Dapagliflozin added to sepsis care did not change SOFA scores in a small trial

Design

Single-centre, prospective, randomised controlled trial, preliminary

Population

136 ICU patients with sepsis randomised; 50 per group analysed

Primary outcome

Change in SOFA score after 5 days of dapagliflozin 10 mg

Effect

No significant difference in SOFA change; 28-day mortality 26% vs 36% (P = 0.387)

This single-centre randomised trial in an ICU in China enrolled 136 patients with sepsis, of whom 100 (50 per group) were analysed. All received standard sepsis treatment, and the intervention group also received dapagliflozin 10 mg for 5 days. The primary outcome was change in SOFA (sequential organ failure assessment) score.

SOFA changes did not differ between groups. By day 5, median left ventricular ejection fraction was higher with dapagliflozin (60.0% vs 55.0%; P < 0.001). Twenty-eight-day mortality was 26% (13 of 50) with dapagliflozin and 36% (18 of 50) without, a difference that was not significant (P = 0.387). Adverse events were similar, but urine ketone levels were higher in the dapagliflozin group (P = 0.040), although no ketoacidosis occurred. Kidney function was a secondary outcome, and the abstract does not report its results.

The trial was preliminary, with a primary outcome that was negative, 100 patients analysed, and secondary findings that are hypothesis-generating at best. The ketone finding is worth noting because sepsis and SGLT2 inhibitors can both predispose to ketosis. It does not support using dapagliflozin in sepsis.

• Do not use dapagliflozin for sepsis outside a trial on the basis of this study.
• The primary outcome, SOFA score, did not change.
• The mortality difference (26% vs 36%) was not significant with only 50 patients per group.
• Urine ketones were higher with dapagliflozin; keep ketone monitoring in mind when SGLT2 inhibitors are used in acute illness.

Why it matters

It reflects a growing interest in SGLT2 inhibitors beyond diabetes, heart failure and CKD, and shows how little evidence exists in acute illness.

Don't overread it

The primary outcome was negative and the numbers are small, so the ejection fraction and mortality differences are hypothesis-generating at most.

The statistics, in plain English

A P value of 0.387 for mortality means a difference of this size could easily arise by chance with 50 patients per group. The result neither supports nor excludes a benefit.

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.