JAMA February 19, 2019

Effect of a Resuscitation Strategy Targeting Peripheral Perfusion Status vs Serum Lactate Levels on 28-Day Mortality Among Patients With Septic Shock: The ANDROMEDA-SHOCK Randomized Clinical Trial

Glenn Hernández, Gustavo A Ospina-Tascón, Lucas Petri Damiani et al.

Bottom Line

In adults with early septic shock, a resuscitation strategy targeting the normalization of capillary refill time did not significantly reduce 28-day mortality compared to a strategy targeting serum lactate levels, though it was associated with less early organ dysfunction.

Key Findings

1. At28days, all-causemortalityoccurredin34.9%(74of212)ofpatientsinthecapillaryrefilltime(CRT)targetedgroupversus43.4%(92of212)inthelactate-targetedgroup, failingtoreachstatisticalsignificance(HR0.75;95%CI, 0.55-1.02;P=0.06)[1.1].

2. The absolute risk difference for 28-day mortality was -8.5% (95% CI, -18.2% to 1.2%) in favor of the CRT-guided approach.

3. Patients assigned to the CRT group had significantly less organ dysfunction at 72 hours, reflected by a lower mean SOFA score (5.6 vs 6.6; mean difference, -1.00; 95% CI, -1.97 to -0.02; P=0.045).

4. Within the first 8 hours, the peripheral perfusion-guided group received significantly less fluid and had a lower incidence of epinephrine use compared to the lactate-guided group.

Study Design

Design

RCT

Open-Label

Sample

424

Patients

Duration

28 days

Median

Setting

5 countries

Population Adult patients (mean age 63) with early septic shock, defined as having suspected or confirmed infection, a lactate level >2.0 mmol/L, and requiring vasopressors to maintain a mean arterial pressure (MAP) >=65 mm Hg after initial fluid resuscitation.

Intervention Peripheral perfusion-targeted step-by-step resuscitation protocol aimed at normalizing capillary refill time (<=3 seconds), assessed every 30 minutes during an 8-hour intervention period.

Comparator Lactate level-targeted step-by-step resuscitation protocol aimed at normalizing lactate or decreasing it by >20% every 2 hours during an 8-hour intervention period.

Outcome All-cause mortality at 28 days.

Study Limitations

• The open-label design could not mask clinicians to the assigned resuscitation strategy, introducing potential performance bias in adjunctive treatments.

• The trial was powered to detect an aggressive absolute mortality reduction of 15%; therefore, the non-significant 8.5% absolute risk reduction suggests the study may have been underpowered to detect a clinically meaningful, albeit smaller, difference.

• Assessments of capillary refill time were performed manually using a glass slide, which, despite a standardized protocol, retains an element of subjectivity compared to automated technologies.

Clinical Significance

Although the ANDROMEDA-SHOCK trial formally missed its primary mortality endpoint (P=0.06), the strong trend toward improved survival, combined with significantly less organ dysfunction and reduced fluid administration, challenged the paradigm of strict lactate clearance. It established capillary refill time—a rapid, non-invasive, and cost-free physical exam finding—as a highly viable, safe, and potentially superior clinical target for guiding hemodynamic resuscitation and fluid stewardship in early septic shock.

Historical Context

For nearly two decades following the early goal-directed therapy (EGDT) era, the Surviving Sepsis Campaign strongly endorsed trending serum lactate to gauge tissue hypoxia and guide fluid resuscitation. However, recognizing that hyperlactatemia in sepsis often results from adrenergic stimulation, mitochondrial dysfunction, or impaired hepatic clearance rather than pure macro-hemodynamic hypoperfusion, investigators designed ANDROMEDA-SHOCK. It was the first large-scale randomized trial to rigorously evaluate a 'low-tech' bedside marker of peripheral perfusion (capillary refill) against a 'high-tech' biochemical target (lactate clearance), paving the way for personalized, less fluid-heavy resuscitation strategies.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

How does capillary refill time (CRT) reflect peripheral perfusion and microcirculatory status differently than serum lactate in a patient with early septic shock?

Key Response

Lactate can be elevated due to Type B mechanisms like epinephrine-driven aerobic glycolysis or impaired hepatic clearance, rather than pure tissue hypoxia. CRT directly assesses microvascular flow and sympathetic tone, making it a more specific, real-time marker for macro- and micro-hemodynamic coherence.

Resident

If a septic shock patient's CRT normalizes but their lactate remains elevated at 4.0 mmol/L after initial fluid resuscitation, how does the ANDROMEDA-SHOCK protocol alter your next management steps compared to traditional lactate-clearance algorithms?

Key Response

Traditional algorithms might prompt further fluid boluses or inotropes to clear the lactate. The ANDROMEDA-SHOCK approach suggests that if CRT is normal, the microcirculation is adequately perfused, and further fluids or vasoactive agents might cause harm (e.g., volume overload), as the elevated lactate is likely due to impaired clearance or beta-adrenergic drive.

Fellow

The ANDROMEDA-SHOCK protocol utilized a specific sequential algorithm (fluid responsiveness testing, vasopressors, then inodilators) to normalize CRT. How does the integration of strict fluid responsiveness testing within this algorithm explain the lower SOFA scores observed in the CRT group?

Key Response

The CRT group received significantly less intravenous fluid overall. By combining a rapidly responsive microcirculatory target (CRT) with strict fluid tolerance tests, the protocol prevented venous congestion, endothelial glycocalyx degradation, and subsequent organ edema, thereby reducing early organ dysfunction compared to the slower-to-respond lactate group.

Attending

Given the primary outcome for 28-day mortality narrowly missed statistical significance (p=0.06) but secondary outcomes strongly favored the CRT group, how do you incorporate these findings into bedside teaching regarding 'negative' trials and the risks of over-resuscitation?

Key Response

While a strict frequentist interpretation labels this a 'negative' trial for mortality, the secondary outcomes (less organ dysfunction, lower fluid balance) and subsequent Bayesian analyses suggest a high probability of benefit. Bedside teaching should emphasize avoiding binary thinking; CRT is a free, non-invasive, real-time parameter that is at least non-inferior to lactate and protects patients from the iatrogenic harms of over-resuscitation.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The ANDROMEDA-SHOCK trial was powered for a 15% absolute risk reduction in mortality. How does this highly optimistic powering assumption impact the frequentist interpretation of the p=0.06 result, and how would a Bayesian framework alter the conclusions drawn?

Key Response

Powering for a 15% absolute risk reduction in modern critical care trials almost guarantees underpowering, as true effect sizes are typically much smaller. In a frequentist model, this leads to a Type II error. A Bayesian re-analysis demonstrated a >90% probability that CRT-targeted therapy improves survival, highlighting the limitations of frequentist thresholds when baseline assumptions are overly optimistic.

Journal Editor

As a peer reviewer evaluating this unblinded pragmatic trial, how would you scrutinize the potential for performance bias given that treating clinicians were aware they were targeting a novel clinical sign (CRT) versus a standard lab value (lactate)?

Key Response

Unblinded trials risk performance bias, where enthusiasm for the novel intervention might lead to closer overall bedside attention. A rigorous reviewer would scrutinize protocol adherence rates, the frequency of bedside assessments, and co-interventions to ensure the control group received equitable, protocolized standard-of-care attention.

Guideline Committee

The Surviving Sepsis Campaign historically strongly recommended targeting lactate normalization. Based on ANDROMEDA-SHOCK, how should guidelines grade the recommendation for using CRT as a resuscitation target, and what specific modifications should be made to existing algorithms?

Key Response

The 2021 Surviving Sepsis Campaign guidelines updated their stance to suggest using CRT as an adjunct or alternative to lactate (weak recommendation, low quality of evidence), explicitly citing ANDROMEDA-SHOCK. The rationale is that while the primary outcome was non-significant, the physiological rationale, lack of harm, reduced fluid volume, and feasibility in resource-limited settings strongly support its inclusion in initial algorithms.

Clinical Landscape

Noteworthy Related Trials

2001

Rivers Trial

n = 263 · NEJM

Tested

Early goal-directed therapy (EGDT)

Population

Patients with severe sepsis or septic shock

Comparator

Standard therapy

Endpoint

In-hospital mortality

Key result: EGDT significantly reduced in-hospital mortality compared to standard therapy (30.5% vs 46.5%).

2014

ProCESS Trial

n = 1,341 · NEJM

Tested

Protocol-based EGDT or protocol-based standard therapy

Population

Patients with septic shock in the emergency department

Comparator

Usual care

Endpoint

60-day in-hospital mortality

Key result: There was no significant difference in 60-day mortality between protocol-based EGDT, protocol-based standard therapy, and usual care.

2014

SEPSISPAM Trial

n = 776 · NEJM

Tested

High mean arterial pressure (MAP) target (80-85 mmHg)

Population

Patients with septic shock requiring vasopressors

Comparator

Low MAP target (65-70 mmHg)

Endpoint

28-day mortality

Key result: Targeting a higher MAP did not reduce 28-day mortality compared to a lower MAP target, though it reduced renal replacement therapy needs in chronic hypertension patients.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis