Thread by @WellingtonICU

#twitterjournalclub
The ANDROMEDA- SHOCK Randomised Clinical Trial
Thoughts by @z_cepillo
jamanetwork.com/journals/jama/…

Multicentre (28) open label RCT to determine if peripheral perfusion targeted resuscitation during early septic shock in adults is more effective than a lactate-targeted resuscitation for reducing mortality.

Inclusion criteria: adult patients with septic shock (infection, lactate >2, vasopressors to maintain MAP >65mmHg after IV fluid of 20ml/kg over 60 mins.

Exclusion criteria: bleeding, severe ARDS, and do-not-resuscitate status.

Sample size calculated: 90% power to reduce 28 day mortality from 45% (lactate group) to 30% (CRT group)

Patients randomly assigned in 1:1 ratio to either peripheral perfusion or lactate targeted resuscitation groups. (Note randomisation was NOT stratified according to different research centres).

The goals of the peripheral perfusion and lactate group were: to achieve a normal CRT (<3seconds), and normalise or decrease the lactate level by 20% every two hours respectively.

Targeted resuscitation was performed using a standardised algorithm:
i) Check for fluid responsiveness. If fluid responsive, the patient received 500ml boluses of fluid until resuscitation targets were met or the patient became unresponsive to fluid...

ii) If the previous interventions did not meet the target goals; in patients with chronic hypertension a trial of vasopressors to maintain a MAP of 80-85mmHg was initiated...

iii) If both the previous interventions did not meet the target goals a trial of low dose ino-dilator was commenced. If the inodilator did not have any significant effect, it was discontinued

Primary outcome:
All-cause mortality at 28 days.

Secondary outcomes
Death within 90 days, organ dysfunction in the first 72 hours after randomisation, mechanical ventilation-free days within 28 days; vasopressor-free days within 28 days, RRT-free days within 28days, and length of stay (ICU & Hospital).

424 patients randomised (212 in peripheral perfusion group. 212 in lactate group). By day twenty-eight, 74 patients (34.9%) in the peripheral perfusion group and 92 (43.4% in the lactate group had died (hazard ratio 0.7 [95% CI, 0.55 to 1.02] P=0.06.

Significant secondary outcomes: there was significantly less organ dysfunction at 72 hours after randomisation in the peripheral perfusion group. The peripheral perfusion group received significantly less resuscitation fluid in the first 8 hours.

In septic shock, a resuscitation strategy targeting normalization of CRT compared to a strategy targeting lactate levels, did not reduce all-cause 28 day mortality.

Here are some limitations:

The implementation of the author’s resuscitation protocol could lead to fluid overload as well as complications associated with fluid resuscitation (e.g. glycocalyx loss).
link.springer.com/article/10.100…

Protocolised use of an inodilator (in failing goal directed therapy) may appear somewhat unusual to practitioners because of the widespread availability focused cardiac ECHO.

The authors have powered to study to demonstrate that CRT would have a very large treatment effect when compared to the lactate group. Unfortunately the study was underpowered to prove any smaller mortality effect.

Despite the study being grossly underpowered for a realistic treatment effect, the P value was 'almost' statistically significant.

This is a great situation for a Bayesian analysis that takes into account prior beliefs. Check this out:
benjamin-andrew.shinyapps.io/andromeda_shoc…

...and for more discussion on Bayesian analysis of this trial have a look at this:
discourse.datamethods.org/t/andromeda-sh…

Overall I feel like there is enough evidence to incorporate CRT into my routine examination of the shocked patient to help guide management. It is cheap, easy, quick, informed and will bring a clinician back to the bedside to re-examine a patient after an intervention

Lactate’s role in the body may be too complex to serve as a tool for resuscitation but it still has prognostic value and frequently acts as a ‘call to arms’ for clinicians. ncbi.nlm.nih.gov/pmc/articles/P…

I am biased by my own scepticisms of frequentist analyses but at the very least we must all agree that ANDROMEDA-SHOCK suggests a robust hypothesis for future studies.
@z_cepillo @JAMAnetwork #sepsis #ICU #criticalcare #FOAMed #andromedashock #CCC48 #survivingsepsis

@threadreaderapp unroll