Three in ten children develop kidney injury after a liver transplant

Design

systematic review and meta-analysis of 17 retrospective cohort studies, Newcastle-Ottawa quality assessment, sensitivity and publication bias analyses

Population

2,986 paediatric liver transplant recipients

Primary outcome

pooled incidence of postoperative acute kidney injury and associated risk factors

Effect

pooled incidence 30.7% (95% CI 25.0-36.3); risk factors including preoperative hepatic impairment, coagulation abnormality, intraoperative blood loss, transfusion burden and ischaemia-reperfusion injury, most from few studies each

Seventeen retrospective cohorts covering 2,986 paediatric liver transplant recipients were pooled to estimate the incidence of postoperative acute kidney injury and identify its risk factors. The pooled incidence was 30.7% (95% CI 25.0-36.3), robust on sensitivity analysis with minimal evidence of publication bias.

The risk factors named are perioperative and largely familiar: preoperative hepatic impairment, coagulation abnormality, inadequate nutrition and oxygen delivery, intraoperative blood loss, transfusion burden, ischaemia-reperfusion injury, and the need for continuing organ support afterwards. The authors twice caution that most individual factors were assessed in only a handful of studies, and that the pooled estimates for them should be read cautiously.

So the incidence figure is the finding and the risk factor list is a hypothesis set. Thirty per cent is high enough that acute kidney injury should be treated as an expected complication with a defined monitoring protocol rather than a complication to detect. For a nephrologist covering a transplant service, the practical consequence is being involved before the operation in the children with pre-existing hepatic dysfunction and coagulopathy, rather than being called when the creatinine has already risen.

• Build acute kidney injury monitoring into the standard post-transplant protocol rather than responding to a rise
• Identify children with preoperative hepatic impairment and coagulopathy for closer perioperative renal input
• Creatinine is a poor marker in a malnourished child with low muscle mass — judge on trajectory and urine output too
• Risk factor estimates here rest on few studies each; do not build a scoring system from them
• Prospective data are needed before any of these associations guide an intervention

Why it matters

It sets the expected rate high enough that detection should be protocol-driven rather than reactive.

Don't overread it

Seventeen retrospective cohorts with varying AKI definitions — the incidence is an average across heterogeneous practice, not a rate for any one centre.

The statistics, in plain English

A pooled incidence of 30.7% with an interval of 25.0-36.3% is well estimated because incidence needs only a numerator and a denominator from each study. The risk factor estimates are much weaker: pooling two or three retrospective cohorts per factor produces numbers that look comparable to the incidence figure but are not, which is why the authors flag them separately.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.