A nest of nephrons
The kidneys are responsible for filtering toxins and by-products of metabolism out of the blood. At the heart of this purification system are hundreds of thousands of tubular structures known as nephrons.
Nephron count
Everybody is born with a finite number of nephrons, which then declines as we age. This means that older people are more vulnerable to the effects of nephron damage, because they typically have fewer healthy nephrons at the onset of injury.
Injury and resilience
Acute kidney injury (AKI) describes a range of conditions in which nephrons are damaged, impairing their function and reducing the efficiency of filtration through the glomerulus.
Mortality risk
AKI is classified as mild, moderate or severe on the basis of the contents and quantity of urine that a person produces.
Even mild AKI might increase the risk of death for people in hospital. In the most severe cases, the odds of death have been estimated to be around six-times higher2. AKI also increases the chance of stroke and heart attack.
An unwanted gift
AKI can be community acquired — as a complication of existing health problems, for instance — but in high-income countries it is thought to occur most commonly in hospitals as a consequence of factors such as surgery or sepsis.
A helping hand
Currently, the only way to treat AKI is renal replacement therapy, which usually involves dialysis. This is generally effective, but it is costly and onerous for patients and is typically reserved for people who are severely ill. A number of clinical trials are now exploring alternative approaches for treatment and prevention.
An earlier start
Without intervention, AKI can give rise to chronic kidney disease (see go.nature.com/3s83am6). Measures such as fluid management, blood-pressure monitoring and changes in medication can avert this crisis, but many cases of AKI are identified too late. Diagnostic tools that are currently in development could make a difference.
In 2019, an algorithm was able to predict 90% of AKI cases that would later require dialysis, up to 48 hours in advance8. However, prediction in women is worse than in men, probably because of a lack of representation in the training data9.