Persistent poor perfusion, not pressor dose, marks the deadliest septic shock

Design

Exploratory secondary analysis of a randomised trial (ANDROMEDA-SHOCK-2)

Population

1,363 adults with septic shock after protocolised resuscitation

Primary outcome

28-day mortality by refractoriness definition

Effect

Two-hypoperfusion criteria: 73.6% vs 23.7% mortality, adjusted HR 4.68 (3.31–6.64)

A secondary analysis of the ANDROMEDA-SHOCK-2 trial asked how best to define refractory septic shock. In 1,363 patients, 13.8% were on a high vasopressor dose (norepinephrine-equivalent above 0.5 µg/kg/min) six hours into resuscitation, and nearly half of them died.

Adding signs of persistent tissue hypoperfusion sharpened the picture. Patients who also had a capillary refill time over three seconds and a non-falling lactate — 3.9% of the cohort — had 73.6% mortality, against 23.7% in those not meeting the criteria (adjusted hazard ratio 4.68). This two-sign construct stratified mortality far better than vasopressor dose alone (positive likelihood ratio 8.12 vs 2.71).

The appeal for any ICU, and especially a resource-limited one, is that the extra information is free: capillary refill needs no device and lactate is widely available. It flags the patient who needs escalation or a frank conversation about ceilings of care — though as an exploratory analysis it should inform judgement, not a protocol, until validated.

• In septic shock on high-dose vasopressors, check capillary refill and the lactate trend, not the pressor dose alone.
• A refill time over 3 seconds plus non-falling lactate marked a group with about 74% mortality.
• This combination predicted death far better than vasopressor dose alone.
• Both signs are bedside and cheap, suiting resource-limited intensive care.

Why it matters

It defines refractory shock by perfusion, not just drug dose, flagging who needs escalation or a goals-of-care conversation.

Don't overread it

This is an exploratory, hypothesis-generating secondary analysis of a small refractory subgroup (53 patients) — a prognostic signal to validate, not a protocol.

The statistics, in plain English

A positive likelihood ratio of 8.12 versus 2.71 means the two-sign construct shifts the probability of death far more than pressor dose alone; but with only 53 patients meeting it, the exact mortality figure is imprecise.

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.