patients with advanced CKD have a future care plan
• Remember to restart critical drugs such as ACE inhibitors, when possible, after an admission for AKI
• Symptoms of kidney failure only usually develop with an eGFR <15 ml/min/1.73m2 and dialysis is not usually needed until the eGFR is 5-7 ml/min/1.73m2
• Anaemia of CKD is a diagnosis of exclusion
Dr Colin Jones is a consultant nephrologist at York Teaching Hospital NHS Foundation Trust
References
1
NICE. Chronic kidney disease in adults: assessment and management. CG182 nice.org.uk/guidance/cg182
2
Jones C. When reduced renal function diagnosis may be beneficial. BMJ 2013;347:21
3
da Silva Selistre L, Rech D, de Souza V et al. Diagnostic Performance of Creatinine-Based Equations for Estimating Glomerular Filtration Rate in Adults 65 Years and Older. JAMA Intern Med 2019;179:796-804
4
NICE. Hypertension in adults: diagnosis and management. NG136 nice.org.uk/guidance/ng136
5
NICE. Chronic kidney disease: managing anaemia. NG8 nice.org.uk/guidance/ng8
6
Pfeffer M, Burdmann E, Chen C et al. A Trial of Darbepoetin Alfa in Type 2 Diabetes and Chronic Kidney Disease. N Engl J Med 2009;361:2019-32 tinyurl.com/y2kloqd6
7
Wang C, Zhou Y, Kong Z et al. The renoprotective effects of sodium-glucose cotransporter 2 inhibitors versus placebo in patients with type 2 diabetes with or without prevalent kidney disease: A systematic review and meta-analysis. Diabetes Obesity Metabolism 2019;21:1018-26. tinyurl.com/y243p95w
8
Brar S, Ye F, James M et al. Association of Angiotensin-Converting Enzyme Inhibitor or Angiotensin Receptor Blocker Use With Outcomes After Acute Kidney Injury. JAMA Intern Med 2018;178:1681-90 tinyurl.com/y2ee8k4z
9
RCGP Acute kidney injury. tinyurl.com/y3xt3mj4
