Hydroxyethyl starch and the kidney: 114 trials, no signal, and a boxed warning that predates them

Design

systematic review and meta-analysis with trial sequential analysis, 114 randomised trials from 2009 to 2026, six languages

Population

13,951 surgical patients — 46% major abdominal, 30% cardiac, 24% other — nearly all given HES for under 24 hours

Primary outcome

acute kidney injury and preoperative-to-postoperative change in creatinine concentration

Effect

AKI RR 1.02 (95% CI 0.91 to 1.16); creatinine -0.62 micromol/l (95% CI -3.82 to 2.57)

This meta-analysis pooled 114 randomised trials published between 2009 and 2026, covering 13,951 surgical patients given HES 130/0.4 or 130/0.42. Nearly all trials confined administration to under 24 hours. Acute kidney injury did not differ from comparators (RR 1.02, 95% CI 0.91 to 1.16), nor did the change in creatinine from before to after surgery (-0.62 micromol/l, 95% CI -3.82 to 2.57). Trial sequential analysis reported adequate information size and strong non-inferiority for the creatinine outcome. Mortality and severe adverse events were no worse.

The point of the paper is the mismatch between this evidence and the regulatory record. The FDA's boxed warning came from renal injury in critically ill and septic patients in 2013, and was extended to surgical and trauma patients in 2021; the EMA began withdrawal in 2022. Those decisions rest on a population given starch for days in intensive care, not on patients given a bounded intraoperative volume. This analysis says the surgical case was never the same case.

What it does not do is remove the warning. A prescriber acting on this has to reconcile a favourable evidence base with an unfavourable label, and that is a governance question as much as a clinical one. If your department has already moved to balanced crystalloid and is content, nothing here obliges a change. If starch is still stocked and its use has been quietly discouraged, this is the evidence that the discouragement was borrowed from a different population.

• Check what your hospital's fluid policy actually says about HES, and on which authority's warning it is based
• Distinguish bounded intraoperative use from prolonged ICU resuscitation whenever this evidence is cited
• Renal outcomes here are AKI incidence and creatinine change, not dialysis or long-term function
• The regulatory positions quoted are the FDA's and the EMA's; what CDSCO currently permits is not addressed in this paper and is worth confirming locally
• Author competing interests matter on a fluid question with a commercial sponsor behind the product — read the declaration before quoting the paper

The statistics, in plain English

A relative risk of 1.02 with an interval from 0.91 to 1.16 straddles 1.0, meaning no difference was found; the narrowness of that interval, plus a trial sequential analysis reporting adequate information size, makes this a reasonably confident absence of harm rather than an underpowered null. That is a stronger claim than most negative meta-analyses can make. It remains an absence of evidence of renal harm at the doses and durations studied, not proof of safety in every setting.

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.