UMEM Educational Pearls - By Jessica Downing

Last month, Mark Sutherland posted an overview of a new article investigating the use of personalized MAP targets in resuscitation for septic shock (1). Now, the authors of ANDROMEDA-SHOCK-2 (2) suggest a new multimodal approach to personalize resuscitation in septic shock that largely operates outside of the traditional focus on MAP and lactate.

In 2019, the ANDROMEDA-SHOCK Trial (3) suggested that capillary refill time (CRT) may be a better resuscitation in septic shock than lactate. Now, the same group is suggesting that a stepwise algorithm to guide resuscitation may provide more optimal and “personalized” results when compared to usual care for patients with abnormal CRT:

Tier 1: If CRT is abnormal, assess pulse pressure (PP) and DBP: 

• PP (<40mmHg) OR DBP (>50mmHg)?  Assess for fluid responsiveness and challenge with up to 1L IVF if fluid responsive.
• PP (>40mmHg) AND DBP (<50mmHg)? Increase norepinephrine (NE) for DBP >50mmHg, followed by assessment for fluid responsiveness and possible fluid challenge if CRT remains abnormal.

Tier 2: If CRT remains abnormal despite the above, use POCUS to assess for cardiac dysfunction.

• LV dysfunction? Trial dobutamine @ 5-75 mcg/kg/min (stop for HR > 120 or tachyarrhythmia, or if it doesnt help CRT)
• RV dysfunction? The authors recommend avoiding fluids, increasing pressors if needed, as well as decreasing PEEP, limiting plateau pressures, and/or proning the patient if they have ARDS
• If there is no cardiac dysfunction, assess for fluid responsiveness and fluid challenge if fluid responsive. Continue this cycle until CRT normalizes or there is evidence of harm (evidence of pulmonary edema, worsening oxygenation/ventilation, or high central venous pressure).
  • If the patient is not fluid responsive, investigate for a history of HTN
    • If they have a history of HTN, push MAP to 80-85 for 1h and see if it improves CRT (if not, revert back)
    • If they do not, trial dobutamine @ 5 for 1h and see if it improves CRT (if not, DC dobutamine)
• If CRT remains abnormal after all of this, move on to “rescue therapies” (high dose steroids, hemofiltration, ECMO).

The authors found that at 6 hours, following the protocol resulted in increased use of dobutamine, lower fluid balance, and similar CVP and MAP with lower lactate levels and CRT. They reported an improvement in their composite hierarchical outcome at 28 days, primarily driven by a shorter duration of organ support (vasoactives, mechanical ventilation, renal replacement therapy) and among sicker patients. No difference in mortality was observed between groups.

Food for Thought:

• CRT is a subjective assessment, and all participating clinicians in this study underwent mandatory training. Other, more objective measures that can be used to assess organ perfusion (lactate clearance, urine output, MAP) were excluded from this protocol.
• Fluid responsiveness was assessed using the “preferred technique by each center.” Suggested techniques included pulse pressure/stroke volume variation, change in VTI with passive leg raise, IVC variability, or change in CO with end expiratory pause.
• There was no standardization regarding the integration of vasopressin, and steroids were reserved as “rescue therapies” and considered at the same point as mechanical hemodynamic support. 
• The effect of other inotropic agents (like low dose epinephrine) was not discussed.

Study Details:

• Setting: multicenter randomized control trial conducted in 86 ICUs across 19 countries
• Patients:
  • >1450 adults from the ED, ICU, OR, or floor with septic shock based on Sepsis-3 criteria - suspected/confirmed infection + lactate >2 + pressor requirement to maintain MAP> 65 despite 1L+ IVF bolus.
  • Patients with Child B or C cirrhosis, acute hematologic malignancy, severe ARDS, or anticipated surgery or HD within 8h of being diagnosed with septic shock, or who could not be enrolled within 4h of diagnosis were excluded. Pregnant patients were also excluded.
  • Almost half had an abdominal source of infection, followed by respiratory and urinary.
  • The median time to enrollment was 2h from meeting sepsis criteria
• Outcome: a hierarchical composite outcome using all-cause mortality, duration of vital support (vasoactives, invasive mechanical ventilation, or renal replacement therapy), and hospital LOS at 28d.
  • The primary outcome was assessed in a hierarchical fashion using “wins” and “losses.” The intervention group “won” in 48.9% of cases, while the usual care group “won” in 42.1%, for a Stratified Win Ration of 1.16 (95% CI 1.02-1.33).

Show References

The role of sodium bicarbonate in the treatment of severe acidemia has been controversial, with some studies suggesting no benefit, and others indicating that it may help reduce need for renal replacement therapy (RRT) and even improve mortality. The BICARICU-2 Trial was an open-label multicenter RCT conducted in France that evaluated the impact of a bicarb infusion among patients with metabolic acidosis and moderate to severe AKI. 

There was no difference in 90 day mortality, but patients in the bicarb group were less likely to be started on RRT (38% vs 47% in the control group) using pre-defined criteria for RRT initiation, and had a 50% lower rate of bloodstream infections. Patients in the bicarb group who were started on RRT met criteria for RRT later than those in the control group (median 31h vs 15.5h).

Study Details:

Patient Population: 

• SOFA score >4 OR arterial lactate > 2mmol/L within 48h of ICU admission
• Metabolic acidosis, defined by pH < 7.2, HCO3- < 20mEq/L, and PaCO2  < 45mmHg
• Moderate to severe AKI, defined as Cr >2.0 x baseline or UOP < 0.5 mL/kg/h for >12h. 
• Patients with severe baseline CKD, ketoacidosis, intoxication with exogenous acids (metformin, salicylate, methanol, ethylene glycol), or ongoing bicarb losses via GI or urinary tracts were excluded.
• The presumed etiology of acidemia was septic shock in over half of included patients, and over 75% were on vasopressors.

Intervention: 

• 4.2% bicarb infusion administered in 125-250 aliquots with a target pH >7.3, though not to exceed 1L/500mEq within 24h. 
• The intervention continued for a maximum of 28d or until ICU DC. 
• Patients in the intervention group received a median of 750mL in the first 48h.

RRT Triggers:

• Immediate: K > 6.5mEq/L with EKG changes or cardiogenic pulmonary edema with no UOP and hypoxia
• 24h after enrollment: UOP <0.3 Ml/kg/h over 24h, pH <7.2 despite resuscitation, K > 6.5 MEq/L.

Show References

Norepinephrine (NE) is widely accepted as the first-line vasopressor for the management of septic shock, supported by the Surviving Sepsis Guidelines (1). The use of vasopressin as a second-line agent is also supported by the Surviving Sepsis Campaign, although the appropriate “triggers” for its addition remain vague. The SSG recommend adding vasopressin when NE infusion rates reach 0.25-0.6 mcg/kg/min, citing a catecholamine-sparing effect and potentially improved mortality (1, 2, 3).

What’s New?

The OVISS study (“Optimal vasopressin initiation in septic shock. The OVISS reinforcement learning study”) used machine learning to derive and internally validate a set of rules guiding the addition of vasopressin to NE for patients with septic shock using multiple databases of patient encounters across multiple institutions (4). 

The machine learning model suggested initiation of vasopressin in more patients (87% vs 31%), earlier,  and in less sick patients than was seen to be common practice:

• Timing: 4h after diagnosis of shock (vs. 5h)
• NE dose: 0.2 mcg/kg/min (vs. 0.37mcg/kg/min)
• Serum lactate: 2.5 mmol/L (vs. 3.6 mmol/L)
• SOFA score: 7 (vs. 9)

Practice consistent with the above triggers was associated with decreased odds of in-hospital mortality (AOR 0.81, 95% CI 0.73-0.91).

Limitations

This was not a prospective study or RCT and was only internally validated. Using databases may limit the number of clinical variables available for analysis, and clinical judgment (how the patient looks) is not reflected.

Bottom Line

Consider adding vasopressin for patients with vasodilatory shock with low MAP despite NE >0.2mcg/kg/min and adequate fluid resuscitation, though more evidence is needed for a strong recommendation. As dual-pressor therapy may be riskier via peripheral IV and vasopressin does not have a direct antidote for extravasation, consider central line placement when adding vasopressin (5,6)

Show References