Microalbuminuria as predictor of outcomeBMJ 1999; 318 doi: https://doi.org/10.1136/bmj.318.7178.207 (Published 23 January 1999) Cite this as: BMJ 1999;318:207
Shows promise but large prospective trials are needed
- G Evans, Research fellow.,
- I Greaves, Research fellow.
- Trauma and Surgery Group, Chemical and Biological Defence Sector, Defence Evaluation and Research Agency, Porton Down, Wiltshire SP4 0JQ British Crown Copyright
Microalbuminuria is used clinically to monitor incipient diabetic nephropathy, but it is known also to be a non-specific marker of inflammation, both systemic and local,1–3 and appears to be useful as a predictor of outcome in several clinical situations.4–8 The equipment required to perform the test is small and mobile, each test taking two minutes to perform. Might a test for microalbuminuria provide a simple means of identifying patients whose condition will deteriorate across a range of clinical conditions?
Microalbuminuria is defined as a urinary albumin concentration of 30-200mg/l. These values are currently detectable by semiquantitative dipstick tests. In healthy volunteers spot testing of urinary albumin shows the mean concentration to be 5.1mg/l, with 95% of samples having a concentration of less than 29.6mg/l, regardless of time of collection.9
Microalbuminuria is thought to reflect the glomerular component of a systemic capillary leak10 that is fundamental to the pathogenesis of multiple organ failure.11 In the healthy kidney over 99% of filtered albumin is resorbed by mechanisms that are close to saturation. A small increase in glomerular vascular permeability results in an increase in filtered albumin presented to the renal tubules. This cannot be resorbed and results in a large increase in urinary albumin. This amplification results in a rapid rise in albumin excretion after the onset of systemic inflammation, peak microalbuminuria being detectable up to two days earlier than other markers of systemic inflammation such as C reactive protein. 1 4 In conditions such as acute pancreatitis and bacterial meningitis microalbuminuria may be detected immediately on hospital admission.
Two criteria must be met for a test to be useful clinically to predict outcome and enable resources to be targeted at patients likely to develop a more serious clinical course. Firstly, a measurable rise in microalbuminuria must occur early enough to permit clinical intervention. Secondly, this rise should correlate with outcome. Does microalbuminuria meet these criteria?
In elective surgery for aortic aneurysm all patients displayed the predicted rise in microalbuminuria.6 This rise was slight and transient in patients who subsequently recovered uneventfully. Those patients who developed serious complications, especially pulmonary dysfunction, displayed a larger and more persistent rise. The ratio of urinary albumin:creatinine measured four hours after induction of anaesthesia distinguished the two groups of patients, even though it was not always possible to predict outcome clinically.
The sensitivity and specificity of microalbuminuria as a predictor of outcome has been measured in several conditions. Gosling et al found an early rise in urinary albumin concentration useful in distinguishing myocardial infarction from angina.12 Unfortunately, it was not sufficiently sensitive or specific unless combined with measurement of urinary gammaglobulin (IgG), when the sensitivity rose to 79.6% and specificity to 96.2%.
In bacterial meningitis Roine showed that microalbuminuria distinguished bacterial meningitis from aseptic meningitis with a specificity of 94%.2 The same study also showed that the concentration of urinary albumin correlated with clinical outcome, but this relation did not approach statistical significance.
In a study of acute pancreatitis Shearman et al found that microalbuminuria peaked at 36hours after admission and that serious complications developed later only in those patients with the highest rates of microalbuminuria excretion.4 All patients with lower urinary albumin excretion rates (less than 200µg/min) recovered without serious sequelae, regardless of serum amylase concentration.
Recently, Pallister et al examined the development of adult respiratory distress syndrome after major trauma.7 This built on previous work in trauma victims8 and showed that measurement of microalbuminuria eight hours after admission predicted the development of the syndrome with a positive predictive value of 85% and a negative predictive value of 95%. This was therefore useful in identifying patients who, although seriously injured, were unlikely to develop the syndrome. This prediction facilitated a more rational use of high dependency beds and allowed earlier and more aggressive treatment to be targeted at those most likely to benefit.
Microalbuminuria has been shown to correlate with mortality from all causes in patients aged over 60,even when other known risk factors are taken into account.5 This raises the possibility that concentrations of urinary albumin too low to be detected by currently popular dipstick urinalysis may be useful in “well person” clinics.
The measurement of microalbuminuria is a simple, quick, and non-invasive technique which is showing promise as an early and sensitive indicator of disease severity in a range of clinical conditions. If this promise is confirmed microalbuminuria should prove to be useful in targeting early and aggressive treatment at those patients most at risk of serious complications and adverse outcomes. The next step is to confirm the role of microalbuminuria in a wider range of conditions and then to examine the predictive value in large scale prospective trials.