Intended for healthcare professionals


Glomerular filtration rate

BMJ 2006; 333 doi: (Published 16 November 2006) Cite this as: BMJ 2006;333:1030

Screening cannot be recommended on the basis of current knowledge

  1. Catherine M Clase, associate professor of medicine (clase{at}
  1. 1Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, L8N 1Y2 Canada

Low glomerular filtration rate predicts cardiovascular disease, end stage renal disease (a requirement for dialysis or transplantation), and death.1 In a population based study reported in this week's issue, Hallan and colleagues describe the yield of different screening strategies aimed at identifying people with low glomerular filtration rates.1 The guidelines of the United States kidney disease outcomes quality initiative define chronic kidney disease as glomerular filtration rate persistently lower than 60 ml/min/1.73m2, persistent urine abnormalities, or clinically relevant anatomical abnormalities, and they recommend measuring serum creatinine to determine estimated glomerular filtration rate in high risk groups.2 The Kidney Disease Improving Global Outcomes Group (an international organisation that seeks to improve care and outcomes for people with kidney disease) and draft UK guidelines endorse the classification but do not, at present, look at screening.3 4

All screening programmes cause physical, psychological, and social harm through the screening test itself, and through the investigation and treatment of abnormal results. No trials to show whether the benefits of screening for low glomerular filtration rate outweigh the harms have been conducted. So how should we proceed in the absence of such a trial?

The figure shows some areas in which the information needed to define a screening strategy is missing. Because low glomerular filtration rate is associated with cardiovascular disease, screening patients with established cardiovascular disease or risks for disease rather than unselected populations will lead to lower numbers needed to screen to identify one person.1 However, many people identified this way will have indications for interventions that randomised controlled trials have shown to prevent cardiovascular disease, end stage renal disease, or death, which limits the potential for differential intervention (application of treatments that would not have been used otherwise). Lower numbers needed to screen can also be achieved by screening older people,1 but the same consideration applies.


Information needed to define a screening strategy for low glomerular filtration rate

The screening test and “screen positive” threshold also need to be defined. Many authorities recommend using the estimated glomerular filtration rate.2 3 4 In the study by Hallan and colleagues, 20% of people with estimated glomerular filtration rates lower than 30 ml/min/1.73m2, but only 1-2% of those with values of 30-60 ml/min/1.73m2, progressed to end stage renal disease over eight years. Many people in the community (especially elderly people) with low glomerular filtration rate have disease that does not progress, which further limits the potential benefit of screening.1 5

Identifying people with a low glomerular filtration rate may lead to harm from labelling. For example, in people screened for hypertension at work, positive results can result in absenteeism and reduced income.6

Little research has been conducted into which tests should be done to diagnose possible underlying kidney diseases in people with low glomerular filtration rate.

Because estimated glomerular filtration rate and serum creatinine are continuous variables, and because of uncertainty about normal values and how they change with age, many people with results just above the screening threshold cannot be told that they have normal values. These people cannot be fully reassured, but do not benefit from additional treatment, and they will only be harmed by screening.

The purpose of screening is to identify people who should be treated differently. Angiotensin converting enzyme inhibitors,7 angiotensin II receptor blockers (in people with diabetes8), and low protein diets9 help prevent end stage renal disease in people with low glomerular filtration rate. Evidence of the benefit from angiotensin converting enzyme inhibitors and low protein diets comes largely from studies of people referred to nephrologists and may not generalise to people at lower risk.10 Two further issues exist—additional benefits from angiotensin converting enzyme inhibitors and angiotensin II receptor blockers will be limited to people not already taking them for other indications; and the intensive dietary education to prevent malnutrition that has been used in studies on low protein diets is not practical for large numbers.

Given the obstacles to the design of a screening strategy, what are the research priorities? More information is needed on how age, proteinuria, kidney function, diabetes, and blood pressure combine to predict progression to end stage renal disease, cardiovascular disease, and death in the general population. The proportion of people in candidate risk groups in whom treatment would be modified on the basis of data on kidney function must be defined. Modelling studies would identify whether potentially cost effective screening strategies exist; if so, these should be tested in randomised clinical trials.

In the meantime what should clinicians do? The study by Hallan and colleagues found that screening people with hypertension, diabetes, or age over 55 was the most effective strategy—93.2% (95% confidence interval 92.4% to 94.0%) of all cases of chronic kidney disease were detected, and the number needed to screen to find one case was 8.7 (8.5 to 9.0). Rather than screening, case finding should therefore continue in patients who have risks for or evidence of kidney disease, who are taking drugs that affect kidney function, or whose comorbidities make knowledge of kidney function important. Wide use of angiotensin converting enzyme inhibitors in people who have cardiovascular disease or diabetes and another cardiovascular disease risk factor will automatically treat many people with low glomerular filtration rate and will prevent cardiovascular disease and death.11 Maximising the proportion of hypertensive people who attain conventional blood pressure targets will prevent cardiovascular disease and death and is also likely to prevent end stage renal disease.


  • Competing interests: None declared.


View Abstract