Perfusion-guided resuscitation in septic shock: a mortality signal that just missed significance

Design

Systematic review with frequentist and Bayesian meta-analysis of RCTs

Population

7 RCTs, 2,408 adults with septic shock

Primary outcome

28-day mortality

Effect

RR 0.87 (0.76–1.01), I² 0%; posterior probability of benefit 97.2%

A meta-analysis in Critical Care Medicine (11 September) pooled seven randomised trials, 2,408 adults with septic shock, comparing resuscitation guided by peripheral perfusion — mostly capillary refill time — with standard care or lactate-guided resuscitation.

28-day mortality had a risk ratio of 0.87 (95% CI 0.76 to 1.01; p = 0.06), with no heterogeneity. The prespecified Bayesian analysis put the probability of any mortality benefit at 97%, stable across different prior assumptions. But the two largest, lowest-risk trials were each non-significant on their own, and the size of any benefit is uncertain.

For the emergency department this supports capillary refill time as a legitimate resuscitation target: it is quick, free, repeatable at the bedside and does not wait for a lab. It does not show that perfusion targets are superior to lactate, and it does not replace clinical judgement about fluid responsiveness. Where lactate is slow or unavailable — common in smaller Indian emergency departments — capillary refill is a reasonable guide in its own right.

• Use capillary refill time as a bedside resuscitation target in septic shock, alongside other measures.
• Recheck capillary refill every 30–60 minutes during early resuscitation.
• A normalised capillary refill may be a reason to pause further fluid, not push more.
• Where lactate is delayed or unavailable, capillary refill is a reasonable guide.

Why it matters

A free, instant bedside sign performed at least as well as lactate-guided care, with a lean toward benefit.

Don't overread it

The primary analysis was not statistically significant; this is a signal, not proof of lower mortality.

The statistics, in plain English

A risk ratio of 0.87 means 13% fewer deaths, but the confidence interval reaches 1.01, which includes no effect, so the frequentist result is not significant. The Bayesian analysis asks a different question — how likely is any benefit — and answered 97%. Both agree the direction favours perfusion guidance; neither pins down the size.

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.