Intended for healthcare professionals

Practice Rational Testing

Estimated glomerular filtration rate

BMJ 2014; 348 doi: (Published 24 January 2014) Cite this as: BMJ 2014;348:g264
  1. Philly O’Riordan, general practitioner1,
  2. Paul E Stevens, consultant nephrologist and medical director2,
  3. Edmund J Lamb, consultant clinical scientist3
  1. 1City Road Medical Centre, London, UK
  2. 2Kent Kidney Care Centre, East Kent Hospitals University NHS Foundation Trust, Canterbury, UK
  3. 3Clinical Biochemistry, East Kent Hospitals University NHS Foundation Trust, Canterbury CT1 3NG, UK
  1. Correspondence to: E Lamb elamb{at}

What should you do when confronted with a patient with previously unrecognised moderate chronic kidney disease? This article gives helpful advice in this not uncommon scenario

Learning points

  • Confirm a finding of low estimated glomerular filtration rate (eGFR) with a repeat sample taken after avoidance of meat for at least 12 hours. Consider spurious causes of low eGFR such as high muscle mass. Take into account normal biological and analytical variability when interpreting eGFR results (and most other laboratory tests)

  • A low eGFR result should prompt a check for proteinuria by measuring urinary albumin to creatinine ratio

  • Stage 3 (moderate) chronic kidney disease, especially in the absence of proteinuria or haematuria, is typically stable and can be safely managed in primary care

  • In some instances, the cause for a patient’s chronic kidney disease may be uncertain, but this does not affect the management

A 75 year old white woman presented to the emergency department after a fall with a displaced Colles’ fracture requiring surgical management. She was euvolaemic, her blood pressure was 158/89 mm Hg, and her body mass index was 19. She consumed no alcohol, smoked 20 cigarettes a day, and was not taking regular medication. Preoperative blood tests showed a serum creatinine concentration of 140 μmol/L (reference interval 49-90 μmol/L) and estimated glomerular filtration rate (eGFR) 32 mL/min/1.73 m2. Blood electrolytes, liver function tests, and glucose concentration were normal. One day postoperatively her creatinine level remained increased (125 μmol/L, eGFR 36 mL/min/1.73 m2). On discharge, the orthopaedic team asked her general practitioner to investigate her renal function further.

What are the next investigations?

Estimated glomerular filtration rate

In the UK approximately 1 in 20 people are thought to have stage 3 (moderate) and later chronic kidney disease,1 2 which is generally identified by means of estimation of glomerular filtration rate3 and detection of protein in the urine (albuminuria or proteinuria). Case finding in high risk individuals is recommended (figure),4 but opportunistic detection, as here, is common.


Investigation and management of chronic kidney disease in people without diabetes (adapted from National Institute for Health and Care Excellence (NICE) clinical guideline4)

Glomerular filtration rate is the best overall index of kidney function and ideally would be measured with reference procedures that follow the clearance of an infused exogenous substance (such as EDTA labelled with chromium-51). However, these methods are cumbersome and impractical for general detection and management of kidney disease. Estimation of glomerular filtration rate by means of equations is widely used as a surrogate measure of glomerular filtration rate. Such equations use the inverse relationship between serum creatinine concentration and glomerular filtration rate, making adjustments for the non-renal influences of age, gender, and ethnicity on that relationship.

Estimates of glomerular filtration rate are normally expressed per 1.73 m2 body surface area, which is an average surface area for an adult. Hence such estimates might overestimate true glomerular filtration rate in small individuals and vice versa. Estimated glomerular filtration rate (eGFR) may be falsely low after a meal of high meat content, as blood creatinine concentration increases after meat intake, potentially misclassifying patients as having chronic kidney disease.5 Thus patients should avoid eating meat for at least 12 hours before repeat testing.5 Blood creatinine is predominantly derived from muscle. In this patient’s case the low body mass index could suggest poor nutrition: consequently her creatinine concentration may reflect reduced muscle mass and her true glomerular filtration rate may be lower. (Note that the converse can also be true: a muscular young man may have increased serum creatinine concentration and a falsely lowered eGFR). Table 1 lists the situations in which serum creatinine may give rise to unreliable estimates of glomerular filtration rate.

Table 1

 Situations in which glomerular filtration rate (GFR) estimation by means of serum creatinine concentration may be unreliable (adapted from KDIGO chronic kidney disease guideline7)

View this table:

When is a change in eGFR significant?

Serum creatinine concentration, as with all biomarkers, has a natural biological variation, to which laboratory variation should be added: a true change in glomerular filtration rate cannot be inferred with any degree of certainty unless the so called reference change value is exceeded.6 Based on the biological and analytical variation of serum creatinine, the reference change value for eGFR is about 14%: in this patient’s case the eGFR would need to change by approximately 5 mL/min/1.73 m2 for that change to be considered significant with 95% certainty.

Although this woman’s eGFR is low, acute kidney injury is unlikely given that the preoperative and postoperative serum creatinine results were essentially unchanged. Nevertheless, the patient’s general practitioner should request a repeat eGFR within two weeks of discharge to exclude acute kidney injury.

Urinary albumin:creatinine ratio

In addition to repeating serum creatinine measurement to obtain the eGFR, the general practitioner will need to check for proteinuria. The presence of substantial proteinuria implies that the patient is at increased risk of adverse outcome, including progressive kidney failure, acute kidney injury, cardiovascular disease, and death.7 Proteinuria is most effectively checked by requesting a urinary albumin:creatinine ratio on a random or early morning urine sample. Protein reagent strips (“dipsticks”) are notoriously unreliable, and their cost effectiveness has never been established.8 9 Laboratory measurement of urine total protein lacks sensitivity and standardisation. Urinary albumin is the main urinary protein in the vast majority of nephropathies and can be measured with sensitive, specific, and quantitative assays.8 10 In the setting of non-diabetic chronic kidney disease a urinary albumin:creatinine ratio ≥30 mg/mmol denotes significant proteinuria requiring further treatment.4 However, it is increasingly clear that morbidity and mortality risk increase at much lower levels of albuminuria,11 and recent guidance has proposed a more complex staging system in which both diabetic and non-diabetic individuals will be considered proteinuric when albumin:creatinine ratio exceeds 3 mg/mmol.7

Additional tests if chronic kidney disease is confirmed

Chronic kidney disease is defined as abnormalities of kidney structure or function, present for at least three months, with implications for health.7 The abnormalities most often encountered clinically that meet this definition are a sustained low glomerular filtration rate (<60 mL/min/1.73 m2) and albuminuria, and current classification systems base the staging of chronic kidney disease on glomerular filtration rate and albuminuria (table 2). Common causes of chronic kidney disease include diabetes and vascular disease.

Table 2

 Stages of chronic kidney disease (CKD) as defined by NICE4

View this table:

In the present case, although no glucose abnormality was detected while the patient was in hospital, it would be prudent to recheck her blood glucose and to recheck her blood pressure. The ratio of serum cholesterol to high density lipoprotein (HDL) cholesterol should be requested, in addition to baseline blood tests for metabolic abnormalities related to chronic kidney disease (blood haemoglobin, calcium, phosphate, potassium).

Although derangements of bone mineral metabolism are uncommon in the early stages of chronic kidney disease, these become more common at lower levels of glomerular filtration rate and are not unexpected in late stage 3B chronic kidney disease. Given that this patient had a low impact fracture, it would be reasonable to measure her blood 25-hydroxyvitamin D and parathyroid hormone concentrations. Although later stages of kidney failure are characterised by failure of vitamin D activation (renal 1-α hydroxylation), rendering 25-hydroxyvitamin D concentrations difficult to interpret, substrate vitamin D deficiency is common in stage 3B. Parathyroid hormone concentrations may reflect vitamin D deficiency but are altered by many factors, including kidney disease, and a mildly elevated concentration (<4 times the upper limit of normal) would be common with this patient’s glomerular filtration rate.

Urine should be checked for haematuria with a reagent strip test. If positive, this may indicate occult malignancy, but in the presence of substantial proteinuria would be more suggestive of underlying glomerular disease.

A renal ultrasound scan is of value to exclude structural abnormalities of the kidney or obstructive uropathy and to assess renal morphology. Bilateral small kidneys would suggest longstanding chronic kidney disease.


The patient’s repeat eGFR was 33 mL/min/1.73 m2, which excluded acute kidney injury, and she had neither haematuria or proteinuria (albumin:creatinine ratio 2.0 mg/mmol). An eGFR blood test repeated three months later was essentially unchanged, consistent with stage 3B (moderate) chronic kidney disease.7 12 Her general practitioner reviewed her social, family, and drug history. As chronic kidney disease confers increased vascular risk, the importance of smoking cessation was emphasised and a calcium channel blocker was prescribed as she remained hypertensive (figure). Blood glucose, calcium, phosphate, potassium, and cholesterol concentrations were normal. Although her cholesterol:HDL cholesterol ratio was not increased (4.0), given her vascular risk she was offered a hydroxymethyl glutaryl coenzyme A reductase inhibitor for primary prevention of cardiovascular disease.13 Her haemoglobin concentration was 110 g/L, which was consistent with her stage of chronic kidney disease and did not require treatment.14

Renal ultrasound showed kidneys of normal morphology save for a bipolar length of 8 cm bilaterally. The cause of chronic kidney disease is uncertain in this case but may be related to hypertension or prior unrecognised renal insult. This is not uncommon, and an extensive search for less common causes of chronic kidney disease (such as vasculitis) is not warranted. The patient’s 25-hydroxyvitamin D concentration demonstrated insufficiency,15 and she was offered vitamin D replacement. Her general practitioner also undertook further assessment of her fracture risk.

In England and Wales it is recommended4 that glomerular filtration rate should be estimated every six months in people with stage 3 chronic kidney disease, who comprise about 6-7% of the overall UK population.2 Currently the Quality and Outcomes Framework suggests that the urinary albumin:creatinine ratio should be monitored annually. The aim of disease monitoring is to identify and manage those most likely to progress to kidney failure or death. Most people with stage 3 chronic kidney disease, especially those with stage 3A chronic kidney disease (glomerular filtration rate 45-59 mL/min/1.73 m2) and no proteinuria, are not at increased risk of progressive disease, but they are at increased risk of future adverse events (acute kidney injury, cardiovascular disease, mortality, fractures, infections).16 17 Nevertheless, their increased risk can be safely managed in primary care.18 19 Indications for consultation with the local renal service or referral are given in box 1.

Box 1: Recommended criteria for referral to a renal specialist (adapted from KDIGO chronic kidney disease guideline7)

We recommend referral to specialist kidney care services for people with chronic kidney disease in the following circumstances:

  • Acute kidney injury or abrupt sustained fall in glomerular filtration rate (GFR)

  • Glomerular filtration rate <30 mL/min/1.73 m2*

  • Consistent albuminuria (urine albumin:creatinine ratio >70 mg/mmol†)

  • Progression of chronic kidney disease (defined below‡)

  • Urinary red cell casts, red blood cell count >20 per high power field (sustained and not readily explained)

  • Chronic kidney disease and hypertension refractory to treatment with four or more antihypertensive agents

  • Persistent abnormalities of serum potassium concentration

  • Recurrent or extensive nephrolithiasis

  • Hereditary kidney disease.

We recommend timely referral for planning renal replacement therapy in people with progressive chronic kidney disease in whom the risk of kidney failure within 1 year is 10-20% or higher§, as determined by validated risk prediction tools.

  • *If this is a stable isolated finding, formal referral (formal consultation and ongoing care management) may not be necessary, and advice from specialist services may be all that is required to facilitate best care for the patients (This will depend on the healthcare system)

  • †The KDIGO guideline7 suggests a threshold albumin:creatinine ratio of 30 mg/mmol: we have used 70 mg/mmol for consistency with the NICE guideline4

  • ‡Chronic kidney disease progression defined as a decline in glomerular filtration rate category (>90, 60-89, 45-59, 30-44, 15-29, <15 mL/min/1.73 m2). A certain drop in eGFR is defined as a drop in GFR category accompanied by a ≥25% drop in eGFR from baseline. Rapid progression is defined as a sustained decline in eGFR of >5 mL/min/1.73 m2/year.

  • §The aim is to avoid late referral (referral to specialist services <1 year before start of renal replacement therapy)


Cite this as: BMJ 2014;348:g264


  • This series of occasional articles provides an update on the best use of key diagnostic tests in the initial investigation of common or important clinical presentations. The series advisers are Steve Atkin, professor, head of department of academic endocrinology, diabetes, and metabolism, Hull York Medical School; and Eric Kilpatrick, honorary professor, department of clinical biochemistry, Hull Royal Infirmary, Hull York Medical School. To suggest a topic for this series, please email us at

  • Contributors: All authors contributed to the conception and design of the article, EJL produced an initial draft and PES and POR revised the article for intellectual content. All authors approved the final version of the article.

  • Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: PES was clinical advisor of the 2008 NICE guideline on chronic kidney disease and clinical chair of the forthcoming 2014 guideline. He was co-chair of the KDIGO 2012 guideline on chronic kidney disease. EJL was a member of the Guideline Development Group of the 2008 and the forthcoming 2014 NICE guideline on chronic kidney disease and of the KDIGO 2012 guideline.

  • Provenance and peer review: Commissioned, externally peer reviewed.

  • Patient consent not required (patient anonymised, dead, or hypothetical).


View Abstract

Log in

Log in through your institution


* For online subscription