A double-blind, placebo-controlled trial across seven Dutch hospitals randomised 784 adults undergoing elective cardiac surgery to dapagliflozin 10 mg or placebo, starting the day before surgery and continuing to the second postoperative day — four doses in total. The primary outcome was acute kidney injury by KDIGO criteria within seven days.
Acute kidney injury occurred in 28% of the dapagliflozin group and 52% of the placebo group, a relative risk of 0.54 with a confidence interval from 0.45 to 0.65. Follow-up was 99% complete. Atrial fibrillation occurred in 45% of both arms and reoperation in 11% versus 10%, so the benefit did not come with an obvious safety cost in this population.
Two cautions before this changes anything. The placebo AKI rate of 52% is at the very top of the 2% to 50% range the authors themselves cite, which suggests a sensitive definition — creatinine rises of 0.3 mg/dL and short periods of low urine output both count. A halving of a soft, common endpoint is not the same as a halving of dialysis or death, and this trial was not powered for those. The cohort was also 97% White with a median eGFR of 80, so it says little about the patient with established chronic kidney disease in whom you would most want an answer.
- Elective surgery only, and started the day before — not applicable to an emergency case
- KDIGO stage 1 dominates this endpoint; harder renal outcomes were not powered for
- No excess atrial fibrillation or reoperation, which is the safety question that mattered here
- Median eGFR was 80, so the established-CKD patient remains unstudied
- Consider whether your unit's own AKI rate resembles the 52% seen in the placebo arm
The statistics, in plain English
A relative risk of 0.54 with an interval of 0.45 to 0.65 is a precise estimate — the interval is narrow and nowhere near 1.0 — so the effect on this endpoint is real. Precision is not the same as importance: the endpoint counts a creatinine rise of 0.3 mg/dL, which most patients will never notice, alongside injury that changes a hospital stay. With 784 patients the trial can measure that composite well and cannot say anything about dialysis or death.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for cardiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free